Desalination 343 (2014) 187–193

Contents lists available at ScienceDirect

Desalination

j ourna l homepage: www.e lsev ie r .com/ locate /desa l

A new commercial thin film composite membrane for forward osmosis

Jian Ren, Jeffrey R. McCutcheon ⁎Department of Chemical and Biomolecular Engineering, Center for Environmental Science and Engineering University of Connecticut, 191 Auditorium Rd. Unit 3222, Storrs, CT 06269-3222, USA

H I G H L I G H T S

• A new thin film composite forward osmosis membrane was evaluated.• The membrane was made by Hydration Technology Innovations.• This membrane was produced on a 40-inch continuous line.• The membrane was shown to have excellent performance under FO conditions.

⁎ Corresponding author. Tel.: +1 860 486 4601.E-mail address: jeff@engr.uconn.edu (J.R. McCutcheon

0011-9164/$ – see front matter © 2013 Elsevier B.V. All rihttp://dx.doi.org/10.1016/j.desal.2013.11.026

a b s t r a c t

a r t i c l e i n f o

Article history:Received 6 August 2013Received in revised form 28 October 2013Accepted 4 November 2013Available online 19 December 2013

Keywords:Forward osmosisPressure retarded osmosisThin film composite

New membranes for forward osmosis (FO) have been made by numerous academic groups around the world.Few of these designs, however, have made it to full-scale production. For two decades, the only FO membranemade on a full-scale production line was a cellulose acetate membrane from Hydration Technology Innovations(HTI). Only recently have other companies designed new membranes and produced them on a large scale, butthose membranes are still largely unavailable to academic researchers. In this study, we report on a newlylaunched forward osmosis membrane from HTI. This thin film composite (TFC) membrane is a departure fromtheir cellulose acetate platform and is among if not the first TFC membrane to be made on a 40-inch line. TheTFC membrane tested, which is their first generation TFC membrane, exhibited high water permeance andgood mechanical strength relative to other membranes discussed in the academic literature. Under FO tests,the membrane achieved high water flux of 46.4 and 22.9 L m−2 h−1 with a modest salt flux of 24.9 and6.4 g m−2 h−1 using 1 M sodium chloride against deionized water in pressure retarded osmosis (PRO) and FOmodes, respectively.

© 2013 Elsevier B.V. All rights reserved.

1. Introduction

Seawater desalination and wastewater reuse have received world-wide attention to alleviate the water stress caused by populationgrowth and increasing resource consumption [1,2]. Forward osmosis(FO) has been touted as a high water recovery and low cost desali-nation option [2–4]. FO utilizes osmotic pressure difference todrive water across a semipermeable membrane from a dilute feed solu-tion to a concentrated draw solution while rejecting most solutes [5].Over the past decade, FO has attracted considerable attention in a num-ber of fields in both industrial application and academic research [3,4].

However, the development of FO has been hampered by the lack ofeffective membranes. Early studies were limited to existing commercialmembranes, most of whichwere designed for reverse osmosis (RO) [5].There have been a number of studies focusing on the development of

).

ghts reserved.

high performance membranes specifically for FO [6–8], but they are alllimited to lab scale fabrication techniques.

Since FO saw its resurgence in themiddle of the last decade, the onlycommercially available FO membrane has been the asymmetric cellu-lose acetate (CA) membrane from Hydration Technology Innovations(HTI, Albany, OR). The HTI CA membrane has an optimized structurefor FO consisting of a thin selective layer followed by a relatively looseand thin support layer embeddedwith amesh for strength [3,9]. The hy-drophilic nature of cellulose acetate as thematrixmaterial favors properwetting compared with hydrophobic membranes, but is susceptible tohydrolysis [10,11]. Moreover, concerns about low water flux and highsaltflux due to the relatively poorwater permeability and selectivity, re-spectively, of CA membranes have limited the use of FO to niche appli-cations [12].

In this study, a newly designed TFC membrane from HTI is intro-duced as a commercially available FO membrane which is made in acontinuous process on a 40-inch production line. This TFC membraneinherits the mesh-embedded structure from the CA membrane but

188 J. Ren, J.R. McCutcheon / Desalination 343 (2014) 187–193

surpasses it by tailoring a porous support layer that promotes highwater flux, low salt crossover, and hydrolytic resistance.

2. Experimental

2.1. Materials

Thin film composite (TFC)membranes and asymmetric cellulose ac-etate (CA) membranes were provided by HTI. The TFC membrane isconsidered by HTI to be their early generation membrane. Both the CAand TFC membranes are fabricated on a 40-inch continuous productionline. 2-Propanol (isopropyl alcohol, IPA, anhydrous, 99%)waspurchasedfrom J.T.Baker (Center Valley, PA). Red food coloring fromMcCormick &Company Inc. (Sparks, MD) was used to ensure integrity of the mem-brane. For the osmotic flux tests, sodium chloride (NaCl, crystalline, cer-tified ACS, Fisher Scientific) and deionized (DI) water from a MilliporeIntegral 10 water system (Millipore, USA) were used.

2.2. Membrane preparation

Someof the TFCmembraneswerewetted using a 50 wt.% solution ofisopropyl alcohol (IPA) for 5 min at room temperature [9,13]. The IPAwas then thoroughly rinsed out of the membranes using DI water andstored at 5 °C in DI water. These are referred to as prewetted TFC inthis study. CAmembraneswere not prewetted since they easily saturatewhen exposed to water.

2.3. Membrane characterization

2.3.1. Scanning electron microscopyThe surface morphology and cross-sectional structure of the TFC

membranes were imaged with scanning electron microscopy (SEM). Acold cathode field emission scanning electron microscope JSM-6335F(FESEM, JEOL Ltd., Japan) and a FEI Phenomdesktop SEM (FEI Company,OR) were used for surface and cross-sectional morphology imaging, re-spectively. To view the cross sections of the membranes, the sampleswere submerged in liquid nitrogen to preserve the pore structure andcutwith a razor blade. Prior to imaging, the sampleswere sputter coatedwith a thin layer of gold.

2.3.2. Contact angleThe contact angles of the selective and support layers of the TFC

membrane were measured using the sessile drop method on a CAM101 series contact angle goniometer (KSV Company Linthicum Heights,MD). The values were taken as an average of six points with a dropletvolume of 10 ± 1 μL. All measurements were taken at roomtemperature.

2.3.3. Attenuated total reflection Fourier-transform infrared spectroscopyAttenuated total reflection Fourier-transform infrared (ATR-FTIR)

spectroscopy was used to study the materials of the selective and sup-port layers of the TFC membrane. ATR-FTIR spectra were obtainedusing a Jasco 670 plus FTIR spectrometer equippedwith an ATR element(45° multi-reflection germanium crystal).

2.4. Water permeance and salt permeability

Pure water permeance, salt permeability and salt rejection – of theCA and TFC membranes – were evaluated in a laboratory-scale cross-flow RO test unit described elsewhere [6,14]. McCormick™ red foodgrade dye (1 mL) was added into feed water (10 mL) to detect pin-holes. Pure water permeance was measured at 20 ± 0.5 °C and aver-aged over four pressures ranging from 100 to 250 psi. Pure water flux(Jw) was calculated by dividing the volumetric permeate rate by themembrane area andmeasured from at least four samples. Salt rejection(R) tests were conducted using a feed solution of 2000 mg/L NaCl and a

feed pressure of 125 psi. Intrinsic water permeance and salt permeabil-ity were derived by Yip et al. [7] and assumed to be constant and inde-pendent of pressure and salt concentration. The salt permeability (B)was determined from [7,10,15]

B ¼ Jw1−RR

� �exp − Jw

k

� �ð1Þ

where k is themass transfer coefficient for the cross-flow channel of theRO membrane cell [12].

2.5. Osmotic flux testing

Osmotic water flux and reverse salt flux through CA and TFC mem-branes were characterized using a custom lab-scale cross-flow forwardosmosis system. The experimental setupwas described in earlier inves-tigations [9,12]. Osmotic flux tests were carried out with themembraneoriented in both FOmode (the membrane selective layer faces the feedsolution) and PRO mode (the membrane selective layer faces the drawsolution). Two testing conditions – a recently published standardmeth-odology [13] and one suggested by HTI – were used.

2.5.1. Standard methodAs new membranes are developed, especially commercial mem-

branes, it is necessary to test performance under a standard protocolto make reasonable comparisons with other membranes. Recently,Cath et al. developed a method that was intended to standardize FOmembrane testing [13]. In this method, water and salt fluxes weremea-sured at 20 ± 0.5 °C using DI feed and 1 M NaCl draw solution. Thecross flow velocities were maintained at 0.25 m/s on both sides of themembrane and the Reynolds number in both channels was set to1125. No hydraulic transmembrane pressure or channel spacers wereused. As in previously described methods for testing performance ofFO membranes, the mass of draw solution reservoir was constantlymonitored on a scale which outputs data to a computer. The osmoticwater flux (Jw) was calculated by normalizing the volumetric flow rateby the effective membrane area [9,16]. Similarly, the salt flux (Js) wascalculated by dividing the NaCl mass flow rate by the membrane areaand was accomplished by measuring the conductivity of the feed solu-tions at certain time points during the tests.

2.5.2. HTI methodFor full scale FO operations, the HTI TFC membrane is most likely to

be used in a spiral wound element. An 8-inch diameter spiral woundTFC membrane element was developed by HTI and has been commer-cially available since August, 2012. To date only a handful of investiga-tions have been done on spiral wound FO membrane modules. Theeffect of draw solution concentration and operating conditions onwater flux was discussed by Xu et al in 2009 [17] while the effects ofstructural features were investigated by Kim and Park in 2011 [18]. Asnew membranes are developed, testing in elements is expensive andimpractical. Also, for comparison to other membranes developed inthe academic space, flat sheet studies are more appropriate. However,when conducting flat sheet studies, we can try to mimic the operatingconditions in a spiral wound element. HTI provided such a method.

For the HTI method, the temperature and draw solution concentra-tion were kept same as those in standard method at 20 ± 0.5 °C and1 M NaCl, respectively. To simulate the mass transfer near the mem-brane surface in spiral wound FO elements, turbulence enhancingspacers (diamond pattern, ~0.8 mm in thickness and ~2.5 mm in spac-ing) were used to fill the flow channel on both feed and draw sides ofthe membrane [18]. Furthermore, a cross flow velocity of 0.30 m/s(Reynolds number 1350) and a small transmembrane hydraulic pres-sure of 4 psiwere used aswell to better simulate the conditions in a typ-ical HTI element. Such a low pressure differential is not anticipated to

189J. Ren, J.R. McCutcheon / Desalination 343 (2014) 187–193

cause substantial water flux. McCormick™ red food grade dye wasadded to the high pressure side to detect the pin-holes.

2.6. Membrane structural parameter

As an asymmetric membrane, the HTI TFC membrane comprises athin selective layer supported by a porous support layer. Both experi-mental and modeling studies have shown that this support layer im-parts a resistance to solute diffusion and causes internal concentrationpolarization (ICP) [7,9,19–21]. The membrane structural contributionsto this phenomenon are defined using what is known as the structuralparameter, S. It is defined as the product of the thickness (t) and tortu-osity (τ), divided by the porosity (ε) (i.e., S = tτ ∕ ε) of themembranesupport layer.

In the experimental tests, the membrane effective structural param-eter can be determined using the empirical equation previously de-scribed [22],

S ¼ DJw

� �ln

Bþ AπD;b

Bþ Jw þ Aπ F;mð2Þ

where D is the diffusion coefficient of the draw solute, Jw is the mea-sured water flux in FO mode, πD,b is the bulk osmotic pressure of thedraw solution, and πF,m is the osmotic pressure at themembrane surfaceon the feed side (0 for DI feed).

3. Results and discussion

3.1. Characterization of membrane

3.1.1. Scanning electron microscopyThe FESEM images of the top (selective) and bottom (support) sur-

faces of the TFC membrane are shown in Fig. 1. The selective layer hasa ridge and valleymorphologywhich is a typical characteristic of a poly-amide layer formed via interfacial polymerization [7,20]. The selectivelayer shows a uniform and continuous morphology, without defects orpin-holes. The FESEM images of the bottom surface of the support

Fig. 1. Top surface SEM images (a, b and c) and bottom surface FESEM images (d, e and f) of TFC

layer as shown in Fig. 1d, e, and f show a porous structure with poresize ranging from 100 to 600 nm.

Fig. 2 shows the cross-sectional SEM images of the TFC membrane.The thickness of the membrane is uniform at ~115 μmwhile the diam-eters of the polyester fibers in the embedded mesh are ~50 μm. Thismesh provides most of the mechanical strength to the membrane,thereby eliminating the need for a thick porous support layer. A similarapproach was used in the design of their CA membrane. Fig. 2c and dshow that the selective layer adheres well to the denser part of the sup-port layer. This layer accounts for the integrity and uniformity of thepolyamide layer [7].

3.1.2. Contact angleThe relative hydrophilicity of both the selective layer and the sup-

port layer was measured by contact angle (Table 1). The selectivelayer showed a low contact angle (~14°) which implies a polyamidelayer that is more hydrophilic than the CAmembrane and other report-ed high performance FO membranes [7,9,20]. The contact angle of thesupport layer of the TFC membrane was not shown here because thewater droplet was absorbed during the measurement. This occurs onporous materials that are hydrophilic. We can assume that the materialis hydrophilic, but a comparative contact angle cannot bemeasured.Wecan say, however, that the support layer will wet easily when exposedto water.

3.1.3. Attenuated total reflection Fourier-transform infrared spectroscopyFig. 3 shows the ATR-FTIR spectra of both the support and selective

layers of the TFC membrane. The spectrum of the selective layershows peaks attributed to both the support and selective layermaterials. Arrows indicate the peaks that are specific to the selectivelayer [16,23]. The selective layer spectrum displays the characteristicpeaks of polyamide such as 1655 cm−1 (C_O stretching of amide),1610 cm−1 (aromatic ring), and 1545 cm−1 (C\N stretching ofamide) [16]. These peaks strongly suggest the likelihood that polyamideserves as the functional selective layer material.

membrane atmagnifications of (a and d) 2000×, (b and e) 10,000×, and (c and f) 50,000×.

Fig. 2. Cross-sectional SEM images of TFC membrane at magnifications of (a) 2000×, (b) 2020×, (c) 5800× and (d) 11,200×. Dotted boxes show the zooming sections.

190 J. Ren, J.R. McCutcheon / Desalination 343 (2014) 187–193

3.2. Intrinsic separation properties

The intrinsic water permeance (A), salt permeability (B), and salt re-jection of the CA, TFC and prewetted TFCmembranes selective layer arereported in Figs. 4 and 5. Fig. 4 shows that the salt rejections of the TFCmembranes are slightly lower than that of the CAmembrane. This is aninteresting result as most TFC type membranes normally exhibit far su-perior selectivity than their CA membrane counterparts. It is possiblethat the membrane properties have been optimized for FO and there-fore not designed to be tested under the relatively high pressure of RO[6,14]. However, it is still worth noting the potential of the TFC mem-brane for PRO applications since it was able to withstand the hydraulicpressure of 250 psi.

Table 1Measured contact angles of the selective and support layers of the membranes used.

Membranes Contact angle (°)

Selective Support

CA 62.0 ± 7.2 63.6 ± 13.0TFC 14.3 ± 1.6 N/A

Selective and support layer contact angles are included for the CA [9] and TFCmembranes.Temperature during themeasurementswas 21 ± 0.5 °C. The contact angle of TFC supportlayer was not measureable using the sessile drop method because the droplet wasadsorbed into the support layer during the measurement.

The water permeance of the TFC membrane is about two times thatof the CA membrane. This is consistent with other TFC membranes forRO exhibiting higher water permeance compared to CA ROmembranes.For the prewetted TFC, the water permeance is even higher, almost 50%higher than the virgin TFC and three times of the CA membrane. Alongwith this increase in water permeance, the salt permeability of theprewetted TFC membrane also increased. This is likely due to two pos-sible effects that IPA has on the polyamide selective layer [24–27].First, unreacted amine and lowmolecularweight products of the conden-sation reaction can be extracted by IPA from the selective layer. Removalof these smallmolecules resulted in amore open structure in the polyam-ide layer [26]. Second, the physical swelling of the polyamide chains wasexacerbated by the presence of IPA molecules. The lower polarity of IPAcomparedwithwater engaged in hydrogen bonding and non-polar inter-actionswith polyamide [26,27]. Thus, theweaker andmore flexible chaininteractionswithin the polyamide caused the enlargement of pore, or freeelement, size. Meanwhile, it is not surprising to observe significant vari-ability in the salt rejection and salt permeability of the prewetted TFC. Be-cause small membrane samples were tested and DI water feeds wereused, thedeviation is large since the sensitivity of a conductivitymeasure-ment is relatively high. Furthermore, deviation in selectivity from couponto coupon can cause substantial variability in salt flux, especially whenthese coupons are prewetted with an agent that may change membraneproperties. Large variability in salt flux has been observed in some previ-ous work on FO membrane investigations [6,7].

Fig. 3. ATR-FTIR spectrum of the TFC membrane support (black curve) and selective layer (gray curve). Arrows indicate peaks specific to the selective layer.

191J. Ren, J.R. McCutcheon / Desalination 343 (2014) 187–193

3.3. Osmotic flux results

The osmotic water fluxes of TFC membranes are shown in Fig. 6 forboth FO and PRO modes. In the standard method tests, the two TFCmembranes achieved nearly two times higher water fluxes than theCA membrane (Fig. 6). This is consistent with the higher A values ofthe TFC membranes and suggests a support structure that is moreopen and/or hydrophilic.

Prewetting preparation is normally used to ensure that the mem-brane porous support is fully water saturated [9,13]. This is especiallytrue with more conventional TFC chemistries that use polysulfone orother hydrophobic polymers as support materials. Generally, prewettedmembranes with hydrophobic supports show a higher water flux thanthe virgin membrane, consistent with other prewetting studies[24,25]. In our study,we note thatwater fluxes are generally unchangedafter prewetting with the one exception of the PRO mode testing usingthe HTI method. It is likely that the virgin TFC support is already easilysaturated in water and the prewetting preparation is unnecessary. An-other possible reason might be the negative effects of prewetting prep-aration on themembrane selective layer. As discussed in Section 3.2, theprewetting proceduremight enlarge the pore or free element size of thepolyamide layer. The resulting lower selectivity results in higher reversesalt flux and decreases the osmotic pressure difference across themem-brane. It can also worsen ICP. Therefore, the unchanged water flux is a

CA TFC Prewetted TFC70

75

80

85

90

95

100

Sal

t R

ejec

tio

n, R

, %

Membranes

Fig. 4. Salt rejections (%R) for the three membranes. Results are an average of four exper-iments with different coupons. Error bars indicate standard deviation. Operating condi-tions: feed pressure 8.62 bars (125 psi), feed temperature 20 °C, and feed flowvelocities 0.25 m/s (for CAmembrane) and 0.30 m/s (for TFC membranes). The Reynoldsnumbers for the flows in CA and TFC membranes were 1125 and 1350, respectively.

result of enhanced water permeance balanced with lower effective os-motic pressure and enhanced ICP.

Fig. 7 shows the reverse saltfluxes of the threemembranes in FO andPRO modes. It is worth noting that the TFC membrane, despite a two-fold higher water flux than the CAmembrane, has a comparable reversesalt flux. For the prewetted TFC membrane, it is not surprising to see amuch higher salt flux since it was shown to have a lower rejection inFig. 4.

By comparing the water flux tested by the two testing methods(Fig. 6), the HTI method gives a higher water flux in both FO and PROmodes. While the slight transmembrane pressure difference in the HTImethodmight be part of the reason, themore likely cause is the reducedECP in the method. The HTI method uses a higher cross-flow velocityand incorporates turbulence promoting spacers into the channel, bothof which facilitate mass transfer and reduce ECP.

3.4. Structural parameters

The structural parameters for the three membranes were calculatedaccording to Eq. (2) using the corresponding measured water flux datain FOmode. The S value for CAmembranes was calculated as the lowest~465 μm. The TFCmembrane, despite a thickness of more than twice ofthe CAmembrane (~100 μmvs ~50 μm), shows a comparable structureparameter of ~533 μm with CA. This suggests that the TFC membrane

TFC Prewetted TFC0.0

0.5

1.0

1.5

2.0

2.5

3.0Water Permeance, A, Lm-2h-1bar-1

Salt Permeability, B, Lm-2h-1

MembranesCA

Co

effi

cien

ts

Fig. 5. Pure water permeance (A) and salt permeability (B) for the three membranes. Re-sults are an average of four experiments with different coupons. Error bars indicate stan-dard deviation. Operating conditions: feed pressure 8.62 bars (125 psi), feed temperature20 °C, and feed flow velocities 0.25 m/s (for CA membrane) and 0.30 m/s (for TFC mem-branes). The Reynolds numbers for the flows in CA and TFC membranes were 1125 and1350, respectively.

0

10

20

30

40

50

60Standard methodHTI method

MembranesCA

Wat

er F

lux

(FO

mo

de)

, Lm

-2h

-1

TFC Prewetted TFC

TFC Prewetted TFC0

10

20

30

40

50

60Standard methodHTI method

Membranes

CA

Wat

er F

lux

(PR

Om

od

e),L

m-2

h-1

Fig. 6. Water flux of FO and PRO tests with three membranes. Results are an average ofthree experiments with different coupons. Error bars indicate standard deviation. Stan-dard method operating conditions: 1 M NaCl draw solution, deionized water feed, 20 °Cfeed and draw solution temperature, 0.25 m/s feed and draw solution cross flow velocity,0 transmembrane pressure and no spacers [13]. HTI method operating conditions: 1 MNaCl draw solution, deionized water feed, 0.30 m/s feed and draw solution cross flow ve-locity, and 20 °C feed and draw solution temperature. Spacers were used on both sides. 1and 5 psi hydraulic pressures (gage pressures) on draw and feed sides in FO mode. 5 and1 psi hydraulic pressures on draw and feed sides in PRO mode.

0

5

10

15

20

25

30Standard methodHTI method

MembranesCA

Sal

t F

lux

(FO

mo

de)

, gm

-2h

-1

TFC Prewetted TFC

TFC Prewetted TFC0

5

10

15

20

25

30Standard methodHTI method

Membranes

CA

Sal

t F

lux

(PR

O m

od

e), g

m-2

h-1

Fig. 7. Salt flux of FO and PRO tests with threemembranes. Results are an average of threeexperiments with different coupons. Error bars indicate standard deviation. Standardmethod operating conditions: 1 M NaCl draw solution, deionized water feed, 20 °C feedand draw solution temperature, 0.25 m/s feed and draw solution cross flow velocity, 0transmembrane pressure and no spacers [13]. HTI method operating conditions: 1 MNaCl draw solution, deionized water feed, 0.30 m/s feed and draw solution cross flow ve-locity, and 20 °C feed and draw solution temperature. Spacers were used on both sides. 1and 5 psi hydraulic pressures (gauge pressures) on draw and feed sides in FOmode. 5 and1 psi hydraulic pressures on draw and feed sides in PRO mode.

192 J. Ren, J.R. McCutcheon / Desalination 343 (2014) 187–193

support has a higher porosity, lower tortuosity, or allows for more com-plete wetting of the structure.

The prewetted membrane exhibits a structural parameter of~620 μm. This is an interesting result since the structural parameter,which characterizes the ICP in support layer, should be decreased bythe wetting procedure [9]. It is likely due to the increase of the supportthickness which was caused by the swelling of the support layer in theIPA. The pores in support layer may also shrink as the polymer swells,increasing the tortuosity and decreasing porosity. The IPA may also af-fect selective layer properties (A and B) and thus will impact the calcu-lation of structural parameter when using empirical data.

4. Conclusions

In this study, we report the performance of an early generation TFCFOmembrane fromHTI. This membrane incorporates a selective barrierwith a hydrophilic support structure with a low structural parameter,giving it improved performance over their existing CA membrane.This membrane represents the first TFC FO membrane manufacturedon a 40-inch continuous production line and was shown to have supe-rior performance when compared to the cellulose acetate membranethat has been often used in recent FO studies. Later generations of theTFC membrane platform will likely replace the CA membrane as a

benchmark for FO, further pushing the bar higher for improving FOmembrane performance with new membrane designs.

Acknowledgments

We acknowledge the funding from the U.S. Environmental ProtectionAgency (#R834872), the Department of Energy (DE-EE00003226), andthe National Science Foundation (CBET #1160098). We thank HydrationTechnology Innovations for providing the membranes for this work.

References

[1] M. Elimelech, W.A. Phillip, The future of seawater desalination: energy, technology,and the environment, Science 333 (2011) 712–717.

[2] R.L. McGinnis, M. Elimelech, Global challenges in energy and water supply: thepromise of engineered osmosis, Environ. Sci. Technol. 42 (2008) 8625–8629.

[3] S. Zhao, L. Zou, C.Y. Tang, D. Mulcahy, Recent developments in forward osmosis: op-portunities and challenges, J. Membr. Sci. 396 (2012) 1–21.

[4] T.-S. Chung, S. Zhang, K.Y. Wang, J. Su, M.M. Ling, Forward osmosis processes: yes-terday, today and tomorrow, Desalination 287 (2012) 78–81.

[5] T. Cath, A. Childress, M. Elimelech, Forward osmosis: principles, applications, and re-cent developments, J. Membr. Sci. 281 (2006) 70–87.

[6] N.-N. Bui, J.R. McCutcheon, Hydrophilic Nanofibers as new supports for thin filmcomposite membranes for engineered osmosis, Environ. Sci. Technol. 47 (2013)1761–1769.

193J. Ren, J.R. McCutcheon / Desalination 343 (2014) 187–193

[7] N.Y. Yip, A. Tiraferri, W.A. Phillip, J.D. Schiffman, M. Elimelech, High performancethin-film composite forward osmosis membrane, Environ. Sci. Technol. 44 (2010)3812–3818.

[8] R. Wang, L. Shi, C.Y. Tang, S. Chou, C. Qiu, A.G. Fane, Characterization of novel for-ward osmosis hollow fiber membranes, J. Membr. Sci. 355 (2010) 158–167.

[9] J.R. McCutcheon, M. Elimelech, Influence of membrane support layer hydrophobic-ity on water flux in osmotically driven membrane processes, J. Membr. Sci. 318(2008) 458–466.

[10] R. Baker, Membrane Technology and Applications, Membrane Technology and Re-search Inc., Menlo Park, CA, 2004.

[11] K.D. Vos, Kinetic study of the hydrolysis of cellulose acetate in the pH range of 2–10,J. Appl. Polym. Sci. 10 (1966) 825–832.

[12] J.R. McCutcheon, M. Elimelech, Influence of concentrative and dilutive internal con-centration polarization on flux behavior in forward osmosis, J. Membr. Sci. 284(2006) 237–247.

[13] 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,Standardmethodology for evaluating membrane performance in osmotically drivenmembrane processes, Desalination 312 (2013) 31–38.

[14] J.T. Arena, B. McCloskey, B.D. Freeman, J.R. McCutcheon, Surface modification of thinfilm composite membrane support layers with polydopamine: enabling use of re-verse osmosis membranes in pressure retarded osmosis, J. Membr. Sci. 375 (2011)55–62.

[15] M. Mulder, Basic Principles of Membrane Technology, Springer, 1996.[16] N.-N. Bui, M.L. Lind, E.M.V. Hoek, J.R. McCutcheon, Electrospun nanofiber supported

thin film composite membranes for engineered osmosis, J. Membr. Sci. 385–386(2011) 10–19.

[17] Y. Xu, X. Peng, C.Y. Tang, Q.S. Fu, S. Nie, Effect of draw solution concentration and op-erating conditions on forward osmosis and pressure retarded osmosis performancein a spiral wound module, J. Membr. Sci. 348 (2010) 298–309.

[18] Y.C. Kim, S.-J. Park, Experimental study of a 4040 spiral-wound forward-osmosismembrane module, Environ. Sci. Technol. 45 (2011) 7737–7745.

[19] J.R. McCutcheon, M. Elimelech, Modeling water flux in forward osmosis: implica-tions for improved membrane design, AIChE J 53 (2007) 1736–1744.

[20] J. Wei, C. Qiu, C.Y. Tang, R.Wang, A.G. Fane, Synthesis and characterization of flat-sheetthin film composite forward osmosis membranes, J. Membr. Sci. 372 (2011) 292–302.

[21] C.H. Tan, H.Y. Ng, Modified models to predict flux behavior in forward osmosis inconsideration of external and internal concentration polarizations, J. Membr. Sci.324 (2008) 209–219.

[22] S. Loeb, L. Titelman, E. Korngold, J. Freiman, Effect of porous support fabric on osmosisthrough a Loeb–Sourirajan type asymmetric membrane, J. Membr. Sci. 129 (1997)243–249.

[23] C.Y. Tang, Y.-N. Kwon, J.O. Leckie, Effect of membrane chemistry and coating layeron physiochemical properties of thin film composite polyamide RO and NF mem-branes: I. FTIR and XPS characterization of polyamide and coating layer chemistry,Desalination 242 (2009) 149–167.

[24] A. Kulkarni, D. Mukherjee, W.N. Gill, Flux enhancement by hydrophilization of thinfilm composite reverse osmosis membranes, J. Membr. Sci. 114 (1996) 39–50.

[25] J. Kochan, T.Wintgens, R. Hochstrat, T. Melin, Impact of wetting agents on the filtrationperformance of polymeric ultrafiltration membranes, Desalination 241 (2009) 34–42.

[26] S. Zhang, F. Fu, T.-S. Chung, Substrate modifications and alcohol treatment on thinfilm composite membranes for osmotic power, Chem. Eng. Sci. 87 (2013) 40–50.

[27] V. Freger, Swelling and morphology of the skin layer of polyamide composite mem-branes: an atomic forcemicroscopy study, Environ. Sci. Technol. 38 (2004) 3168–3175.