Enhanced desalination performance of forward osmosis membranes based on reduced graphene oxide laminates coated with hydrophilic polydopamine

Accepted Manuscript Enhanced desalination performance of forward osmosis membranes based on reduced graphene oxide laminates coated with hydrophilic polydopamine Euntae Yang, Chang-Min Kim, Jun-ho Song, Hangil Ki, Moon-Ho Ham, In S. Kim PII:

S0008-6223(17)30246-4

DOI:

10.1016/j.carbon.2017.03.005

Reference:

CARBON 11812

To appear in:

Carbon

Received Date: 5 December 2016 Revised Date:

1 March 2017

Accepted Date: 2 March 2017

Please cite this article as: E. Yang, C.-M. Kim, J.-h. Song, H. Ki, M.-H. Ham, I.S. Kim, Enhanced desalination performance of forward osmosis membranes based on reduced graphene oxide laminates coated with hydrophilic polydopamine, Carbon (2017), doi: 10.1016/j.carbon.2017.03.005. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

Enhanced desalination performance of forward osmosis membranes based on

reduced

graphene

oxide

laminates

coated

with

hydrophilic

RI PT

polydopamine

Euntae Yang1, Chang-Min Kim1, Jun-ho Song1, Hangil Ki2, Moon-Ho Ham2, In S.

1

SC

Kim1,*

Global Desalination Research Center (GDRC), School of Environmental Sciences and

Buk-gu, Gwangju 61005, South Korea 2

M AN U

Engineering, Gwangju Institute of Science and Technology (GIST), 123 Cheomdangwagi-ro,

School of Materials Science and Engineering, Gwangju Institute of Science and Technology

(GIST), 123 Cheomdangwagi-ro, Buk-gu, Gwangju 61005, South Korea *

TE D

Corresponding author. Tel: +82 62 715 2460. Fax: +82 62 715 2434.

E-mail address: [email protected] (In S. Kim)

EP

ABSTRACT

Forward osmosis (FO) processes have been recognized as a low energy required next-

AC C

generation desalination technology, but the low performance of polymeric membranes remains a bottleneck in the practical application of FO. Graphene oxide (GO) membranes possess a huge potential as an alternative to polymeric membranes because of their facile fabrication procedure, ultra-thin thickness, controllable pore size, and outperformance of competing materials in terms of water transport rate; the low stability of GO laminates under water and their hydrophobic property (if GO laminates are reduced) remain as problems requiring solution prior to their practical implementation. To solve these problems, herein, the chemical reduction of GO laminates and a hydrophilic adhesive polydopamine (pDA) layer

ACCEPTED MANUSCRIPT were applied. Reduced GO (rGO) laminates sustainably retained their compacted nanochannels (3.45 Å) compared to pristine GO laminates, which increased the selectivity of hydrated ions. In addition, adding a pDA coating onto the rGO laminates improved the

RI PT

hydrophilicity of the rGO laminate surface, which accelerated the water absorption speed. As a result of these synergistic effects, pDA-coated rGO (pDA-rGO) membranes achieved an outstanding water flux of 36.6 L/m2·h, with a reverse solute flux of 0.042 mol/L·m2 and a

SC

high salt rejection rate of 92.0% in FO.

M AN U

Keywords: Membrane, forward osmosis, reduced graphene oxide, polydopamine, hydrophilicity

TE D

1. Introduction

Forward osmosis (FO) is driven by the osmotic pressure that is naturally generated by the concentration gradient across a semipermeable membrane for water desalination; FO

EP

produces water at a lower energy consumption than reverse osmosis (RO), which uses hydraulic pressure as the driving force. Consequently, FO has emerged as a future

AC C

desalination technology, and has been in rapid development over the last decade. However, several challenges with FO remain (e.g., membrane fouling, low water flux, and reverse solute flux) prior to its practical application in the field of desalination [1]. These challenges are mainly related to the fact that only a few types of polymeric FO membranes are commercially available. Moreover, state-of-the-art polymeric FO membranes do not offer sufficient water permeability and salt rejection. Therefore, alternative membranes for current polymeric FO membranes need to be developed to further improve the performance of FO processes.

ACCEPTED MANUSCRIPT Recently, the potential of nanoporous graphene membranes, having high mechanical strength, atomic thickness, and fine-tuned pores, has been identified for use in a nextgeneration membrane that has ultrafast water flux and almost flawless salt rejection [2–5].

RI PT

Using molecular dynamics (MD) simulation, Cohen-Tanugi et al. [2] predicted that singlelayer graphene membranes having functionalized nanopores would have hundreds of times higher water permeability, with almost 100% salt rejection, than commercial polymeric RO

SC

membranes. Gai et al. [4] exhibited functionalized porous graphene membranes displaying amazing water flux and salt rejection during FO processes. Furthermore, actual porous

M AN U

graphene membrane fabrications have been attempted by forming nanopores in a single-layer graphene sheet using various methods (e.g., combination of ion bombardment and chemical etching [6], focused ion beam irradiation [7], and oxygen plasma etching [8]). In particular, a nanoporous graphene membrane perforated by oxygen plasma etching showed an excellent

TE D

water flux of 106 g/m2/s with almost 100% salt rejection [8]. However, synthesis of scalable defect-free single-layer and formation of high-density well-defined nanopores in graphene sheets remains significantly challenging. Moreover, the high price of scalable defect-free

EP

single-layer graphene also inhibits the commercialization of porous graphene membranes [9]. As an alternative for nanoporous graphene membranes, graphene oxide (GO)

AC C

membranes have attracted the attention of researchers [9–11]. GO is a single-layer graphene based nanomaterial having plentiful oxygen functional groups in the form of epoxy, hydroxyl, carboxyl, and others [12]. As a novel membrane material, GO has excellent characteristics (e.g., inexpensive mass-production cost, high chemical stability, high hydrophilicity, negatively charged surface and strong anti-bacterial propensity) [10,13,14]. By stacking GO nanosheets, GO membranes can be easily fabricated via several facile methods, such as vacuum-assisted filtration and drop casting [10,15,16]. Nanochannels formed between stacked GO nanosheets offer a low-friction pathway for the ultra-fast transport of water

ACCEPTED MANUSCRIPT molecules. Due to non-oxidized areas on the surface of GO membranes, water molecules could also be transported at ultrafast speeds with low friction [17,18]. In addition, the GO nanochannels naturally made in a laminated structure provide precise selectivity for ionic and

RI PT

molecular species, based on size exclusion [10,18–20]. However, serious drawbacks prior to the real application of GO membranes remain. In particular, GO laminates are highly susceptible to redispersal into water, and the size of

SC

GO nanochannels becomes enlarged due to the high hydrophilicity of GO nanosheets. These phenomena negatively affect the stability and selectivity of GO membranes [20].

M AN U

Herein, to overcome these challenges and finally use GO membranes in FO desalination processes, two technical strategies are adopted. Initially, GO laminates are chemically reduced in order to remove hydrophilic functional groups from GO. Consequently, the stability of GO laminates can be improved, and the nanochannels can be narrowed, which

TE D

increases the rejection of ionic species for desalination [15], though the tightened rGO nanochannels and the hydrophobic surface of rGO laminates would be obstructive factors against water transport. However, water permeability across rGO laminates can be enhanced

EP

by reducing the thickness of rGO laminates within a nanometer scale. Han et al. [21] reported that the water permeability of rGO membranes increased with decreases in the thickness of

AC C

rGO laminates. In their study, the thinnest rGO laminates achieved the highest water permeability of 21.81 L/m2·h·bar. In addition, it was reported that the ultrafast water flow in the core of carbon nanotubes (CNTs) is due to the hydrophobic inner wall of the CNTs, though the high hydrophobic inlet of CNTs cannot allow water to readily enter the core of CNTs [22]. To reduce the entrance resistance, previous studies have introduced hydrophilic functional groups [23,24]. The water transport in the interior of CNTs is similar to that in GO nanochannels [10]; if the surface of rGO laminates is more hydrophilic, water could more readily flow into the nanochannels formed in rGO laminates. Therefore, a hydrophilic

ACCEPTED MANUSCRIPT polydopamine (pDA) thin layer was finally deposited onto the rGO laminates to increase the efficiency of the water uptake into rGO nanochannels. pDA is an adhesive mussel-foot-protein-inspired polymeric substance that can be

RI PT

easily and strongly deposited on any substrate and thereby increase the hydrophilicity, as pDA contains catechol and amine functional groups [25]. In this study, pDA-coated rGO (pDA-rGO) membranes exhibit superior FO desalination performance, with improved water

SC

flux and lowered reverse solute permeation compared to either a commercial cellulose triacetate embedded support-type (CTA-ES) membrane or a pristine rGO membrane,

M AN U

indicating that pDA-rGO membranes have great potential as a novel membrane for highperformance FO desalination processes.

2. Experimental

TE D

2.1. Synthesis of pDA-rGO membrane

The fabrication procedure of the pDA-rGO membrane is described in Fig. 1. First, an ultra-high concentration single-layer GO aqueous solution (6.2 mg/mL) was purchased from

EP

Graphene Supermarket (Calverton, US), and was diluted with deionized (DI) water to prepare a GO aqueous solution of 0.006 mg/mL. The prepared GO aqueous solution (0.006 mg/mL)

AC C

was then sonicated for 30 min, and then filtered through a mixed cellulose ester membrane having a 0.2 µm pore (MCE; 47 mm area, Advantec MFS, Inc., Japan) using a vacuumassisted filtration system. The GO laminates on the MCE membranes were dried at 40 °C for more than 24 h. To reduce the resultant GO laminates on the MCE membrane, the GO laminates were exposed to hydriodic acid vapor (HI; Sigma Aldrich, USA) [15]. The rGO membrane was left at room temperature for 1 day. Next, the rGO membrane was fully dipped in a 2.0 g/L dopamine solution (pH 8.5) for 1 h. The 2.0 g/L dopamine solution was prepared by adding dopamine hydrochloride (Sigma Aldrich Co. LCC., USA) to a Tris-HCl buffer (15

ACCEPTED MANUSCRIPT mM, pH 8.5). The resulting pDA-rGO membrane was dried at 40 °C and stored in a desiccator prior to use.

RI PT

Fig. 1.

2.2. Membrane characterization

SC

The characteristics (e.g., surface morphology, thickness, hydrophilicity, chemical structure, nanochannel size) of the prepared GO, rGO and pDA-rGO membranes were

M AN U

investigated. The membrane morphology and thickness were observed using scanning electron microscopy (SEM; S-4700, Hitachi, Japan) and a stylus profiler system (Dektak XT, Bruker Cooperation, US), respectively. The membrane hydrophilicity was determined by measuring the contact angle using a goniometer (Phoenix 300, Surface Electro Optics Co.,

TE D

South Korea). The chemical structure of the membranes was analyzed using X-ray photoelectron spectroscopy (XPS; VG Microtech MultiLab ESCA 2000, Thermo VG Scientific, United Kingdom) and Fourier transform infrared spectroscopy (FT-IR; Frontier

EP

FT-IR/NIR, Perkin Elmer, USA). The nanochannels formed between GO (or rGO) nanosheets in the membranes were characterized using X-ray diffraction (XRD; X’pert PRO, PANalitical,

AC C

Netherlands).

2.3. Ion transport and salt rejection A two-compartment diffusion cell was prepared to observe the ion transport across a

membrane (Fig. 2(a)). A test membrane (effective area: 0.73 cm2) was installed between two compartments. For the real-time monitoring of ion permeability, the probe of a total dissolved solids (TDS) meter was inserted into one compartment. The compartment having the TDS probe (COND610, Eutech Instruments, Singapore) was filled with 100 mL of DI water, and

ACCEPTED MANUSCRIPT the other compartment was filled with 100 mL of 0.6 M NaCl solution.

2.4. Membrane performance evaluation in an FO system

RI PT

To evaluate the membrane performance in FO, a lab-scale FO system was set up (Fig. 2(b)). The system has a two-compartment membrane cell, separated equally using a test membrane (effective area: 0.73 cm2). The compartment of the membrane cell facing the

SC

rejection layer of the membrane (or GO laminate) was connected to an external feed solution reservoir filled with 1 L of DI water. The DI water was circulated from the feed solution

M AN U

reservoir, through the membrane cell compartment facing the GO laminate, and back to the feed solution reservoir using a gear pump (Model 7527-15, Cole-Parmer, USA) at a flow rate of 200 mL/min. The other compartment of the membrane cell facing the MCE support layer of the GO membrane was connected to an external draw solution reservoir filled with 1 L of

TE D

0.6 M NaCl solution. The 0.6 M NaCl solution continuously flowed from the compartment facing the MCE support layer of the membrane to the draw solution reservoir at a flow rate of 200 mL/min. To determine the water flux, weight changes of the draw solution reservoir were

EP

obtained using a digital balance (GF-6100, A&D Company, Japan) at 1 min intervals. The water flux (Jw) was computed using [26]: ∆ (1)  ∆

AC C

J =

where ∆m is the weight change of the draw solution over time ∆t, and A is the effective membrane area.

To monitor the reverse solute permeate from the NaCl solution to DI water across the membrane, the TDS of the DI water (feed solution) was measured using a TDS meter. The reverse solute permeate flux (Js) was calculated using [26]: J =

∆(   ) (2)  ∆

ACCEPTED MANUSCRIPT where Ct and Vt are the salt concentration and the volume, respectively, of the feed solution over the operation time ∆t. To determine the salt rejection efficiencies of the membranes, the same FO system

RI PT

was used, though 0.1 M NaCl and 1 M dextrose solutions were used as the feed and draw solutions, respectively. Eight hours after each salt rejection test started, volume changes, caused by water flux across the membrane due to the osmotic pressure difference between the

SC

two solutions, and TDS changes of the dextrose solution were measured. The salt rejection (R) was calculated according to:

  × 100 (3)



M AN U

 = 1 −

where Cf is the initial salt concentration of the feed solution, and Cp is the permeate salt concentration calculated based on the quantity of water migrated from the feed to the draw solutions and the final salt concentration of the dextrose solution.

TE D

During FO system operation, the temperature was maintained at 25 °C using a

AC C

Fig. 2.

EP

temperature control system.

3. Results and discussion

3.1. Physicochemical properties of laminated graphene membranes To synthesize the pDA-rGO membrane, a GO membrane was fabricated from GO

solutions using a vacuum-assisted filtration method. The rGO membrane was then prepared through the reduction of the resulting GO membrane using HI steam. Finally, the surface of the resultant rGO membrane was coated with a pDA polymeric layer. Digital images of the prepared membranes are shown in Fig. 3(a–c). The pDA-rGO

ACCEPTED MANUSCRIPT membrane exhibited a polished dark black-colored surface (Fig. 3(c)), whereas the GO (Fig. 3(a)) and rGO (Fig. 3(b)) membranes have a glossy yellow-colored surface and a graphitelike lustrous black colored surface, respectively. For both the rGO and pDA-rGO membranes,

RI PT

graphitic sp2-bondings recovered by the HI vapor reduction process normally absorb visiblerange light, such that the colors of the prepared membranes are dark and black; GO was not black due to the sp2-bondings destroyed by the oxidation process.

SC

Nanochannels (i.e., inter-layer spacings between the laminated GO nanosheets) of the GO, rGO and pDA-rGO laminates were examined by determining d-spacing using an XRD

M AN U

analysis. Fig. 3(d) shows the XRD pattern of the prepared membranes. Nanochannels play critical roles as pathways for ultra-fast water molecule transport and as selective barriers for ionic and molecular species; the width of nanochannels (i.e., inter-layer distance) is a crucial factor affecting the performance of GO membranes. The GO laminates exhibited one XRD

TE D

peak at 9.60°; the nanochannel width of the GO laminates is 8.02 Å, consistent with the XRD peak of GO in a previous study [27]. However, peaks at 12.52°, 23.25°, and 25.80° were observed in the XRD pattern of the rGO laminates, corresponding to nanochannel widths of

EP

7.06 Å, 3.82 Å, and 3.45 Å. The reason for the decreased nanochannel widths of the rGO laminates is that oxygenated functional groups attached on GO nanosheets are removed by

AC C

the reduction treatment using HI vapor, which is supported by the results of the FT-IR analyses. The pDA-rGO membrane showed a peak at the same position as the rGO membrane (25.80°), indicating that the nanochannel width of pDA-rGO is 3.45 Å. However, the peak intensity is very weak; because the amorphous pDA coating reduced the crystallographic quality of the rGO laminate [28]. FT-IR data (Fig. 3(e)) confirm the chemical structure difference between the fabricated membranes, as shown in Fig. 3(d). In the FT-IR spectrum of the GO membrane, peaks for -OH stretching at around 3390 cm-1, C=O peak at around 1720 cm-1, C=C bond

ACCEPTED MANUSCRIPT stretching at 1580 cm-1, C-O vibrational band at around 1367 cm-1, C-O epoxy at around 1220 cm-1, and C-O alkoxy at 1060 cm-1 were observed [29–31]. However, in the FT-IR spectrum of the rGO membrane, all peaks disappeared, demonstrating that the HI vapor

RI PT

treatment can successfully remove oxygenated functional groups. One the other hand, in the FT-IR spectrum of the pDA-rGO membrane, new peaks were observed, such as the C=N stretching peak at 1510 cm-1, N-H vibration peak at 1501 cm-1, phenolic C-OH stretching

SC

vibration peak at 1280-1, which originated from the pDA coating layer [32–34].

As a result of the chemical reduction and pDA coating, the GO, rGO, and pDA-rGO

M AN U

laminates also have different chemical element distributions, as derived from the XPS spectra of the GO, rGO, and pDA-rGO laminates (Figs. 4(a) and S1). The carbon-to-oxygen (C/O) ratio of the pristine GO laminates was around 1.2, but that of the rGO laminates was about 4.1. This result is in agreement with the chemical structure of the resultant laminates analyzed

TE D

by FT-IR. Therefore, the different chemical element distributions between the GO and rGO laminates can be attributed to existence of oxygenated functional groups such as hydroxyl, carboxylic, and epoxide [35]. However, the pDA-rGO laminates had a C/O ratio similar to

EP

the pristine GO laminates, and contained nitrogen atoms. These atomic contents are derived from the pDA coating layer, which is comprised of dopamine, dihydroxyindole, and

AC C

indoledione units [36].

The water contact angle (Figs. 4(b) and S2) and thickness (Fig. S3) were also

analyzed, as critical factors influencing the mass transport across membranes [37]. The water contact angle of the rGO membrane (61.2 ± 3.9°) was higher than that of the GO membrane (39.6 ± 5.8°), indicating that the reduction of GO laminates decreases the hydrophilicity of GO laminates by eliminating hydrophilic oxygen functional groups. On the other hand, the pDA-rGO membrane markedly reduced the water contact angle to 26.6 ± 12.9°, which means that the pDA-rGO is highly hydrophilic, because the pDA coating layer possesses abundant

ACCEPTED MANUSCRIPT hydrophilic catechol and amine functional groups [38] In terms of thickness, the rGO laminate (159.7 ± 16.6 nm) was thinnest, followed by the pDA-rGO (170.4 ± 20.2 nm) and GO (210.9 ± 21.6 nm) laminates. As illustrated in Fig.

RI PT

S3, the thinness of the rGO laminates can be attributed to the chemical reduction process, which removes the oxygen containing functional groups attached to the GO nanosheets; subsequently, the interlayer spaces between the GO nanosheets become more compact. The

SC

pDA coating leads to a slight increase in the pDA-rGO laminate thickness.

M AN U

Fig. 3

Fig. 4

TE D

3.2. Ion permeability of laminated graphene membranes

Ion permeation across the resultant membranes was evaluated in a two-chamber diffusion cell, in which the chamber facing the GO laminate was filled with DI water and the

EP

chamber facing the MCE substrate was filled with a 0.6 M NaCl solution. Fig. 5 presents the concentration of ion permeation through the membranes over time. As expected, among the

AC C

fabricated membranes the GO membrane was the most permeable to sodium and chloride ions even though, compared to the bare MCE filter, the GO membrane displayed an approximately 70% decrease in the ion permeation rate. The rGO membrane was found to have an extremely lower ion permeation rate due to the tightened nanochannels of the rGO laminates, and a further slight decline in the ion permeation rate was obtained during the test using the pDA-rGO membrane. The most likely reason for the further decrease in ion permeation for the pDA-rGO membrane is due to the additional pDA layer on graphene membrane, which has a nanometer-scale or a few nanometer thickness, which could retard

ACCEPTED MANUSCRIPT the ion permeation. The results of the ion permeability test suggest that the pDA-rGO laminates can provide a more effective ion rejection barrier for desalination than the GO and

RI PT

rGO laminates.

Fig. 5

SC

3.3. Desalination performance of graphene oxide membranes in FO

Taking the results of the ion permeation tests into account, we subsequently

M AN U

evaluated the water flux and reverse solute flux of the rGO and pDA-rGO membranes in an FO process, using DI water and 0.6 M NaCl solution for the feed and draw solutions, respectively. For comparison, a commercial CTA ES FO membrane was also tested under the same conditions. Fig. 6(a) shows the water flux across each membrane, where the rGO

TE D

membrane is seen to achieve a higher water flux (average value: 20.1 LMH) than the commercial CTA ES membrane (5.96 LMH). This result, in which the rGO membrane outperformed the CTA ES, is consistent with a previous report [15]. However, the previous

EP

study only tested freestanding rGO laminates, which were detached from the MCE filter, and reported that the main cause for the higher water flux across the freestanding rGO laminates

AC C

in FO was the significant decrease in internal concentration polarization due to an absence of support layer [15].

The rGO membrane used in this study included a support layer (i.e., MCE filter),

indicating that the fast water flux through rGO is not only because of the omission of a support layer, but is also due to the low-friction hydrophobic nanochannels [18]. Moreover, the water flux was further improved by embedding a hydrophilic layer onto the rGO laminates, compared to the rGO membrane. Fig. 6(b) presents the reverse solute flux across each membrane tested using DI water as the feed solution and 0.6 M NaCl solution as the

ACCEPTED MANUSCRIPT draw solution. In contrast to the water flux, the reverse solute flux of the pDA-rGO membrane was the lowest at 0.04 mol/m2·h, followed by the rGO membrane (0.08 mol/m2·h) and the commercial CTA ES membrane (0.21 mol/m2·h). Note that the reverse solute

RI PT

diffusion exerts a baneful influence on FO processes, such as reducing the osmotic driving force, contaminating the feed solution, and causing a loss of the draw solute. Therefore, the pDA-rGO membrane could provide a solution for overcoming this bottleneck.

SC

To assess the salt rejection of the membranes in FO, we operated the lab-scale FO system using a 0.1 M NaCl solution for the feed solution and 1 M dextrose for the draw

M AN U

solution (Fig. 6(c)). The pDA-rGO membrane exhibited a slight decrease in salt rejection (92.0%), compared to the rGO membrane (96.0%), which means more ions passed through the pDA-rGO membrane than the rGO membrane. However, in the ion diffusion test and during FO operation using DI water as the feed solution and 0.6 M NaCl as the draw solution,

TE D

the pDA-rGO membrane displayed a lower ion diffusion rate. Note that the salt rejection was determined based on the forward solute flux from the feed solution to the draw solution, i.e., that the water and ions flow through a membrane (e.g., GO laminate and MCE support filter,

EP

in series) in the same direction. In contrast to the salt rejection measurement, the ion diffusion test and FO operation using DI water as the feed solution and 0.6 M NaCl as the draw

AC C

solution, ions diffuse from the draw to feed solutions across a membrane (e.g., from MCE substrate to GO laminate), but water flows through the membrane in an opposite direction as the ions.

For a clearer determination for the declined salt rejection of the pDA-rGO membrane,

further investigation is required, but there are two potential reasons. First, the pDA coating may diminish the electrostatic repulsion force between the membrane surface and anionic species by decreasing the negative surface charge of rGO laminates. Compared to the surface charge of the pDA coating layer and the rGO based on previously reported data, the pDA

ACCEPTED MANUSCRIPT coating layer may have a much weaker negative surface charge (about -2.3 mV) than rGO (near -35.0 mV) at pH 7 [39,40]. In addition, a previous study reported that a nanocomposite membrane having an enhanced negative surface charge due to carbon nanotube positioning

RI PT

had a superior salt rejection capacity because a stronger electrostatic repulsion force can be generated between the more negatively charged membrane surface and anions [41]. Second, cake-enhanced concentration polarization induced by the pDA coating layer may decrease the

SC

salt rejection of the pDA-rGO membrane [42,43]. Salts rejected by rGO laminates may accumulate in a salt-permeable pDA coating layer because the pDA coating layer may hinder

within the pDA coating layer.

M AN U

back-diffusion of the rejected salts. Thus, severe concentration polarization may be generated

In terms of water flux, the pDA-rGO membrane also displayed the best performance in the salt rejection test (feed solution: 0.1 M NaCl solution; draw solution: 1.0 M dextrose).

TE D

Despite the lowered salt rejection, however, the pDA-rGO membrane can still be competitive in FO because of its more than 90% salt rejection rate and excellent water flux. The possible mechanism for improving the water transport across the pDA-rGO

EP

membrane is illustrated in Fig. 7. A high hydrophilicity pDA coating layer can improve the wettability of the membrane surface [44]; the hydrophilic pDA coating layer can make water

AC C

more readily enter the rGO nanochannels, and consequently enhance water permeation across the pDA-rGO membrane.

Fig. 6

Fig. 7

4. Conclusion

ACCEPTED MANUSCRIPT We fabricated pDA-rGO membranes through two simple strategies, including the chemical reduction of GO laminates and adding a hydrophilic pDA coating onto the rGO surface to overcome typical drawbacks of GO laminates. The chemical reduction of GO

RI PT

laminates can drastically increase its high salt rejection efficiency by providing highly stable and tightened nanochannels between GO nanosheets. In addition, adding a hydrophilic pDA coating onto the rGO surface could further improve the water flux by enhancing the water

SC

absorption rate into rGO laminates. In FO tests, pDA-rGO membranes revealed about a 1.8 times and 5 times higher rate than rGO and commercial CTA membranes, respectively, with

M AN U

greatly reduced reverse solute flux. Concurrently, pDA-rGO membranes achieved an approximately 92.0% salt rejection rate. Therefore, these results indicate that although further investigation into pDA-rGO membranes regarding long-term stability and scale-up feasibility is required prior to their practical application, excellent lab-scale FO desalination

TE D

performances have been achieved by the synergistic effect of rGO laminates having ultra-fast water transport and the hydrophilic pDA polymeric layer having great water absorption

EP

ability.

ACKNOWLEDGEMENTS

AC C

This research was supported by a grant (code 16IFIP-B088091-03) from the Industrial Facilities & Infrastructure Research Program funded by Ministry of Land, Infrastructure and Transport of Korean government and by a National Research Foundation of Korea grant funded by the Korean Government (NRF-2013R1A1A2066114).

References [1] N. Akther, A. Sodiq, A. Giwa, S. Daer, H.A. Arafat, S.W. Hasan, Recent advancements in forward osmosis desalination: A review, Chem. Eng. J. 281 (2015) 502-522.

ACCEPTED MANUSCRIPT [2] D. Cohen-Tanugi, J.C. Grossman, Water desalination across nanoporous graphene, Nano Lett. 12 (2012) 3602-3608. [3] J.-G. Gai, X.-L. Gong, W.-W. Wang, X. Zhang, W.-L. Kang, An ultrafast water transport

RI PT

forward osmosis membrane: porous graphene, J. Mater. Chem. A 2 (2014) 4023-4028. [4] J.-G. Gai, X.-L. Gong, Zero internal concentration polarization FO membrane: functionalized graphene, J. Mater. Chem. A 2 (2014) 425-429.

SC

[5] D. Cohen-Tanugi, J.C. Grossman, Mechanical Strength of Nanoporous Graphene as a Desalination Membrane, Nano lett. 14 (2014) 6171-6178.

M AN U

[6] S.C. O’Hern, M.S.H. Boutilier, J.-C. Idrobo, Y. Song, J. Kong, T. Laoui, M. Atieh, R. Karnik, Selective Ionic Transport through Tunable Subnanometer Pores in Single-Layer Graphene Membranes, Nano lett. 14 (2014) 1234-1241.

[7] K. Celebi, J. Buchheim, R.M. Wyss, A. Droudian, P. Gasser, I. Shorubalko, J.-I. Kye, C.

(2014) 289-292.

TE D

Lee, H.G. Park, Ultimate permeation across atomically thin porous graphene, Science 344

[8] S.P. Surwade, S.N. Smirnov, I.V. Vlassiouk, R.R. Unocic, G.M. Veith, S. Dai, S.M.

459-464.

EP

Mahurin, Water desalination using nanoporous single-layer graphene, Nat. Nano. 10 (2015)

AC C

[9] F. Perreault, A.F. de Faria, M. Elimelech, Environmental applications of graphene-based nanomaterials, Chem. Soc. Rev. 44 (2015) 5861-5896. [10] M. Hu, B. Mi, Enabling graphene oxide nanosheets as water separation membranes, Environ. Sci. Technol. 47 (2013) 3715-3723. [11] G. Liu, W. Jin, N. Xu, Graphene-based membranes, Chem. Soc. Rev. 44 (2015) 50165030. [12] S. Stankovich, D.A. Dikin, R.D. Piner, K.A. Kohlhaas, A. Kleinhammes, Y. Jia, Y. Wu, S.T. Nguyen, R.S. Ruoff, Synthesis of graphene-based nanosheets via chemical reduction of

ACCEPTED MANUSCRIPT exfoliated graphite oxide, Carbon 45 (2007) 1558-1565. [13] D.R. Dreyer, S. Park, C.W. Bielawski, R.S. Ruoff, The chemistry of graphene oxide, Chem. Soc. Rev. 39 (2010) 228-240.

RI PT

[14] S. Liu, T.H. Zeng, M. Hofmann, E. Burcombe, J. Wei, R. Jiang, J. Kong, Y. Chen, Antibacterial activity of graphite, graphite oxide, graphene oxide, and reduced graphene oxide: membrane and oxidative stress, ACS Nano 5 (2011) 6971-6980.

SC

[15] H. Liu, H. Wang, X. Zhang, Facile Fabrication of Freestanding Ultrathin Reduced Graphene Oxide Membranes for Water Purification, Adv. Mater. 27 (2015) 249-254.

M AN U

[16] L. Huang, Y. Li, Q. Zhou, W. Yuan, G. Shi, Graphene Oxide Membranes with Tunable Semipermeability in Organic Solvents, Adv. Mater. 27 (2015) 3797-3802. [17] R.R. Nair, H.A. Wu, P.N. Jayaram, I.V. Grigorieva, A.K. Geim, Unimpeded permeation of water through helium-leak-tight graphene-based membranes, Science 335 (2012) 442-444.

TE D

[18] H. Huang, Z. Song, N. Wei, L. Shi, Y. Mao, Y. Ying, L. Sun, Z. Xu, X. Peng, Ultrafast viscous water flow through nanostrand-channelled graphene oxide membranes, Nat. commun. 4 (2013) 1-9.

EP

[19] K. Huang, G. Liu, Y. Lou, Z. Dong, J. Shen, W. Jin, A Graphene Oxide Membrane with Highly Selective Molecular Separation of Aqueous Organic Solution, Angew. Chem. Int. Edit.

AC C

53 (2014) 6929-6932.

[20] R. Joshi, P. Carbone, F. Wang, V. Kravets, Y. Su, I. Grigorieva, H. Wu, A. Geim, R. Nair, Precise and ultrafast molecular sieving through graphene oxide membranes, Science 343 (2014) 752-754.

[21] Y. Han, Z. Xu, C. Gao, Ultrathin Graphene Nanofiltration Membrane for Water Purification, Adv. Funct. Mater. 23 (2013) 3693-3700. [22] J.H. Walther, K. Ritos, E.R. Cruz-Chu, C.M. Megaridis, P. Koumoutsakos, Barriers to Superfast Water Transport in Carbon Nanotube Membranes. Nano letters. 13 (2013) 1910-

ACCEPTED MANUSCRIPT 1914. [23] H.D. Lee, H.W. Kim, Y.H. Cho, H.B. Park, Experimental evidence of rapid water transport through carbon nanotubes embedded in polymeric desalination membranes. Small.

RI PT

10 (2014) 2653-2660. [24] M. Son, V. Novotny, H. Choi, Thin-film nanocomposite membrane with vertically embedded carbon nanotube for forward osmosis. Desalination and Water Treatment. 57 (2016)

SC

26670-26679.

[25] H. Lee, S.M. Dellatore, W.M. Miller, P.B. Messersmith, Mussel-Inspired Surface

M AN U

Chemistry for Multifunctional Coatings, Science 318 (2007) 426-430.

[26] Q. Yang, K.Y. Wang, T.-S. Chung, Dual-layer hollow fibers with enhanced flux as novel forward osmosis membranes for water production, Environ. Sci. Technol. 43 (2009) 28002805.

TE D

[27] C.-N. Yeh, K. Raidongia, J. Shao, Q.-H. Yang, J. Huang, On the origin of the stability of graphene oxide membranes in water, Nat. Chem. 7 (2015) 166-170. [28] Y. Su, V.G. Kravets, S.L. Wong, J. Waters, A.K. Geim, R.R. Nair, Impermeable barrier

EP

films and protective coatings based on reduced graphene oxide, Nat. Commun. 5 (2014) 1-5. [29] X. Sun, Z. Liu, K. Welsher, J.T. Robinson, A. Goodwin, S. Zaric, H. Dai, Nano-graphene

AC C

oxide for cellular imaging and drug delivery, Nano Res. 1 (2008) 203-212. [30] G. Wang, B. Wang, J. Park, J. Yang, X. Shen, J. Yao, Synthesis of enhanced hydrophilic and hydrophobic graphene oxide nanosheets by a solvothermal method, Carbon, 47 (2009) 68-72.

[31] N.A. Kumar, H.-J. Choi, Y.R. Shin, D.W. Chang, L. Dai, J.-B. Baek, Polyaniline-Grafted Reduced Graphene Oxide for Efficient Electrochemical Supercapacitors, ACS Nano. 6 (2012) 1715-1723. [32] Y. He, J. Wang, H. Zhang, T. Zhang, B. Zhang, S. Cao, J. Liu, Polydopamine-modified

ACCEPTED MANUSCRIPT graphene oxide nanocomposite membrane for proton exchange membrane fuel cell under anhydrous conditions, J. Mater. Chem. A 2 (2014) 9548-9558. [33] N. Liu, M. Zhang, W. Zhang, Y. Cao, Y. Chen, X. Lin, L. Xu, L. Feng, C. Li, Y. Wei,

RI PT

Ultralight Free-Standing Reduced Graphene Oxide Membranes for Oil-in-Water Emulsions Separation, J. Mater. Chem. A 3 (2015) 20113-20117.

[34] R.A. Zangmeister, T.A. Morris, M.J. Tarlov, Characterization of polydopamine thin films

SC

deposited at short times by autoxidation of dopamine, Langmuir 29 (2013) 8619-8628.

[35] S. Stankovich, D.A. Dikin, G.H. Dommett, K.M. Kohlhaas, E.J. Zimney, E.A. Stach,

M AN U

R.D. Piner, S.T. Nguyen, R.S. Ruoff, Graphene-based composite materials, Nature 442 (2006) 282-286.

[36] J. Liebscher, R. Mrówczyński, H.A. Scheidt, C. Filip, N.D. Hădade, R. Turcu, A. Bende, S. Beck, Structure of Polydopamine: A Never-Ending Story?, Langmuir 29 (2013) 10539-

TE D

10548.

[37] W. Tang, H.Y. Ng, Concentration of brine by forward osmosis: Performance and influence of membrane structure, Desalination 224 (2008) 143-153.

EP

[38] M.H. Ryou, Y.M. Lee, J.K. Park, J.W. Choi, Mussel‐Inspired Polydopamine‐Treated Polyethylene Separators for High‐Power Li‐Ion Batteries, Adv. Mater. 23 (2011) 3066-3070.

AC C

[39] B. Yu, J. Liu, S. Liu, F. Zhou, Pdop layer exhibiting zwitterionicity: a simple electrochemical interface for governing ion permeability, Chem. Commun. 46 (2010) 59005902.

[40] D. Li, M.B. Mueller, S. Gilje, R.B. Kaner, G.G. Wallace, Processable aqueous dispersions of graphene nanosheets, Nat. Nanotechnol. 3 (2008) 101-105. [41] M. Son, H.-g. Choi, L. Liu, E. Celik, H. Park, H. Choi, Efficacy of carbon nanotube positioning in the polyethersulfone support layer on the performance of thin-film composite membrane for desalination, Chem. Eng. J. 266 (2015) 376-384.

ACCEPTED MANUSCRIPT [42] Z. Yang, Y. Wu, J. Wang, B. Cao, C.Y. Tang, In Situ Reduction of Silver by Polydopamine: A Novel Antimicrobial Modification of a Thin-Film Composite Polyamide Membrane, Environ. Sci. Technol. 50 (2016) 9543-9550.

RI PT

[43] E.M.V. Hoek, M. Elimelech, Cake-Enhanced Concentration Polarization:  A New Fouling Mechanism for Salt-Rejecting Membranes, Environ. Sci. Technol. 37 (2003) 55815588.

SC

[44] R. Yang, J. Xu, G. Ozaydin-Ince, S.Y. Wong, K.K. Gleason, Surface-tethered zwitterionic ultrathin antifouling coatings on reverse osmosis membranes by initiated

AC C

EP

TE D

M AN U

chemical vapor deposition, Chemistry of Materials, 23 (2011) 1263-1272.

ACCEPTED MANUSCRIPT Figure captions

Fig. 1. Schematic diagram of fabrication procedure for a pDA-rGO membrane.

RI PT

Fig. 2. (a) Two-compartment diffusion cell used in ion permeation tests, and (b) set-up of labscale FO system having a membrane cell divided by a membrane (projected area: 0.73 cm2)

SC

into feed and draw sides.

Fig. 3. Photographs of a (a) GO membrane, (b) rGO membrane, and (c) pDA-rGO membrane.

M AN U

(d) XRD analyses spectra, and (e) FT-IR spectra of GO, rGO and pDA-rGO membranes.

Fig. 4. (a) Atomic percentage of elements obtained from XPS patterns, and (b) water contact

TE D

angle.

Fig. 5. (a) Ion permeation across membranes (bare MCE, GO, rGO, and pDA-rGO) from 0.6 M NaCl solution (0.1 L) to DI water (0.1 L) in a diffusion cell. (b) Magnified graph lines for

EP

rGO and pDA-rGO.

AC C

Fig. 6. (a) Water flux and (b) reverse solute permeation of commercial CTA ES, rGO, and pDA-rGO membranes. DI water was used as the feed solution, and 0.6 M NaCl solution (1 L) was chosen as the draw solution (1 L). (c) Salt rejection and water flux of commercial CTA ES, rGO, and pDA-rGO membranes. The feed solution was 0.1 M NaCl solution (1 L), and the draw solution was 1.0 M dextrose solution (1 L).

Fig. 7. Conceptual mechanism for enhancing water permeability across pDA-rGO

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

membranes.

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

Fig. 1

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

Fig. 2

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

Fig. 3

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

Fig. 4

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

Fig. 5

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

Fig. 6

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

Fig. 7