Enhanced osmotic energy generation from salinity gradients by modifying thin film composite membranes

Chemical Engineering Journal 242 (2014) 195–203

Contents lists available at ScienceDirect

Chemical Engineering Journal journal homepage: www.elsevier.com/locate/cej

Enhanced osmotic energy generation from salinity gradients by modifying thin film composite membranes Yue Cui a,b, Xiang-Yang Liu a,c, Tai-Shung Chung a,⇑ a

Department of Chemistry, National University of Singapore, Singapore 117542, Singapore Department of Chemical and Biomolecular Engineering, National University of Singapore, Singapore 117542, Singapore c Department of Physics, National University of Singapore, Singapore 117542, Singapore b

h i g h l i g h t s  Modified thin film composite (TFC)

membrane can sustain 22 bar in PRO process.  Sodium dodecyl sulfate modification increases free volume intensity of TFC layer.  N,N-dimethylformamide treatment dissolves less crosslinked parts of the TFC layer.  With both treatments, the membrane harvests superior energy generation in PRO process.

a r t i c l e

i n f o

Article history: Received 10 October 2013 Received in revised form 16 December 2013 Accepted 26 December 2013 Available online 2 January 2014 Keywords: Salinity-gradient osmotic power Pressure retarded osmosis Surfactant treatment Solvent treatment

g r a p h i c a l a b s t r a c t Macrovoid-free Matrimid substrate

Sodium dodecyl sulfate (SDS)

High water flux TFC membrane

M-phenylenediamine (MPD) Trimesoyl chloride ((TMC))

Pronouncedly enhanced energy gy generation g

20

2% SDS + 50% DMF for 00.5 5 hr 16

Dimethylformamide (DMF) treatment

Power Density12 (W/m2) 8 4 0 0

5

10

15

20

25

Pressure (Bar)

a b s t r a c t Renewable osmotic energy from salinity gradients can be harvested from pressure retarded osmosis (PRO) processes. However, more effective PRO membranes with high power density and pressure resistance are needed to commercialize PRO technologies. In this study, we fabricated thin-film composite (TFC) membranes consisting of a polyamide thin film layer via interfacial polymerization (IP) and a macrovoid-free polyimide support. Three different treatments were explored to improve water flux as well as power density. Firstly, a surfactant of sodium dodecyl sulfate (SDS) was added into the amine IP solution. It resulted in an increase in power density from 8.65 W/m2 of the pristine membrane to 15.79 W/m2 of the modified one. Data from positron annihilation spectroscopy (PAS) for the first time confirmed that SDS significantly affected the thin film formation and thus led to a higher power density. The second treatment was conducted by immersing the TFC membranes in N,N-dimethylformamide (DMF) that resulted in a further increase in power density to 16.87 W/m2. Finally, a combination of both pre- and post-treatments on TFC membranes synergistically enhanced the harvested power density to 18.09 W/ m2, which surpasses all flat-sheet PRO membranes reported in literatures. In addition, the proposed treatments did not sacrifice the robustness of the membranes as they were able to withstand a trans-membrane hydraulic pressure of 22 bar. The newly developed membranes with such mechanical robustness and power density show great potential to practically harvest osmotic power through salinity gradients. Ó 2013 Elsevier B.V. All rights reserved.

1. Introduction Energy security is a global issue due to the explosion of population and over exploitation of fossil fuel [1,2]. Meanwhile, sustain⇑ Corresponding author. Tel.: +65 65166645; fax: +65 67791936. E-mail address: [email protected] (T.-S. Chung). 1385-8947/$ - see front matter Ó 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.cej.2013.12.078

able and renewable energy sources are in high demand because of global warming. The osmotic energy harvested from the salinity gradient between seawater and fresh water as electricity [3] has recently received worldwide attention due to its sustainability and environmental-friendliness with great energy potential [4]. Ideally, around 0.8 kW energy can be extracted when one cubic meter of fresh water mixed with seawater [5]. If the retentate of

196

Y. Cui et al. / Chemical Engineering Journal 242 (2014) 195–203

reverse osmosis (RO) plants is used as a replacement of seawater, more energy can be generated. In addition, it eliminates the disposal issue of environmental-unfriendly RO retentate [6]. The term of ‘‘pressure retarded osmosis (PRO)’’ was firstly proposed by Loeb in 1975 [7,8]. Basically, it harvests the salinity-gradient energy by allowing water transport from a low osmotic pressure side (i.e., the feed solution) to a high osmotic pressure side (i.e., the draw solution) through a semi-permeable membrane naturally. Since the compartment of the draw solution (i.e., seawater) has a fixed volume, its hydraulic pressure increases with the additional water inflow. As water continuously flows in, this pressure is further raised enough to drive a turbine to generate electricity. According to the prototype design by Statkraft, the power density acquired from flat membranes must be equal or great than 5 W/m2 in order to have commercial values for PRO applications [9,10]. However, the current PRO technology is hindered by the lack of suitable semi-permeable membranes that have desirable power output [6,9,11–13]. The ideal PRO membrane should have a high water permeability, a low salt permeability and robust mechanical strength to withstand high pressures [6,9,12,14,15]. So far, the thin film composite (TFC) polyamide membranes fabricated via interfacial polymerization (IP) exhibit promising potential. For the development of flat-sheet PRO membranes, some advancements have been achieved. Zhang et al. [16] fabricated TFC flat-sheet membranes on polyacrylonitrile supports with different post-treatments. They reported a power density of 2.6 W/m2 at 10 bar using 3.5 wt% NaCl as the draw solution. Li et al. [12] fabricated TFC membranes on Torlon substrates with various morphology and compared their structural deformation under high-pressure PRO tests. They also modified the membranes with the aid of pre-compression, polydopamine and chlorine as well as alcohol treatments. The resultant membranes were able to withstand 12 bar and harvested a power density of 2.84 W/m2 using seawater and de-ionized water as feeds. Achilli et al. [17] studied flat-sheet cellulose triacetate (CTA) FO membranes and obtained power densities of 2.7 and 5.1 W/m2 respectively using 35 and 60 g/L NaCl as draw solutions. Meanwhile, the effects of internal and external concentration polarization (ICP and ECP) were explored. Li et al. [18] further modified the free volume of TFC membranes by incorporating a bulky monomer into the interfacial polymerization followed by a methanol treatment, their membranes can harvest a power density of 6 W/m2 and withstand 9 bar using 1 M NaCl as the draw solution. Song et al. [19] fabricated TFC membranes on nano-fiber fabrics and obtained a power density of 15. 4 W/m2 at 15.2 bar by employing 1.06 M NaCl and 0.9 mM NaCl as feeds, while Han et al. [9] synthesized TFC membranes on macrovoid-free Matrimid substrates which can withstand 15 bar and display a power density of 12 W/m2 using 1 M NaCl as the feed. For the development of hollow fiber PRO membranes, Chou et al. [20] developed a TFC hollow fiber membrane on a polyethersulfone (PES) substrate which can harvest a power density of 10.6 W/m2 using seawater brine and wastewater brine as feeds. However, the membrane can only withstand a trans-membrane pressure of 9 bar. Han et al. [14] fabricated TFC hollow fiber membranes with enhanced mechanical properties and water permeability. Their membrane can produce a power density of 14 W/m2 at 16 bar from seawater brine and river water. Recently, Zhang et al. [15] synthesized TFC hollow fiber membranes on specially designed PES substrates. The resultant membrane with the minimal salt permeability can harvest a power density of 24.3 W/m2 and withstand a hydraulic pressure of 20 bar using 1 M NaCl as the feed. Since the power density of current state-of-art flat-sheet membranes is much lower than that of hollow fiber membranes and more energy loss would be encountered when using flat-sheet

membranes in PRO processes due to the flow friction across spacers [21–23], one must further enhance the PRO performance of flat-sheet TFC membranes. Other design and modification strategies should be explored. Surfactants have been reported to improve the formation of the interfacially polymerized layer of TFC membranes [24–26]. Mansourpanah et al. [24,25] added the surfactants into the organic phase as well as the aqueous phase to fabricate TFC nanofiltration (NF) membranes. They found that there were a change in membrane morphology and an improvement on NF performance. This was due to the fact that the amphiphilicity of surfactants acts as a pre-wetting agent that allows a better contact between the amine solution and the hydrophobic substrate [27]. Moreover, the presence of surfactants in the aqueous phase will influence the diffusion of amine into the organic phase. A similar study using polyethylene glycol (PEG) as an additive in amine solutions has also been reported [28]. Hence, we added sodium dodecyl sulfate (SDS) as a surfactant to the amine aqueous solution prior to the interfacial polymerization reaction. In addition to study the SDS influences on FO and PRO performance of TFC membranes, we aim to examine its effects on membrane morphology and other physicochemical properties. Many chemical and physical modification methods have been proposed on the TFC membrane to enhance its performance. The organic solvent treatment method involves the use of solvents but some of these solvents cause membrane morphological changes permanently [29]. Alcohols are considered to be good treatment agents for polyamide membranes because a higher water flux and a higher power density have been reported [16]. This was due to the fact that alcohols swollen up the membranes and washed away some of low molecular fragments; hence, additional free volumes were generated [16,30]. However, alcohols are not the best solvents to dissolve the polyamide fragments. According to the solubility parameter [29], some stronger polar aprotic solvents such as N,N-dimethylformamide (DMF) and dimethylsulfoxide (DMSO) are more suitable for polyamide membranes. Solomon et al. [29,31] reported a new generation of TFC membranes for organic solvent nanofiltration (OSN) fabricated via interfacial polymerization. The membranes possessed high water permeability without compromising rejection after the DMF, DMSO, etc. treatment. In this study, mechanically robust macrovoid-free Matrimid membranes were employed as the support substrate for the formation of TFC flat sheet membranes via interfacial polymerization. Since we aimed to enhance TFC membrane performance, the IP selective layer was firstly synthesized with the addition of SDS into the amine aqueous solution at various concentrations and/or the freshly fabricated TFC membrane was post-treated by the DMF solution at various conditions. The physicochemical changes of TFC membranes and their effects on FO and PRO performance were systematically analyzed with the aid of advanced analytic tools. Since positively results on membrane performance and power density have been observed, we believe that this study would offer some inspirations to design suitable PRO membranes for osmotic power generation in the near future. 2. Materials and methods 2.1. Materials The commercially available polyimide polymer, MatrimidÒ 5218 (Vantico Inc.) was utilized to fabricate the membrane substrate. The solvent N-methyl-2-pyrrolidinone (NMP, >99.5%) and the non-solvent polyethylene glycol 400 (PEG 400, Mw = 400 g/ mol) were purchased from Merck. M-phenylenediamine (MPD, >99%) and trimesoyl chloride (TMC, >98%) were ordered from Sigma–Aldrich and used as the monomers for the interfacial

Y. Cui et al. / Chemical Engineering Journal 242 (2014) 195–203

polymerization (IP) reaction. NaCl (Merck, Germany) was used to determine membrane FO or PRO performance and characterize transport properties. N,N-dimethylformamide (DMF, >99.8%) and sodium dodecyl sulfate (SDS, >99%) acquired from Sigma–Aldrich were used for the post-treatment of TFC membranes. The deionized (DI) water was produced by a Milli-Q ultrapure water system (Millipore, USA). All chemicals were used as received. 2.2. Fabrication of macrovoid-free Matrimid substrate The flat sheet membranes were prepared by a casting process, followed by the non-solvent induced phase inversion [9,12,16,32] with a dope formulation of Matrimid/polyethylene glycol 400 (PEG 400)/N-methyl-2-pyrrolidinone (NMP) at the weight ratio of 18/16/66. The MatrimidÒ 5218 polymer was dried overnight at 80 °C in a vacuum oven to remove moisture prior to preparing the dope solution. The dried MatrimidÒ 5218 (18 wt%) and PEG 400 (16 wt%) were dissolved in NMP at 70 °C overnight. Subsequently, the polymer solution was cool down naturally to room temperature and degased overnight before usage. The membrane was prepared by casting the polymer solution on a glass plate using a casting knife, followed by immersing the membrane into DI water to form an asymmetric structure. The as-cast membranes were removed from the glass plate and preserved in tap water to remove the residual NMP and PEG. 2.3. Interfacial polymerization of thin-film-composite (TFC) membranes The formation of a thin polyamide layer on top of Matrimid substrates was achieved by an IP reaction between MPD in the aqueous phase and TMC in the organic phase. The as-cast Matrimid substrate was firstly immersed in a 2% MPD aqueous solution for 120 s. Then the extra MPD solution was removed and the membrane was thoroughly dried with the filter paper. Then a 0.1% TMC hexane solution was deposited on the top of the MPD-saturated substrate for 60 s, followed by air-dry for 5 min to allow a complete reaction and form the dense TFC layer. At last, the formed TFC membrane was preserved into DI water for further usage. 2.4. Pre-treatment and post-treatment of TFC membranes The surfactant, SDS was used as the pre-treatment agent. It was added into the MPD aqueous solution with various concentrations ranging from 0.2% to 5% based on weight ratio. The TFC membrane was fabricated with the IP process as aforementioned, and the membrane was preserved in DI water overnight to remove the SDS residues. The DMF solvent, which has a similar Hildebrand solubility parameter to the polyamide top layer [29,31], was used for the membrane post-treatment. However, due to its powerful ability to dissolve the non-crosslinked substrate, an aqueous DMF solution was used instead of pure DMF. For the post-treatment, the fabricated TFC membranes were immersed into aqueous DMF solutions with a DMF concentration of 10%, 30% or 50% for a duration varying from 5 min to 24 h. Finally, the treated TFC membranes were rinsed thoroughly by DI water to remove excess treatment agents and any dissolved portions of the membranes, and then they were kept in DI water for further usage. 2.5. Water reclamation through forward osmosis tests of TFC membranes Forward osmosis (FO) experiments were conducted on a labscale FO unit [32–34], as shown in Fig. S1(a). The volumetric flows of both draw solution (1 M NaCl solution) and feed solution (DI water) were 0.2 L/min. They flowed co-currently through the FO

197

cell and were circulated in the setup. The water flux (Jw, L m2 h1, LMH) was calculated as Eq. (1):

Jw ¼

Dm 1 Dt A m

ð1Þ

where Am is the effective area of the FO cell; Dm(g) is the absolute weight loss in the feed solution side, and Dt(h) is the test duration. The reverse salt flux (Js, g m2 h1, gMH) was calculated from the conductivity increment of the feed solution side, as Eq. (2):

Js ¼

DC t V 1 Dt A m

ð2Þ

where DCt (g/L) and V (L) are the changes of salt concentration and the volume of the feed solution, respectively. 2.6. Power generation through PRO tests of TFC membranes To assess the osmotic power generation of the modified TFC membranes, a lab-scale cross flow PRO setup was utilized with a detailed design illustrated in Fig. S1(b) or elsewhere [9,12,16]. The permeation cell was similar to the previous FO cell except the whole cell was made of stainless steel and with two porous meshes to support the membranes [12,16]. A variable-speed gear pump (Cole-Palmer, Vernon Hills, IL) and a high-pressure hydra cell pump (Minneapolis, MN) were employed to circulate the feed solution (DI water) and the draw solution (1 M NaCl) at 0.3 L/min, respectively. In PRO tests, no external hydraulic pressure (DP) was applied in the beginning and then DP was gradually increased. Water flux (Jw) was determined by the absolute weight change of the feed solution and reverse salt flux (Js) was calculated based on conductivity increment of the feed side, similar to FO test [12,16]. The power density (W) is defined as the osmotic energy output per membrane area that was obtained from the product of water flux (Jw) and applied hydraulic pressure (DP) in the draw solution [7,35] as following:

W ¼ J w  DP

ð3Þ

2.7. Characterizations of TFC membranes 2.7.1. Membrane morphology The membrane morphology was examined by FESEM. First, the membrane was dried with a freeze-dryer and fractured in liquid nitrogen. After coated with platinum by a Jeol JFC-1100E Ion Sputtering device, membrane morphology was observed using a FESEM (JEOL JSM-6700). 2.7.2. Membrane micro-structure Position annihilation spectroscopy (PAS) was used as an advanced tool to inspect the membrane micro-structure. To examine the membrane micro-structure as a function of depth from the top surface, the doppler broadening energy spectroscopy (DBES) using a PAS coupled with a variable monoenergy slow positron beam was employed to explore the TFC membranes [36–38]. For each sample, a total of 29 DBES spectra were obtained at different incident energies in the range of 0.1–27 keV measured by an HP Ge detector (EG&G Ortec). A counting rate of around 800 counts per second was maintained and each spectrum carried 1.0 million counts in total. The varied incident energy of the PAS beam was correlated to the depth in a membrane by Eq. (4):

ZðEþ Þ ¼

40

q

E1:6 þ

ð4Þ

where Z (nm) is the depth, q (g/cm3) refers to the density of the polymer, and E+ (keV) is the incident positron energy.

198

Y. Cui et al. / Chemical Engineering Journal 242 (2014) 195–203

The DBES spectra are usually interpreted by two parameters; namely, the S parameter and the R parameter. The S parameter refers to the ratio of integrated counts between 510.3 and 511.7 keV, and it is related to the relative value of low momentum part of positron–electron annihilation radiation. It is sensitive to the chemistry change and the free volume change. On the other hand, the R parameter is defined as the ratio of the total counts from the valley region with an energy width between 364.2 and 496.2 keV (from 3c annihilation) to those from the 511 keV peak region with a width between 504.35 and 517.65 (from 2c annihilation) [37,39]. The R parameter presents the states of voids in the range of nanometer to micrometer. Therefore, it usually represents the changes in pore size and pore size distribution of a membrane. 2.7.3. Water and salt transport properties Water permeability, A (L m2 h1 bar1), and salt rejection, R (%), of the TFC membranes were determined by testing the membranes under pressure in dead-end cells at room temperature, as described in our previous studies[33,34]. Briefly, the water permeability A and the salt rejection R were tested under a trans-membrane pressure, DP, of 5.0 bar. The feed solutions contained 1000 ppm NaCl. The concentrations of NaCl in the feed (Cf) and permeate (Cp) were determined by conductivity measurements with a conductivity meter (Metrohm, Swiss). The water permeability was calculated as Eq. (5):

Dm 1 Dt Am DP

3.2. The addition of SDS in TFC membranes To investigate the effects of SDS pretreatment on the IP selective layer, water flux and reverse salt flux of the resultant membranes synthesized from different SDS concentrations were tested and shown in Fig. 2. Without the addition of SDS into the

45 40

R ¼ ð1  C p =C f Þ  100%

ð6Þ

Water flux (LMH)

ð5Þ

where Dm(g) is the absolute weight loss in the feed solution side, and Dt(h) is the test duration, Am is the effective area of the testing cell, DP (bar) is the applied trans-membrane pressure. The salt rejection R was calculated from Eq. (6):

Water flux Reverse salt flux

35

30

30

25

25

20

20

15

15

10

10

5

5

0

0 0

3.1. Membrane morphology of Matrimid substrates and TFC membranes

1

2

3

5

Fig. 2. Water flux and reverse salt flux of TFC membranes fabricated with MPD solutions containing different SDS concentrations under the PRO mode Draw solution: 1 M NaCl; feed solution: DI water.

Top p surface 30K

p ppart)) Cross-section ((top 30K

10 µm

100 nm

4

SDS concentration (%)

The typical surface morphology of the porous Matrimid substrate is shown in Fig. 1. The Matrimid substrate has a nodule-like,

Cross-section Cross section (middle part) 30K

40

35

3. Results and discussion

100 nm

45

Reverse salt flux (gMH)

Jw ¼

relatively dense top skin layer of 150 nm in thickness and a fully porous, macrovoid-free cross-sectional morphology. The top surface is relatively smooth and full of tiny pores with sizes less than 10 nm, while the bottom surface comprises large interconnected pores with sizes of 200 nm. This is the so-called sponge-like cross-sectional structure, which consists of interconnected opencell pores, is critical to minimize the transport resistance for water permeation under high pressures. This substrate morphology may have advantages of (1) minimizing internal concentration polarization (ICP) within the substrate, (2) maintaining high water permeability and (3) possessing robust mechanical properties so that the TFC membrane can withstand high hydraulic pressures and perform in PRO processes [10,15,40].

Cross-section Cross section 1K

100 nm

Bottom surface 30K

100 nm

Fig. 1. FESEM images of different bulk and surface morphologies of Matrimid flat sheet substrates.

Y. Cui et al. / Chemical Engineering Journal 242 (2014) 195–203

0 SDS

0.2% SDS

0.4% SDS

0 .6% SDS

0 8% SDS

1% SDS

2% SDS

3% SDS

5% SDS

199

Fig. 3. Typical surface morphologies of TFC membranes fabricated with MPD solutions containing different SDS concentrations.

MPD solution, the membrane exhibits a water flux of 26.84 LMH with a reverse salt flux of 6.67 gMH. The water flux shows a convex trend with the addition of SDS and reaches a peak when the MPD solution contains 2% SDS. Besides, there is a slight decrease in reverse salt flux which remains relatively low. When the SDS concentration is at 2%, the membrane achieves an impressive FO performance with a water flux as high as 40.08 LMH and a reverse salt flux as low as 3.71 gMH. Clearly, the SDS concentration has a pronounced impact on membrane’s FO performance. Fig. 3 displays the membrane morphology as a function of SDS concentration. For the pristine TFC membrane, the surface exhibits a typical ‘‘ridge and valley’’ structure of polyamide membranes formed via the IP process. The surface morphology gradually turns into a ‘‘flake-like’’ morphology with an increase in SDS concentration. It is also interesting to notice that there is an incremental growth of these bottom flakes that merge with each other and form larger pieces. Furthermore, the surface morphology illustrates a mainly ‘‘nodular’’ structure when the SDS concentration in the MPD solution reaches 2%. However, the change becomes insignificant and the surface morphology changes back to a ‘‘flake-like’’ structure, when the membranes are fabricated with a SDS concentration higher than 2%. The surface morphologies obtained from AFM images also show the similar patterns and trend with FESEM results, as illustrated in Fig. S2. The trend surprisingly coincides with the trend of water flux. Hence, the change in surface morphology clearly affects the membrane’s FO performance. To explore the mechanism behind the changes in surface morphology and membrane performance, PAS was applied to examine the micro-structure change of these TFC membranes. As illustrated in Fig. 4, S and R parameters were used to describe the changes of free-volume and micro-voids, respectively. Since the SDS is non-reactive and would be washed away after the IP reaction, it does not introduce any chemical change that influences the formation of positroniums. Therefore, the S parameter profile represents the free volume status of the membrane and the evolution of S parameter as a function of positron incident energy could elucidate the morphological transition with the

depth. Basically, S parameter increases with an increase in (i) free volume cavities and/or (ii) free volume number [36]. The depth profiles of S parameter for all the membranes present a similar trend. All these curves exhibit a sharp increase, which was resulted from the back diffusion and scattering of positroniums near the membrane surface [36,37]. Then the S parameter decreases gradually after reaching the peak, which indicates the gradual transition from the IP selective layer to the bulk Matrimid substrate. Due to the quenching effect of polyimide groups [37,41], the S parameter decreases. Hence, the micro-structure of the selective IP layer of the TFC membrane can be analyzed by examining the peak. As illustrated in Fig. 4(a), the peaks of the depth profiles showed different positions and heights. This suggests that the thickness and the free volume intensity of the IP layer changes with SDS concentration in the MPD solutions. Initially, the thickness of the IP layer gradually increases with an increase in SDS concentration up till 2%. With a further increment of SDS concentration greater than 2%, a reverse trend is perceived where the IP layer thickness decreases with increasing SDS concentration. Moreover, the free volume intensity follows a similar trend. In other words, the intensity first intensifies and then reduces with the increment of SDS concentration while displaying a highest value when 2% SDS is added in the MPD solution. Since the reverse salt flux shown in Fig. 2 does not amplify but decline in the presence of SDS, the addition of SDS seems to facilitate the creation as well as the downsizing of free volume cavities, thus increasing the free volume intensity. The depth profiles of R parameter also present a similar trend, as indicated in Fig. 4(b). All curves start with a decrease at the surface region and reach the lowest level before giving a sharp increase. R parameter provides information on the relatively large voids (i.e., from a few nm to hundreds nm). As the voids become smaller and/or the quantity is reduced, R parameter decreases. From a macroscopic view, the lowest region represents the relatively dense layer of the membrane; namely, the IP layer. Consistent with the S parameter, the dense layer becomes thicker with

200

Y. Cui et al. / Chemical Engineering Journal 242 (2014) 195–203

0.465

(a)

0.42

0.46

0SDS 0.4%SDS 0.8%SDS 2%SDS 4%SDS

0.455

0.41

0SDS 0 4%SDS 0.4%SDS 0 8%SDS 0.8%SDS 2%SDS 4%SDS

0.4

0.45

R

S

(b)

0.43

0.39 0.38

0.445

0.37 0.44 0.36 0.435

0.35

0

10

5

15

20

25

Positron Incident Energy (keV)

0

5

10

15

20

25

Positron Incident Energy (keV)

Fig. 4. Depth profiles of (a) S parameter and (b) R parameter of TFC membranes fabricated with MPD solutions containing different SDS concentrations.

Pristine TFC 50% DMF for 0.5 hr 2% SDS

20

2% SDS + 50% DMF for 0.5 hr

Power Density (W/m2)

16

12

8

4

0

0

5

10

15

20

25

Pressure (Bar) Fig. 5. Power density of the TFC membranes with different treatments using brine (1 M NaCl) as the draw solution, and DI water as the feed solution.

intensity is increased [43], but the size of the cavities keeps unchanged or even becomes smaller. As a result, the water flux increases while the reverse salt flux decreases slightly with an increase in SDS concentration. However, with a further increase in SDS concentration, the contact between the organic and aqueous phases is somehow inhibited. As a result, the surface morphology turns back into ‘‘flake-like’’ and partially ‘‘ridge and valley’’ structures. Consequently, a drop in free volume intensity is observed, which lead to a decline in water flux. Nevertheless, the reverse salt flux remains low and unaffected possibly due to the fact that a high SDS concentration mainly inhibits the growth of the number of free volume cavities. Since the membrane made of a 2% SDS addition in the MPD solution shows promising FO performance, its energy output under the PRO mode is evaluated. As indicated in Fig. 5, the membrane obtains a power density of 15.79 W/m2 under a trans-membrane pressure (DP) of 22 bar, which is much higher than the energy of 8.65 W/m2 obtained by the pristine membrane. However, a higher power density is still desirable for practical applications. 3.3. DMF post-treatment on TFC membranes

an increase in SDS concentration. However, an opposite trend occurs when the SDS concentration is higher than 2%. The IP layer becomes thinner when the SDS concentration increases to 4%. Since the addition of SDS in monomer solutions influences the contact between the aqueous and organic phases as well as the extent of interfacial polymerization [24,25,27,42], the selective layer thickness, free volume intensity, membrane surface morphology and the FO performance vary with the SDS concentration. With the addition of SDS, it allows a better interaction between the MPD aqueous solution and the Matrimid substrate, which is hydrophobic. This enhances the wetting of the polyimide substrate and leads to a more homogenous spread of MPD on/in the support. In addition, it helps facilitate a better contact and reaction between the MPD aqueous solution and the TMC organic solution. The MPD can be efficiently transported to the organic phase, thus enlarges the contact area and results in a more thorough and complete reaction. As a consequence, the surface morphology evolves from a ‘‘ridge and valley’’ structure into a ‘‘flake-like’’ structure and finally into a ‘‘nodular’’ structure. The ‘‘nodular’’ structure owns a larger surface area than the other two, which possibly leads to a higher water flux [9]. Moreover, more free volume cavities are created and the free volume

To further enhance the energy output, the TFC membranes were post-treated by DMF with various concentrations for different durations. Fig. 6 shows a comparison of the membranes’ performance under the PRO mode with DP = 0 as a function of DMF concentration and immersion duration. Similar trends on water flux, reverse salt flux and Js/Jw are perceived for membranes immersed in 10% and 30% DMF aqueous solutions. As illustrated in Fig. 6(a), the water flux is increased with an increase in treatment duration at the beginning; however, this increment slows down at certain duration and then remains stagnant over the rest of treatment. In contrast, the TFC membranes treated with 50% DMF shows a slightly different trend. Its water flux rises dramatically in the first one hour and reaches a flux value almost threefold of the original one. However, the water flux starts to decline thereafter due to the high reverse salt flux that lowers the overall osmotic driving force across the membrane. It is interesting to notice that these three treated membranes ultimately reach almost the same water flux if the membranes are over-treated in DMF solutions for at least 12 h or longer. Fig. 6(b) exhibits the change of reverse salt flux as a function of DMF concentration and treatment duration. The reverse salt fluxes

201

Y. Cui et al. / Chemical Engineering Journal 242 (2014) 195–203

(a)

70

(b)

100

65

Water Flux (LMH)

55

Reverse Salt Flux (gMH)

10% DMF 30% DMF 50% DMF

60

50 45 40 35

80

10% DMF 30% DMF 50% DMF

60

40

20

30 25

0 0

50

100

800

1000 1200

0

1400

50

100

Duration (mins) 2.0

1000 1200 1400

(c)

1.5

Js/Jw

800

Duration (mins)

10% DMF 30% DMF 50% DMF

1.0

0.5

0.0 0

50

100

800

1000

1200

1400

Duration (mins) Fig. 6. (a) Water flux, (b) reverse salt flux and (c) selectivity of TFC membranes treated with DMF aqueous solutions under the PRO mode (DP = 0).

Table 1 Transport properties of the TFC membranes with different treatments.

100 Pristine TFC 2% SDS

Water Flux (LMH)

80

50% DMF for 0.5 hr 2% SDS + 50% DMF for 0.5 hr

60

Membrane

PWP (LMH/bar)

Salt rejection (%)

TFC 2% SDS 50% DMF for 0.5 h 2% SDS + 50% DMF for 0.5 h

1.19 ± 0.14 1.52 ± 0.11 2.20 ± 0.17 2.77 ± 0.15

82.9 ± 0.006 82.8 ± 0.004 84.9 ± 0.008 90.6 ± 0.012

Salt rejection tests conditions: under 5 bar, the feed solution is 1000 ppm NaCl solution.

40

20

0 5

10

15

20

25

Pressure (Bar) Fig. 7. Water flux of the TFC membranes with different treatments in PRO tests using brine (1 M NaCl) as the draw solution, and fresh water as the feed solution.

for all three membranes increase with treatment duration in all DMF concentrations. However, similar to the water flux profile, the increment slows down with a prolonged treatment. Moreover, the membrane treated by 50% DMF shows the highest reverse salt flux. Generally, under the same treatment duration, the reverse salt flux increases with increasing DMF concentration. Only a slight increment in reverse salt flux was observed if DMF is at low concentrations, but the increment became significant when the DMF concentration was increased from 30% to 50%. This is due to the fact that DMF and the MPD–TMC polyamide have very close solubility parameters (i.e., 24.8 MPa1/2 vs. 23 MPa1/2), thus low molec-

ular weight molecules and loosely crosslinked polyamide fragments would be dissolved by the DMF and washed away during the rinsing process after the post-treatment [29,31]. Additional free volume and voids are therefore created in the selective layer that facilitates higher water flux and salt transport. Consequently, the Js/Jw ratio rises with an increase in DMF concentration and treatment duration as shown in Fig. 6(c). To obtain a high water flux but maintain a relatively low Js/Jw ratio, a DMF concentration of 50% and duration of 30 min were chosen to post-treat the membranes for osmotic power generation. As indicated in Fig. 5, the resultant membrane displays a power density of 16.87 W/m2 at DP = 22 bar in the PRO test, which is better than that (15.79 W/ m2) of the membrane pre-treated by the 2% SDS addition during the IP reaction. 3.4. A combination of pre- and post-treatments on TFC membranes for osmotic power generation Fig. 5 shows and compares the power density of the membranes after pre- or/and post-treatments using 1 M NaCl brine water and

202

Y. Cui et al. / Chemical Engineering Journal 242 (2014) 195–203

DI water as feeds. The power density follows the order of treatment by SDS + DMF > DMF > SDS > pristine under the same DP. Interestingly, all membranes can withstand a trans-membrane pressure difference (DP) of 22 bar, which suggests un-sacrificed membrane robustness and great potential of these membranes for practical PRO processes. To examine the causes of this high performance and its order, Fig. 7 displays the water flux as a function of DP in PRO tests, while Table 1 summarizes the pure water permeability and salt (NaCl) rejection tested under the RO mode at 5 bar. Water flux in PRO tests decreases due to the lower osmotic gradient across the membranes and the reverse salt flux [9,12,14,15]. Since water permeability and salt rejection of the modified membranes are in the orders of SDS + DMF > DMF > SDS > pristine and SDS + DMF > DMF > SDS ffi pristine, respectively, the water flux in PRO tests follows the order of SDS + DMF > DMF > SDS > pristine under the same DP [15]. Because power density (W) is the product of Jw and DP, the magnitude of W follows the same order of Jw, thus the membrane fabricated with both pre- and post-treatments harvests the highest power density at certain DP, as illustrated in Fig. 5. The enhanced osmotic energy may be explainable from various physicochemical aspects. Firstly, the addition of SDS into the MPD solution improves water permeability while maintaining salt rejection in comparison with the pristine TFC membrane. As a result, a higher power density is produced. Secondly, the DMF treatment creates additional free volume and voids that lead to a higher water flux. According to the previous study [44,45], the selective IP layer consists of a sandwich structure, whereby a dense core (i.e., the actual selective barrier) is sandwiched between two looser polymer structures. The dense core has the highest polymer density and is significantly thinner than the other parts of the polyamide. The DMF treatment may dissolve the loosely crosslinked parts of the polyamide selective layer but not the dense core [29,44]. Although being created by DMF, these voids are removed by high pressures. Hence, the membrane treated with DMF exhibits enhancement in water permeability without sacrificing salt rejection. When combining both treatments, the resultant membrane has the highest power density of 18.09 W/m2 under 22 bar, which is more than double of the energy (8.65 W/m2) of the pristine membrane.

4. Conclusion In this study, we have demonstrated three different treatments on TFC membranes to improve their osmotic power generation. The following conclusions can be drawn: (1) By adding SDS into MPD solutions before the interfacial polymerization reaction, the TFC membrane obtains an enhanced power density of 15.79 W/m2. Positron annihilation spectroscopy (PAS) data have confirmed that SDS increases the free volume of the TFC layer. As a result, the water permeability as well as power density enhances without sacrificing the selectivity. (2) By immersing the TFC membrane into the DMF aqueous solutions, the resultant membrane shows a power density of 16.87 W/m2. This arises from the fact that DMF dissolves loosely crosslinked parts of the selective layer without damaging the core part. As a consequence, the membrane maintains high selectivity under pressures with enhanced permeability and power density. (3) The combination of both SDS and DMF treatments on TFC membranes further enhances the harvested power density to 18.09 W/m2, which surpasses all flat-sheet PRO membranes reported in literatures.

(4) All the modified membranes are able to withstand a hydraulic pressure of 22 bar, which suggests desirable mechanical stability of these membranes and shows great potential for practical osmotic power generation.

Acknowledgements This research was funded by the Singapore National Research Foundation under its (1) Competitive Research Program for the project entitled, ‘‘Advanced FO Membranes and Membrane Systems for Wastewater Treatment, Water Reuse and Seawater Desalination’’ (Grant Numbers: R-279-000-336-281 and R-279000-339-281) and (2) Environmental and Water Technologies Strategic Research Programme and administered by the Environment and Water Industry Programme Office (EWI) of the PUB under the project entitled ‘‘Membrane development for osmotic power generation, Part 1. Materials development and membrane fabrication’’ (1102-IRIS-11-01) and NUS Grant Number of R-279-000-381-279. Thanks are also due to Dr. Xue Li, Miss Xiuzhu Fu and Dr. Dave W. Mangindaan for the suggestions and help on this work. Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.cej.2013.12.078. References [1] A. Evans, V. Strezov, T.J. Evans, Assessment of sustainability indicators for renewable energy technologies, Renew. Sustain. Energy Rev. 13 (2009) 1082– 1088. [2] S. Zhao, L. Zou, C.Y. Tang, D. Mulcahy, Recent developments in forward osmosis: opportunities and challenges, J. Membr. Sci. 396 (2012) 1–21. [3] S.E. Skilhagen, J.E. Dugstad, R.J. Aaberg, Osmotic power — power production based on the osmotic pressure difference between waters with varying salt gradients, Desalination 220 (2008) 476–482. [4] C. Klaysom, T.Y. Cath, T. Depuydt, I.F.J. Vankelecom, Forward and pressure retarded osmosis: potential solutions for global challenges in energy and water supply, Chem. Soc. Rev. 42 (2013) 6959–6989. [5] B.E. Logan, M. Elimelech, Membrane-based processes for sustainable power generation using water, Nature 488 (2012) 313–319. [6] T.S. Chung, X. Li, R.C. Ong, Q. Ge, H. Wang, G. Han, Emerging forward osmosis (FO) technologies and challenges ahead for clean water and clean energy applications, Curr. Opin. Chem. Eng. 1 (2012) 246–257. [7] S. Loeb, Osmotic power-plants, Science 189 (1975) 654–655. [8] S. Loeb, F. Van Hessen, D. Shahaf, Production of energy from concentrated brines by pressure-retarded osmosis: II. Experimental results and projected energy costs, J. Membr. Sci. 1 (1976) 249–269. [9] G. Han, S. Zhang, X. Li, T.S. Chung, High performance thin film composite pressure retarded osmosis (PRO) membranes for renewable salinity-gradient energy generation, J. Membr. Sci. 440 (2013) 108–121. [10] T. Thorsen, T. Holt, The potential for power production from salinity gradients by pressure retarded osmosis, J. Membr. Sci. 335 (2009) 103–110. [11] N.Y. Yip, A. Tiraferri, W.A. Phillip, J.D. Schiffrnan, L.A. Hoover, Y.C. Kim, M. Elimelech, Thin-film composite pressure retarded osmosis membranes for sustainable power generation from salinity gradients, Environ. Sci. Technol. 45 (2011) 4360–4369. [12] X. Li, S. Zhang, F. Fu, T.S. Chung, Deformation and reinforcement of thin-film composite (TFC) polyamide-imide (PAI) membranes for osmotic power generation, J. Membr. Sci. 434 (2013) 204–217. [13] I.L. Alsvik, M.B. Hägg, Pressure retarded osmosis and forward osmosis membranes: materials and methods, Polymers 5 (2013) 303–327. [14] G. Han, P. Wang, T.S. Chung, Highly robust thin-film composite pressure retarded osmosis (PRO) hollow fiber membranes with high power densities for renewable salinity-gradient energy generation, Environ. Sci. Technol. 47 (2013) 8070–8077. [15] S. Zhang, T.S. Chung, Minimizing the instant and accumulative effects of salt permeability to sustain ultrahigh osmotic power density, Environ. Sci. Technol. 47 (2013) 10085–10092. [16] S. Zhang, F.J. Fu, T.S. Chung, Substrate modifications and alcohol treatment on thin film composite membranes for osmotic power, Chem. Eng. Sci. 87 (2013) 40–50. [17] A. Achilli, T.Y. Cath, A.E. Childress, Power generation with pressure retarded osmosis: an experimental and theoretical investigation, J. Membr. Sci. 343 (2009) 42–52.

Y. Cui et al. / Chemical Engineering Journal 242 (2014) 195–203 [18] X. Li, T.S. Chung, Effects of free volume in thin-film composite membranes on osmotic power generation, AIChE J. (2013). [19] X. Song, Z. Liu, D.D. Sun, Energy recovery from concentrated seawater brine by thin-film nanofiber composite pressure retarded osmosis membranes with high power density, Energy Environ. Sci. 6 (2013) 1199–1210. [20] S. Chou, R. Wang, L. Shi, Q. She, C. Tang, A.G. Fane, Thin-film composite hollow fiber membranes for pressure retarded osmosis (PRO) process with high power density, J. Membr. Sci. 389 (2012) 25–33. [21] 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) 64–70. [22] E. Sivertsen, T. Holt, W. Thelin, G. Brekke, Pressure retarded osmosis efficiency for different hollow fibre membrane module flow configurations, Desalination 312 (2013) 107–123. [23] M. Park, J.H. Kim, Numerical analysis of spacer impacts on forward osmosis membrane process using concentration polarization index, J. Membr. Sci. 427 (2013) 10–20. [24] Y. Mansourpanah, K. Alizadeh, S.S. Madaeni, A. Rahimpour, H. Soltani Afarani, Using different surfactants for changing the properties of poly(piperazineamide) TFC nanofiltration membranes, Desalination 271 (2011) 169–177. [25] Y. Mansourpanah, S.S. Madaeni, A. Rahimpour, Fabrication and development of interfacial polymerized thin-film composite nanofiltration membrane using different surfactants in organic phase; study of morphology and performance, J. Membr. Sci. 343 (2009) 219–228. [26] I.C. Kim, B.R. Jeong, S.J. Kim, K.H. Lee, Preparation of high flux thin film composite polyamide membrane: the effect of alkyl phosphate additives during interfacial polymerization, Desalination 308 (2013) 111–114. [27] A.K. Ghosh, B.H. Jeong, X. Huang, E.M.V. Hoek, Impacts of reaction and curing conditions on polyamide composite reverse osmosis membrane properties, J. Membr. Sci. 311 (2008) 34–45. [28] G. Kang, M. Liu, B. Lin, Y. Cao, Q. Yuan, A novel method of surface modification on thin-film composite reverse osmosis membrane by grafting poly(ethylene glycol), Polymer 48 (2007) 1165–1170. [29] M.F. Jimenez Solomon, Y. Bhole, A.G. Livingston, High flux membranes for organic solvent nanofiltration (OSN)-interfacial polymerization with solvent activation, J. Membr. Sci. 423 (2012) 371–382. [30] A. Kulkarni, D. Mukherjee, W.N. Gill, Flux enhancement by hydrophilization of thin film composite reverse osmosis membranes, J. Membr. Sci. 114 (1996) 39–50. [31] M.F. Jimenez Solomon, Y. Bhole, A.G. Livingston, High flux hydrophobic membranes for organic solvent nanofiltration (OSN)—Interfacial polymerization, surface modification and solvent activation, J. Membr. Sci. 434 (2013) 193–203. [32] Y. Cui, H. Wang, H. Wang, T.S. Chung, Micro-morphology and formation of layer-by-layer membranes and their performance in osmotically driven processes, Chem. Eng. Sci. 101 (2013) 13–26.

203

[33] J. Su, Q. Yang, J.F. Teo, T.S. Chung, Cellulose acetate nanofiltration hollow fiber membranes for forward osmosis processes, J. Membr. Sci. 355 (2010) 36–44. [34] X. Li, K.Y. Wang, B. Helmer, T.S. Chung, Thin-film composite membranes and formation mechanism of thin-film layers on hydrophilic cellulose acetate propionate substrates for forward osmosis processes, Ind. Eng. Chem. Res. 51 (2012) 10039–10050. [35] S. Loeb, Production of energy from concentrated brines by pressure-retarded osmosis. 1. Preliminary technical and economic correlations, J. Membr. Sci. 1 (1976) 49–63. [36] H.M. Chen, W.S. Hung, C.H. Lo, S.H. Huang, M.L. Cheng, G. Liu, K.R. Lee, J.Y. Lai, Y.M. Sun, C.C. Hu, R. Suzuki, T. Ohdaira, N. Oshima, Y.C. Jean, Free-volume depth profile of polymeric membranes studied by positron annihilation spectroscopy: layer structure from interfacial polymerization, Macromolecules 40 (2007) 7542–7557. [37] Y.C. Jean, P.E. Mallon, D.M. Schrader, Principle and Application of Positron and Positronium Chemistry, Word Scientific Press, 2003. [38] Y.C. Jean, H.M. Chen, S. Zhang, H.Z. Chen, L.J. Lee, S. Awad, J. Huang, C.H. Lau, H. Wang, F.Y. Li, T.S. Chung, Characterizing free volumes and layer structures in polymeric membranes using slow positron annihilation spectroscopy, 12th International Workshop on Slow Positron Beam, Techniques (Slopos12), 262 (2011). [39] T. Shintani, A. Shimazu, S. Yahagi, H. Matsuyama, Characterization of methylsubstituted polyamides used for reverse osmosis membranes by positron annihilation lifetime spectroscopy and MD simulation, J. Appl. Polym. Sci. 113 (2009) 1757–1762. [40] L. Huang, N.N. Bui, M.T. Meyering, T.J. Hamlin, J.R. McCutcheon, Novel hydrophilic nylon 6,6 microfiltration membrane supported thin film composite membranes for engineered osmosis, J. Membr. Sci. 437 (2013) 141–149. [41] V.P. Shantarovich, T. Suzuki, C. He, V.W. Gustov, Inhibition of positronium formation by polar groups in polymers – relation with TSL experiments, Radiat. Phys. Chem. 67 (2003) 15–23. [42] S.H. Kim, S.Y. Kwak, T. Suzuki, Positron annihilation spectroscopic evidence to demonstrate the flux-enhancement mechanism in morphology-controlled thin-film-composite (TFC) membrane, Environ. Sci. Technol. 39 (2005) 1764– 1770. [43] J.M. Gohil, A.K. Suresh, Development of high flux thin-film composite membrane for water desalination: a statistical study using response surface methodology, Desalination Water Treat. (2013) 1–10. [44] V. Freger, Nanoscale heterogeneity of polyamide membranes formed by interfacial polymerization, Langmuir 19 (2003) 4791–4797. [45] F.A. Pacheco, I. Pinnau, M. Reinhard, J.O. Leckie, Characterization of isolated polyamide thin films of RO and NF membranes using novel TEM techniques, J. Membr. Sci. 358 (2010) 51–59.