- Email: [email protected]
Thermo-sensitive polyelectrolytes as draw solutions in forward osmosis process
Desalination 318 (2013) 48–55
Contents lists available at SciVerse ScienceDirect
Desalination journal homepage: www.elsevier.com/locate/desal
Thermo-sensitive polyelectrolytes as draw solutions in forward osmosis process Ranwen Ou a, Yaqin Wang a, Huanting Wang b,⁎, Tongwen Xu a,⁎⁎ a b
Functional Membrane Laboratory, School of Chemistry and Material Science, University of Science and Technology of China, Hefei, Anhui 230026, China Department of Chemical Engineering, Monash University, Clayton, Victoria 3800, Australia
H I G H L I G H T S • New thermo-sensitive polyelectrolytes were prepared and evaluated as draw solutions for FO. • Hot ultrafiltration was used as a low-energy method to recover the water from draw agents. • The ease of water recovery and reuse make the polyelectrolyte a good candidate for FO.
a r t i c l e
i n f o
Article history: Received 2 December 2012 Received in revised form 20 March 2013 Accepted 22 March 2013 Available online 17 April 2013 Keywords: Polyelectrolyte Draw solution Forward osmosis PNIPAM-SA Hot ultrafiltration
a b s t r a c t A series of polyelectrolytes were evaluated as draw solutions for the forward osmosis (FO) process. Such polyelectrolytes were synthesized by copolymerization of N-isopropylacrylamide with different amounts of sodium acrylate. These polyelectrolytes were thermo-sensitive and water soluble. Hot ultrafiltration (HUF) operated at 45 °C and 2 bar was used as a low-energy method to recover the water from the polyelectrolyte draw solutions. The results showed that 4%PNIPAM-SA solution worked best among nine polyelectrolytes in the forward osmosis process and HUF process, and its FO water flux was 0.347 LMH while the feed solution was pure water and its water recovery fraction was 65.2%. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Forward osmosis (FO) is a process in which a semi-permeable membrane is used as a separation medium, and the differential of osmotic pressure of two sides of membrane acts as driving force [1]. On the permeate side of membrane is the draw solution with higher chemical potential, while on the other side is the feed solution with lower chemical potential flow through. The difference of chemical potential between the draw solution and the feed solution drives the pure water from the feed solution to the draw solution; meanwhile ions are rejected by the semi-permeable membrane. Different terms are used in literature to name the higher chemical potential solution, such as draw solution, draw agent, osmotic agent, osmotic media, driving solution, osmotic engine and so on [2]. Due to its high rejection and operating at low pressure or without additional pressure, the forward ⁎ Corresponding author. Tel.: +61 3 9908 3449. ⁎⁎ Corresponding author. Tel.: +86 551 63601587. E-mail addresses: [email protected] (H. Wang), [email protected] (T. Xu). 0011-9164/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.desal.2013.03.022
osmosis process has found a wide range of applications such as in wastewater treatment, water purification, seawater desalination, food processing, and pharmaceutical industry [2–5]. The performance of semi-permeable membrane and draw solution has a great effect on the forward osmosis process. The chemical potential of draw solution is the driving force. Many different draw solutions with high osmotic pressures have been studied, including salts, fertilizer, saccharide, hydrogel, hydrophilic nano-particle, polyelectrolyte etc. Salts can generate high osmotic pressures, but it is difficult to separate the water from salt solutions. McGinnis in 2002 disclosed that combination of ammonia and carbon dioxide gases in a specific ratio could produce a draw solution with high osmotic pressure in excess of 250 atm. Elimelech and co-workers have intensively investigated ammonium bicarbonate as draw solute for seawater desalination. The draw solute could be decomposed to ammonia and carbon dioxide gases at 65 °C by heating, and readily recovered [6]. Since ammonia is highly soluble in water, thus the product water is unsuitable for drinking directly [7]. Further processing is needed to eliminate ammonia in water to reach the 4th edition Guidelines for Drinkingwater Quality (1.5 mg/l for ammonia, World Health Organization),
R. Ou et al. / Desalination 318 (2013) 48–55
which increase the total cost of desalination. In 2012, Wilson explored switchable polarity solvents as draw solutes for forward osmosis, switchable polarity solvent can be mechanically separated from the purified water after polar to nonpolar phase shift induced by introduction of 1 atm carbon dioxide to 1 atm of air or nitrogen with mild heating. However, the switchable polarity solvent was found to degrade the commercially available FO membrane [8]. In 2011, a fertilizer was evaluated as draw solutions for FO desalination, and 1 kg of fertilizer could extract 11 to 29 l of water from seawater; the diluted fertilizer solution was directly applied for fertigation [9]. Saccharide solution can be used directly without requiring separation in some applications such as in hydration bags, and emergency lifeboats after the forward osmosis step since saccharide solution is directly drinkable [2,10]. The water flux induced by hydrogel draw agents was relatively lower, but it required less energy to recover water from the draw agents [1,11]. In addition, the use of hydrophilic nanoparticles as draw solutes resulted in a moderate water flux and could be separated by a magnetic field, but nanoparticles tend to aggregate after recovery, which has an adverse effect on reuse of nanoparticles [12–14]. In 2011, an integrated FO–UF (forward osmosis– ultrafiltration) system was studied as a potentially sustainable way to recover the hydrophilic nanoparticles. The novel FO–UF process was tested for 5 continuous runs for the purpose of desalination without increasing nanoparticle draw solute size or reducing osmotic functionality [15]. At the same time, thermo-sensitive magnetic nanoparticle was investigated as smart draw solutes in FO without particle size changes upon magnetic separation [16]. Very recently, polyelectrolyte (PSA) was used as draw solute; the molecular weights of PSA used were 1200, 1800 and 5000, and the water flux produced by these PSA solutions was 17 LMH, 15 LMH, 12 LMH, respectively when the PSA concentration was 0.48 g/ml [17]. The polyelectrolyte solution was recovered using with 1 kDa and 3 kDa ultrafiltration membranes at a feed pressure of 10 bar. Even though PSA could be recovered by ultrafiltration, the process consumed a significant amount of energy. Therefore there is still a need for development of new draw solutions for the FO process. Stimuli-responsive polymer hydrogels are a class of hydrogels whose structure, physical property and chemical property change with external environment. The environmental stimuli include pH, ionic strength, temperature, light, electric field, specific chemicals etc. Crosslinked poly (N-isopropylacrylamide) (PNIPAM) is a classical thermo-sensitive hydrogel with both hydrophilic groups and hydrophobic groups in the network. Simultaneously, when it is heated to over 32 °C, it undergoes a reversible phase transition, transferring from hydrophilic to hydrophobic and gradually releasing the water absorbed. This phase transition temperature is termed as lower critical solution temperature (LCST) [1]. This unique characteristic of PNIPAM makes it effective and convenient to recover water from the hydrogel through heating over LCST. Since PNIPAM hydrogel is a cross-linking polymer, the swelling ratio is restricted by its network. According to the recent study carried out at Wang's group [11], a swelling ratio of PNIPAM of 11.9 was observed in the FO step and a water recovery fraction of 17% was achieved in the dewatering step. By co-polymerization of NIPAM with sodium acrylate (SA), poly (NIPAM–SA) possessed ionic segments of sodium acrylate, and thus showed enhanced swelling ratio. Another study showed that modification of hydrogel by incorporating carbon particles led to improved swelling ratio and water recovery fraction [1]. The objectives of the present work are the preparation of thermosensitive draw solutions with NIPAM and sodium acrylate, and the evaluation of the forward osmosis water flux performance of the synthesized draw solution and the regeneration with hot ultrafiltration.
before use. Acrylic acid (AA, purity ≥ 98%), ammonium persulfate (APS, purity ≥ 98%), and sodium hydroxide (NaOH, purity ≥ 96%) were supplied by Sinopharm Chemical Reagent Co., Ltd. 2.2. Synthesis of polyelectrolytes Sodium acrylate (SA) was prepared from acrylic acid by neutralizing it to pH 7–8 by 30% sodium hydroxide solution. The resulting solution without further purification was dried in a vacuum oven at 35 °C. Nine polyelectrolyte samples were synthesized by free-radical polymerization of SA and NIPAM in aqueous solutions, as shown in Fig. 1 [18–21]; the ratios of sodium acrylate (SA) and Nisopropylacrylamide (NIPAM) of polyelectrolytes were 0:100, 2:98, 4:96, 6:94, 10:90, 20:80, 35:65, 50:50, and 100:0. Typically, sodium acrylate (SA) and N-isopropylacrylamide (NIPAM) were dissolved in deionized water to form a 14.28 wt.% solution. Ammonium persulfate (APS) was used as an initiator without further purification, and the molar ratio of monomer and initiator was fixed at 100:1. Then the solution was deaerated by bubbling with nitrogen for 30 min after the initiator was dissolved in the solution. The resulting solution was stirred in a capped bottle at 70 °C for 2 h to complete polymerization. The polyelectrolytes synthesized with 2%, 4%, 6%, 10%, 20%, 35%, and 50% of sodium acrylate (SA), were denoted as 2%PNIPAM–SA (2-P), 4%PNIPAM–SA (4-P), 6%PNIPAM–SA (6-P), 10%PNIPAM–SA (10-P), 20%PNIPAM–SA (20-P), 35%PNIPAM–SA (35-P) and 50%PNIPAM–SA (50-P), respectively. 2.3. Physical properties of polyelectrolytes 2.3.1. LCST of polyelectrolytes Polyelectrolyte solutions were put in conical flasks, and those conical flasks were placed into an oil bath pan. The temperature of oil rose from 20 °C to 80 °C until the clear solution became milky, and this transition temperature was recorded as the LCST. The temperature interval in this process was 1 °C, and the samples were kept at each temperature for 30 min to reach equilibrium. 2.3.2. pH of polyelectrolytes The pH value of polyelectrolyte solutions was tested with a pH meter (FE20-FiveEasy pH, Mettler Toledo). 1 g of polyelectrolyte was dissolved in 6 g pure water to form the test solution. Before testing,
2. Experiments 2.1. Materials N-isopropyl acrylamide (NIPAM, purity >98%) was supplied by Tokyo Chemical Industry Co., Ltd and recrystallized by n-hexane
49
Fig. 1. Routes of synthesis of thermo-sensitive polyelectrolytes.
50
R. Ou et al. / Desalination 318 (2013) 48–55
Table 1 Physical properties of different polyelectrolyte solution.
PNIPAM 2-P 4-P 6-P 10-P 20-P 35-P 50-P PSA
SA %
NIPAM%
LCST/°C
pH
0 2 4 6 10 20 35 50 100
100 98 96 94 90 80 65 50 0
32 28 28 26 21 18 12 – –
5.5 6.85 5.10 6.68 5.18 6.47 6.96 7.31 8.78
the pH meter was calibrated using a standard buffer solution. The result is listed in Table 1. 2.3.3. Molecular weight of polyelectrolytes The viscosity average molecular weights of polyelectrolytes were determined by dilute solution viscometry. When the solvent and temperature are fixed, the intrinsic viscosity of polymer solution is only affected by the molecular weight of polymer. The relationship between intrinsic viscosity and molecular weight can be expressed as Mark–Houwink empirical formula: α
½η ¼ K M
where K, α were determined by absolute method, and it can be found in the Polymer Handbook or papers. The viscometric relationship of poly (N-isopropylacrylamide) follows, ½η ¼ 4:58 10
−4
0:93
M
in H2 O at 20∘C
[22] where [η] is expressed in cm3·g−1. The viscometric relationship of poly (sodium acrylate) follows, −4
½η ¼ 12:4 10
0:5
M
in 1:5 M NaBr at 15∘C
[23] where [η] is expressed in dl·g −1. Since the viscometric relationship of PNIPAM–SA could not be found, in this work, the viscosity average molecular weights of PNIPAM and PSA and the intrinsic viscosity of PNIPAM–SA (50-P) were determined. PNIPAM was dissolved in pure water to form uniform solution, this solution was tested by a Ubbelohde viscometer at (20 ± 0.05) °C. PNIPAM–SA and PSA were dissolved in 1.5 mol/l NaBr solution to form uniform solutions, and tested at (15 ± 0.05) °C by dilute solution viscometry [22,23]. For easy comparison of data, the same concentration was used for all these polyelectrolyte solutions. 2.3.4. Osmotic pressure of polyelectrolytes The osmotic pressure of polyelectrolytes was calculated by
polyelectrolyte chain was ignored since their molecular weight was too large to generate moderate osmotic pressure. Through the calculation, we assumed that Na+ totally dissociated from polyelectrolyte although it was only appropriated for polyelectrolytes with a small amount of SA, and the temperature was 298 K. 2.4. Forward osmosis experiment 2.4.1. Forward osmosis (FO) process The application of polyelectrolyte solution as draw solutions in FO process and the water recovery by hot ultrafiltration are schematically illustrated in Fig. 2. The FO membrane was immersed into pure water overnight before each test. A 14.28 wt.% polyelectrolyte solution which contains 1 g of pure polyelectrolyte was used as draw solution on the active side of the FO membrane, while pure water as feed solution. During the FO process, the polyelectrolyte solution was stirred by a magnetic stirrer to reduce the influence of concentration polarization as shown in Fig. 3. A commercial hydrophilic cellulose-based FO membrane (provided by Hydration Technologies Inc., Albany, OR) was used, and the membrane area was 4.91 cm2. 2.4.2. Determination of water flux and water recovery fraction Water flux was determined by measuring the weight change of polyelectrolyte solution with time using an analytical balance. It was calculated by F¼
V At
where V (l) is the volume change of the polyelectrolyte solution over a given period of time t (h), and the volume change is determined by the mass increase of polyelectrolyte solution. A (m 2) is the effective area of FO membrane used in the permeation cell. Water recovery fraction was calculated by R¼
Wp −W1 Wp −W0
where Wp (g) and Wl (g) are the weight of polyelectrolyte solution before and after hot ultrafiltration, respectively. W0 (g) is the weight of dry polyelectrolyte. 2.4.3. Hot ultrafiltration (HUF) process After the forward osmosis process, the dilute polyelectrolyte solution was dewatered by hot ultrafiltration. As shown in Fig. 2, the dilute polyelectrolyte solution was put into the stirred cell with a 4000 Da SPES-based ultrafiltration membrane (Shanghai Institute of Nuclear Research, HPM 98-16), and the effective area of membrane was 45.36 cm2. An air compressor was connected to the stirred cell to maintain the pressure at 2 bar. The temperature of stirred cell was maintained at 45 °C by a water bath. The hot ultrafiltration process continued for 1 h for each recovery of polyelectrolyte solution.
Π ¼ iCRT 3. Results and discussion where i is the correction factor, C (mol·kg−1) is the molality of solution, R is 8.314 kPa·l·mol−1·K−1, and T (K) is the absolute temperature. In this paper, the osmotic pressure of different 14.28 wt.% polyelectrolyte solutions was calculated using the above formula, and the projected osmotic pressure was based exclusively on the Na+ ions which have the potential to dissociate and create osmotic pressure. Wilson et. al. recommended the use of molality when working with and evaluating feed and draw solutions, and the osmotic pressure might be underestimated due to effects of waters of hydration or overestimated due to “ion pairing” [24]. Thus C (mol·kg −1) is the molality of Na+ ions. And the osmotic pressure which was generated by
3.1. FO process In this work, nine different polyelectrolytes (PNIPAM, PSA, 2-P, 4-P, 6-P, 10-P, 20-P, 35-P, 50-P) were prepared and applied to the FO process, and the FO membrane tested in this work was supplied by Hydration Technologies Inc. (Albany, OR). As shown in Table 1, pH of polyelectrolytes increases with the increasing SA content in general. Table 1 also shows the LCST of different polyelectrolytes in this work. There wasn't an obvious LCST for PSA and 50-P in the temperature range of 20 °C–80 °C. The LCSTs for PNIPAM, 2-P, 4-P, 6-P, 10-P, 20-P, and 35-P decrease successively; this phenomenon is likely
R. Ou et al. / Desalination 318 (2013) 48–55
51
Fig. 2. Schematic diagram of the application of polyelectrolyte solution as draw agent in FO process and recovery of the water by hot ultrafiltration.
due to two reasons: first, the appearance of carboxylate groups destroy the highly ordered solvation shell around polyelectrolyte chains, that increase the system entropy and decrease the energy consumption of phase transition; second, the addition of sodium acrylate can increase the flexibility of polyelectrolyte chain, so it costs less energy to form intramolecular hydrogen bond and achieve phase transition [25,26]. But for 50-P, a large number of sodium acrylate occupy polyelectrolyte chains that will hinder the combination of acylaminos and isopropyls, and go against phase transition of polyelectrolytes. Thus nine types of polyelectrolytes except 50-P and PSA are thermo-sensitive material. Our preliminary tests showed that the 14.28 wt.% PNIPAM solution without chemical crosslinking could not drive pure water permeate through the semi-permeable membrane after several tests. The weight increases of other eight polyelectrolytes in the FO process are shown in
Fig. 4. In this figure, the initial concentration for eight polyelectrolyte solutions was 14.28 wt.% with 1 g of pure polyelectrolyte. As shown in Fig. 4, the weight of polyelectrolytes increases linearly, and performance of 50-P, PSA, 35-P, 4-P, 20-P, 6-P, 2-P, and 10-P decreases successively. After 24 h FO process, 8.789 g, 7.576 g, 1.311 g, 4.088 g, 2.562 g, 0.828 g, 3.08 g, and 4.333 g of water permeated into 50-P solution, PSA solution, 2-P solution, 4-P solution, 6-P solution, 10-P solution, 20-P solution and 35-P solution respectively. Fig. 5 shows the osmotic pressure of different polyelectrolytes, which was calculated at the ideal condition. The trend of weight increase corresponds with the increase in osmotic pressure. The dissociation of sodium ions has an important influence on the osmotic pressure of polyelectrolyte solution. In order to further compare the FO performance of polyelectrolytes, the average water flux of polyelectrolyte solution in 24 h FO process is
Fig. 3. Schematic diagram of forward osmosis process.
Fig. 4. The weight increase of polyelectrolyte solutions in 24 h FO process. Pure water was used as feed solution. The initial concentration for five polyelectrolyte solutions was 14.28 wt.% with 1 g pure polyelectrolyte.
52
R. Ou et al. / Desalination 318 (2013) 48–55 Table 2 Data of dilute solution viscometry.
PNIPAM 50-P PSA
K
α
[η]/dl/g
Mw/g·mol−1
4.58·10−4 ml/g – 12.4·10−4 dl/g
0.93 – 0.5
0.3358 0.6492 0.7912
170400 – 413242
3.2. Hot ultrafiltration (HUF) process
Fig. 5. Osmotic pressure of 14.28 wt.% polyelectrolyte solutions with different SA content. The osmotic pressure was calculated at the ideal condition.
presented in Fig. 6. The water flux of 50-P solution is higher than PSA solution. Chung's study showed that water flux of PSA solution decreased with increasing the molecular weight of polyelectrolyte; this phenomenon may arise from the onset of higher viscosity and lower dissociation of polyelectrolyte with higher molecular weight [17]. 50-P and PSA were both polymerized at 70 °C for 2 h, and poly (sodium acrylate) was expected to be a high molecular weight polymer under this synthesis condition. As shown in Table 2, the intrinsic viscosity of 50-P is lower than that of PSA while they were tested at the same condition. The intrinsic viscosity of 50-P is larger than that of PNIPAM. When the solvent and temperature are determined, the intrinsic viscosity of polymer solution is only affected by the molecular weight of polymer. So the molecular weight of 50-P may be between 170,400 g/mol to 413,242 g/mol. Thus dissociation of 50-P may be larger than that of PSA. Obviously, 50-P and PSA solutions are preferred in the FO process in terms of their ability to generate higher osmotic pressures. But 50-P and PSA are not thermo-sensitive, so it would cost more energy to recover the water by ultrafiltration. The FO performance of 4-P and 35-P is closer to PSA than other thermo-sensitive polyelectrolyte solutions at the same working conditions; therefore, 4-P and 35-P are more appropriate for use as a draw solution in FO process among these polyelectrolytes.
Fig. 6. The average water flux induced by polyelectrolyte solutions in 24 h FO process.
Since 2-P, 4-P, 6-P, 10-P, 20-P and 35-P are thermo-sensitive, when heated to about their LCST, polyelectrolyte chains transfer from hydrophilic to hydrophobic, and gradually separate from water, so it consumes less energy to recover water from these polyelectrolyte solutions, energy consumption will be discussed below. In this work, the water recovery fraction of the thermo-sensitive polyelectrolytes' solution was investigated at different temperatures and the results are shown in Fig. 7. The concentration of test solution was kept at 5 ± 0.5%. It can be observed that the water recovery fraction of 4-P solution was larger than that of other tested solution at all testing temperatures. As mentioned above, the addition of sodium acrylate would improve flexibility of polyelectrolyte chains which benefited the formation of hydrogen bonds among acrylamides, while the improvement of hydrophilicity kept the hydrone staying around carboxylate ions. Thus an appropriate concentration of sodium acrylate enhanced water recovery fraction of polyelectrolyte solution. As shown in Fig. 7, the water recovery fraction of different polyelectrolyte solutions decreases as the SA content increases, and the water recovery fraction increases logarithmically as the temperature increases. The water recovery fraction of 35-P is the lowest. For 2-P or 4-P solution, 38 °C is the desirable temperature for HUF while the polyelectrolyte solution recovered from the stirred cell is about 14.28 wt.%. Though this solution can be reused directly, it is too dilute to collect polyelectrolyte as a solid, which easily causes the loss of polyelectrolyte. By comparison, 45 °C was chosen for ultrafiltration process since the resulting product was very viscous and could be conveniently collected and recycled. After 1 h HUF process at 45 °C and 2 bar, 65% water of 4-P solution was separated from the polyelectrolyte and collected by a beaker as shown in Fig. 2. The remaining solution could be easily recovered using a medicine spoon. 3.3. Repeatability of 4-P solution Compared with other polyelectrolyte solutions, 4-P exhibited the best performance in the FO process and HUF process. So the reuse of 4-P was studied in FO process and the result is shown in Fig. 8.
Fig. 7. Water recovery fraction for 2-P, 4-P, 6-P, 10-P, 20-P, and 35-P solutions at different temperatures.
R. Ou et al. / Desalination 318 (2013) 48–55
Fig. 8. Comparison of the weight increase of 4-P solution after reuse in 24 h FO process. The concentration of 4-P solutions was kept at 14.28 wt.% with 1 g of pure polyelectrolyte (assuming there was no loss of polyelectrolyte during HUF process).
After 24 h, the weight increase of 4-P solution was 4.088 g, 3.774 g and 3.545 g for first, second and third FO processes, respectively. The decline in FO performance was likely caused by the loss of polyelectrolyte during the recovery process. As shown in Fig. 8, the line of weight increase for second FO process (line 2) is similar to that of first FO process (line 1), while the line of weight increase for third FO process (line 3) is different from both of them. It can be observed from line 1 that the flux of 16–24 h is larger than that of 1–16 h, and the same condition is happened at line 2 while its critical point is 8 h. We conjectured reasonably from this phenomenon that polyelectrolyte solution had an optimal concentration for FO process. The loss of polyelectrolyte during FO and HUF processes reduces the original concentration for second and third FO processes, thus the critical point appeared earlier. Because of this, line 3 is different from others. What's more, the weight of 4-P solution increases linearly in 24 h FO process, which means that the dilution of polyelectrolyte solution has a little influence on the osmotic pressure of polyelectrolyte solution in this concentration range. As shown in Fig. 9, 4-P, 50-P and PSA solutions perform well during longer FO operating times (50 h). To further compare the reuse performance of 4-P polyelectrolyte, the average water flux and water recovery fraction of three tests are shown in Fig. 10. The hot ultrafiltration was operated at 45 °C and
53
Fig. 10. Average water flux and water recovery fraction for 4-P solution in three tests. The condition of HUF was 45 °C, and 2 bar.
2 bar. After 1 h HUF process, 65.23%, 64.77% and 64.40% of water were recovered from 4-P solution, which was used in first FO process, second FO process and third FO process, respectively. Reduction of water recovery fraction was mainly due to the reduction of weight increase since water content of polyelectrolyte solution decreased. Although both weight increase and water recovery fraction became lower gradually, the amount of reduction had little impact on FO process and HUF process. The concentration of the polyelectrolyte solution recovered from stirred cup was greater than 14.28 wt.%, and some water was added to achieve a certain concentration before reuse. The average water flux of first, second and third FO processes was 0.347 LMH, 0.320 LMH, and 0.301 LMH, respectively. 3.4. Economic evaluation of HUF process The energy consumption of the hot ultrafiltration process includes energy of heat for heating the solution and the power consumed by the ultrafiltration system. In this study, a normally unutilized heat source, such as low grade geothermal heat or industrial waste heat is assumed as the source for heating polyelectrolyte solution. Thus only the energy required for the ultrafiltration system will be considered. When the thermo-sensitive polyelectrolytes are heated to their LCST, polyelectrolytes separate from water gradually and most of polymer chains hold together, thus polyelectrolyte solutions became two phases which are the organic phase and the milky aqueous solution. The milky aqueous solution will be used as feed solution at a fullscale ultrafiltration process, since the organic phase may easily plug pores of the membrane. This will improve the flux and water recovery fraction efficiently because of lower viscosity. A pilot-scale plan assumed for economic calculation is shown in Fig. 11 [27]. The feed pump power Wfeed can be expressed as: Wfeed ¼ Q p P1 where Qp is the permeate flow rate (m 3·s −1) of feed solution and P1 is the pressure at module inlet (Pa). A 60% efficiency for the feed pump (screw pump) has been assumed. If the air is assumed to be an ideal gas and the compression adiabatic, then the air compressor power Wcomp which is the power necessary to compress the air from an initial state (Patm) to a final state (P1) can be expressed as:
Fig. 9. The weight increase of polyelectrolyte solutions in 50 h FO process.
Wcomp ¼
γ Q gas Patm γ1
P1 Patm
γ1 γ
1
54
R. Ou et al. / Desalination 318 (2013) 48–55
solution which can be applied in many FO process such as concentration of dilute industrial wastewater and source water purification. Though the viscosity of 4-P solution would increase and might affect the performance of 4-P solution, solid hydrogels even can be used as a draw agent and generate a moderate flux [1,11]. Thus 4-P solution has the potential to be applied in desalination. 4. Conclusions
Fig. 11. Pilot-scale ultrafiltration process assumed for economic estimation.
where γ is 1.395 for air, Qgas is gas flow rate measured at atmospheric pressure (m3·s−1). 85% efficiency for the compressor has been assumed. The concentrated solution will be mixed with organic phase to reuse. Then a total power WT is calculated as the sum of the different powers: WT ¼ Wfeed þ Wcomp The energy consumed per m 3 of permeate produced (Wh·m −3) is given by: Ec ¼
In this work, nine polyelectrolytes (PNIPAM, 2-P, 4-P, 6-P, 10-P, 20-P, 35-P, 50-P, PSA) were synthesized and studied as draw solutions in the forward osmosis process. The water permeated into polyelectrolyte solution was recovered by hot ultrafiltration which was operated at 45 °C and 2 bar for 1 h. By contrast, 2-P, 4-P, 6-P, 10-P, 20-P and 35-P solutions were more suitable for used in the FO process than PSA and 50-P, since the formers were more thermo-sensitive, they were expected to transfer from hydrophilic to hydrophobic when heated to about their LCST, and thus gradually separated from water, reducing the energy required to recover the water from former polyelectrolyte solutions. Further investigation showed that the forward osmosis performance and water recovery efficiency of 4-P and 35-P solutions were better than those of other solutions. For 4-P solution, the average water flux of first, second and third FO processes were 0.347 LMH, 0.320 LMH, and 0.301 LMH, respectively, while their water recovery fractions were 65.23%, 64.77%, and 64.40%. The water flux induced by 14.28 wt.% 4-P solution was smaller than that by other draw solutions used in FO process, but the ease of water recovery and reused would make it potentially a good candidate for the FO process. Acknowledgments
WT FP ⋅S
where Fp is permeate flux (m 3·m −2·h −l), and S is filtration area (m 2). Since operated condition has an important effect on capital investment and cost, the economic evaluation here is calculated from assumed parameters in Table 3. At the assumed condition, it costs around 1.18 RMB to recover 1 ton product water. Overall, 14.28 wt.% 4-P solution has a number of desirable properties among the polyelectrolyte solutions besides the calculated osmotic pressure (3.5 atm). The osmotic pressure of 70,000 ppm NaCl solution is about 3.5 atm. Thus, 14.28 wt.% 4-P solution can be used as the draw solution for brackish water desalination and food processing with TDS ranging from thousands to ten thousands [2,28]. In addition, the osmotic pressure of polyelectrolyte solution increases with its increasing concentration theoretically. Increasing the concentration to 60 wt.% and using the molal projection would produce a 31.2 atm solution which is similar to seawater, and 75 wt.% would produce a 62.4 atm
Table 3 Parameters for economic evaluation. Items
Parameters
Ultrafiltration membrane module Membrane area (m2) Inlet pressure P1 (Pa) Atmospheric pressure Patm (Pa) Water recovery fraction (%) Permeate flow rate Qp (m3·h−1) Gas flow rate Qgas (m3·min−1) Permeate flux Fp (m3·m−2·h−1) Price of electricity (RMB/KWH) Totally power WT (W) Energy consumption EC (Wh·m−3) Cost (RMB·ton−1)
Minipore HM-160 50.5 200000 101325 80 0.625 0.11 0.01 0.6 992.39 1965.13 1.18
We thank the financial support from the National Natural Science Foundation of China (nos. 21128004, 21025626). References [1] D. Li, X.Y. Zhang, J.F. Yao, Y. Zeng, G.P. Simon, H.T. Wang, Composite polymer hydrogels as draw agents in forward osmosis and solar dewatering, Soft Matter 7 (2011) 10048–10056. [2] T.Y. Cath, A.E. Childress, M. Elimelech, Forward osmosis: principles, applications, and recent developments, J. Membr. Sci. 281 (2006) 70–87. [3] A. Achilli, T.Y. Cath, A.E. Childress, Selection of inorganic-based draw solutions for forward osmosis applications, J. Membr. Sci. 364 (2010) 233–241. [4] R.W. Holloway, A.E. Childress, K.E. Dennett, T.Y. Cath, Forward osmosis for concentration of anaerobic digester centrate, Water Res. 41 (2007) 4005–4014. [5] N.Y. Yip, M. Elimelech, et al., High performance thin-film composite forward osmosis membrane, Environ. Sci. Technol. 44 (2010) 3812–3818. [6] J.R. McCutcheon, R.L. McGinnis, M. Elimelech, A novel ammonia–carbon dioxide forward (direct) osmosis desalination process, Desalination 174 (2005) 1–11. [7] Z.Y. Liu, H.W. Bai, J. Lee, D.D. Sun, A low-energy forward osmosis process to produce drinking water, Energy Environ. Sci. 4 (2011) 2582–2585. [8] M.L. Stone, C. Rae, F.F. Stewart, A.D. Wlison, Switchable polarity solvents as draw solutes for forward osmosis, Desalination 312 (2013) 124–129. [9] S. Phuntsho, H.K. Shon, et al., A novel low energy fertilizer driven forward osmosis desalination for direct fertigation: evaluating the performance of fertilizer draw solutions, J. Membr. Sci. 375 (2011) 172–181. [10] J.S. Yong, W.A. Philip, M. Elimelech, Coupled reverse draw solute permeation and water flux in forward osmosis with neutral draw solutes, J. Membr. Sci. 392 (2012) 9–17. [11] D. Li, X.Y. Zhang, J.F. Yao, G.P. Simon, H.T. Wang, Stimuli-responsive polymer hydrogels as a new class of draw agent for forward osmosis desalination, Chem. Commun. 47 (2011) 1710–1712. [12] T.S. Chung, S. Zhang, K.Y. Wang, J.C. Su, M.M. Ling, Forward osmosis processes: yesterday, today and tomorrow, Desalination 287 (2012) 78–81. [13] P. Wang, Q.H. Shi, Y.F. Shi, K.K. Clark, G.D. Stucky, A.A. Keller, Magnetic permanently confined micelle arrays for treating hydrophobic organic compound contamination, J. Am. Chem. Soc. 131 (2009) 182–188. [14] M.M. Ling, K.Y. Wang, T.S. Chung, Highly water-soluble magnetic nanoparticles as novel draw solutes in forward osmosis for water reuse, Ind. Eng. Chem. Res. 49 (2010) 5869–5876. [15] M.M. Ling, T.S. Chung, Desalination process using super hydrophilic nanoparticles via forward osmosis integrated with ultrafiltration regeneration, Desalination 278 (2011) 194–202.
R. Ou et al. / Desalination 318 (2013) 48–55 [16] M.M. Ling, T.S. Chung, X.M. Lu, Facile synthesis of thermosensitive magnetic nanoparticles as “smart” draw solutes in forward osmosis, Chem. Commun. 47 (2011) 10788–10790. [17] Q.C. Ge, J.C. Su, G.L. Amy, T.S. Chung, Exploration of polyelectrolytes as draw solutes in forward osmosis processes, Water Res. 46 (2012) 1318–1326. [18] B. Jeong, S.W. Kim, Y.H. Bae, Thermosensitive sol–gel reversible hydrogels, Adv. Drug Deliv. Rev. 64 (2012) 154–162. [19] R.M.K. Ramanan, P. Chellamuthu, L.P. Tang, K.T. Nguyen, Development of a temperature-sensitive composite hydrogel for drug delivery applications, Biotechnol. Prog. 22 (2006) 118–125. [20] R. Pelton, Temperature-sensitive aqueous microgels, Adv. Colloid Interface Sci. 85 (2000) 1–33. [21] J. Zhang, N.A. Peppas, Synthesis and characterization of pH- and temperaturesensitive poly(methacrylic acid)/poly(n-isopropylacrylamide) interpenetrating polymeric networks, Macromolecules 33 (2000) 102–107. [22] O. Chiantore, M. Guaita, L. Trosarelli, Solution properties of poly (n-isopropylacrylamide), Makromol. Chem. 180 (1979) 969–973.
55
[23] A. Takahashi, M. Nagasawa, Excluded volume of polyelectrolyte in salt solutions, J. Am. Chem. Soc. 86 (4) (1964) 543–548. [24] A.D. Wilson, F.F. Stewart, Deriving osmotic pressures of draw solutes used in osmotically driven membrane process, J. Membr. Sci. 431 (2013) 205–211. [25] P.W. Zhu, D.H. Napper, Volume phase transitions of poly (N-isopropylacrylamide) latex particles in mixed water-N, N-dimethylformamide solutions, Chem. Phys. Lett. 256 (1996) 51–56. [26] H.M. Crowther, B. Vincent, Swelling behavior of poly-N-isopropylacrylamide microgel particles in alcoholic solutions, Colloid Polym. Sci. 276 (1998) 46–51. [27] S. Laborie, C. Cabassud, L. Durand-Bourlier, J.M. Laine, Fouling control by air sparging inside hollow fibre membrane effects on energy consumption, Desalination 118 (1998) 189–196. [28] M.S. Mohsen, O.R. Al-Jayyousi, Brackish water desalination: an alternative for water supply enhancement in Jordan, Desalination 124 (1999) 163–174.
Contents lists available at SciVerse ScienceDirect
Desalination journal homepage: www.elsevier.com/locate/desal
Thermo-sensitive polyelectrolytes as draw solutions in forward osmosis process Ranwen Ou a, Yaqin Wang a, Huanting Wang b,⁎, Tongwen Xu a,⁎⁎ a b
Functional Membrane Laboratory, School of Chemistry and Material Science, University of Science and Technology of China, Hefei, Anhui 230026, China Department of Chemical Engineering, Monash University, Clayton, Victoria 3800, Australia
H I G H L I G H T S • New thermo-sensitive polyelectrolytes were prepared and evaluated as draw solutions for FO. • Hot ultrafiltration was used as a low-energy method to recover the water from draw agents. • The ease of water recovery and reuse make the polyelectrolyte a good candidate for FO.
a r t i c l e
i n f o
Article history: Received 2 December 2012 Received in revised form 20 March 2013 Accepted 22 March 2013 Available online 17 April 2013 Keywords: Polyelectrolyte Draw solution Forward osmosis PNIPAM-SA Hot ultrafiltration
a b s t r a c t A series of polyelectrolytes were evaluated as draw solutions for the forward osmosis (FO) process. Such polyelectrolytes were synthesized by copolymerization of N-isopropylacrylamide with different amounts of sodium acrylate. These polyelectrolytes were thermo-sensitive and water soluble. Hot ultrafiltration (HUF) operated at 45 °C and 2 bar was used as a low-energy method to recover the water from the polyelectrolyte draw solutions. The results showed that 4%PNIPAM-SA solution worked best among nine polyelectrolytes in the forward osmosis process and HUF process, and its FO water flux was 0.347 LMH while the feed solution was pure water and its water recovery fraction was 65.2%. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Forward osmosis (FO) is a process in which a semi-permeable membrane is used as a separation medium, and the differential of osmotic pressure of two sides of membrane acts as driving force [1]. On the permeate side of membrane is the draw solution with higher chemical potential, while on the other side is the feed solution with lower chemical potential flow through. The difference of chemical potential between the draw solution and the feed solution drives the pure water from the feed solution to the draw solution; meanwhile ions are rejected by the semi-permeable membrane. Different terms are used in literature to name the higher chemical potential solution, such as draw solution, draw agent, osmotic agent, osmotic media, driving solution, osmotic engine and so on [2]. Due to its high rejection and operating at low pressure or without additional pressure, the forward ⁎ Corresponding author. Tel.: +61 3 9908 3449. ⁎⁎ Corresponding author. Tel.: +86 551 63601587. E-mail addresses: [email protected] (H. Wang), [email protected] (T. Xu). 0011-9164/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.desal.2013.03.022
osmosis process has found a wide range of applications such as in wastewater treatment, water purification, seawater desalination, food processing, and pharmaceutical industry [2–5]. The performance of semi-permeable membrane and draw solution has a great effect on the forward osmosis process. The chemical potential of draw solution is the driving force. Many different draw solutions with high osmotic pressures have been studied, including salts, fertilizer, saccharide, hydrogel, hydrophilic nano-particle, polyelectrolyte etc. Salts can generate high osmotic pressures, but it is difficult to separate the water from salt solutions. McGinnis in 2002 disclosed that combination of ammonia and carbon dioxide gases in a specific ratio could produce a draw solution with high osmotic pressure in excess of 250 atm. Elimelech and co-workers have intensively investigated ammonium bicarbonate as draw solute for seawater desalination. The draw solute could be decomposed to ammonia and carbon dioxide gases at 65 °C by heating, and readily recovered [6]. Since ammonia is highly soluble in water, thus the product water is unsuitable for drinking directly [7]. Further processing is needed to eliminate ammonia in water to reach the 4th edition Guidelines for Drinkingwater Quality (1.5 mg/l for ammonia, World Health Organization),
R. Ou et al. / Desalination 318 (2013) 48–55
which increase the total cost of desalination. In 2012, Wilson explored switchable polarity solvents as draw solutes for forward osmosis, switchable polarity solvent can be mechanically separated from the purified water after polar to nonpolar phase shift induced by introduction of 1 atm carbon dioxide to 1 atm of air or nitrogen with mild heating. However, the switchable polarity solvent was found to degrade the commercially available FO membrane [8]. In 2011, a fertilizer was evaluated as draw solutions for FO desalination, and 1 kg of fertilizer could extract 11 to 29 l of water from seawater; the diluted fertilizer solution was directly applied for fertigation [9]. Saccharide solution can be used directly without requiring separation in some applications such as in hydration bags, and emergency lifeboats after the forward osmosis step since saccharide solution is directly drinkable [2,10]. The water flux induced by hydrogel draw agents was relatively lower, but it required less energy to recover water from the draw agents [1,11]. In addition, the use of hydrophilic nanoparticles as draw solutes resulted in a moderate water flux and could be separated by a magnetic field, but nanoparticles tend to aggregate after recovery, which has an adverse effect on reuse of nanoparticles [12–14]. In 2011, an integrated FO–UF (forward osmosis– ultrafiltration) system was studied as a potentially sustainable way to recover the hydrophilic nanoparticles. The novel FO–UF process was tested for 5 continuous runs for the purpose of desalination without increasing nanoparticle draw solute size or reducing osmotic functionality [15]. At the same time, thermo-sensitive magnetic nanoparticle was investigated as smart draw solutes in FO without particle size changes upon magnetic separation [16]. Very recently, polyelectrolyte (PSA) was used as draw solute; the molecular weights of PSA used were 1200, 1800 and 5000, and the water flux produced by these PSA solutions was 17 LMH, 15 LMH, 12 LMH, respectively when the PSA concentration was 0.48 g/ml [17]. The polyelectrolyte solution was recovered using with 1 kDa and 3 kDa ultrafiltration membranes at a feed pressure of 10 bar. Even though PSA could be recovered by ultrafiltration, the process consumed a significant amount of energy. Therefore there is still a need for development of new draw solutions for the FO process. Stimuli-responsive polymer hydrogels are a class of hydrogels whose structure, physical property and chemical property change with external environment. The environmental stimuli include pH, ionic strength, temperature, light, electric field, specific chemicals etc. Crosslinked poly (N-isopropylacrylamide) (PNIPAM) is a classical thermo-sensitive hydrogel with both hydrophilic groups and hydrophobic groups in the network. Simultaneously, when it is heated to over 32 °C, it undergoes a reversible phase transition, transferring from hydrophilic to hydrophobic and gradually releasing the water absorbed. This phase transition temperature is termed as lower critical solution temperature (LCST) [1]. This unique characteristic of PNIPAM makes it effective and convenient to recover water from the hydrogel through heating over LCST. Since PNIPAM hydrogel is a cross-linking polymer, the swelling ratio is restricted by its network. According to the recent study carried out at Wang's group [11], a swelling ratio of PNIPAM of 11.9 was observed in the FO step and a water recovery fraction of 17% was achieved in the dewatering step. By co-polymerization of NIPAM with sodium acrylate (SA), poly (NIPAM–SA) possessed ionic segments of sodium acrylate, and thus showed enhanced swelling ratio. Another study showed that modification of hydrogel by incorporating carbon particles led to improved swelling ratio and water recovery fraction [1]. The objectives of the present work are the preparation of thermosensitive draw solutions with NIPAM and sodium acrylate, and the evaluation of the forward osmosis water flux performance of the synthesized draw solution and the regeneration with hot ultrafiltration.
before use. Acrylic acid (AA, purity ≥ 98%), ammonium persulfate (APS, purity ≥ 98%), and sodium hydroxide (NaOH, purity ≥ 96%) were supplied by Sinopharm Chemical Reagent Co., Ltd. 2.2. Synthesis of polyelectrolytes Sodium acrylate (SA) was prepared from acrylic acid by neutralizing it to pH 7–8 by 30% sodium hydroxide solution. The resulting solution without further purification was dried in a vacuum oven at 35 °C. Nine polyelectrolyte samples were synthesized by free-radical polymerization of SA and NIPAM in aqueous solutions, as shown in Fig. 1 [18–21]; the ratios of sodium acrylate (SA) and Nisopropylacrylamide (NIPAM) of polyelectrolytes were 0:100, 2:98, 4:96, 6:94, 10:90, 20:80, 35:65, 50:50, and 100:0. Typically, sodium acrylate (SA) and N-isopropylacrylamide (NIPAM) were dissolved in deionized water to form a 14.28 wt.% solution. Ammonium persulfate (APS) was used as an initiator without further purification, and the molar ratio of monomer and initiator was fixed at 100:1. Then the solution was deaerated by bubbling with nitrogen for 30 min after the initiator was dissolved in the solution. The resulting solution was stirred in a capped bottle at 70 °C for 2 h to complete polymerization. The polyelectrolytes synthesized with 2%, 4%, 6%, 10%, 20%, 35%, and 50% of sodium acrylate (SA), were denoted as 2%PNIPAM–SA (2-P), 4%PNIPAM–SA (4-P), 6%PNIPAM–SA (6-P), 10%PNIPAM–SA (10-P), 20%PNIPAM–SA (20-P), 35%PNIPAM–SA (35-P) and 50%PNIPAM–SA (50-P), respectively. 2.3. Physical properties of polyelectrolytes 2.3.1. LCST of polyelectrolytes Polyelectrolyte solutions were put in conical flasks, and those conical flasks were placed into an oil bath pan. The temperature of oil rose from 20 °C to 80 °C until the clear solution became milky, and this transition temperature was recorded as the LCST. The temperature interval in this process was 1 °C, and the samples were kept at each temperature for 30 min to reach equilibrium. 2.3.2. pH of polyelectrolytes The pH value of polyelectrolyte solutions was tested with a pH meter (FE20-FiveEasy pH, Mettler Toledo). 1 g of polyelectrolyte was dissolved in 6 g pure water to form the test solution. Before testing,
2. Experiments 2.1. Materials N-isopropyl acrylamide (NIPAM, purity >98%) was supplied by Tokyo Chemical Industry Co., Ltd and recrystallized by n-hexane
49
Fig. 1. Routes of synthesis of thermo-sensitive polyelectrolytes.
50
R. Ou et al. / Desalination 318 (2013) 48–55
Table 1 Physical properties of different polyelectrolyte solution.
PNIPAM 2-P 4-P 6-P 10-P 20-P 35-P 50-P PSA
SA %
NIPAM%
LCST/°C
pH
0 2 4 6 10 20 35 50 100
100 98 96 94 90 80 65 50 0
32 28 28 26 21 18 12 – –
5.5 6.85 5.10 6.68 5.18 6.47 6.96 7.31 8.78
the pH meter was calibrated using a standard buffer solution. The result is listed in Table 1. 2.3.3. Molecular weight of polyelectrolytes The viscosity average molecular weights of polyelectrolytes were determined by dilute solution viscometry. When the solvent and temperature are fixed, the intrinsic viscosity of polymer solution is only affected by the molecular weight of polymer. The relationship between intrinsic viscosity and molecular weight can be expressed as Mark–Houwink empirical formula: α
½η ¼ K M
where K, α were determined by absolute method, and it can be found in the Polymer Handbook or papers. The viscometric relationship of poly (N-isopropylacrylamide) follows, ½η ¼ 4:58 10
−4
0:93
M
in H2 O at 20∘C
[22] where [η] is expressed in cm3·g−1. The viscometric relationship of poly (sodium acrylate) follows, −4
½η ¼ 12:4 10
0:5
M
in 1:5 M NaBr at 15∘C
[23] where [η] is expressed in dl·g −1. Since the viscometric relationship of PNIPAM–SA could not be found, in this work, the viscosity average molecular weights of PNIPAM and PSA and the intrinsic viscosity of PNIPAM–SA (50-P) were determined. PNIPAM was dissolved in pure water to form uniform solution, this solution was tested by a Ubbelohde viscometer at (20 ± 0.05) °C. PNIPAM–SA and PSA were dissolved in 1.5 mol/l NaBr solution to form uniform solutions, and tested at (15 ± 0.05) °C by dilute solution viscometry [22,23]. For easy comparison of data, the same concentration was used for all these polyelectrolyte solutions. 2.3.4. Osmotic pressure of polyelectrolytes The osmotic pressure of polyelectrolytes was calculated by
polyelectrolyte chain was ignored since their molecular weight was too large to generate moderate osmotic pressure. Through the calculation, we assumed that Na+ totally dissociated from polyelectrolyte although it was only appropriated for polyelectrolytes with a small amount of SA, and the temperature was 298 K. 2.4. Forward osmosis experiment 2.4.1. Forward osmosis (FO) process The application of polyelectrolyte solution as draw solutions in FO process and the water recovery by hot ultrafiltration are schematically illustrated in Fig. 2. The FO membrane was immersed into pure water overnight before each test. A 14.28 wt.% polyelectrolyte solution which contains 1 g of pure polyelectrolyte was used as draw solution on the active side of the FO membrane, while pure water as feed solution. During the FO process, the polyelectrolyte solution was stirred by a magnetic stirrer to reduce the influence of concentration polarization as shown in Fig. 3. A commercial hydrophilic cellulose-based FO membrane (provided by Hydration Technologies Inc., Albany, OR) was used, and the membrane area was 4.91 cm2. 2.4.2. Determination of water flux and water recovery fraction Water flux was determined by measuring the weight change of polyelectrolyte solution with time using an analytical balance. It was calculated by F¼
V At
where V (l) is the volume change of the polyelectrolyte solution over a given period of time t (h), and the volume change is determined by the mass increase of polyelectrolyte solution. A (m 2) is the effective area of FO membrane used in the permeation cell. Water recovery fraction was calculated by R¼
Wp −W1 Wp −W0
where Wp (g) and Wl (g) are the weight of polyelectrolyte solution before and after hot ultrafiltration, respectively. W0 (g) is the weight of dry polyelectrolyte. 2.4.3. Hot ultrafiltration (HUF) process After the forward osmosis process, the dilute polyelectrolyte solution was dewatered by hot ultrafiltration. As shown in Fig. 2, the dilute polyelectrolyte solution was put into the stirred cell with a 4000 Da SPES-based ultrafiltration membrane (Shanghai Institute of Nuclear Research, HPM 98-16), and the effective area of membrane was 45.36 cm2. An air compressor was connected to the stirred cell to maintain the pressure at 2 bar. The temperature of stirred cell was maintained at 45 °C by a water bath. The hot ultrafiltration process continued for 1 h for each recovery of polyelectrolyte solution.
Π ¼ iCRT 3. Results and discussion where i is the correction factor, C (mol·kg−1) is the molality of solution, R is 8.314 kPa·l·mol−1·K−1, and T (K) is the absolute temperature. In this paper, the osmotic pressure of different 14.28 wt.% polyelectrolyte solutions was calculated using the above formula, and the projected osmotic pressure was based exclusively on the Na+ ions which have the potential to dissociate and create osmotic pressure. Wilson et. al. recommended the use of molality when working with and evaluating feed and draw solutions, and the osmotic pressure might be underestimated due to effects of waters of hydration or overestimated due to “ion pairing” [24]. Thus C (mol·kg −1) is the molality of Na+ ions. And the osmotic pressure which was generated by
3.1. FO process In this work, nine different polyelectrolytes (PNIPAM, PSA, 2-P, 4-P, 6-P, 10-P, 20-P, 35-P, 50-P) were prepared and applied to the FO process, and the FO membrane tested in this work was supplied by Hydration Technologies Inc. (Albany, OR). As shown in Table 1, pH of polyelectrolytes increases with the increasing SA content in general. Table 1 also shows the LCST of different polyelectrolytes in this work. There wasn't an obvious LCST for PSA and 50-P in the temperature range of 20 °C–80 °C. The LCSTs for PNIPAM, 2-P, 4-P, 6-P, 10-P, 20-P, and 35-P decrease successively; this phenomenon is likely
R. Ou et al. / Desalination 318 (2013) 48–55
51
Fig. 2. Schematic diagram of the application of polyelectrolyte solution as draw agent in FO process and recovery of the water by hot ultrafiltration.
due to two reasons: first, the appearance of carboxylate groups destroy the highly ordered solvation shell around polyelectrolyte chains, that increase the system entropy and decrease the energy consumption of phase transition; second, the addition of sodium acrylate can increase the flexibility of polyelectrolyte chain, so it costs less energy to form intramolecular hydrogen bond and achieve phase transition [25,26]. But for 50-P, a large number of sodium acrylate occupy polyelectrolyte chains that will hinder the combination of acylaminos and isopropyls, and go against phase transition of polyelectrolytes. Thus nine types of polyelectrolytes except 50-P and PSA are thermo-sensitive material. Our preliminary tests showed that the 14.28 wt.% PNIPAM solution without chemical crosslinking could not drive pure water permeate through the semi-permeable membrane after several tests. The weight increases of other eight polyelectrolytes in the FO process are shown in
Fig. 4. In this figure, the initial concentration for eight polyelectrolyte solutions was 14.28 wt.% with 1 g of pure polyelectrolyte. As shown in Fig. 4, the weight of polyelectrolytes increases linearly, and performance of 50-P, PSA, 35-P, 4-P, 20-P, 6-P, 2-P, and 10-P decreases successively. After 24 h FO process, 8.789 g, 7.576 g, 1.311 g, 4.088 g, 2.562 g, 0.828 g, 3.08 g, and 4.333 g of water permeated into 50-P solution, PSA solution, 2-P solution, 4-P solution, 6-P solution, 10-P solution, 20-P solution and 35-P solution respectively. Fig. 5 shows the osmotic pressure of different polyelectrolytes, which was calculated at the ideal condition. The trend of weight increase corresponds with the increase in osmotic pressure. The dissociation of sodium ions has an important influence on the osmotic pressure of polyelectrolyte solution. In order to further compare the FO performance of polyelectrolytes, the average water flux of polyelectrolyte solution in 24 h FO process is
Fig. 3. Schematic diagram of forward osmosis process.
Fig. 4. The weight increase of polyelectrolyte solutions in 24 h FO process. Pure water was used as feed solution. The initial concentration for five polyelectrolyte solutions was 14.28 wt.% with 1 g pure polyelectrolyte.
52
R. Ou et al. / Desalination 318 (2013) 48–55 Table 2 Data of dilute solution viscometry.
PNIPAM 50-P PSA
K
α
[η]/dl/g
Mw/g·mol−1
4.58·10−4 ml/g – 12.4·10−4 dl/g
0.93 – 0.5
0.3358 0.6492 0.7912
170400 – 413242
3.2. Hot ultrafiltration (HUF) process
Fig. 5. Osmotic pressure of 14.28 wt.% polyelectrolyte solutions with different SA content. The osmotic pressure was calculated at the ideal condition.
presented in Fig. 6. The water flux of 50-P solution is higher than PSA solution. Chung's study showed that water flux of PSA solution decreased with increasing the molecular weight of polyelectrolyte; this phenomenon may arise from the onset of higher viscosity and lower dissociation of polyelectrolyte with higher molecular weight [17]. 50-P and PSA were both polymerized at 70 °C for 2 h, and poly (sodium acrylate) was expected to be a high molecular weight polymer under this synthesis condition. As shown in Table 2, the intrinsic viscosity of 50-P is lower than that of PSA while they were tested at the same condition. The intrinsic viscosity of 50-P is larger than that of PNIPAM. When the solvent and temperature are determined, the intrinsic viscosity of polymer solution is only affected by the molecular weight of polymer. So the molecular weight of 50-P may be between 170,400 g/mol to 413,242 g/mol. Thus dissociation of 50-P may be larger than that of PSA. Obviously, 50-P and PSA solutions are preferred in the FO process in terms of their ability to generate higher osmotic pressures. But 50-P and PSA are not thermo-sensitive, so it would cost more energy to recover the water by ultrafiltration. The FO performance of 4-P and 35-P is closer to PSA than other thermo-sensitive polyelectrolyte solutions at the same working conditions; therefore, 4-P and 35-P are more appropriate for use as a draw solution in FO process among these polyelectrolytes.
Fig. 6. The average water flux induced by polyelectrolyte solutions in 24 h FO process.
Since 2-P, 4-P, 6-P, 10-P, 20-P and 35-P are thermo-sensitive, when heated to about their LCST, polyelectrolyte chains transfer from hydrophilic to hydrophobic, and gradually separate from water, so it consumes less energy to recover water from these polyelectrolyte solutions, energy consumption will be discussed below. In this work, the water recovery fraction of the thermo-sensitive polyelectrolytes' solution was investigated at different temperatures and the results are shown in Fig. 7. The concentration of test solution was kept at 5 ± 0.5%. It can be observed that the water recovery fraction of 4-P solution was larger than that of other tested solution at all testing temperatures. As mentioned above, the addition of sodium acrylate would improve flexibility of polyelectrolyte chains which benefited the formation of hydrogen bonds among acrylamides, while the improvement of hydrophilicity kept the hydrone staying around carboxylate ions. Thus an appropriate concentration of sodium acrylate enhanced water recovery fraction of polyelectrolyte solution. As shown in Fig. 7, the water recovery fraction of different polyelectrolyte solutions decreases as the SA content increases, and the water recovery fraction increases logarithmically as the temperature increases. The water recovery fraction of 35-P is the lowest. For 2-P or 4-P solution, 38 °C is the desirable temperature for HUF while the polyelectrolyte solution recovered from the stirred cell is about 14.28 wt.%. Though this solution can be reused directly, it is too dilute to collect polyelectrolyte as a solid, which easily causes the loss of polyelectrolyte. By comparison, 45 °C was chosen for ultrafiltration process since the resulting product was very viscous and could be conveniently collected and recycled. After 1 h HUF process at 45 °C and 2 bar, 65% water of 4-P solution was separated from the polyelectrolyte and collected by a beaker as shown in Fig. 2. The remaining solution could be easily recovered using a medicine spoon. 3.3. Repeatability of 4-P solution Compared with other polyelectrolyte solutions, 4-P exhibited the best performance in the FO process and HUF process. So the reuse of 4-P was studied in FO process and the result is shown in Fig. 8.
Fig. 7. Water recovery fraction for 2-P, 4-P, 6-P, 10-P, 20-P, and 35-P solutions at different temperatures.
R. Ou et al. / Desalination 318 (2013) 48–55
Fig. 8. Comparison of the weight increase of 4-P solution after reuse in 24 h FO process. The concentration of 4-P solutions was kept at 14.28 wt.% with 1 g of pure polyelectrolyte (assuming there was no loss of polyelectrolyte during HUF process).
After 24 h, the weight increase of 4-P solution was 4.088 g, 3.774 g and 3.545 g for first, second and third FO processes, respectively. The decline in FO performance was likely caused by the loss of polyelectrolyte during the recovery process. As shown in Fig. 8, the line of weight increase for second FO process (line 2) is similar to that of first FO process (line 1), while the line of weight increase for third FO process (line 3) is different from both of them. It can be observed from line 1 that the flux of 16–24 h is larger than that of 1–16 h, and the same condition is happened at line 2 while its critical point is 8 h. We conjectured reasonably from this phenomenon that polyelectrolyte solution had an optimal concentration for FO process. The loss of polyelectrolyte during FO and HUF processes reduces the original concentration for second and third FO processes, thus the critical point appeared earlier. Because of this, line 3 is different from others. What's more, the weight of 4-P solution increases linearly in 24 h FO process, which means that the dilution of polyelectrolyte solution has a little influence on the osmotic pressure of polyelectrolyte solution in this concentration range. As shown in Fig. 9, 4-P, 50-P and PSA solutions perform well during longer FO operating times (50 h). To further compare the reuse performance of 4-P polyelectrolyte, the average water flux and water recovery fraction of three tests are shown in Fig. 10. The hot ultrafiltration was operated at 45 °C and
53
Fig. 10. Average water flux and water recovery fraction for 4-P solution in three tests. The condition of HUF was 45 °C, and 2 bar.
2 bar. After 1 h HUF process, 65.23%, 64.77% and 64.40% of water were recovered from 4-P solution, which was used in first FO process, second FO process and third FO process, respectively. Reduction of water recovery fraction was mainly due to the reduction of weight increase since water content of polyelectrolyte solution decreased. Although both weight increase and water recovery fraction became lower gradually, the amount of reduction had little impact on FO process and HUF process. The concentration of the polyelectrolyte solution recovered from stirred cup was greater than 14.28 wt.%, and some water was added to achieve a certain concentration before reuse. The average water flux of first, second and third FO processes was 0.347 LMH, 0.320 LMH, and 0.301 LMH, respectively. 3.4. Economic evaluation of HUF process The energy consumption of the hot ultrafiltration process includes energy of heat for heating the solution and the power consumed by the ultrafiltration system. In this study, a normally unutilized heat source, such as low grade geothermal heat or industrial waste heat is assumed as the source for heating polyelectrolyte solution. Thus only the energy required for the ultrafiltration system will be considered. When the thermo-sensitive polyelectrolytes are heated to their LCST, polyelectrolytes separate from water gradually and most of polymer chains hold together, thus polyelectrolyte solutions became two phases which are the organic phase and the milky aqueous solution. The milky aqueous solution will be used as feed solution at a fullscale ultrafiltration process, since the organic phase may easily plug pores of the membrane. This will improve the flux and water recovery fraction efficiently because of lower viscosity. A pilot-scale plan assumed for economic calculation is shown in Fig. 11 [27]. The feed pump power Wfeed can be expressed as: Wfeed ¼ Q p P1 where Qp is the permeate flow rate (m 3·s −1) of feed solution and P1 is the pressure at module inlet (Pa). A 60% efficiency for the feed pump (screw pump) has been assumed. If the air is assumed to be an ideal gas and the compression adiabatic, then the air compressor power Wcomp which is the power necessary to compress the air from an initial state (Patm) to a final state (P1) can be expressed as:
Fig. 9. The weight increase of polyelectrolyte solutions in 50 h FO process.
Wcomp ¼
γ Q gas Patm γ1
P1 Patm
γ1 γ
1
54
R. Ou et al. / Desalination 318 (2013) 48–55
solution which can be applied in many FO process such as concentration of dilute industrial wastewater and source water purification. Though the viscosity of 4-P solution would increase and might affect the performance of 4-P solution, solid hydrogels even can be used as a draw agent and generate a moderate flux [1,11]. Thus 4-P solution has the potential to be applied in desalination. 4. Conclusions
Fig. 11. Pilot-scale ultrafiltration process assumed for economic estimation.
where γ is 1.395 for air, Qgas is gas flow rate measured at atmospheric pressure (m3·s−1). 85% efficiency for the compressor has been assumed. The concentrated solution will be mixed with organic phase to reuse. Then a total power WT is calculated as the sum of the different powers: WT ¼ Wfeed þ Wcomp The energy consumed per m 3 of permeate produced (Wh·m −3) is given by: Ec ¼
In this work, nine polyelectrolytes (PNIPAM, 2-P, 4-P, 6-P, 10-P, 20-P, 35-P, 50-P, PSA) were synthesized and studied as draw solutions in the forward osmosis process. The water permeated into polyelectrolyte solution was recovered by hot ultrafiltration which was operated at 45 °C and 2 bar for 1 h. By contrast, 2-P, 4-P, 6-P, 10-P, 20-P and 35-P solutions were more suitable for used in the FO process than PSA and 50-P, since the formers were more thermo-sensitive, they were expected to transfer from hydrophilic to hydrophobic when heated to about their LCST, and thus gradually separated from water, reducing the energy required to recover the water from former polyelectrolyte solutions. Further investigation showed that the forward osmosis performance and water recovery efficiency of 4-P and 35-P solutions were better than those of other solutions. For 4-P solution, the average water flux of first, second and third FO processes were 0.347 LMH, 0.320 LMH, and 0.301 LMH, respectively, while their water recovery fractions were 65.23%, 64.77%, and 64.40%. The water flux induced by 14.28 wt.% 4-P solution was smaller than that by other draw solutions used in FO process, but the ease of water recovery and reused would make it potentially a good candidate for the FO process. Acknowledgments
WT FP ⋅S
where Fp is permeate flux (m 3·m −2·h −l), and S is filtration area (m 2). Since operated condition has an important effect on capital investment and cost, the economic evaluation here is calculated from assumed parameters in Table 3. At the assumed condition, it costs around 1.18 RMB to recover 1 ton product water. Overall, 14.28 wt.% 4-P solution has a number of desirable properties among the polyelectrolyte solutions besides the calculated osmotic pressure (3.5 atm). The osmotic pressure of 70,000 ppm NaCl solution is about 3.5 atm. Thus, 14.28 wt.% 4-P solution can be used as the draw solution for brackish water desalination and food processing with TDS ranging from thousands to ten thousands [2,28]. In addition, the osmotic pressure of polyelectrolyte solution increases with its increasing concentration theoretically. Increasing the concentration to 60 wt.% and using the molal projection would produce a 31.2 atm solution which is similar to seawater, and 75 wt.% would produce a 62.4 atm
Table 3 Parameters for economic evaluation. Items
Parameters
Ultrafiltration membrane module Membrane area (m2) Inlet pressure P1 (Pa) Atmospheric pressure Patm (Pa) Water recovery fraction (%) Permeate flow rate Qp (m3·h−1) Gas flow rate Qgas (m3·min−1) Permeate flux Fp (m3·m−2·h−1) Price of electricity (RMB/KWH) Totally power WT (W) Energy consumption EC (Wh·m−3) Cost (RMB·ton−1)
Minipore HM-160 50.5 200000 101325 80 0.625 0.11 0.01 0.6 992.39 1965.13 1.18
We thank the financial support from the National Natural Science Foundation of China (nos. 21128004, 21025626). References [1] D. Li, X.Y. Zhang, J.F. Yao, Y. Zeng, G.P. Simon, H.T. Wang, Composite polymer hydrogels as draw agents in forward osmosis and solar dewatering, Soft Matter 7 (2011) 10048–10056. [2] T.Y. Cath, A.E. Childress, M. Elimelech, Forward osmosis: principles, applications, and recent developments, J. Membr. Sci. 281 (2006) 70–87. [3] A. Achilli, T.Y. Cath, A.E. Childress, Selection of inorganic-based draw solutions for forward osmosis applications, J. Membr. Sci. 364 (2010) 233–241. [4] R.W. Holloway, A.E. Childress, K.E. Dennett, T.Y. Cath, Forward osmosis for concentration of anaerobic digester centrate, Water Res. 41 (2007) 4005–4014. [5] N.Y. Yip, M. Elimelech, et al., High performance thin-film composite forward osmosis membrane, Environ. Sci. Technol. 44 (2010) 3812–3818. [6] J.R. McCutcheon, R.L. McGinnis, M. Elimelech, A novel ammonia–carbon dioxide forward (direct) osmosis desalination process, Desalination 174 (2005) 1–11. [7] Z.Y. Liu, H.W. Bai, J. Lee, D.D. Sun, A low-energy forward osmosis process to produce drinking water, Energy Environ. Sci. 4 (2011) 2582–2585. [8] M.L. Stone, C. Rae, F.F. Stewart, A.D. Wlison, Switchable polarity solvents as draw solutes for forward osmosis, Desalination 312 (2013) 124–129. [9] S. Phuntsho, H.K. Shon, et al., A novel low energy fertilizer driven forward osmosis desalination for direct fertigation: evaluating the performance of fertilizer draw solutions, J. Membr. Sci. 375 (2011) 172–181. [10] J.S. Yong, W.A. Philip, M. Elimelech, Coupled reverse draw solute permeation and water flux in forward osmosis with neutral draw solutes, J. Membr. Sci. 392 (2012) 9–17. [11] D. Li, X.Y. Zhang, J.F. Yao, G.P. Simon, H.T. Wang, Stimuli-responsive polymer hydrogels as a new class of draw agent for forward osmosis desalination, Chem. Commun. 47 (2011) 1710–1712. [12] T.S. Chung, S. Zhang, K.Y. Wang, J.C. Su, M.M. Ling, Forward osmosis processes: yesterday, today and tomorrow, Desalination 287 (2012) 78–81. [13] P. Wang, Q.H. Shi, Y.F. Shi, K.K. Clark, G.D. Stucky, A.A. Keller, Magnetic permanently confined micelle arrays for treating hydrophobic organic compound contamination, J. Am. Chem. Soc. 131 (2009) 182–188. [14] M.M. Ling, K.Y. Wang, T.S. Chung, Highly water-soluble magnetic nanoparticles as novel draw solutes in forward osmosis for water reuse, Ind. Eng. Chem. Res. 49 (2010) 5869–5876. [15] M.M. Ling, T.S. Chung, Desalination process using super hydrophilic nanoparticles via forward osmosis integrated with ultrafiltration regeneration, Desalination 278 (2011) 194–202.
R. Ou et al. / Desalination 318 (2013) 48–55 [16] M.M. Ling, T.S. Chung, X.M. Lu, Facile synthesis of thermosensitive magnetic nanoparticles as “smart” draw solutes in forward osmosis, Chem. Commun. 47 (2011) 10788–10790. [17] Q.C. Ge, J.C. Su, G.L. Amy, T.S. Chung, Exploration of polyelectrolytes as draw solutes in forward osmosis processes, Water Res. 46 (2012) 1318–1326. [18] B. Jeong, S.W. Kim, Y.H. Bae, Thermosensitive sol–gel reversible hydrogels, Adv. Drug Deliv. Rev. 64 (2012) 154–162. [19] R.M.K. Ramanan, P. Chellamuthu, L.P. Tang, K.T. Nguyen, Development of a temperature-sensitive composite hydrogel for drug delivery applications, Biotechnol. Prog. 22 (2006) 118–125. [20] R. Pelton, Temperature-sensitive aqueous microgels, Adv. Colloid Interface Sci. 85 (2000) 1–33. [21] J. Zhang, N.A. Peppas, Synthesis and characterization of pH- and temperaturesensitive poly(methacrylic acid)/poly(n-isopropylacrylamide) interpenetrating polymeric networks, Macromolecules 33 (2000) 102–107. [22] O. Chiantore, M. Guaita, L. Trosarelli, Solution properties of poly (n-isopropylacrylamide), Makromol. Chem. 180 (1979) 969–973.
55
[23] A. Takahashi, M. Nagasawa, Excluded volume of polyelectrolyte in salt solutions, J. Am. Chem. Soc. 86 (4) (1964) 543–548. [24] A.D. Wilson, F.F. Stewart, Deriving osmotic pressures of draw solutes used in osmotically driven membrane process, J. Membr. Sci. 431 (2013) 205–211. [25] P.W. Zhu, D.H. Napper, Volume phase transitions of poly (N-isopropylacrylamide) latex particles in mixed water-N, N-dimethylformamide solutions, Chem. Phys. Lett. 256 (1996) 51–56. [26] H.M. Crowther, B. Vincent, Swelling behavior of poly-N-isopropylacrylamide microgel particles in alcoholic solutions, Colloid Polym. Sci. 276 (1998) 46–51. [27] S. Laborie, C. Cabassud, L. Durand-Bourlier, J.M. Laine, Fouling control by air sparging inside hollow fibre membrane effects on energy consumption, Desalination 118 (1998) 189–196. [28] M.S. Mohsen, O.R. Al-Jayyousi, Brackish water desalination: an alternative for water supply enhancement in Jordan, Desalination 124 (1999) 163–174.