Water Research 91 (2016) 361e370
Contents lists available at ScienceDirect
Water Research journal homepage: www.elsevier.com/locate/watres
Combination of forward osmosis (FO) process with coagulation/ flocculation (CF) for potential treatment of textile wastewater Gang Han a, Can-Zeng Liang a, Tai-Shung Chung a, *, Martin Weber b, Claudia Staudt b, Christian Maletzko c a b c
Department of Chemical and Biomolecular Engineering, National University of Singapore, 4 Engineering Drive 4, Singapore 117585, Singapore Advanced Materials & Systems Research, BASF SE, GM-B001, 67056 Ludwigshafen, Germany Performance Materials, BASF SE, G-PM/PU-F206, 67056 Ludwigshafen, Germany
a r t i c l e i n f o
a b s t r a c t
Article history: Received 14 November 2015 Received in revised form 22 December 2015 Accepted 14 January 2016 Available online 16 January 2016
A novel combination of forward osmosis (FO) process with coagulation/flocculation (CF) (FOeCF) has been experimentally conceived for the treatment and reuse of textile wastewater. FO is employed to spontaneously recover water from the wastewater via osmosis and thus effectively reduces its volume with a dramatically enhanced dye concentration. CF is then applied to precipitate and remove dyes from the FO concentrated stream with much improved efficiency and reduced chemical dosage. The FOeCF hybrid system exhibits unique advantages of high water flux and recovery rate, well controlled membrane fouling, high efficiency, and minimal environmental impact. Using a lab-made thin-film composite (TFC) FO membrane, an initial water flux (Jw) of 36.0 L m2 h1 with a dye rejection of 99.9% has been demonstrated by using 2 M NaCl as the draw solution and synthetic textile wastewater containing multiple textile dyes, inorganic salts and organic additives as the feed under the FO mode. The Jw could be maintained at a high value of 12.0 L m2 h1 even when the recovery rate of the wastewater reaches 90%. Remarkable reverse fouling behavior has also been observed where the Jw of the fouled membrane can be almost fully restored to the initial value by physical flushing without using any chemicals. Due to the great dye concentration in the FO concentrated wastewater stream, the CF process could achieve more than 95% dye removal with a small dosage of coagulants and flocculants at 500e1000 ppm. The newly developed FOeCF hybrid process may open up new exploration of alternative technologies for the effective treatment and reuse of textile effluents. © 2016 Elsevier Ltd. All rights reserved.
Keywords: Forward osmosis Coagulation and flocculation Textile wastewater treatment TFC membrane Water reuse
1. Introduction More than 700,000 ton high water-content textile wastewater is projected to be produced by the conventional printing and dyeing industry each year (Robinson et al., 2001). Textile effluents generally comprise multiple textile dyes, electrolytes, mineral oils, suspended solids, etc. They have been identified as one of the most polluting wastewaters (Lau and Ismail, 2009). New technologies that can treat and reuse textile effluents have received great attention because of the global concerns about water scarcity and stringent government regulations to dispose impaired water resources. Currently, the methods to treat textile wastewater generally
* Corresponding author. E-mail address: [email protected] (T.-S. Chung). http://dx.doi.org/10.1016/j.watres.2016.01.031 0043-1354/© 2016 Elsevier Ltd. All rights reserved.
include biological, physicochemical and filtration processes, during which the dye molecules are destructed and/or separated from water (Hao et al., 2000; Lipscomb, 2008). The biological treatment is an effective process to reduce the chemical oxygen demand (COD) of the wastewater but is unable to fully decolor the effluents (Dafale et al., 2008; Kapdan and Kargi, 2002). Physicochemical approaches, such as oxidation (Nidheesh et al., 2013; Pan et al., 2011), adsorption (Rafatullah et al., 2010), and coagulation and flocculation (CF) (Verma et al., 2012), show promising potentials in decolorization of textile wastewater. Among these methods, CF is widely used because of its simple operation, relatively low capital cost and environmental impact (Chen et al., 2010; Verma et al., 2012; Zahrim and Hilal, 2013). However, the efficiency of CF is relative low when the dye concentration is not high enough (Liang et al., 2014), which results in high chemical usage and large amount of sludge. Membrane based pressure-driven filtration processes such as ultrafiltration (UF) (Fersi and Dhahbi, 2008; Marcucci et al.,
362
G. Han et al. / Water Research 91 (2016) 361e370
2002), nanofiltration (NF) (Cheng et al., 2012; Lin et al., 2015a, 2015b; Ong et al., 2014), and reverse osmosis (RO) (Ciardelli et al., 2001; Treffry-Goatley et al., 1983) are emerging technologies for treating textile wastewater. However, the rejection of UF membrane toward low molecular weight dyes is inefficient (Marcucci et al., 2002), and the main hurdle of NF and RO lies in their high energy requirement, low water flux and recovery rate, and severe flux decline due to membrane fouling (Van der Bruggen et al., 2005; Verliefde et al., 2008). Recently, forward osmosis (FO) has shown great promise in the treatment and reuse of impaired water streams (Chung et al., 2012; Han et al., 2015a, 2015b, 2016; Linares et al., 2011). In lieu of driven by the hydraulic pressure, FO utilizes the osmotic pressure gradient between the feed and the more concentrated draw solution to spontaneously induce a water flow across a semipermeable membrane. FO possesses several merits over the conventional pressuredriven processes in terms of significantly lowered energy consumption, reduced cake layer formation and high fouling reversibility, great rejections toward a wide range of contaminants, and high water recovery rate. However, FO itself is not a separation process and the concentrated feed stream is general an unwanted by-product that should be further processed during the subsequent wastewater treatment. In the current work, a novel hybrid system has been demonstrated by integrating forward osmosis (FO) and coagulation/ flocculation (CF) for the treatment of synthetic textile wastewater. FO is used to dewater the wastewater stream and thus reduces its volume with a high water flux and recovery as well as low energy consumption and fouling; while CF is employed to remove the dyes from the FO concentrate with much improved efficiency and reduced dosage. The feasibility of the FOeCF hybrid system was examined by using synthetic dye solutions and textile wastewater containing multiple dyes, salts and additives as feed solutions. The objectives of this study are to investigate: (1) the effects of FO operation modes, dye chemistry and concentration on FO performance; (2) membrane rejections toward salts and dye compounds; (3) membrane fouling behavior and regeneration; (4) the efficiency of CF treatment, and (5) the performance of the FOeCF hybrid system to treat the synthetic textile wastewater. The newly developed FOeCF hybrid process may open up an alternative technology for the treatment and reuse of textile effluents. 2. Experimental and methods 2.1. Materials and chemicals Indigo carmine (INCA), Reactive black 5 (RB-5) and Alcian blue 8GX (AB-8) ordered from SigmaeAldrich were employed to represent the acid, reactive and basic dyes, respectively. Table 1 summarizes their chemical structures and molecular weights. Acetic acid, magnesium sulfate, magnesium chloride, monosodium phosphate, calcium chloride, and sodium chloride purchased from SigmaeAldrich were used as the additives to prepare the textile wastewater. A lab-made thin-film composite (TFC) flat-sheet membrane consisting of a woven-embedded sulfonated polyphenylenesulfone substrate and a polyamide selective skin was employed as the FO membrane (Han et al., 2016). The detailed specification of membrane fabrication and characterization is disclosed in the supporting information (SI) and our previous work (Han et al., 2016). Sodium chloride (NaCl) was acquired to prepare the draw solutions for FO performance tests. A mixture containing various inorganic salts (i.e., Iron (II) chloride-anhydrous, calcium oxide, aluminum sulfate-octadecahydrate, and polyaluminum chloride) and macromolecules (i.e., polydiallyldimethyl ammonium
chloride, cationic polyacrylamide, and cyanoguanidine) was utilized as the coagulants and flocculants during the CF treatment (Liang et al., 2014). Deionized water was produced by a Milli-Q unit (Millipore) with a resistivity of 15 MU cm. 2.2. Forward osmosis (FO) tests for water reclamation from textile solutions A lab-scale FO system was utilized to study the FO dewatering performance where two variable speed peristaltic pumps were used to circulate the feed and draw solutions through the membrane cell counter currently (Han et al., 2012). The membrane permeation cell was a plate-and-frame design with a spacer-free rectangular channel at both sides, and the effective membrane surface area was 10 cm2. The cross-flow velocity was kept at 0.2 L min1 (0.017 m s1) at both sides of the membrane and there was no hydraulic pressure difference across the membrane. The performances were measured at room temperature of 22 ± 0.5 C. The weight change of the feed solution over a certain duration was automatically collected via a digital balance (EK-4100i, A&D Company Ltd., Japan) that was connected to a computer. The water permeation flux, Jw, (in L m2 h1, abbreviated as LMH) was then calculated as:
Jw ¼
DV Am Dt
(1)
where DV (L) is the permeate water over a predetermined time Dt (h) and Am is the effective membrane surface area (m2). The salt concentration in the feed solution was obtained via measuring its conductivity, and the reverse solute flux, Js (in g m2 h1, abbreviated as gMH), was calculated as (Han et al., 2015a, 2016, 2012):
Js ¼
Cf ;t Vf ;t Cf ;i Vf ;i Am Dt
(2)
where Cf,t and Vf,t are the salt concentration and feed volume at the end of tests, respectively, while Cf,i and Vf,i refer to the salt concentration and total feed volume at the beginning, respectively. In terms of Jw and Js, the basic osmosis performance of the TFCFO membrane were first evaluated by using deionized water as the feed and NaCl solutions with various molar concentrations as the draw solutions. Then, the applicability of FO for water reclamation from synthetic dye solutions with various types of dyes (i.e., acid, reactive and basic) and/or concentrations (100e500 ppm) against recovery rate was investigated. The FO tests were operated under two different modes depending on the membrane orientation: (1) pressure retarded osmosis (PRO) mode where the draw solution faces against the polyamide selective layer, and (2) FO mode where the feed faces against the selective layer. For the short-term performance tests, the average data over the first 10 min of each test was reported. For the long-term performance tests, 0.5 L dye solutions were used as the feed and continuously run through the FO membrane until the predetermined recovery rate was reached. The feed water recovery rate, Re (%), was calculated by the following equation:
Re ¼
DV 100% Vf ;i
(3)
where Vf,i (L) is the initial volume of the feed solution. The draw solution concentration was maintained constant throughout the tests by conductivity control to exclude the effects of draw solution dilution effects. The membrane cleaning was
G. Han et al. / Water Research 91 (2016) 361e370
363
Table 1 Chemical structures and molecular weights of the organic dyes used in this study. Name of dye
Type of dye
Indigo carmine (INCA)
Acid
Molecular structure
466
Reactive black 5 (RB-5)
Reactive
992
Alcian blue 8GX (AB-8)
Basic
conducted by physically flushing freshwater through the surface of the fouled FO membrane at a flow rate of 0.4 L min1 (0.034 m s1) for 1 h and the total water volume used in each rinse was around 3 L. 2.3. Coagulation/flocculation (CF) tests for dye removal The jar-test experiment was adopted to perform the coagulation/flocculation (CF) operation during which an appropriate amount of coagulants & flocculants solution was added into the wastewater solutions under stirring (Liang et al., 2014). After settling for 1 h, the treated solution was filtered by using a vacuum filtration apparatus. The filtrates were collected for further analyses. 2.4. Hybrid FOeCF process to treat synthetic textile wastewater By integrating forward osmosis (FO) and coagulation/flocculation (CF) processes, one hybrid FOeCF system was investigated for the treatment of textile wastewater. As illustrated in Fig. 1, FO is employed to dewater the wastewater stream and CF is followed to remove the dyes. Compared to conventional pressure driven membrane processes such as NF and RO, FO possesses a higher water permeation flux yet very low energy consumption, greater recovery rate and dye rejection, less and more reversible fouling (Liu and Mi, 2012; Jin et al., 2012; Zhao et al., 2012). After FO dewatering, the dye concentration in the wastewater stream is significantly enhanced while the volume is dramatically reduced. As a result, the efficiency of CF could be improved and the chemical dosage is reduced so that less cost and environmental impacts are expected. In order to mimic the real textile wastewater, INCA, RB-5 and AB-8 were dissolved in deionized water with an equivalent weight ratio, and then acetic acid, magnesium sulfate, magnesium chloride, monosodium phosphate, calcium chloride and sodium
Molecular weight (g/mol)
1299
chloride were added to represent the chemical additives (Doumic et al., 2015). The main characteristics of the synthetic textile wastewater are tabulated in Table 2. In the current study, a batch mode was chosen to operate the FOeCF system since it provides easy operation, minimal attention, and is close to practical applications. Specifically, 0.5 L synthetic textile wastewater was continuously pumped into the FO membrane module as the feed until reaching a recovery rate of 90% (or 450 ml water was recovered). NaCl solutions with a concentration of 1 M and 2 M were used as the draw solutions, respectively. After FO, the highly concentrated wastewater stream was treated by CF with a dosage of coagulants/flocculants at 500e1000 ppm (or 0.5e1.0 g per liter textile wastewater). After settlement and filtration, the obtained sludge could be dried and burned, while the filtrates that contain trace amounts of dyes and coagulates/flocculates can be discharged or returned to the FOeCF system to further improve the overall recovery rate. 2.5. Physicochemical analyses A UVevis integral method and a total organic carbon (TOC) analyzer were employed to determine the dye concentration in solutions (Liang et al., 2014; Ong et al., 2014). The former was based on the UVevis integral between 350 and 650 nm obtained by scanning the sample solution via a UVevis spectrophotometer (Pharo 300, Merck), while the latter was relied on the concentration measurement determined by a TOC analyzer (TOC, ASI 5000A, Shimazu). As shown in Fig. S2 and our previous work (Liang et al., 2014), the TOC value has an almost linear relationship with dye concentration in the range of 0e1000 ppm for the three dyes. The UVevis integral displays a good linear relationship with dye concentration when it is less than 100 ppm. In order to achieve reasonable accuracy, the highly concentrated dye solutions had to be diluted to lower concentrations (i.e., less than 100 ppm) when the UVevis integral method was used.
364
G. Han et al. / Water Research 91 (2016) 361e370
Fig. 1. The schematic of the coagulation/flocculation (CF) promoted forward osmosis (FO) process to treat textile wastewater.
Table 2 Key characteristic parameters of the synthetic textile wastewater, pH ¼ 3.1 ± 0.3 Characteristics
INCA
RB-5
AB-8
Acetic acid
Naþ
Ca2þ
Mg2þ
PO4 3
Cl1
SO4 2
Unit
Concentration
100 ± 5
100 ± 5
100 ± 5
60 ± 3
1000 ± 10
72 ± 2
12 ± 2
0.3 ± 2
1700 ± 10
3.0 ± 1
mg/L
The dye rejection, Rd, obtained from the TOC measurement can be calculated as:
Cp 1 Cf ;i
Rd ¼
3.1. Characteristics of the TFC-FO membrane
! 100%
(4)
where Cf,i (ppm) is the initial dye concentration in the feed, and Cp (ppm) is the concentration in the permeate or treated solution. Using the UVevis method, Rd could be alternatively calculated as:
Ip 1 If ;i
Rd ¼
! 100%
(5)
where If,i and Ip are the UVevis integral values of the feed and the permeate or treated solutions, respectively. It is worthy to point out that Cp in FO should be obtained from the change of the dye concentration in the draw solution as follows (Han et al., 2015a):
Cp ¼
Cd;t Vd;i þ DV Cd;i Vd;i DV
(6)
where Cd,t (ppm) is the dye concentration at the end of tests and Cd,i (ppm) is the concentration in the initial draw solution with a initial volume of Vd,i (L). The dye rejection in FO, Rf,d, was then calculated as:
Rf ;d ¼ ¼
1
Cp Cf ;i
3. Results and discussion
! 100%
! Cd;t Vd;i þ DV Cd;i Vd;i 100% 1 Cf ;i DV
(7)
In order to ensure the reproducibility of the obtained data, three tests were carried out for each condition and the averaged value was reported.
A lab-made thin film composite (TFC) flat-sheet membrane consisting of a polyamide selective layer and a woven-embedded sulfonated polyphenylenesulfone (sPPSU) substrate was used as the FO membrane (Han et al., 2016). As shown in Fig. 2, the polyamide skin has a typical “ridge-and-valley” surface morphology with a very thin thickness of several hundred nanometers (Veríssimo et al., 2005). The sPPSU membrane substrate shows a fully sponge-like structure across the whole cross-section even though a woven is incorporated. Table 3 presents the transport properties and structural parameter of the TFC-FO membrane (Han et al., 2016). It possesses a high pure water permeability, A, of 3.4 L m2 h1 bar1 accompanying with a low salt permeability coefficient, B, of 0.103 L m2 h1. The dye rejections of the TFC-FO membrane under the RO operation at 1 bar are always greater than 99.9% (Table 4), signifying the outstanding selectivity of the FO membrane toward a relatively wide range of dyes and salts. In addition, a small structural parameter, S, of 300 mm is achieved. These are mainly due to the hydrophilic nature of the sPPSU material and the small membrane thickness (~67 mm). In FO, this structure can not only provide robust mechanical strength but also effectively minimize the transport resistance and internal concentration polarization (ICP). As a result, the TFC-FO membrane shows outstanding FO performance. Fig. 3 shows the water flux, Jw, and specific reverse salt flux, Js/Jw, of the membrane as a function of draw solution concentration using deionized water as the feed under PRO and FO modes, respectively. High Jw values of 32.7e69.3 LMH and 23.3e38.7 LMH were obtained under PRO and FO modes, respectively, using 0.6e2.0 M NaCl as draw solutions. Simultaneously, a small Js/Jw of less than 0.2 g/L was achieved even when the draw solution concentration is increased to 2 M, verifying the low salt permeation of the FO membrane to NaCl. In terms of membrane
G. Han et al. / Water Research 91 (2016) 361e370
365
Fig. 2. FESEM images of the top surface, bottom surface, cross-section of the polyamide layer, and the overall cross-section of the flat-sheet TFC-FO membrane (Han et al., 2016).
Table 3 Transport properties, structural parameter and dye rejection of the flat-sheet TFC-FO membrane (Han et al., 2016). Membrane
Water permeability, A (L m2 h1 bar1)
Salt permeability, B (L m2 h1)
Km ( 105 s m1)
Sa ( 106 m)
Dye rejection,b Rd (%)
TFC-FO
3.40 ± 0.26
0.103 ± 0.008
1.97
300
>99.9
a b
S is calculated based on the membrane performance under FO mode using 1 M NaCl as the draw solution and de-ionized water as feed. Rd is tested at 1 bar under RO operation using 100e500 ppm dye solutions containing three different dyes as the feeds, respectively.
Table 4 Initial water permeation flux, Jw, and dye rejections, Rf,d, of the TFC-FO membrane as a function of feed solution, draw solution and operation mode. Feed solution code
Dye concentration (ppm)
Draw solution (M)
Initial water flux, Jw (LMH)a
Dye rejection, Rf,d (%)b
PRO mode
FO mode
PRO mode
FO mode
Deionized water INCA INCA INCA RB-5 AB-8 Synthetic textile wastewater Synthetic textile wastewater
e 100 300 500 300 300 300 300
1.0 1.0 1.0 1.0 1.0 1.0 1.0 2.0
47.7 e 27.1 e e e e e
30.5 30.0 29.5 26.7 29.0 30.0 24.0 36.0
e e >99.9 e e e e e
e >99.9 >99.9 >99.9 >99.9 >99.9 >99.9 >99.9
a b
The initial Jw is the averaged value during the first 10 min of each test. The Rf,d of the FO process is calculated after both short and long-term tests.
morphology, mechanical strength, water flux, and salt and dye rejection, the newly fabricated TFC membrane is an effective membrane to treat textile wastewater under FO operations. 3.2. FO for water reclamation from dye solutions The feasibility of applying FO to reclaim water from textile effluents was firstly investigated using synthetic dye solutions as feeds. Acid (INCA), reactive (RB-5), and basic (AB-8) dyes were chosen in this study because of their popularity in textile and dye
industries. As summarized in Table 1, each INCA and RB-5 molecule possesses two and four sulfonic acid groups, respectively, suggesting that they remain in the anionic state in the solutions due to the low pKa value of the sulfonic groups (Pirillo et al., 2009). On the contrary, AB-8 is a cationic dye and it contains four isothiouronium groups and one copper nucleus per molecule. In other words, INCA and RB-5 have negative charges while AB-8 possesses positive charges in the solutions. In addition, AB-8 has the largest molecular weight (Mw) of 1299 g/mol, RB-5 possesses a medium Mw of 992 g/ mol, and INCA has the smallest Mw of 466 g/mol.
G. Han et al. / Water Research 91 (2016) 361e370
Water flux, Jw (LMH)
100
PRO mode FO mode
(a)
80
60
40
20
0 0.0
0.5
1.0
1.5
2.0
NaCl concentration (M)
Specific reverse salt flux, Js/Jw (g/L)
366
1.0
(b)
0.8
0.6
0.4
0.2
0.0 0.0
0.5
1.0
1.5
2.0
NaCl concentration (M)
Fig. 3. Water flux, Jw, and specific reverse salt flux, Js/Jw of the TFC-FO membrane as a function of draw solution concentration and operation mode. The feed is deionized water.
3.2.1. Short-term FO performance for water reclamation from dye solutions In terms of water permeation flux, Jw, the short-term FO performance of the TFC membrane was first investigated under both PRO and FO modes using NaCl solutions with various concentrations as draw solutions and dye solutions and synthetic textile wastewater as feeds. As tabulated in Table 4, the TFC-FO membrane shows high initial Jw values of 47.7 LMH and 30.5 LMH under PRO and FO modes, respectively, when using deionized water as the feed and 1 M NaCl as the draw solution. Under the PRO mode, however, the initial Jw drops from 47.7 LMH to 27.1 LMH even during a 10-min test when the feed is changed to a 300 ppm INCA solution. The performance under the FO mode is quite different from the PRO mode. The initial Jw slightly decreases from 30.5 LMH to 29.5 LMH when using an INCA solution of 300 ppm as the feed. An increase in INCA concentration does not show much effect on the initial Jw under the FO mode (i.e., 26.7e30.0 LMH for 100e500 ppm INCA vs. 30.5 LMH for deionized water). This might be because membrane fouling has not been developed on the polyamide surface during such short duration tests. Similar performances under the FO mode are observed for 300 ppm RB-5 and AB-8 dye solutions with initial Jw values of 29.0 LMH and 30.0 LMH, respectively. In the treatment of synthetic textile wastewater, the initial Jw drops to 24.0 LMH because of its higher salt concentration; however, Jw could be increased to 36 LMH when the draw solution concentration is raised to 2 M. In addition, the dye content in draw solutions is not detectable by both TOC and UVevis measurements, which indicates that the highly selective TFC-FO membrane has a high dye rejection, Rf,d, of larger than 99.9%. Clearly, the FO process has exhibited promising for water reclamation from textile effluents with high water fluxes, excellent salt and dye rejections. 3.2.2. Long-term FO performance for water reclamation from dye solutions 3.2.2.1. The effects of FO operation modes. Fig. 4 portrays the reduction of Jw of the TFC-FO membrane as a function of feed recovery rate running under (a) PRO and (b) FO modes, respectively. Since the draw solution concentration was maintained at 1 M throughout the tests, the baseline curves under both modes slightly decrease with an increase in recovery of deionized water due to the reverse salt flux and ICP effects. However, in the PRO mode, the initial Jw immediately drops to 27.1 LMH when the feed is changed to an INCA solution of 300 ppm (Table 4 and Fig. 4 (a)). Then, Jw rapidly decreases with an increase in recovery rate and drops to 17.6 LMH when the recovery reaches 90%. This suggests that membrane fouling under the PRO mode induced by the dye molecules is very fast and significant. Such a dramatic flux decline
could be attributed to two reasons: one is that the dyes in the feed would accumulate within the porous structure and on the membrane surface of the substrate and thus enhances the transport resistance and increases the structural parameter; the other is due to the back diffusion of the salts from the draw solution which enlarges the ICP effects (McCutcheon and Elimelech, 2007). Conversely, a much milder decrease in Jw is observed in the FO mode. As displayed in Fig. 4 (b) and Table 4, almost the same initial Jw values (i.e., 29.5 vs. 30.5 LMH) are achieved when using 300 ppm INCA instead of deionized water as the feed, and then Jw slightly decreases to 23.7 LMH when the recovery rate goes to 90%. In other words, the reduction of Jw under the FO mode is only 20% even at such a high feed recovery. Interestingly, the decline of Jw is characterized by two stages; a relatively fast and significant drop occurs at a low recovery of less than 5%, and followed by a mild and stable decay till to 90% recovery. The first flux reduction is most likely resulted from the deposition of dye molecules on the polyamide surface. Once the fouling is gradually developed and stabilized, the flux decline becomes milder and slower (Zhang et al., 2014). A comparison of the membrane performance in PRO and FO modes indicates that FO mode is more effective for textile solution separation in terms of better transport stability and less membrane fouling despite of its lower initial Jw. 3.2.2.2. The effects of dye concentration and chemistry. The effects of dye chemistry and concentration on FO performance under the FO mode are investigated in the next step. Fig. 5 (a) shows the Jw reduction as a function of feed recovery rate using different dye solutions with the same dye concentration of 300 ppm as feeds and 1 M NaCl as the draw solution. The decrease of Jw follows a trend of AB-8 [ RB-5 > INCA. INCA shows the lowest reduction in Jw which is only about 18.5% of the initial value at a high recovery rate of 90%. RB-5 displays a slightly larger flux decrease where its Jw drops from 29.0 LMH to 19.7 LMH at the same recovery. Since both INCA and RB-5 possess negative charges, the greater Jw decline of the later is probably due to its larger molecular weight (Table 1). However, AB8 exhibits a much faster and more severe Jw decline along with an increase in water recovery. Its Jw drops from 30 LMH to 10.6 LMH when the recovery reaches 90%, suggesting a 65% reduction. In addition to its large molecular weight, this large drop might be caused by its positive charges that induce strong interaction between the dye molecules and the negative charged polyamide layer. By using INCA solutions with a concentration ranging from 100 to 500 ppm as feed solutions, the effects of dye concentration on membrane Jw under the FO mode are studied. As depicted in Fig. 5 (b), a small drop in Jw is observed when the feed is changed from deionized water to 100 ppm INCA. The Jw reduction undergoes a
G. Han et al. / Water Research 91 (2016) 361e370
367
Fig. 4. The reduction of water flux as a function of feed water recovery rate under (a) PRO and (b) FO modes. Deionized water and 300 ppm INCA dye solution were the feeds, while 1 M NaCl solution was the draw solution and its concentration was maintained constantly throughout the tests.
Fig. 5. The reduction of water flux as a function of feed water recovery rate under the FO mode: (a) using various dye solutions as feeds with the same dye concentration of 300 ppm, and (b) using various INCA solutions as feeds with different dye concentrations. 1 M NaCl solution was the draw solution and its concentration was maintained constantly throughout the tests.
3.2.2.3. FO membrane fouling and cleaning. After the performance tests shown in Fig. 5 (a), a thin gel-like dye layer is found on the top surface of the polyamide skin which could explain the decline in water flux. However, the fouling layer can be easily washed away via physically flushing freshwater without using any chemicals. As displayed in Fig. 6, the initial Jw of the fouled TFC-FO membrane by three different dyes could be almost fully recovered with recovery rates of approximately 97% or above. Although AB-8 induces the most severe fouling where Jw drops from 30.0 LMH to 10.6 LMH when the feed recovery rate reaches 90% (Fig. 5), its Jw can be restored to 29.0 LMH (i.e., a 97% recovery) after rinse. These results demonstrate that the membrane fouling of the TFC-FO membrane under the FO mode is not only mild but also loose and reversible (Mi and Elimelech, 2010). The low and reversible fouling feature
Feed: Deionized water
100
Normalized initial Jw, %
slight increase when the INCA concentration is increased to 300 ppm. This phenomenon is possibly caused by two factors; namely, the enhanced attachment of dyes on the polyamide surface and the higher osmotic pressure of the feed solution at a larger dye concentration. An increase in INCA concentration to 500 ppm does not further induce a significant drop in Jw, implying that dye fouling under the FO mode is mild and the polyamide surface is probably saturated at a low concentration. According to the results illustrated in Fig. 5, it is concluded that membrane fouling under the FO mode is closely related to dye chemistry and molecule weight but is less sensitive to dye concentration.
80 60 40 20 0
Fesh
INCA
RB-5
AB-8
Fig. 6. The normalized initial water flux, Jw, of the regenerated TFC-FO membrane over the initial Jw of the fresh membrane. The membrane is fouled by INCA, RB-5 and AB-8, respectively. Membranes are cleaned by physically flushing freshwater through the fouled polyamide surface for 1 h. 1 M NaCl solution was the draw solution and its salt concentration was maintained constantly throughout the tests.
makes FO superior to other membrane separation processes such as NF for textile wastewater treatment (Liang et al., 2014; Ong et al., 2014; Quintanilla et al., 2010).
368
G. Han et al. / Water Research 91 (2016) 361e370
3.3. Hybrid FOeCF process for the treatment of synthetic textile wastewater FO has shown promising performance in dewatering the dye solutions; however, it is not able to achieve a complete dye removal. Therefore, CF is applied after FO to further treat the highly concentrated stream. To mimic the real applications, one synthetic textile wastewater containing 300 ppm multi-dyes, inorganic salts and organic additives was used as the feed in FO (Table 2). Fig. 7 shows the change of membrane flux as a function of synthetic textile wastewater recovery rate under the FO mode using (a) 1 M and (b) 2 M NaCl as draw solutions. Since the draw solution concentration is maintained constant, the base-line curves using deionized water as the feed shows a small variation in Jw within the tests. When changing the feed to the synthetic textile wastewater, the initial Jw decreases to 24 LMH when using 1 M NaCl as the draw solution. The initial Jw increases to 36.0 LMH when the draw solution concentration is increased to 2 M due to the increased osmotic pressure gradient. However, because of the presence of dyes and the inorganic salt in the feed, Jw for both draw solutions undergoes significant declines against recovery rate with similar trends. In general, a rapid flux decline at a relatively low recovery of less than 40% is firstly observed, followed by a mild decay until reaching a high recovery rate. As illustrated in Fig. 7, Jw drops to 9.8 LMH and 12.0 LMH for 1 M and 2 M NaCl draw solutions, respectively, when the recovery rate reaches 90%. The first rapid drop is mainly attributed to fouling while the next mild decline is due to the effects of the concentrated textile solution and the back diffusion of salts as the fouling has been gradually stabilized. It is also found that the Jw decline is faster and more dramatic for the 2 M draw solution than the 1 M one even though the former has a higher initial Jw. As observed in other FO processes (Li et al., 2014; Liu and Mi, 2012; Jin et al., 2012; Zhang et al., 2014), a larger initial Jw would accelerate the accumulation of species on the polyamide surface, resulting in faster flux decrease. Nevertheless, the Jw values of the TFC-FO membrane always maintain 41% and 33% of the initial values, respectively, even at a high recovery of 90%. The long-term stability of the FO performance for dewatering the synthetic textile wastewater is also investigated. Fig. 8 shows the Jw profiles of the fresh and the regenerated TFC-FO membranes as a function of feed recovery rate and testing duration. Three FO tests are performed using the same membrane but the fouled membrane is cleaned by physically flushing freshwater through the fouled polyamide surface for 1 h after each test. As displayed in Fig. 8, quite stable and reproducible performance is achieved. The Jw
reduction during each test at a recovery of 90% or within a testing duration of 168 h is almost the same which is around 67% of the initial water flux (i.e., from 35e36 LMH to 11.6e12 LMH). In addition, water rinses can effectively rejuvenate the fouled membrane and the initial Jw is able to be restored by 98% even after 3 cycles. This again demonstrates the potential of using FO for textile wastewater reclamation. After FO, the highly concentrated textile wastewater stream is further treated by CF for dye removal. Fig. 9 clearly shows that CF becomes very effective in treating this highly concentrated textile effluent. When the coagulants & flocculants with a dosage of 500e1000 ppm are added into the solution under stirring, the formation of particles/flocs immediately happens. After 1 h settling, the suspensions are spontaneously phase-separated with water and stay at the bottom of the container. Since ferrous ions are used as one of the coagulants, the water solution shows a light green color. After filtration, a gel-like sticky film is left on the filter paper which could be further treated easily. This indicates that the dye removal efficiency of CF is significantly enhanced by FO preconcentration, while the amount of chemical usage and the produced sludge volume are much reduced. The UVevis spectra of the synthetic textile wastewater, FO concentrated wastewater, and filtrate of the CF treated wastewater further confirms the above conclusions. As shown in Fig. 10, the UV absorbance increases dramatically after FO treatment because the textile wastewater is highly concentrated at a high recovery rate of 90%. In other words, the dye concentration is around 10 times of the original value. After CF treatment, the absorbance drops significantly and the measured dye concentration in the filtrate is less than 150 ppm. This suggests that the removal efficiency of the CF process is larger than 95%. 4. Conclusions This study has experimentally conceived the applicability of the combination of forward osmosis (FO) process with coagulation/ flocculation (CF) for the treatment and reuse of textile wastewater. In the hybrid FOeCF system, FO is employed to dewatering the wastewater with high water flux and recovery yet low membrane fouling; and CF is applied for dye precipitation and removal from the FO concentrated stream with high efficiency, low chemical dosages, and minimal environmental impact. By employing a labfabricated thin-film composite (TFC) FO membrane and synthetic dye solutions, the effects of membrane operation mode (FO and PRO modes), dye chemistry (acid, basic and reactive dyes) and concentration (100e500 ppm) on FO performance were firstly studied, which revealed that 1) dye fouling and flux reduction in
Fig. 7. The reduction of water flux as a function of feed water recovery rate under the FO mode by using deionized water and synthetic textile wastewater as feeds and (a) 1 M NaCl and (b) 2 M NaCl as draw solutions. The salt concentration of the draw solution was maintained constantly throughout the tests.
G. Han et al. / Water Research 91 (2016) 361e370
369
Fig. 8. The reduction of water flux as a function of feed water recovery rate (a) and testing duration (b) under the FO mode by using the synthetic textile wastewater as the feed and 2 M NaCl as the draw solution. Membranes were cleaned by physically flushing freshwater through the fouled polyamide surface for 1 h and then reused. The salt concentration of the draw solution was maintained constantly throughout the tests.
Fig. 9. Photos of the coagulation/flocculation (CF) process and the obtained sludge after filtration.
the PRO mode is much more severe than the FO mode; 2) dye fouling on the polyamide selective layer under the FO mode is mild and slightly increases with dye concentration; 3) dye fouling and flux decline is fast but no further significant fouling occurs in the extended recovery rate; 4) dye chemistry significantly influences the membrane fouling tendency and cationic dyes normally show more dramatic fouling because of the strong electrostatic interaction with the polyamide layer; 5) membrane fouling under the FO mode is loose and highly reversible; the water flux could be almost fully recovered by physically flushing; 6) the extent of dye fouling is closely related to the initial water flux; and 7) FO shows excellent rejections toward dyes and salts. The batch-mode FOeCF integrated system was then studied for treating synthetic textile wastewater containing multi-dyes, inorganic salts and organic additives. With the increase of feed recovery, flux reduction was observed during the FO operation; however, the flux was larger than 9.8 and 12.0
Fig. 10. The UVevis absorbance spectra of the original synthetic textile wastewater, the concentrated wastewater stream by FO dewatering at a recovery of 90%, and the filtrate after the CF treatment and filtration. The UVevis spectrum of the FO concentrated solution was obtained via diluting the original solution by 30 times and the diluted sample was measured, then the resultant absorbance was multiplied by a factor of 30.
LMH even at a high recovery rate of 90% using 1 M and 2 M NaCl as the draw solutions, respectively. Since the dye concentration in the FO concentrated wastewater is increased by about 10 times, CF shows outstanding efficiency and is able to remove more than 95% of the dyes with a dosage of 500e1000 ppm. It is believed that the FOeCF hybrid system has great potential for the treatment and reuse of the textile wastewater.
370
G. Han et al. / Water Research 91 (2016) 361e370
Acknowledgments This research was funded by the Singapore National Research Foundation under its Competitive Research Program for the project entitled, “Advanced FO Membranes and Membrane Systems for Wastewater Treatment, Water Reuse and Seawater Desalination” (grant number: R-279-000-336-281). The authors also would like to thank BASF SE, Germany for funding this work with a grant number of R-279-000-411-597. Special thanks are due to Mr. Heyang Wu. Appendix A. Supporting information Supporting information related to this article can be found at http://dx.doi.org/10.1016/j.watres.2016.01.031. References Chen, T., Gao, B., Yue, Q., 2010. Effect of dosing method and pH on color removal performance and floc aggregation of polyferric chlorideepolyamine dualcoagulant in synthetic dyeing wastewater treatment. Colloids Surfaces A Physicochem. Eng. Asp. 355, 121e129. Cheng, S., Oatley, D.L., Williams, P.M., Wright, C.J., 2012. Characterisation and application of a novel positively charged nanofiltration membrane for the treatment of textile industry wastewaters. Water Res. 46, 33e42. Chung, T.S., Li, X., Ong, R.C., Ge, Q., Wang, H., Han, G., 2012. Emerging forward osmosis (FO) technologies and challenges ahead for clean water and clean energy applications. Curr. Opin. Chem. Eng. 1, 246e257. Ciardelli, G., Corsi, L., Marcucci, M., 2001. Membrane separation for wastewater reuse in the textile industry. Resour. Conserv. Recycl. 31, 189e197. Dafale, N., Rao, N.N., Meshram, S.U., Wate, S.R., 2008. Decolorization of azo dyes and simulated dye bath wastewater using acclimatized microbial consortiumebiostimulation and halo tolerance. Bioresour. Technol. 99, 2552e2558. Doumic, L.I., Soares, P.A., Ayude, M.A., Cassanello, M., Boaventura, R.A.R., Vilar, V.J.P., 2015. Enhancement of a solar photo-Fenton reaction by using ferrioxalate complexes for the treatment of a synthetic cotton-textile dyeing wastewater. Chem. Eng. J. 277, 86e96. Fersi, C., Dhahbi, M., 2008. Treatment of textile plant effluent by ultrafiltration and/ or nanofiltration for water reuse. Desalination 222, 263e271. Han, G., Chung, T.S., Toriida, M., Tamai, S., 2012. Thin-film composite forward osmosis membranes with novel hydrophilic supports for desalination. J. Membr. Sci. 423e424, 543e555. Han, G., de Wit, J.S., Chung, T.S., 2015a. Water reclamation from emulsified oily wastewater via effective forward osmosis hollow fiber membranes under the PRO mode. Water Res. 81, 54e63. Han, G., Zuo, J., Wan, C., Chung, T.S., 2015b. Hybrid pressure retarded osmosismembrane distillation (PROMD) process for osmotic power and clean water generation. Environ. Sci. Water Res. Technol. 1, 507e515. Han, G., Zhao, B., Fu, F., Chung, T.S., Weber, M., Staudt, C., Maletzko, C., 2016. High performance thin-film composite membranes with mesh-reinforced hydrophilic sulfonated polyphenylenesulfone (sPPSU) substrates for osmotically driven processes. J. Membr. Sci. 502, 84e93. Hao, O.J., Kim, H., Chiang, P.C., 2000. Decolorization of wastewater. Crit. Rev. Environ. Sci. Technol. 30, 449e505. Jin, X., She, Q., Ang, X., Tang, C.Y., 2012. Removal of boron and arsenic by forward osmosis membrane: influence of membrane orientation and organic fouling. J. Membr. Sci. 389, 182e187. Kapdan, I.K., Kargi, F., 2002. Simultaneous biodegradation and adsorption of textile dyestuff in an activated sludge unit. Process Biochem. 37, 973e981. Lau, W.J., Ismail, A.F., 2009. Polymeric nanofiltration membranes for textile dye wastewater treatment: preparation, performance evaluation, transport modelling, and fouling controlea review. Desalination 245, 321e348. Li, P., Lim, S.S., Neo, J.G., Ong, R.C., Weber, M., Staudt, C., Widjojo, N., Maletzko, C., Chung, T.S., 2014. Short- and long-term performance of the thin-film composite forward osmosis (TFC-FO) hollow fiber membranes for oily wastewater purification. Ind. Eng. Chem. Res. 53, 14056e14064.
Liang, C.Z., Sun, S.P., Li, F.Y., Ong, Y.K., Chung, T.S., 2014. Treatment of highly concentrated wastewater containing multiple synthetic dyes by a combined process of coagulation/flocculation and nanofiltration. J. Membr. Sci. 469, 306e315. Lin, J., Ye, W., Zeng, H., Yang, H., Shen, J., Darvishmanesh, S., Luis, P., Sotto, A., Van der Bruggen, B., 2015a. Fractionation of direct dyes and salts in aqueous solution using loose nanofiltration membranes. J. Membr. Sci. 477, 183e193. Lin, J., Tang, C.Y., Ye, W., Sun, S.P., Hamdan, S.H., Volodin, A., Van Haesendonck, C., Sotto, A., Luis, P., Van der Bruggen, B., 2015b. Unraveling flux behavior of superhydrophilic loose nanofiltration membranes during textile wastewater treatment. J. Membr. Sci. 493, 690e702. Linares, R.V., Quintanilla, V.Y., Li, Z.Y., Amy, G., 2011. Rejection of micropollutants by clean and fouled forward osmosis membrane. Water Res. 45, 6737e6744. Lipscomb, G.G., 2008. In: Flank, W.H., Abraham, M.A., Matthews, M.A. (Eds.), Membrane Separation Technology: Past, Present, and Future, Innovations in Industrial and Engineering Chemistry, ACS Symposium Series 1000. American Chemical Society, pp. 281e333. Liu, Y., Mi, B., 2012. Combined fouling of forward osmosis membranes: synergistic foulant interaction and direct observation of fouling layer formation. J. Membr. Sci. 407e408, 136e144. Marcucci, M., Ciardelli, G., Matteucci, A., Ranieri, L., Russo, M., 2002. Experimental campaigns on textile wastewater for reuse by means of different membrane processes. Desalination 149, 137e143. McCutcheon, J.R., Elimelech, M., 2007. Modeling water flux in forward osmosis: implications for improved membrane design. AIChE J. 53, 1736e1744. Mi, B.X., Elimelech, M., 2010. Organic fouling of forward osmosis membranes: fouling reversibility and cleaning without chemical reagents. J. Membr. Sci. 348, 337e345. Nidheesh, P.V., Gandhimathi, R., Ramesh, S.T., 2013. Degradation of dyes from aqueous solution by Fenton processes: a review. Environ. Sci. Pollut. Res. 20, 2099e2132. Ong, Y.K., Li, F.Y., Sun, S.P., Zhao, B.W., Liang, C.Z., Chung, T.S., 2014. Nanofiltration hollow fiber membranes for textile wastewater treatment: lab-scale and pilotscale studies. Chem. Eng. Sci. 114, 51e57. Pan, J.H., Han, G., Zhou, R., Zhao, X.S., 2011. Hierarchical N-doped TiO2 hollow microspheres consisting of nanothorns with exposed anatase {101} facets. Chem. Commun. 47, 6942e6944. Pirillo, S., Pedroni, V., Rueda, E., Ferreira, M.L., 2009. Elimination of dyes from aqueous solutions using iron oxides and chitosan as adsorbents, a comparative study. Quim. Nova 32, 1239e1244. Quintanilla, V.Y., Maeng, S.K., Fujioka, T., Kennedy, M., Amy, G., 2010. Proposing nanofiltration as acceptable barrier for organic contaminants in water reuse. J. Membr. Sci. 362, 334e345. Rafatullah, M., Sulaiman, O., Hashim, R., Ahmad, A., 2010. Adsorption of methylene blue on low-cost adsorbents: a review. J. Hazard. Mater. 177, 70e80. Robinson, T., McMullan, G., Marchant, R., Nigam, P., 2001. Remediation of dyes in textile effluent: a critical review on current treatment technologies with a proposed alternative. Bioresour. Technol. 77, 247e255. Treffry-Goatley, K., Buckley, C.A., Groves, G.R., 1983. Reverse osmosis treatment and reuse of textile dyehouse effluents. Desalination 47, 313e320. Van der Bruggen, B., Cornelis, G., Vandecasteele, C., Devreese, I., 2005. Fouling of nanofiltration and ultrafiltration membranes applied for wastewater regeneration in the textile industry. Desalination 175, 111e119. Verliefde, A.R.D., Cornelissen, E.R., Heijman, S.G.J., Verberk, J.Q.J.C., Amy, G.L., Van der Bruggen, B., van Dijk, J.C., 2008. The role of electrostatic interactions on the rejection of organic solutes in aqueous solutions with nanofiltration. J. Membr. Sci. 322, 52e66. Verma, A.K., Dash, R.R., Bhunia, P., 2012. A review on chemical coagulation/flocculation technologies for removal of colour from textile wastewaters. J. Environ. Manag. 93, 154e168. Veríssimo, S., Peinemann, K.V., Bordado, J., 2005. Thin-film composite hollow fiber membranes: an optimized manufacturing method. J. Membr. Sci. 264, 48e55. Zahrim, A.Y., Hilal, N., 2013. Treatment of highly concentrated dye solution by coagulation/flocculationesand filtration and nanofiltration. Water Resour. Ind. 3, 23e34. Zhao, S., Zou, L., Tang, C.Y., Mulcahy, D., 2012. Recent developments in forward osmosis: opportunities and challenges. J. Membr. Sci. 396, 1e21. Zhang, S., Wang, P., Fu, X., Chung, T.S., 2014. Sustainable water recovery from oily wastewater via forward osmosis-membrane distillation (FO-MD). Water. Res. 52, 112e121.
Contents lists available at ScienceDirect
Water Research journal homepage: www.elsevier.com/locate/watres
Combination of forward osmosis (FO) process with coagulation/ flocculation (CF) for potential treatment of textile wastewater Gang Han a, Can-Zeng Liang a, Tai-Shung Chung a, *, Martin Weber b, Claudia Staudt b, Christian Maletzko c a b c
Department of Chemical and Biomolecular Engineering, National University of Singapore, 4 Engineering Drive 4, Singapore 117585, Singapore Advanced Materials & Systems Research, BASF SE, GM-B001, 67056 Ludwigshafen, Germany Performance Materials, BASF SE, G-PM/PU-F206, 67056 Ludwigshafen, Germany
a r t i c l e i n f o
a b s t r a c t
Article history: Received 14 November 2015 Received in revised form 22 December 2015 Accepted 14 January 2016 Available online 16 January 2016
A novel combination of forward osmosis (FO) process with coagulation/flocculation (CF) (FOeCF) has been experimentally conceived for the treatment and reuse of textile wastewater. FO is employed to spontaneously recover water from the wastewater via osmosis and thus effectively reduces its volume with a dramatically enhanced dye concentration. CF is then applied to precipitate and remove dyes from the FO concentrated stream with much improved efficiency and reduced chemical dosage. The FOeCF hybrid system exhibits unique advantages of high water flux and recovery rate, well controlled membrane fouling, high efficiency, and minimal environmental impact. Using a lab-made thin-film composite (TFC) FO membrane, an initial water flux (Jw) of 36.0 L m2 h1 with a dye rejection of 99.9% has been demonstrated by using 2 M NaCl as the draw solution and synthetic textile wastewater containing multiple textile dyes, inorganic salts and organic additives as the feed under the FO mode. The Jw could be maintained at a high value of 12.0 L m2 h1 even when the recovery rate of the wastewater reaches 90%. Remarkable reverse fouling behavior has also been observed where the Jw of the fouled membrane can be almost fully restored to the initial value by physical flushing without using any chemicals. Due to the great dye concentration in the FO concentrated wastewater stream, the CF process could achieve more than 95% dye removal with a small dosage of coagulants and flocculants at 500e1000 ppm. The newly developed FOeCF hybrid process may open up new exploration of alternative technologies for the effective treatment and reuse of textile effluents. © 2016 Elsevier Ltd. All rights reserved.
Keywords: Forward osmosis Coagulation and flocculation Textile wastewater treatment TFC membrane Water reuse
1. Introduction More than 700,000 ton high water-content textile wastewater is projected to be produced by the conventional printing and dyeing industry each year (Robinson et al., 2001). Textile effluents generally comprise multiple textile dyes, electrolytes, mineral oils, suspended solids, etc. They have been identified as one of the most polluting wastewaters (Lau and Ismail, 2009). New technologies that can treat and reuse textile effluents have received great attention because of the global concerns about water scarcity and stringent government regulations to dispose impaired water resources. Currently, the methods to treat textile wastewater generally
* Corresponding author. E-mail address: [email protected] (T.-S. Chung). http://dx.doi.org/10.1016/j.watres.2016.01.031 0043-1354/© 2016 Elsevier Ltd. All rights reserved.
include biological, physicochemical and filtration processes, during which the dye molecules are destructed and/or separated from water (Hao et al., 2000; Lipscomb, 2008). The biological treatment is an effective process to reduce the chemical oxygen demand (COD) of the wastewater but is unable to fully decolor the effluents (Dafale et al., 2008; Kapdan and Kargi, 2002). Physicochemical approaches, such as oxidation (Nidheesh et al., 2013; Pan et al., 2011), adsorption (Rafatullah et al., 2010), and coagulation and flocculation (CF) (Verma et al., 2012), show promising potentials in decolorization of textile wastewater. Among these methods, CF is widely used because of its simple operation, relatively low capital cost and environmental impact (Chen et al., 2010; Verma et al., 2012; Zahrim and Hilal, 2013). However, the efficiency of CF is relative low when the dye concentration is not high enough (Liang et al., 2014), which results in high chemical usage and large amount of sludge. Membrane based pressure-driven filtration processes such as ultrafiltration (UF) (Fersi and Dhahbi, 2008; Marcucci et al.,
362
G. Han et al. / Water Research 91 (2016) 361e370
2002), nanofiltration (NF) (Cheng et al., 2012; Lin et al., 2015a, 2015b; Ong et al., 2014), and reverse osmosis (RO) (Ciardelli et al., 2001; Treffry-Goatley et al., 1983) are emerging technologies for treating textile wastewater. However, the rejection of UF membrane toward low molecular weight dyes is inefficient (Marcucci et al., 2002), and the main hurdle of NF and RO lies in their high energy requirement, low water flux and recovery rate, and severe flux decline due to membrane fouling (Van der Bruggen et al., 2005; Verliefde et al., 2008). Recently, forward osmosis (FO) has shown great promise in the treatment and reuse of impaired water streams (Chung et al., 2012; Han et al., 2015a, 2015b, 2016; Linares et al., 2011). In lieu of driven by the hydraulic pressure, FO utilizes the osmotic pressure gradient between the feed and the more concentrated draw solution to spontaneously induce a water flow across a semipermeable membrane. FO possesses several merits over the conventional pressuredriven processes in terms of significantly lowered energy consumption, reduced cake layer formation and high fouling reversibility, great rejections toward a wide range of contaminants, and high water recovery rate. However, FO itself is not a separation process and the concentrated feed stream is general an unwanted by-product that should be further processed during the subsequent wastewater treatment. In the current work, a novel hybrid system has been demonstrated by integrating forward osmosis (FO) and coagulation/ flocculation (CF) for the treatment of synthetic textile wastewater. FO is used to dewater the wastewater stream and thus reduces its volume with a high water flux and recovery as well as low energy consumption and fouling; while CF is employed to remove the dyes from the FO concentrate with much improved efficiency and reduced dosage. The feasibility of the FOeCF hybrid system was examined by using synthetic dye solutions and textile wastewater containing multiple dyes, salts and additives as feed solutions. The objectives of this study are to investigate: (1) the effects of FO operation modes, dye chemistry and concentration on FO performance; (2) membrane rejections toward salts and dye compounds; (3) membrane fouling behavior and regeneration; (4) the efficiency of CF treatment, and (5) the performance of the FOeCF hybrid system to treat the synthetic textile wastewater. The newly developed FOeCF hybrid process may open up an alternative technology for the treatment and reuse of textile effluents. 2. Experimental and methods 2.1. Materials and chemicals Indigo carmine (INCA), Reactive black 5 (RB-5) and Alcian blue 8GX (AB-8) ordered from SigmaeAldrich were employed to represent the acid, reactive and basic dyes, respectively. Table 1 summarizes their chemical structures and molecular weights. Acetic acid, magnesium sulfate, magnesium chloride, monosodium phosphate, calcium chloride, and sodium chloride purchased from SigmaeAldrich were used as the additives to prepare the textile wastewater. A lab-made thin-film composite (TFC) flat-sheet membrane consisting of a woven-embedded sulfonated polyphenylenesulfone substrate and a polyamide selective skin was employed as the FO membrane (Han et al., 2016). The detailed specification of membrane fabrication and characterization is disclosed in the supporting information (SI) and our previous work (Han et al., 2016). Sodium chloride (NaCl) was acquired to prepare the draw solutions for FO performance tests. A mixture containing various inorganic salts (i.e., Iron (II) chloride-anhydrous, calcium oxide, aluminum sulfate-octadecahydrate, and polyaluminum chloride) and macromolecules (i.e., polydiallyldimethyl ammonium
chloride, cationic polyacrylamide, and cyanoguanidine) was utilized as the coagulants and flocculants during the CF treatment (Liang et al., 2014). Deionized water was produced by a Milli-Q unit (Millipore) with a resistivity of 15 MU cm. 2.2. Forward osmosis (FO) tests for water reclamation from textile solutions A lab-scale FO system was utilized to study the FO dewatering performance where two variable speed peristaltic pumps were used to circulate the feed and draw solutions through the membrane cell counter currently (Han et al., 2012). The membrane permeation cell was a plate-and-frame design with a spacer-free rectangular channel at both sides, and the effective membrane surface area was 10 cm2. The cross-flow velocity was kept at 0.2 L min1 (0.017 m s1) at both sides of the membrane and there was no hydraulic pressure difference across the membrane. The performances were measured at room temperature of 22 ± 0.5 C. The weight change of the feed solution over a certain duration was automatically collected via a digital balance (EK-4100i, A&D Company Ltd., Japan) that was connected to a computer. The water permeation flux, Jw, (in L m2 h1, abbreviated as LMH) was then calculated as:
Jw ¼
DV Am Dt
(1)
where DV (L) is the permeate water over a predetermined time Dt (h) and Am is the effective membrane surface area (m2). The salt concentration in the feed solution was obtained via measuring its conductivity, and the reverse solute flux, Js (in g m2 h1, abbreviated as gMH), was calculated as (Han et al., 2015a, 2016, 2012):
Js ¼
Cf ;t Vf ;t Cf ;i Vf ;i Am Dt
(2)
where Cf,t and Vf,t are the salt concentration and feed volume at the end of tests, respectively, while Cf,i and Vf,i refer to the salt concentration and total feed volume at the beginning, respectively. In terms of Jw and Js, the basic osmosis performance of the TFCFO membrane were first evaluated by using deionized water as the feed and NaCl solutions with various molar concentrations as the draw solutions. Then, the applicability of FO for water reclamation from synthetic dye solutions with various types of dyes (i.e., acid, reactive and basic) and/or concentrations (100e500 ppm) against recovery rate was investigated. The FO tests were operated under two different modes depending on the membrane orientation: (1) pressure retarded osmosis (PRO) mode where the draw solution faces against the polyamide selective layer, and (2) FO mode where the feed faces against the selective layer. For the short-term performance tests, the average data over the first 10 min of each test was reported. For the long-term performance tests, 0.5 L dye solutions were used as the feed and continuously run through the FO membrane until the predetermined recovery rate was reached. The feed water recovery rate, Re (%), was calculated by the following equation:
Re ¼
DV 100% Vf ;i
(3)
where Vf,i (L) is the initial volume of the feed solution. The draw solution concentration was maintained constant throughout the tests by conductivity control to exclude the effects of draw solution dilution effects. The membrane cleaning was
G. Han et al. / Water Research 91 (2016) 361e370
363
Table 1 Chemical structures and molecular weights of the organic dyes used in this study. Name of dye
Type of dye
Indigo carmine (INCA)
Acid
Molecular structure
466
Reactive black 5 (RB-5)
Reactive
992
Alcian blue 8GX (AB-8)
Basic
conducted by physically flushing freshwater through the surface of the fouled FO membrane at a flow rate of 0.4 L min1 (0.034 m s1) for 1 h and the total water volume used in each rinse was around 3 L. 2.3. Coagulation/flocculation (CF) tests for dye removal The jar-test experiment was adopted to perform the coagulation/flocculation (CF) operation during which an appropriate amount of coagulants & flocculants solution was added into the wastewater solutions under stirring (Liang et al., 2014). After settling for 1 h, the treated solution was filtered by using a vacuum filtration apparatus. The filtrates were collected for further analyses. 2.4. Hybrid FOeCF process to treat synthetic textile wastewater By integrating forward osmosis (FO) and coagulation/flocculation (CF) processes, one hybrid FOeCF system was investigated for the treatment of textile wastewater. As illustrated in Fig. 1, FO is employed to dewater the wastewater stream and CF is followed to remove the dyes. Compared to conventional pressure driven membrane processes such as NF and RO, FO possesses a higher water permeation flux yet very low energy consumption, greater recovery rate and dye rejection, less and more reversible fouling (Liu and Mi, 2012; Jin et al., 2012; Zhao et al., 2012). After FO dewatering, the dye concentration in the wastewater stream is significantly enhanced while the volume is dramatically reduced. As a result, the efficiency of CF could be improved and the chemical dosage is reduced so that less cost and environmental impacts are expected. In order to mimic the real textile wastewater, INCA, RB-5 and AB-8 were dissolved in deionized water with an equivalent weight ratio, and then acetic acid, magnesium sulfate, magnesium chloride, monosodium phosphate, calcium chloride and sodium
Molecular weight (g/mol)
1299
chloride were added to represent the chemical additives (Doumic et al., 2015). The main characteristics of the synthetic textile wastewater are tabulated in Table 2. In the current study, a batch mode was chosen to operate the FOeCF system since it provides easy operation, minimal attention, and is close to practical applications. Specifically, 0.5 L synthetic textile wastewater was continuously pumped into the FO membrane module as the feed until reaching a recovery rate of 90% (or 450 ml water was recovered). NaCl solutions with a concentration of 1 M and 2 M were used as the draw solutions, respectively. After FO, the highly concentrated wastewater stream was treated by CF with a dosage of coagulants/flocculants at 500e1000 ppm (or 0.5e1.0 g per liter textile wastewater). After settlement and filtration, the obtained sludge could be dried and burned, while the filtrates that contain trace amounts of dyes and coagulates/flocculates can be discharged or returned to the FOeCF system to further improve the overall recovery rate. 2.5. Physicochemical analyses A UVevis integral method and a total organic carbon (TOC) analyzer were employed to determine the dye concentration in solutions (Liang et al., 2014; Ong et al., 2014). The former was based on the UVevis integral between 350 and 650 nm obtained by scanning the sample solution via a UVevis spectrophotometer (Pharo 300, Merck), while the latter was relied on the concentration measurement determined by a TOC analyzer (TOC, ASI 5000A, Shimazu). As shown in Fig. S2 and our previous work (Liang et al., 2014), the TOC value has an almost linear relationship with dye concentration in the range of 0e1000 ppm for the three dyes. The UVevis integral displays a good linear relationship with dye concentration when it is less than 100 ppm. In order to achieve reasonable accuracy, the highly concentrated dye solutions had to be diluted to lower concentrations (i.e., less than 100 ppm) when the UVevis integral method was used.
364
G. Han et al. / Water Research 91 (2016) 361e370
Fig. 1. The schematic of the coagulation/flocculation (CF) promoted forward osmosis (FO) process to treat textile wastewater.
Table 2 Key characteristic parameters of the synthetic textile wastewater, pH ¼ 3.1 ± 0.3 Characteristics
INCA
RB-5
AB-8
Acetic acid
Naþ
Ca2þ
Mg2þ
PO4 3
Cl1
SO4 2
Unit
Concentration
100 ± 5
100 ± 5
100 ± 5
60 ± 3
1000 ± 10
72 ± 2
12 ± 2
0.3 ± 2
1700 ± 10
3.0 ± 1
mg/L
The dye rejection, Rd, obtained from the TOC measurement can be calculated as:
Cp 1 Cf ;i
Rd ¼
3.1. Characteristics of the TFC-FO membrane
! 100%
(4)
where Cf,i (ppm) is the initial dye concentration in the feed, and Cp (ppm) is the concentration in the permeate or treated solution. Using the UVevis method, Rd could be alternatively calculated as:
Ip 1 If ;i
Rd ¼
! 100%
(5)
where If,i and Ip are the UVevis integral values of the feed and the permeate or treated solutions, respectively. It is worthy to point out that Cp in FO should be obtained from the change of the dye concentration in the draw solution as follows (Han et al., 2015a):
Cp ¼
Cd;t Vd;i þ DV Cd;i Vd;i DV
(6)
where Cd,t (ppm) is the dye concentration at the end of tests and Cd,i (ppm) is the concentration in the initial draw solution with a initial volume of Vd,i (L). The dye rejection in FO, Rf,d, was then calculated as:
Rf ;d ¼ ¼
1
Cp Cf ;i
3. Results and discussion
! 100%
! Cd;t Vd;i þ DV Cd;i Vd;i 100% 1 Cf ;i DV
(7)
In order to ensure the reproducibility of the obtained data, three tests were carried out for each condition and the averaged value was reported.
A lab-made thin film composite (TFC) flat-sheet membrane consisting of a polyamide selective layer and a woven-embedded sulfonated polyphenylenesulfone (sPPSU) substrate was used as the FO membrane (Han et al., 2016). As shown in Fig. 2, the polyamide skin has a typical “ridge-and-valley” surface morphology with a very thin thickness of several hundred nanometers (Veríssimo et al., 2005). The sPPSU membrane substrate shows a fully sponge-like structure across the whole cross-section even though a woven is incorporated. Table 3 presents the transport properties and structural parameter of the TFC-FO membrane (Han et al., 2016). It possesses a high pure water permeability, A, of 3.4 L m2 h1 bar1 accompanying with a low salt permeability coefficient, B, of 0.103 L m2 h1. The dye rejections of the TFC-FO membrane under the RO operation at 1 bar are always greater than 99.9% (Table 4), signifying the outstanding selectivity of the FO membrane toward a relatively wide range of dyes and salts. In addition, a small structural parameter, S, of 300 mm is achieved. These are mainly due to the hydrophilic nature of the sPPSU material and the small membrane thickness (~67 mm). In FO, this structure can not only provide robust mechanical strength but also effectively minimize the transport resistance and internal concentration polarization (ICP). As a result, the TFC-FO membrane shows outstanding FO performance. Fig. 3 shows the water flux, Jw, and specific reverse salt flux, Js/Jw, of the membrane as a function of draw solution concentration using deionized water as the feed under PRO and FO modes, respectively. High Jw values of 32.7e69.3 LMH and 23.3e38.7 LMH were obtained under PRO and FO modes, respectively, using 0.6e2.0 M NaCl as draw solutions. Simultaneously, a small Js/Jw of less than 0.2 g/L was achieved even when the draw solution concentration is increased to 2 M, verifying the low salt permeation of the FO membrane to NaCl. In terms of membrane
G. Han et al. / Water Research 91 (2016) 361e370
365
Fig. 2. FESEM images of the top surface, bottom surface, cross-section of the polyamide layer, and the overall cross-section of the flat-sheet TFC-FO membrane (Han et al., 2016).
Table 3 Transport properties, structural parameter and dye rejection of the flat-sheet TFC-FO membrane (Han et al., 2016). Membrane
Water permeability, A (L m2 h1 bar1)
Salt permeability, B (L m2 h1)
Km ( 105 s m1)
Sa ( 106 m)
Dye rejection,b Rd (%)
TFC-FO
3.40 ± 0.26
0.103 ± 0.008
1.97
300
>99.9
a b
S is calculated based on the membrane performance under FO mode using 1 M NaCl as the draw solution and de-ionized water as feed. Rd is tested at 1 bar under RO operation using 100e500 ppm dye solutions containing three different dyes as the feeds, respectively.
Table 4 Initial water permeation flux, Jw, and dye rejections, Rf,d, of the TFC-FO membrane as a function of feed solution, draw solution and operation mode. Feed solution code
Dye concentration (ppm)
Draw solution (M)
Initial water flux, Jw (LMH)a
Dye rejection, Rf,d (%)b
PRO mode
FO mode
PRO mode
FO mode
Deionized water INCA INCA INCA RB-5 AB-8 Synthetic textile wastewater Synthetic textile wastewater
e 100 300 500 300 300 300 300
1.0 1.0 1.0 1.0 1.0 1.0 1.0 2.0
47.7 e 27.1 e e e e e
30.5 30.0 29.5 26.7 29.0 30.0 24.0 36.0
e e >99.9 e e e e e
e >99.9 >99.9 >99.9 >99.9 >99.9 >99.9 >99.9
a b
The initial Jw is the averaged value during the first 10 min of each test. The Rf,d of the FO process is calculated after both short and long-term tests.
morphology, mechanical strength, water flux, and salt and dye rejection, the newly fabricated TFC membrane is an effective membrane to treat textile wastewater under FO operations. 3.2. FO for water reclamation from dye solutions The feasibility of applying FO to reclaim water from textile effluents was firstly investigated using synthetic dye solutions as feeds. Acid (INCA), reactive (RB-5), and basic (AB-8) dyes were chosen in this study because of their popularity in textile and dye
industries. As summarized in Table 1, each INCA and RB-5 molecule possesses two and four sulfonic acid groups, respectively, suggesting that they remain in the anionic state in the solutions due to the low pKa value of the sulfonic groups (Pirillo et al., 2009). On the contrary, AB-8 is a cationic dye and it contains four isothiouronium groups and one copper nucleus per molecule. In other words, INCA and RB-5 have negative charges while AB-8 possesses positive charges in the solutions. In addition, AB-8 has the largest molecular weight (Mw) of 1299 g/mol, RB-5 possesses a medium Mw of 992 g/ mol, and INCA has the smallest Mw of 466 g/mol.
G. Han et al. / Water Research 91 (2016) 361e370
Water flux, Jw (LMH)
100
PRO mode FO mode
(a)
80
60
40
20
0 0.0
0.5
1.0
1.5
2.0
NaCl concentration (M)
Specific reverse salt flux, Js/Jw (g/L)
366
1.0
(b)
0.8
0.6
0.4
0.2
0.0 0.0
0.5
1.0
1.5
2.0
NaCl concentration (M)
Fig. 3. Water flux, Jw, and specific reverse salt flux, Js/Jw of the TFC-FO membrane as a function of draw solution concentration and operation mode. The feed is deionized water.
3.2.1. Short-term FO performance for water reclamation from dye solutions In terms of water permeation flux, Jw, the short-term FO performance of the TFC membrane was first investigated under both PRO and FO modes using NaCl solutions with various concentrations as draw solutions and dye solutions and synthetic textile wastewater as feeds. As tabulated in Table 4, the TFC-FO membrane shows high initial Jw values of 47.7 LMH and 30.5 LMH under PRO and FO modes, respectively, when using deionized water as the feed and 1 M NaCl as the draw solution. Under the PRO mode, however, the initial Jw drops from 47.7 LMH to 27.1 LMH even during a 10-min test when the feed is changed to a 300 ppm INCA solution. The performance under the FO mode is quite different from the PRO mode. The initial Jw slightly decreases from 30.5 LMH to 29.5 LMH when using an INCA solution of 300 ppm as the feed. An increase in INCA concentration does not show much effect on the initial Jw under the FO mode (i.e., 26.7e30.0 LMH for 100e500 ppm INCA vs. 30.5 LMH for deionized water). This might be because membrane fouling has not been developed on the polyamide surface during such short duration tests. Similar performances under the FO mode are observed for 300 ppm RB-5 and AB-8 dye solutions with initial Jw values of 29.0 LMH and 30.0 LMH, respectively. In the treatment of synthetic textile wastewater, the initial Jw drops to 24.0 LMH because of its higher salt concentration; however, Jw could be increased to 36 LMH when the draw solution concentration is raised to 2 M. In addition, the dye content in draw solutions is not detectable by both TOC and UVevis measurements, which indicates that the highly selective TFC-FO membrane has a high dye rejection, Rf,d, of larger than 99.9%. Clearly, the FO process has exhibited promising for water reclamation from textile effluents with high water fluxes, excellent salt and dye rejections. 3.2.2. Long-term FO performance for water reclamation from dye solutions 3.2.2.1. The effects of FO operation modes. Fig. 4 portrays the reduction of Jw of the TFC-FO membrane as a function of feed recovery rate running under (a) PRO and (b) FO modes, respectively. Since the draw solution concentration was maintained at 1 M throughout the tests, the baseline curves under both modes slightly decrease with an increase in recovery of deionized water due to the reverse salt flux and ICP effects. However, in the PRO mode, the initial Jw immediately drops to 27.1 LMH when the feed is changed to an INCA solution of 300 ppm (Table 4 and Fig. 4 (a)). Then, Jw rapidly decreases with an increase in recovery rate and drops to 17.6 LMH when the recovery reaches 90%. This suggests that membrane fouling under the PRO mode induced by the dye molecules is very fast and significant. Such a dramatic flux decline
could be attributed to two reasons: one is that the dyes in the feed would accumulate within the porous structure and on the membrane surface of the substrate and thus enhances the transport resistance and increases the structural parameter; the other is due to the back diffusion of the salts from the draw solution which enlarges the ICP effects (McCutcheon and Elimelech, 2007). Conversely, a much milder decrease in Jw is observed in the FO mode. As displayed in Fig. 4 (b) and Table 4, almost the same initial Jw values (i.e., 29.5 vs. 30.5 LMH) are achieved when using 300 ppm INCA instead of deionized water as the feed, and then Jw slightly decreases to 23.7 LMH when the recovery rate goes to 90%. In other words, the reduction of Jw under the FO mode is only 20% even at such a high feed recovery. Interestingly, the decline of Jw is characterized by two stages; a relatively fast and significant drop occurs at a low recovery of less than 5%, and followed by a mild and stable decay till to 90% recovery. The first flux reduction is most likely resulted from the deposition of dye molecules on the polyamide surface. Once the fouling is gradually developed and stabilized, the flux decline becomes milder and slower (Zhang et al., 2014). A comparison of the membrane performance in PRO and FO modes indicates that FO mode is more effective for textile solution separation in terms of better transport stability and less membrane fouling despite of its lower initial Jw. 3.2.2.2. The effects of dye concentration and chemistry. The effects of dye chemistry and concentration on FO performance under the FO mode are investigated in the next step. Fig. 5 (a) shows the Jw reduction as a function of feed recovery rate using different dye solutions with the same dye concentration of 300 ppm as feeds and 1 M NaCl as the draw solution. The decrease of Jw follows a trend of AB-8 [ RB-5 > INCA. INCA shows the lowest reduction in Jw which is only about 18.5% of the initial value at a high recovery rate of 90%. RB-5 displays a slightly larger flux decrease where its Jw drops from 29.0 LMH to 19.7 LMH at the same recovery. Since both INCA and RB-5 possess negative charges, the greater Jw decline of the later is probably due to its larger molecular weight (Table 1). However, AB8 exhibits a much faster and more severe Jw decline along with an increase in water recovery. Its Jw drops from 30 LMH to 10.6 LMH when the recovery reaches 90%, suggesting a 65% reduction. In addition to its large molecular weight, this large drop might be caused by its positive charges that induce strong interaction between the dye molecules and the negative charged polyamide layer. By using INCA solutions with a concentration ranging from 100 to 500 ppm as feed solutions, the effects of dye concentration on membrane Jw under the FO mode are studied. As depicted in Fig. 5 (b), a small drop in Jw is observed when the feed is changed from deionized water to 100 ppm INCA. The Jw reduction undergoes a
G. Han et al. / Water Research 91 (2016) 361e370
367
Fig. 4. The reduction of water flux as a function of feed water recovery rate under (a) PRO and (b) FO modes. Deionized water and 300 ppm INCA dye solution were the feeds, while 1 M NaCl solution was the draw solution and its concentration was maintained constantly throughout the tests.
Fig. 5. The reduction of water flux as a function of feed water recovery rate under the FO mode: (a) using various dye solutions as feeds with the same dye concentration of 300 ppm, and (b) using various INCA solutions as feeds with different dye concentrations. 1 M NaCl solution was the draw solution and its concentration was maintained constantly throughout the tests.
3.2.2.3. FO membrane fouling and cleaning. After the performance tests shown in Fig. 5 (a), a thin gel-like dye layer is found on the top surface of the polyamide skin which could explain the decline in water flux. However, the fouling layer can be easily washed away via physically flushing freshwater without using any chemicals. As displayed in Fig. 6, the initial Jw of the fouled TFC-FO membrane by three different dyes could be almost fully recovered with recovery rates of approximately 97% or above. Although AB-8 induces the most severe fouling where Jw drops from 30.0 LMH to 10.6 LMH when the feed recovery rate reaches 90% (Fig. 5), its Jw can be restored to 29.0 LMH (i.e., a 97% recovery) after rinse. These results demonstrate that the membrane fouling of the TFC-FO membrane under the FO mode is not only mild but also loose and reversible (Mi and Elimelech, 2010). The low and reversible fouling feature
Feed: Deionized water
100
Normalized initial Jw, %
slight increase when the INCA concentration is increased to 300 ppm. This phenomenon is possibly caused by two factors; namely, the enhanced attachment of dyes on the polyamide surface and the higher osmotic pressure of the feed solution at a larger dye concentration. An increase in INCA concentration to 500 ppm does not further induce a significant drop in Jw, implying that dye fouling under the FO mode is mild and the polyamide surface is probably saturated at a low concentration. According to the results illustrated in Fig. 5, it is concluded that membrane fouling under the FO mode is closely related to dye chemistry and molecule weight but is less sensitive to dye concentration.
80 60 40 20 0
Fesh
INCA
RB-5
AB-8
Fig. 6. The normalized initial water flux, Jw, of the regenerated TFC-FO membrane over the initial Jw of the fresh membrane. The membrane is fouled by INCA, RB-5 and AB-8, respectively. Membranes are cleaned by physically flushing freshwater through the fouled polyamide surface for 1 h. 1 M NaCl solution was the draw solution and its salt concentration was maintained constantly throughout the tests.
makes FO superior to other membrane separation processes such as NF for textile wastewater treatment (Liang et al., 2014; Ong et al., 2014; Quintanilla et al., 2010).
368
G. Han et al. / Water Research 91 (2016) 361e370
3.3. Hybrid FOeCF process for the treatment of synthetic textile wastewater FO has shown promising performance in dewatering the dye solutions; however, it is not able to achieve a complete dye removal. Therefore, CF is applied after FO to further treat the highly concentrated stream. To mimic the real applications, one synthetic textile wastewater containing 300 ppm multi-dyes, inorganic salts and organic additives was used as the feed in FO (Table 2). Fig. 7 shows the change of membrane flux as a function of synthetic textile wastewater recovery rate under the FO mode using (a) 1 M and (b) 2 M NaCl as draw solutions. Since the draw solution concentration is maintained constant, the base-line curves using deionized water as the feed shows a small variation in Jw within the tests. When changing the feed to the synthetic textile wastewater, the initial Jw decreases to 24 LMH when using 1 M NaCl as the draw solution. The initial Jw increases to 36.0 LMH when the draw solution concentration is increased to 2 M due to the increased osmotic pressure gradient. However, because of the presence of dyes and the inorganic salt in the feed, Jw for both draw solutions undergoes significant declines against recovery rate with similar trends. In general, a rapid flux decline at a relatively low recovery of less than 40% is firstly observed, followed by a mild decay until reaching a high recovery rate. As illustrated in Fig. 7, Jw drops to 9.8 LMH and 12.0 LMH for 1 M and 2 M NaCl draw solutions, respectively, when the recovery rate reaches 90%. The first rapid drop is mainly attributed to fouling while the next mild decline is due to the effects of the concentrated textile solution and the back diffusion of salts as the fouling has been gradually stabilized. It is also found that the Jw decline is faster and more dramatic for the 2 M draw solution than the 1 M one even though the former has a higher initial Jw. As observed in other FO processes (Li et al., 2014; Liu and Mi, 2012; Jin et al., 2012; Zhang et al., 2014), a larger initial Jw would accelerate the accumulation of species on the polyamide surface, resulting in faster flux decrease. Nevertheless, the Jw values of the TFC-FO membrane always maintain 41% and 33% of the initial values, respectively, even at a high recovery of 90%. The long-term stability of the FO performance for dewatering the synthetic textile wastewater is also investigated. Fig. 8 shows the Jw profiles of the fresh and the regenerated TFC-FO membranes as a function of feed recovery rate and testing duration. Three FO tests are performed using the same membrane but the fouled membrane is cleaned by physically flushing freshwater through the fouled polyamide surface for 1 h after each test. As displayed in Fig. 8, quite stable and reproducible performance is achieved. The Jw
reduction during each test at a recovery of 90% or within a testing duration of 168 h is almost the same which is around 67% of the initial water flux (i.e., from 35e36 LMH to 11.6e12 LMH). In addition, water rinses can effectively rejuvenate the fouled membrane and the initial Jw is able to be restored by 98% even after 3 cycles. This again demonstrates the potential of using FO for textile wastewater reclamation. After FO, the highly concentrated textile wastewater stream is further treated by CF for dye removal. Fig. 9 clearly shows that CF becomes very effective in treating this highly concentrated textile effluent. When the coagulants & flocculants with a dosage of 500e1000 ppm are added into the solution under stirring, the formation of particles/flocs immediately happens. After 1 h settling, the suspensions are spontaneously phase-separated with water and stay at the bottom of the container. Since ferrous ions are used as one of the coagulants, the water solution shows a light green color. After filtration, a gel-like sticky film is left on the filter paper which could be further treated easily. This indicates that the dye removal efficiency of CF is significantly enhanced by FO preconcentration, while the amount of chemical usage and the produced sludge volume are much reduced. The UVevis spectra of the synthetic textile wastewater, FO concentrated wastewater, and filtrate of the CF treated wastewater further confirms the above conclusions. As shown in Fig. 10, the UV absorbance increases dramatically after FO treatment because the textile wastewater is highly concentrated at a high recovery rate of 90%. In other words, the dye concentration is around 10 times of the original value. After CF treatment, the absorbance drops significantly and the measured dye concentration in the filtrate is less than 150 ppm. This suggests that the removal efficiency of the CF process is larger than 95%. 4. Conclusions This study has experimentally conceived the applicability of the combination of forward osmosis (FO) process with coagulation/ flocculation (CF) for the treatment and reuse of textile wastewater. In the hybrid FOeCF system, FO is employed to dewatering the wastewater with high water flux and recovery yet low membrane fouling; and CF is applied for dye precipitation and removal from the FO concentrated stream with high efficiency, low chemical dosages, and minimal environmental impact. By employing a labfabricated thin-film composite (TFC) FO membrane and synthetic dye solutions, the effects of membrane operation mode (FO and PRO modes), dye chemistry (acid, basic and reactive dyes) and concentration (100e500 ppm) on FO performance were firstly studied, which revealed that 1) dye fouling and flux reduction in
Fig. 7. The reduction of water flux as a function of feed water recovery rate under the FO mode by using deionized water and synthetic textile wastewater as feeds and (a) 1 M NaCl and (b) 2 M NaCl as draw solutions. The salt concentration of the draw solution was maintained constantly throughout the tests.
G. Han et al. / Water Research 91 (2016) 361e370
369
Fig. 8. The reduction of water flux as a function of feed water recovery rate (a) and testing duration (b) under the FO mode by using the synthetic textile wastewater as the feed and 2 M NaCl as the draw solution. Membranes were cleaned by physically flushing freshwater through the fouled polyamide surface for 1 h and then reused. The salt concentration of the draw solution was maintained constantly throughout the tests.
Fig. 9. Photos of the coagulation/flocculation (CF) process and the obtained sludge after filtration.
the PRO mode is much more severe than the FO mode; 2) dye fouling on the polyamide selective layer under the FO mode is mild and slightly increases with dye concentration; 3) dye fouling and flux decline is fast but no further significant fouling occurs in the extended recovery rate; 4) dye chemistry significantly influences the membrane fouling tendency and cationic dyes normally show more dramatic fouling because of the strong electrostatic interaction with the polyamide layer; 5) membrane fouling under the FO mode is loose and highly reversible; the water flux could be almost fully recovered by physically flushing; 6) the extent of dye fouling is closely related to the initial water flux; and 7) FO shows excellent rejections toward dyes and salts. The batch-mode FOeCF integrated system was then studied for treating synthetic textile wastewater containing multi-dyes, inorganic salts and organic additives. With the increase of feed recovery, flux reduction was observed during the FO operation; however, the flux was larger than 9.8 and 12.0
Fig. 10. The UVevis absorbance spectra of the original synthetic textile wastewater, the concentrated wastewater stream by FO dewatering at a recovery of 90%, and the filtrate after the CF treatment and filtration. The UVevis spectrum of the FO concentrated solution was obtained via diluting the original solution by 30 times and the diluted sample was measured, then the resultant absorbance was multiplied by a factor of 30.
LMH even at a high recovery rate of 90% using 1 M and 2 M NaCl as the draw solutions, respectively. Since the dye concentration in the FO concentrated wastewater is increased by about 10 times, CF shows outstanding efficiency and is able to remove more than 95% of the dyes with a dosage of 500e1000 ppm. It is believed that the FOeCF hybrid system has great potential for the treatment and reuse of the textile wastewater.
370
G. Han et al. / Water Research 91 (2016) 361e370
Acknowledgments This research was funded by the Singapore National Research Foundation under its Competitive Research Program for the project entitled, “Advanced FO Membranes and Membrane Systems for Wastewater Treatment, Water Reuse and Seawater Desalination” (grant number: R-279-000-336-281). The authors also would like to thank BASF SE, Germany for funding this work with a grant number of R-279-000-411-597. Special thanks are due to Mr. Heyang Wu. Appendix A. Supporting information Supporting information related to this article can be found at http://dx.doi.org/10.1016/j.watres.2016.01.031. References Chen, T., Gao, B., Yue, Q., 2010. Effect of dosing method and pH on color removal performance and floc aggregation of polyferric chlorideepolyamine dualcoagulant in synthetic dyeing wastewater treatment. Colloids Surfaces A Physicochem. Eng. Asp. 355, 121e129. Cheng, S., Oatley, D.L., Williams, P.M., Wright, C.J., 2012. Characterisation and application of a novel positively charged nanofiltration membrane for the treatment of textile industry wastewaters. Water Res. 46, 33e42. Chung, T.S., Li, X., Ong, R.C., Ge, Q., Wang, H., Han, G., 2012. Emerging forward osmosis (FO) technologies and challenges ahead for clean water and clean energy applications. Curr. Opin. Chem. Eng. 1, 246e257. Ciardelli, G., Corsi, L., Marcucci, M., 2001. Membrane separation for wastewater reuse in the textile industry. Resour. Conserv. Recycl. 31, 189e197. Dafale, N., Rao, N.N., Meshram, S.U., Wate, S.R., 2008. Decolorization of azo dyes and simulated dye bath wastewater using acclimatized microbial consortiumebiostimulation and halo tolerance. Bioresour. Technol. 99, 2552e2558. Doumic, L.I., Soares, P.A., Ayude, M.A., Cassanello, M., Boaventura, R.A.R., Vilar, V.J.P., 2015. Enhancement of a solar photo-Fenton reaction by using ferrioxalate complexes for the treatment of a synthetic cotton-textile dyeing wastewater. Chem. Eng. J. 277, 86e96. Fersi, C., Dhahbi, M., 2008. Treatment of textile plant effluent by ultrafiltration and/ or nanofiltration for water reuse. Desalination 222, 263e271. Han, G., Chung, T.S., Toriida, M., Tamai, S., 2012. Thin-film composite forward osmosis membranes with novel hydrophilic supports for desalination. J. Membr. Sci. 423e424, 543e555. Han, G., de Wit, J.S., Chung, T.S., 2015a. Water reclamation from emulsified oily wastewater via effective forward osmosis hollow fiber membranes under the PRO mode. Water Res. 81, 54e63. Han, G., Zuo, J., Wan, C., Chung, T.S., 2015b. Hybrid pressure retarded osmosismembrane distillation (PROMD) process for osmotic power and clean water generation. Environ. Sci. Water Res. Technol. 1, 507e515. Han, G., Zhao, B., Fu, F., Chung, T.S., Weber, M., Staudt, C., Maletzko, C., 2016. High performance thin-film composite membranes with mesh-reinforced hydrophilic sulfonated polyphenylenesulfone (sPPSU) substrates for osmotically driven processes. J. Membr. Sci. 502, 84e93. Hao, O.J., Kim, H., Chiang, P.C., 2000. Decolorization of wastewater. Crit. Rev. Environ. Sci. Technol. 30, 449e505. Jin, X., She, Q., Ang, X., Tang, C.Y., 2012. Removal of boron and arsenic by forward osmosis membrane: influence of membrane orientation and organic fouling. J. Membr. Sci. 389, 182e187. Kapdan, I.K., Kargi, F., 2002. Simultaneous biodegradation and adsorption of textile dyestuff in an activated sludge unit. Process Biochem. 37, 973e981. Lau, W.J., Ismail, A.F., 2009. Polymeric nanofiltration membranes for textile dye wastewater treatment: preparation, performance evaluation, transport modelling, and fouling controlea review. Desalination 245, 321e348. Li, P., Lim, S.S., Neo, J.G., Ong, R.C., Weber, M., Staudt, C., Widjojo, N., Maletzko, C., Chung, T.S., 2014. Short- and long-term performance of the thin-film composite forward osmosis (TFC-FO) hollow fiber membranes for oily wastewater purification. Ind. Eng. Chem. Res. 53, 14056e14064.
Liang, C.Z., Sun, S.P., Li, F.Y., Ong, Y.K., Chung, T.S., 2014. Treatment of highly concentrated wastewater containing multiple synthetic dyes by a combined process of coagulation/flocculation and nanofiltration. J. Membr. Sci. 469, 306e315. Lin, J., Ye, W., Zeng, H., Yang, H., Shen, J., Darvishmanesh, S., Luis, P., Sotto, A., Van der Bruggen, B., 2015a. Fractionation of direct dyes and salts in aqueous solution using loose nanofiltration membranes. J. Membr. Sci. 477, 183e193. Lin, J., Tang, C.Y., Ye, W., Sun, S.P., Hamdan, S.H., Volodin, A., Van Haesendonck, C., Sotto, A., Luis, P., Van der Bruggen, B., 2015b. Unraveling flux behavior of superhydrophilic loose nanofiltration membranes during textile wastewater treatment. J. Membr. Sci. 493, 690e702. Linares, R.V., Quintanilla, V.Y., Li, Z.Y., Amy, G., 2011. Rejection of micropollutants by clean and fouled forward osmosis membrane. Water Res. 45, 6737e6744. Lipscomb, G.G., 2008. In: Flank, W.H., Abraham, M.A., Matthews, M.A. (Eds.), Membrane Separation Technology: Past, Present, and Future, Innovations in Industrial and Engineering Chemistry, ACS Symposium Series 1000. American Chemical Society, pp. 281e333. Liu, Y., Mi, B., 2012. Combined fouling of forward osmosis membranes: synergistic foulant interaction and direct observation of fouling layer formation. J. Membr. Sci. 407e408, 136e144. Marcucci, M., Ciardelli, G., Matteucci, A., Ranieri, L., Russo, M., 2002. Experimental campaigns on textile wastewater for reuse by means of different membrane processes. Desalination 149, 137e143. McCutcheon, J.R., Elimelech, M., 2007. Modeling water flux in forward osmosis: implications for improved membrane design. AIChE J. 53, 1736e1744. Mi, B.X., Elimelech, M., 2010. Organic fouling of forward osmosis membranes: fouling reversibility and cleaning without chemical reagents. J. Membr. Sci. 348, 337e345. Nidheesh, P.V., Gandhimathi, R., Ramesh, S.T., 2013. Degradation of dyes from aqueous solution by Fenton processes: a review. Environ. Sci. Pollut. Res. 20, 2099e2132. Ong, Y.K., Li, F.Y., Sun, S.P., Zhao, B.W., Liang, C.Z., Chung, T.S., 2014. Nanofiltration hollow fiber membranes for textile wastewater treatment: lab-scale and pilotscale studies. Chem. Eng. Sci. 114, 51e57. Pan, J.H., Han, G., Zhou, R., Zhao, X.S., 2011. Hierarchical N-doped TiO2 hollow microspheres consisting of nanothorns with exposed anatase {101} facets. Chem. Commun. 47, 6942e6944. Pirillo, S., Pedroni, V., Rueda, E., Ferreira, M.L., 2009. Elimination of dyes from aqueous solutions using iron oxides and chitosan as adsorbents, a comparative study. Quim. Nova 32, 1239e1244. Quintanilla, V.Y., Maeng, S.K., Fujioka, T., Kennedy, M., Amy, G., 2010. Proposing nanofiltration as acceptable barrier for organic contaminants in water reuse. J. Membr. Sci. 362, 334e345. Rafatullah, M., Sulaiman, O., Hashim, R., Ahmad, A., 2010. Adsorption of methylene blue on low-cost adsorbents: a review. J. Hazard. Mater. 177, 70e80. Robinson, T., McMullan, G., Marchant, R., Nigam, P., 2001. Remediation of dyes in textile effluent: a critical review on current treatment technologies with a proposed alternative. Bioresour. Technol. 77, 247e255. Treffry-Goatley, K., Buckley, C.A., Groves, G.R., 1983. Reverse osmosis treatment and reuse of textile dyehouse effluents. Desalination 47, 313e320. Van der Bruggen, B., Cornelis, G., Vandecasteele, C., Devreese, I., 2005. Fouling of nanofiltration and ultrafiltration membranes applied for wastewater regeneration in the textile industry. Desalination 175, 111e119. Verliefde, A.R.D., Cornelissen, E.R., Heijman, S.G.J., Verberk, J.Q.J.C., Amy, G.L., Van der Bruggen, B., van Dijk, J.C., 2008. The role of electrostatic interactions on the rejection of organic solutes in aqueous solutions with nanofiltration. J. Membr. Sci. 322, 52e66. Verma, A.K., Dash, R.R., Bhunia, P., 2012. A review on chemical coagulation/flocculation technologies for removal of colour from textile wastewaters. J. Environ. Manag. 93, 154e168. Veríssimo, S., Peinemann, K.V., Bordado, J., 2005. Thin-film composite hollow fiber membranes: an optimized manufacturing method. J. Membr. Sci. 264, 48e55. Zahrim, A.Y., Hilal, N., 2013. Treatment of highly concentrated dye solution by coagulation/flocculationesand filtration and nanofiltration. Water Resour. Ind. 3, 23e34. Zhao, S., Zou, L., Tang, C.Y., Mulcahy, D., 2012. Recent developments in forward osmosis: opportunities and challenges. J. Membr. Sci. 396, 1e21. Zhang, S., Wang, P., Fu, X., Chung, T.S., 2014. Sustainable water recovery from oily wastewater via forward osmosis-membrane distillation (FO-MD). Water. Res. 52, 112e121.
