- Email: [email protected]
Efficient removal of heavy metal ions by forward osmosis membrane with a polydopamine modified zeolitic imidazolate framework incorporated selective layer
Journal of Hazardous Materials 367 (2019) 339–347
Contents lists available at ScienceDirect
Journal of Hazardous Materials journal homepage: www.elsevier.com/locate/jhazmat
Efficient removal of heavy metal ions by forward osmosis membrane with a polydopamine modified zeolitic imidazolate framework incorporated selective layer
T
⁎
Ming Qiu, Chunju He
State Key Laboratory for Modification of Chemical Fibers and Polymer Materials, College of Materials Science and Engineering, 2999 North Renmin Road, Songjiang District, Donghua University, Shanghai 201620, China
A R T I C LE I N FO
A B S T R A C T
Keywords: Forward osmosis Selective layer Polydopamine ZIF-8 Heavy metal ions
A novel thin film nanocomposite (TFN) forward osmosis (FO) membrane with a positively charged and nanofunctional selective layer has been developed for effective heavy metal ions removal. The selective layer is constructed by penetrating the polydopamine modified zeolitic imidazolate framework (ZIF-8@PDA) in the poly (ethyleneimine)/1,3,5-benzenetricarboxylic acid chloride (PEI/TMC) crosslinked matrix. Compared with the pristine thin film composite (TFC) membrane, the thin film nanocomposite membrane (0.05 wt % nanofillers loading) exhibits a higher water flux (20.8 vs12.8 LMH) without losing of selectivity in terms of Js/Jw ratio (0.25 vs 0.20 g L−1) in FO mode. This improvement of the permeability is mainly attributed to the optimized selective layer with good wettability and loose structure. Besides, the modified PDA layer facilitates the affinity between the nanofillers and selective layer, which results in an ideal selectivity. In addition, this modified membrane shows a high heavy metal ion (Cu2+, and Ni2+ and Pb2+) rejection (> 96%) in FO mode. Our finding offers a simple and efficient method to enhance the FO performance of membrane by designing the selective layer for treating heavy metal wastewater.
1. Introduction Wastewater with heavy metal ions are produced by diverse industrial processes, such as mining, machinery and electronics, which is harmful to human health and ecological environment [1,2]. Therefore, it is necessary to remove heavy metal ions from the wastewater before discharging. Traditional pressure driven membrane technologies, such as ultrafiltration and nanofiltration have been increasingly developed as a promising approach for heavy metal ions removal [3–6]. However, these membrane technologies own some drawbacks, such as fouling problem and high energy consumption. In the recent years, forward osmosis (FO) technology receives lots of attention for water treatment [7,8], including heavy metal ion removal [9,10], since its low polluted tendency, reduced energy cost and environmental friendly. FO membrane currently mainly studied is thin film composite (TFC) membrane, which is prepared by forming a dense m-Phenylenediamine/1,3,5-benzenetricarboxylic acid chloride (MPD/ TMC) cross-linked layer on a substrate [11,12]. Although the dense selective layer endows the membrane with an excellent solute rejection, the cross-linked networks limits the penetration of water molecule, thus
⁎
reduces the separation efficiency [13,14]. Hence, FO membranes with nanofiltration-like selective layer have been developed in some application areas, such as treating wastewater containing heavy metal ions and organic contaminant. However, there is needing further exploration before the process is effective by these membranes. Introduction of nanofillers in the selective layer, naming as thin film nanocomposite (TFN) [15,16], has been widely used to improve the separation efficiency of membrane. Metal organic frameworks (MOFs), a class of porous materials, has been explored for preparation of TFN membrane [17,18]. Compared with other nanofillers, such as silica, TiO2 and ZnO, the porous MOFs provide additional channels (pore volume of MOFs) for the passage of water molecules, reduce the mass transfer resistance, and thus increase the flux. Among MOF materials, zeolitic imidazolate framework-8 (ZIF-8) seems to be suitable for preparing TFN membrane due to its excellent water stability and welldefined sub-nanometer pores [19]. However, there are two challenges should be addressed to fabricate TNF membranes with ZIF-8, namely (1) dispersing ZIF-8 in polyamide polymer matrix uniformly without agglomeration; (2) improving the compatibility and interactions of ZIF8 with polymer matrix. Recently, polydopamine (PDA) has been used to
Corresponding author. E-mail address: [email protected] (C. He).
https://doi.org/10.1016/j.jhazmat.2018.12.096 Received 9 September 2018; Received in revised form 9 December 2018; Accepted 24 December 2018 Available online 26 December 2018 0304-3894/ © 2018 Elsevier B.V. All rights reserved.
Journal of Hazardous Materials 367 (2019) 339–347
M. Qiu, C. He
Fig. 1. Fabrication process of the ZIF-8@PDA modified TFN membrane.
2.2. Synthesis of nanofillers
modify ZIF-8 to improve the affinity between nanofillers and the membrane matrix for the gas separation [20,21]. The hydrophilic PDA layer can strongly adhere on ZIF-8 surface to enhance the dispersion of ZIF-8 in water and its porous structure would not block the pores of ZIF8 [22,23]. In addition, PDA layer can react with polymer that containing amine groups [24], which provide a feasible way to introduce this nanofillers in the aqueous solution during the interfacial polymerization. Nevertheless, there is few studies focusing on the preparation of FO membrane with a NF-like and nano-functional selective layer for removal of heavy metal ions. Here, a novel TFN FO membrane was developed for efficiency heavy metal ions removal based on a ZIF-8@PDA incorporated selective layer. As illustrated in Fig. 1, the substrate surface was firstly contacted with poly(ethyleneimine) (PEI) aqueous solution containing nanofillers with different concentration, followed by contacting with 1,3,5-benzenetricarboxylic acid chloride (TMC) to take place the interfacial polymerization. Finally, PEI was introduced on the membrane surface to produce a positively charged surface. The introduction of ZIF-8@PDA enhances the water permeability without sacrificing the selectivity. Moreover, the modified membrane shows high separation efficiency for removing heavy metal ions. This study may offer a convenient route to improve FO performance by designing the selective layer of membrane.
ZIF-8 was synthesized by following steps: 0.05 mol/L Zn (NO3)2·6H2O methanol solution (50.0 mL) was poured into 0.4 mol/L 2methylimidazole methanol solution (50.0 mL) with constant stirring (200 rpm). After stirring for 60 min, the crude product was obtained by centrifugation (10,000 rpm, 15 min). Then the nanofillers were purified with dispersion-centrifugation in methanol three circles, dried at 40 °C for 12 h. The dopamine modified ZIF-8 (ZIF-8@PDA) was prepared by following steps [21]: ZIF-8 (0.1 g) were dispersed in a mixed solution of methanol (50.0 mL) and Tris−HCl buffer solution (0.1 M, pH = 8.5, 50.0 mL). After adding the dopamine (0.4 g), the mixture solution was stirred at 200 rpm for 1 h. The crude product was obtained by centrifugation (10,000 rpm, 15 min). Then the nanofillers were purified with dispersion-centrifugation in methanol three circles, dried at 40 °C for 12 h. 2.3. Membrane preparation Firstly, the dense side of substrate was treated in PEI aqueous solution (2.0 wt%) containing different ZIF-8@PDA weight rates for 3 min. Secondly, the dense side of membrane was contacted with TMC hexane solution (0.1 wt%) for 1 min. After removing the unreacted TMC by washing with hexane, the dense skin side was contacted with PEI aqueous solution (0.1 wt%) for 1 min. Finally, the membrane was taken a 60 °C heat treatment for 10 min. The prepared TFN membranes with different ZIF-8@PDA loading (0, 0.025, 0.05, 0.1 wt%) are abbreviated into M0, M1, M2 and M3 membrane, respectively, and the membrane with pristine ZIF-8 loading (0.1 wt%) is denoted as M4.
2. Experimental section 2.1. Materials Polyethersulfone ultrafiltration membrane was obtained from RisingSun membrane technology CO., Ltd. The molecular weight cut off of this membrane is 50,000 Da and water flux is 260.0 LMH (3.5 bar). Hexane, methanol, MgCl2, CuSO4.5H2O, Pb(NO3)2, NiCl2 and Zn (NO3)2.6H2O were obtained from Shanghai chemical reagents medicine group Co., Ltd. Polyethylenimine (PEI, Mw = 10,000) was provided by Shanghai Gobekie New Materials Technology Co., Ltd. Trimesoylchloride (TMC), 2-Methylimidazole, dopamine and tris(hydroxymethyl)aminomethane were obtained from Sigma-Aldrich. Deionized (DI) water was used during the whole process of the experiments.
2.4. Characterizations of nanofillers and membranes Nanofillers were characterized by Fourier transform infrared spectroscopy (FTIR), X-ray diffraction (XRD), particle size distribution, field emission scanning electron microscope (FESEM) and transmission electron microscopy (TEM). Equipment information and measurement procedures were described in the Supporting Information. Attenuated total reflectance Fourier transform infrared spectroscopy 340
Journal of Hazardous Materials 367 (2019) 339–347
M. Qiu, C. He
Fig. 2. (a) FTIR spectra, (b) XRD patterns, and (c) TGA plots of nanofillers; (d) Photo images of nanofillers aqueous suspension at 0 h and 12 h standing.
(DDS-11 A, Leichi, China). PEG concentrations were determined by measuring absorbance at 535 nm after iodine complexation with a UV–Vis spectrophotometer (UV 1800, Shimadzu, Japan). The pore size r (nm) was determined based on the Mw value as shown in Eq. (3).
(ATR-FTIR) of membrane was recorded with a Nicolet 8700 instrument with the wavenumber ranging from 4000 cm−1 to 400 cm−1. X-ray photoelectron spectroscopy (XPS) of membrane surface was collected by a Thermo ESCALAB 250 instrument with a monochromatic Al Kα source (15 kV, 5 mA). The electrical property of membrane surface was collected using an electrokinetic analyzer (SurPASS 3, Anton Paar, Austria) with KCl solution (1.0 mM) as the electrolyte solution. The water contact angle of membrane was performed on a Dataphysics OCA40 micro instrument. The value was recorded every 30 s during a time of 180 s. The membrane intrinsic transport properties (water permeability, solute rejection and salt permeability) were recorded by a self-made cross-flow reverse osmosis device. The membrane with filtration area (S) of 12.5 cm2 was firstly operated at 5.0 bar pressure for 15 min to compact the membrane. Then the pressure (ΔP) was adjusted to 2.0 bar, and collected the passed water (J) every 5 min (t). The water permeability (A, L m−2 h-1 bar-1) was calculated by the Eq. (1).
A=
r = 16.73 × 10−12 × (MWCO)0.557
The salt permeability (B, L m−2 h-1) was determined based on the A and R value as shown in Eq. (4).
1−R B = R A (Δp − Δπ )
(1)
Cf − Cp Cf
× 100%
ΔV 1 Δt Am
(5)
where ΔV (L) is the increased volume of draw solution that passing through the membrane with area Am (m2) during a period of time Δt (h). Reverse salt flux (Js, gMH) was calculated using Eq. (6).
10.0 mM of MgCl2 solution was used as the feed solution to evaluate the rejection performance of the membrane. PEG molecules with different molecular weights (200, 400, 600 and 1000 Da) were used to measure the molecular weight cut-off (MWCO) of the membrane. The MWCO was taken the value of PEG rejection with 90.0%. The feed and permeate concentrations (Cf and Cp, respectively) of salt or PEG were recorded to calculate the solute rejection (R) by the Eq. (2).
R=
(4)
where Δπ is the osmosis pressure difference. The FO performance of as-prepared membrane was measured by a self-made device with an 8.5 cm/s crossflow velocity. DI water and MgCl2 solution (1.0 M) were chosen as the feed and draw solution, respectively. Water flux (Jw, LMH) was calculated using Eq. (5).
Jw =
J S⋅t⋅ΔP
(3)
Js =
Δ (Ct Vt ) ΔtAm
(6) −1
where Ct (g L ) is the MgCl2 concentration at the t time (h) of feed solution, Vt is the decreased volume of feed solution during the t time. The heavy metal ion rejection of prepared membrane was investigated under FO mode. 1.0 L of heavy metal solution (Cu2+, Ni2+, Pb2+) with different concentration was chosen as the feed solution, and the solution pH was adjusted to 6.5 by NaOH (0.1 M), while 1.0 L of
(2)
The MgCl2 concentration was detected using a conductivity meter 341
Journal of Hazardous Materials 367 (2019) 339–347
M. Qiu, C. He
Fig. 3. (a) ATR-FTIR spectra of prepared membranes; (b) XPS survey spectra and (c) Zn 2p core level spectra of M0 and M1 membrane; (d) N 1s core level spectra of M1 membrane.
diameter of about 75.0 nm (Fig. S1(c)). FTIR spectrometry was used to evaluate the synthesis of ZIF-8@PDA. The FTIR spectrum of ZIF-8 is in agreement with the previous literature [25]. For the ZIF-8@PDA, new peaks at approximately 3400–3500, 1630, 1510 and 1270 cm−1 are associated with the OH/NH, indole, indoline and phenolic groups of PDA, respectively [24,26], indicating the successfully modification of PDA on the surface of nanofillers. As shown in Fig. S2 (a), the zeta potential of ZIF-8 varies from positive to negative after modifying, further indicating the existence of PDA on the surface of ZIF-8. TGA results of ZIF-8, PDA and ZIF-8@PDA are shown in Fig. 2 (b). TGA plot shows that no well-defined decomposition temperature can be found for PDA. And PDA has multi-step degradation behavior due to its heterogeneous structure and the presence of different chemical groups [27]. For the ZIF-8@PDA, in addition to the decomposition of pure ZIF-8 phase from 450 °C, the multi-stage over the temperature from 200 to 450 °C should be assigned to the decomposition of PDA. As shown in Fig. 2 (c), the diffraction peaks of ZIF-8 are all found in the XRD patterns of ZIF-8@PDA, suggesting that the modification does not affect the crystal structure of ZIF-8. The dispersion ability of ZIF-8@PDA in water is evaluated and the photo images are shown in Fig. 2(d). The dried ZIF-8 would precipitate quickly when dispersed in water due to its intrinsic hydrophobicity and the formation of strong covalent Zn-methylimidazole-Zn links between different nanoparticles [28]. In contrast, ZIF-8@PDA can be dispersed stably in water for long time due to the improved hydrophilicity of ZIF-8 (Fig.S2 (b)). Although the ZIF-8@ PDA has aggregation after 12 h standing (Fig. S3), it still has a better dispersion stability than the pristine ZIF-8. The good dispersibility would beneficial to the uniformly loading of nanofillers on the
MgCl2 solution with different concentration was chosen as the draw solution. The heavy metal ion rejection Rm was calculated using Eq. (7).
C × Vd ⎞ Rm = ⎜⎛1 − d ⎟ × 100% Cf × Vp ⎠ ⎝
(7)
where Cd (ppm) and Vd (L) are the heavy metal ion concentration and volume of draw solution at last, respectively. Cf (ppm) and Vp (L) are the heavy metal ion concentration and decreased volume of feed solution. The heavy metal concentration was collected by an inductively coupled plasma emission spectrometer (ICP, Prodigy, Leeman, US). The absorption experiment was used to illustrate the chelation reaction between heavy metal ions and PEI. 0.1 g of PES and prepared FO membranes were immersed into a 300.0 ppm of heavy metal solution, respectively. The solution was placed in a water bath oscillator for 24 h. In addition, the membrane was conducted with an X-ray energy dispersive spectrometer (EDS, Oxford) to determine the element on the membrane surface. 3. Results and discussion 3.1. Characterization of nanofillers In this study, ZIF-8@PDA was obtained from ZIF-8 by surface coating PDA. The obvious polyhedral shape of ZIF-8 can be seen from the TEM images (Fig. S1(a)). After PDA coating, the structure of nanoparticles become more random. The average diameter of the ZIF-8 is measured to be about 65.0 nm, while the ZIF-8@PDA has a mean 342
Journal of Hazardous Materials 367 (2019) 339–347
M. Qiu, C. He
Fig. 4. The FESEM surface images and cross-sectional of (a) M0, (b) M1, (c) M2 and (d) M3 membrane. The “1” refers to the surface images with 10.0 K magnification; “2” refers to the surface images with 50.0 K magnification; “3” refers to the cross-sectional images with 50.0 K magnification. The existence of nanoparticles from the cross-sectional images was marked with red arrows (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article).
Fig. 5. AFM images of prepared membranes: (a) M0, (b) M1, (c) M2, and (d) M3.
343
Journal of Hazardous Materials 367 (2019) 339–347
M. Qiu, C. He
Table 1 The root mean square roughness (Rrms) and average roughness (Ra) of prepared membranes. Membrane
Rrms (nm)
Ra (nm)
M0 M1 M2 M3
13.0 15.6 16.4 21.8
10.9 12.5 13.5 17.8
membrane surface. 3.2. Characterization of TFN membranes Fig. 3(a) presents the ATR-FTIR spectra of prepared membranes. Compared with substrate, peak at 1660 cm−1 in the spectra of M0 and M1 membrane corresponds to C]O stretching vibration of amide group (Fig. 3(a)), evincing the formation of polyamide layer through the interfacial polymerization [13]. The broad peak at about 3450 cm−1 is ascribed to the unreacted amine group of PEI, which endows the membranes with a positively charged surface. Meanwhile, the peak intensity at 3450 cm−1 for M1 is apparently higher than that of M0, which can be attributed to the phenolic hydroxyl and amine group of PDA. Fig. 3(b) illustrates the XPS survey spectra of M0 and M1 membrane. Compared with M0 membrane, peaks of Zn (Zn 2p3 at 1021 eV and Zn 2p1 at 1044 eV) appear in the spectrum of M1 membrane (Fig. 3(c)) [29], which further confirms the existence of nanofillers. The high-solution N 1 s XPS spectrum of M1 membrane is shown in Fig. 3(d). Three peaks with binding energy at 398.6, 399.5 and 401.4 eV are associated with C-NH2, NeC = O, and C-N+H3 species [30], respectively, demonstrating the membrane with a positively charged surface. The surface and cross-sectional morphologies of prepared membranes were observed by FESEM. As Fig. 4 (a1-d1) shows, the M0 membrane exhibits a relatively smooth surface, while the M1, M2 and M3 membranes show a lot of grain-like structure on the membrane surface. These grain-like structure should be attributed to the loading of ZIF-8@PDA on the membrane surface. The number of grain-like structure increases with the increasing ZIF-8/PDA loading. In addition, the TFN membranes exhibit irregular wrinkle-shaped structure, especially in M3 membrane, which making the membrane surface rougher (Fig. 4 (a2-d2)). The ZIF-8@PDA provide additional amino group or phenolic hydroxyl group to react with acid chloride group, thus enhancing the cross-linking rate of polyamide layer. Moreover, when the ZIF-8@PDA concentration increases to 0.1 wt%, the aggregation of
Fig. 7. MWCOs of prepared membranes. Table 2 Transport property of as-prepared membranes. Membrane
Water permeabilitya A (L/m2 h bar)
MgCl2 rejectionb R (%)
MgCl2 permeability B (L/m2 h)
B/A (KPa)
M0 M1 M2 M3 M4
2.42 4.88 6.03 8.84 9.46
97.4 97.8 96.9 92.3 80.1
0.09 0.16 0.27 1.10 3.32
3.8 ± 0.5 3.3 ± 0.4 4.5 ± 0.8 12.5 ± 1.6 35.1 ± 3.8
a b
± ± ± ± ±
0.12 0.16 0.15 0.18 0.21
± ± ± ± ±
1.8 0.9 1.3 2.5 4.2
± ± ± ± ±
0.03 0.02 0.06 0.12 0.15
DI water was used as feed solution with 2.0 bar pressure. 10.0 mM of MgCl2 as feed solution with 2.0 bar pressure.
nanoparticles further affect the surface morphology. The cross-sectional SEM images are presented in Fig. 4 (a3-d3). It can be found that nanoparticles are dispersed on the membrane surface evenly for M1 and M2 membrane, while ZIF-8@PDA aggregates are found on the membrane surface for M3 membrane. The membrane surface roughness of prepared membranes was performed by AFM. As shown in Fig. 5, the grain-like structure and the ZIF-8@PDA aggregate on the membrane surface are also well-reflected in AFM images. Table 1 shows the extracted data for root mean square roughness (Rrms) and average roughness (Ra) of prepared membranes. The M0 membrane has root mean square roughness (Rrms) and average
Fig. 6. (a) Measured water contact angles and (b) zeta potential data of prepared membranes. 344
Journal of Hazardous Materials 367 (2019) 339–347
M. Qiu, C. He
Table 3 Performance comparison between the FO membrane with a NF-like selective layer synthesized in this work and other works. Membrane
A (L/m2 h bar)
MgCl2 Rejection (%)
B/A (KPa)
Operating conditions
Ref.
M2
5.95
95.8
3.6
This work
6-PAH/PSS
7.3
80
15
PAI/PEI
4.58
92
–
PVA
1.93
93.5
32
Feed: 5.0 mM MgCl2 Pressure: 1.0 bar Feed: 5.0 mM MgCl2 Pressure: 1.0 bar Feed: 500.0 ppm MgCl2 Pressure: 1.0 bar Feed: 5.0 mM MgCl2 Pressure: 5.0 bar
[37] [38] [39]
Fig. 8. (a) Water flux (Jw), reverse salt flux (Js) and (b) Js/Jw ratios of as-prepared membranes (DI water and 1.0 M MgCl2 were used as feed and draw solution).
molecules by hydrogen bond interactions. In order to reject the heavy metal ions, the nanofiltration-like FO membrane should possess a positively charged surface. The electrical property of membrane surface was determined with zeta potential and the data is shown in Fig. 6(b). All the membranes are found to have a positively charged surface at pH = 6.5, which is attributed to the unreacted primary amine of PEI [24,31,32]. After introduction of nanoparticles, the zeta potential values show a slightly decrease due to the existence of negatively charged nanoparticles (Fig. S2(a)). Overall, these prepared membranes would be suitable for the heavy metal ion removal under FO mode. The molecular weight cut-offs (MWCOs) of the prepared membranes were evaluated by the rejection of polyethylene glycol (PEG) with different molecular weight (200, 400, 600, and 1000 Da). The MWCOs of membranes are taken as the molecular weight of the PEG molecule with 90.0% rejection. As shown in the Fig.7, the MWCOs are 284, 317, 338 and 435 Da for M0, M1, M2 and M3 membrane, respectively. After introduction of nanoparticles, the MWCOs increases from 284 to 435 Da, suggesting that the selective layer becomes loose. This is because the nanoparticles may prevent polymer chains from packing closely together. Interestingly, the increment in MWCOs of M3 membrane are bigger than that of M1 and M2 membrane, which is caused by the aggregation of nanoparticles on the membrane surface. Table 2 shows the intrinsic transport properties of prepared membranes. Note that all TFN membranes have higher water permeability than that of M0 membrane, and the values increase with the increasing loading nanoparticles. The improvement of water permeability should be attributed to the following main factors. Firstly, the good compatibility between ZIF-8@PDA and polyamide layer avoids the formation of interface defects. The ideal generated interface voids between entrapped ZIF-8@PDA nanoparticles and polyamide potentially constitute continuous channels for transporting water molecules [29]. Meanwhile, the porous structure of nanoparticles also provides additional paths for water molecules for water molecules transporting [33]. Secondly, the PDA layer can absorb water molecules due to the strong hydrogen
Fig. 9. Measured water flux and rejection of three heavy metal ions in FO mode (1000 ppm heavy metal salt solution and 1.0 M MgCl2 were used as feed and draw solution, pH = 6.5, 25℃).
roughness (Ra) of 13.0 nm and Ra (average roughness) of 10.9 nm. After the introduction of nanoparticles, the surface morphology of membranes become much rougher with a Rrms value ranging from 15.6 nm to 21.8 nm and a Ra value ranging from 12.5 nm to 17.8 nm. Although the improved roughness of membrane surface may potentially increase the fouling propensity, the enhanced surface area is beneficial for a higher water flux. The hydrophilicity of prepared membranes was measured by the water contact angle (WCAs) and the results are shown in Fig. 6(a). The WCAs of prepared membranes at 0 s are almost same, suggesting that all the membrane surfaces are covered by the PEI layer. Time-dependent of WCAs measurement demonstrate that the TFN membranes have a better wettability than the pristine TFC membrane (Table S1). This result can be attributed to the PDA layer which adsorb the water 345
Journal of Hazardous Materials 367 (2019) 339–347
M. Qiu, C. He
Fig. 10. Schematic image of permeation and rejection mechanism of the prepared FO membranes.
increases to 0.1 wt%, the water flux of M3 membrane begins to decrease. This is caused by the higher salt reverse flux (Fig. 8(a)), which results in a severe ICP and a reduced water flux. In addition, M3 membrane show a higher Js/Jw ratio than that of M1 and M2 membrane (Fig. 8(b)), suggesting that M3 membrane has a poor separation efficiency. Therefore, this result indicates that M2 membrane is suitable for potential application. Fig. 9 shows the result of using M2 membrane to treat the heavy metal wastewater (Cu2+, Ni2+, Pb2+). The water flux of M2 membrane has a decrease when using 1000 ppm heavy metal salt solution as the feed solution. This is because the heavy metal solution increases the feed osmotic pressure, which decrease the driving force across the membrane, hence results in the reduced water flux. Moreover, M2 membrane shows a high rejection to the heavy metal ions in FO mode (99.1% for Cu2+, 98.3% for Ni2+ and 97.7% for Pb2+). Meanwhile, the effect of draw solution and feed solution concentration on the water flux and ions rejection of M2 membrane are presented in Fig. S4. The water flux increases with the draw solution concentration since the increased driving force. In addition, the Cu2+ rejection shows a slightly increasing trend. This is because the Cu2+ concentration across the membrane almost has no changing while the water flux increases, hence reaches a higher rejection. The water flux and Cu2+ rejection all decrease with the feed solution concentration. The Cu2+ flux would increase proportional with the feed solution concentration, while the Cu2+ rejection only has a slightly decrease. This is because the increased feed solution concentration reduces the driving force across the membrane for water and ions transporting. The rejection mechanism of the M2 membrane is presented in Fig. 10. Firstly, the M2 membrane with a MWCO of 338 Da (0.43 nm) can reject the heavy metal ions that have a hydrated radius around 0.4 nm (Table S2) by the size exclusion effect. Besides, the rejection shows a trend of Cu2+ > Ni2+ > Pb2+, which is consistent with the hydrated radius of these ions. Secondly, M2 membrane has a positive charged surface (Fig. 6(b)), which shows an electrostatic repulsion to the heavy metal ions. Thirdly, the residual amino group of PEI can adsorb the metal ions by the chelation reaction (Fig. S5), which is also beneficial to get a high rejection. However, only 6.2% of Cu2+, 6.9% of
bonds and facilitate water transport through the mentioned additional channels [34] (Fig. S2(b) and Table S1). Thirdly, the rougher surface is also beneficial for a higher water flux since an enlargement of the eff ;ective membrane surface area [35] (Fig.5 and Table 1). Last but not least, the pore size of the prepared membranes increases with the increasing nanoparticles concentration, ranged from 0.39 to 0.48 nm (Fig. 6), leading to an increment in permeation of water molecules. The MgCl2 rejection has a decrease with increasing nanoparticles concentration. The salt rejection mainly depends on the Donnan exclusion eff ;ect and size exclusion eff ;ect [36]. The Donnan exclusion eff ;ect is mostly responsible for the repulsion of divalent cations. The strong electrostatic repulsion between Mg2+ and the positively charged surface of prepared membranes, leading to a high repulsion of Mg2+. After introducing of the nanoparticles, the pore size of membranes increases, which causes the decrease of rejection. The B/A value was used to assess the separation efficiency of the FO membrane. Compared with the M1 and M2 membrane, M3 membrane shows a much higher B/A value than that of M0 membrane, which is caused by the decrease of salt rejection. Table 3 shows the water permeability and MgCl2 rejection of the membrane prepared in this work and in other works. All the membranes have a nanofiltration-like selective layer. In order to unify the operation condition, we have added the filtration experiment of M2 membrane using 5.0 mM of MgCl2 as the feed solution under 1.0 bar pressure. Our membrane shows a higher water permeability and salt rejection than other membranes. Although the membrane prepared by layer by layer method has the highest water permeability, it shows a low salt rejection, which would loss the draw solute during FO process, hence causes the serious ICP problem. This result suggests that incorporation of ZIF@PDA in selective layer, the water permeability of membrane can be improved without sacrificing the selectivity. Fig. 8(a) shows the water flux (Jw) and salt reverse flux (Js) of prepared membranes in FO mode. It can be seen that M1 and M2 membrane show a higher water flux than that of M0 membrane, which is consistent with the result of water permeability (A). The improvement of water flux is attributed to the relatively loose nano-functional selective layer. However, when the concentration of nanoparticles 346
Journal of Hazardous Materials 367 (2019) 339–347
M. Qiu, C. He
Ni2+ and 8.1% of Pb2+ were adsorbed by M2 membrane after 24 h, respectively, indicating that the heavy metal ions removal mainly relies on rejection.
performance: a review, Chem. Rev. 113 (2013) 4980–5028. [15] N. Song, X. Gao, Z. Ma, X. Wang, et al., A review of graphene-based separation membrane: materials, characteristics, preparation and applications, Desalination 437 (2018) 59–72. [16] W.J. Lau, S. Gray, T. Matsuura, D. Emadzadeh, et al., A review on polyamide thin film nanocomposite (TFN) membranes: history, applications, challenges and approaches, Water Res. 80 (2015) 306–324. [17] A. Zirehpour, A. Rahimpour, M. Ulbricht, Nano-sized metal organic framework to improve the structural properties and desalination performance of thin film composite forward osmosis membrane, J. Membr. Sci. 531 (2017) 59–67. [18] D. Ma, S.B. Peh, G. Han, S.B. Chen, Thin-film nanocomposite (TFN) membranes incorporated with super-hydrophilic metal–organic framework (MOF) UiO-66: toward enhancement of water flux and salt rejection, ACS Appl. Mater.Inter. 9 (2017) 7523–7534. [19] J. Duan, Y. Pan, F. Pacheco, E. Litwiller, et al., High-performance polyamide thinfilm-nanocomposite reverse osmosis membranes containing hydrophobic zeolitic imidazolate framework-8, J. Membr. Sci. 476 (2015) 303–310. [20] L. Dong, M. Chen, X. Wu, D. Shi, et al., Multi-functional polydopamine coating: simultaneous enhancement of interfacial adhesion and CO2 separation performance of mixed matrix membranes, New J. Chem. 40 (2016) 9148–9159. [21] Z. Wang, D. Wang, S. Zhang, L. Hu, et al., Interfacial design of mixed matrix membranes for improved gas separation performance, Adv. Mater. 28 (2016) 3399–3405. [22] Y. Liang, J. Wei, Y.X. Hu, X.F. Chen, et al., Metal-polydopamine frameworks and their transformation to hollow metal/N-doped carbon particles, Nanoscale 9 (2017) 5323–5328. [23] J.-C. Yang, Y. Chen, Y.-H. Li, X.-B. Yin, Magnetic resonance imaging-guided multidrug chemotherapy and photothermal synergistic therapy with pH and NIRStimulation release, ACS Appl. Mater. Inter. 9 (2017) 22278–22288. [24] F.Y. Zhao, Y.L. Ji, X.D. Weng, Y.F. Mi, et al., High-flux positively charged nanocomposite nanofiltration membranes filled with poly(dopamine) modified multiwall carbon nanotubes, ACS Appl. Mater. Inter. 8 (2016) 6693–6700. [25] J. Im, N. Yim, J. Kim, T. Vogt, et al., High-pressure chemistry of a zeolitic imidazolate framework compound in the presence of different fluids, J. Am. Chem. Soc. 138 (2016) 11477–11480. [26] Y. Wang, C. Hou, Y. Zhang, F. He, et al., Preparation of graphene nano-sheet bonded PDA/MOF microcapsules with immobilized glucose oxidase as a mimetic multienzyme system for electrochemical sensing of glucose, J. Mater. Chem. B Mater. Biol. Med. 4 (2016) 3695–3702. [27] S. Xiong, Y. Wang, J. Yu, L. Chen, et al., Polydopamine particles for next-generation multifunctional biocomposites, J. Mater. Chem. A Mater. Energy Sustain. 2 (2014) 7578–7587. [28] J. Cravillon, S. Münzer, S.-J. Lohmeier, A. Feldhoff, et al., Rapid room-temperature synthesis and characterization of nanocrystals of a prototypical zeolitic imidazolate framework, Chem. Mater. 21 (2009) 1410–1412. [29] J. Zhu, L. Qin, A. Uliana, J. Hou, et al., Elevated performance of thin film nanocomposite membranes enabled by modified hydrophilic MOFs for nanofiltration, ACS Appl. Mater. Inter. 9 (2017) 1975–1986. [30] L. Xu, J. Xu, B. Shan, X. Wang, et al., Novel thin-film composite membranes via manipulating the synergistic interaction of dopamine and m-phenylenediamine for highly efficient forward osmosis desalination, J. Mater. Chem. A Mater. Energy Sustain. 5 (2017) 7920–7932. [31] T. Ma, Y. Su, Y. Li, R. Zhang, et al., Fabrication of electro-neutral nanofiltration membranes at neutral pH with antifouling surface via interfacial polymerization from a novel zwitterionic amine monomer, J. Membr. Sci. 503 (2016) 101–109. [32] D. Wu, S. Yu, D. Lawless, X. Feng, Thin film composite nanofiltration membranes fabricated from polymeric amine polyethylenimine imbedded with monomeric amine piperazine for enhanced salt separations, React. Funct. Polym. 86 (2015) 168–183. [33] H. Sun, B. Tang, P. Wu, Hydrophilic hollow zeolitic imidazolate framework-8 modified ultrafiltration membranes with significantly enhanced water separation properties, J. Membr. Sci. 551 (2018) 283–293. [34] L. Bai, Y. Liu, N. Bossa, A. Ding, et al., Incorporation of cellulose nanocrystals (CNCs) into the polyamide layer of thin-film composite (TFC) nanofiltration membranes for enhanced separation performance and antifouling properties, Environ. Sci. Technol. 52 (2018) 11178–11187. [35] M. Qiu, C. He, Novel zwitterion-silver nanocomposite modified thin-film composite forward osmosis membrane with simultaneous improved water flux and biofouling resistance property, Appl. Surf. Sci. 455 (2018) 492–501. [36] Y. Chen, C. He, High salt permeation nanofiltration membranes based on NMGassisted polydopamine coating for dye/salt fractionation, Desalination 413 (2017) 29–39. [37] Q. Saren, C.Q. Qiu, C.Y. Tang, Synthesis and characterization of novel forward osmosis membranes based on layer-by-Layer assembly, Environ. Sci. Technol. 45 (2011) 5201–5208. [38] C. Qiu, L. Setiawan, R. Wang, C.Y. Tang, et al., High performance flat sheet forward osmosis membrane with an NF-like selective layer on a woven fabric embedded substrate, Desalination 287 (2012) 266–270. [39] D. Qin, Z. Liu, Z. Liu, H. Bai, et al., Superior antifouling capability of hydrogel forward osmosis membrane for treating wastewaters with high concentration of organic foulants, Environ. Sci. Technol. 52 (2018) 1421–1428.
4. Conclusions In summary, a novel TFN FO membrane had been developed via in situ interfacial polymerization for treating heavy metal wastewater. An optimized selective layer was designed by incorporating porous PDA/ ZIF-8 into a PEI/TMC crosslinked network to improve the separation efficiency. The introduction of nanoparticles endows the selective layer with good wettability and loose structure. More importantly, the modified PDA layer improve the affinity between the nanofillers and polyamide matrix, hence maintaining the selectivity of membrane. Furthermore, this membrane exhibits high rejection to heavy metal ions (> 96%) in FO mode. The outstanding performance of these membranes highlight its potential application of TFN FO membranes for heavy metal ion removal. Acknowledgments This work is supported by grants from the Innovation Funds for PhD Students of Donghua University (CUSF-DH-D-2017031) and the National Science Foundation of China (No.51873034). Appendix A. Supplementary data Supplementary material related to this article can be found, in the online version, at doi:https://doi.org/10.1016/j.jhazmat.2018.12.096. References [1] M.A. Alaei Shahmirzadi, S.S. Hosseini, J. Luo, I. Ortiz, Significance, evolution and recent advances in adsorption technology, materials and processes for desalination, water softening and salt removal, J. Environ. Manage. 215 (2018) 324–344. [2] Y. Lv, Y. Du, Z.-X. Chen, W.-Z. Qiu, et al., Nanocomposite membranes of polydopamine/electropositive nanoparticles/polyethyleneimine for nanofiltration, J. Membr. Sci. 545 (2018) 99–106. [3] L. Pino, C. Vargas, A. Schwarz, R. Borquez, Influence of operating conditions on the removal of metals and sulfate from copper acid mine drainage by nanofiltration, Chem. Eng. J. 345 (2018) 114–125. [4] J.E. Efome, D. Rana, T. Matsuura, C.Q. Lan, Insight studies on metal-organic framework nanofibrous membrane adsorption and activation for heavy metal ions removal from aqueous solution, ACS Appl. Mater.Inter. 10 (2018) 18619–18629. [5] M. Chen, K. Shafer-Peltier, S.J. Randtke, E. Peltier, Competitive association of cations with poly(sodium 4-styrenesulfonate) (PSS) and heavy metal removal from water by PSS-assisted ultrafiltration, Chem. Eng. J. 344 (2018) 155–164. [6] P. Roy Choudhury, S. Majumdar, G.C. Sahoo, S. Saha, et al., High pressure ultrafiltration CuO/hydroxyethyl cellulose composite ceramic membrane for separation of Cr (VI) and Pb (II) from contaminated water, Chem. Eng. J. 336 (2018) 570–578. [7] M. Qiu, J. Wang, C. He, A stable and hydrophilic substrate for thin-film composite forward osmosis membrane revealed by in-situ cross-linked polymerization, Desalination 433 (2018) 1–9. [8] S. Adham, A. Hussain, J. Minier-Matar, A. Janson, et al., Membrane applications and opportunities for water management in the oil & gas industry, Desalination 440 (2018) 2–17. [9] Q. Chen, W. Xu, Q. Ge, Novel multicharge hydroacid complexes that effectively remove heavy metal ions from water in forward osmosis processes, Environ. Sci. Technol. 52 (2018) 4464–4471. [10] X. Zhao, C. Liu, Efficient removal of heavy metal ions based on the optimized dissolution-diffusion-flow forward osmosis process, Chem. Eng. J. 334 (2018) 1128–1134. [11] J.M. Gohil, P. Ray, A review on semi-aromatic polyamide TFC membranes prepared by interfacial polymerization: potential for water treatment and desalination, Sep. Purif. Technol. 181 (2017) 159–182. [12] Q. Shi, L. Ni, Y. Zhang, X. Feng, et al., Poly(p-phenylene terephthamide) embedded in a polysulfone as the substrate for improving compaction resistance and adhesion of a thin film composite polyamide membrane, J. Mater. Chem. A Mater. Energy Sustain. 5 (2017) 13610–13624. [13] L. Yang, Z. Wang, J. Zhang, Highly permeable zeolite imidazolate framework composite membranes fabricated via a chelation-assisted interfacial reaction, J. Mater. Chem. A Mater. Energy Sustain. 5 (2017) 15342–15355. [14] H. Vinh-Thang, S. Kaliaguine, Predictive models for mixed-matrix membrane
347
Contents lists available at ScienceDirect
Journal of Hazardous Materials journal homepage: www.elsevier.com/locate/jhazmat
Efficient removal of heavy metal ions by forward osmosis membrane with a polydopamine modified zeolitic imidazolate framework incorporated selective layer
T
⁎
Ming Qiu, Chunju He
State Key Laboratory for Modification of Chemical Fibers and Polymer Materials, College of Materials Science and Engineering, 2999 North Renmin Road, Songjiang District, Donghua University, Shanghai 201620, China
A R T I C LE I N FO
A B S T R A C T
Keywords: Forward osmosis Selective layer Polydopamine ZIF-8 Heavy metal ions
A novel thin film nanocomposite (TFN) forward osmosis (FO) membrane with a positively charged and nanofunctional selective layer has been developed for effective heavy metal ions removal. The selective layer is constructed by penetrating the polydopamine modified zeolitic imidazolate framework (ZIF-8@PDA) in the poly (ethyleneimine)/1,3,5-benzenetricarboxylic acid chloride (PEI/TMC) crosslinked matrix. Compared with the pristine thin film composite (TFC) membrane, the thin film nanocomposite membrane (0.05 wt % nanofillers loading) exhibits a higher water flux (20.8 vs12.8 LMH) without losing of selectivity in terms of Js/Jw ratio (0.25 vs 0.20 g L−1) in FO mode. This improvement of the permeability is mainly attributed to the optimized selective layer with good wettability and loose structure. Besides, the modified PDA layer facilitates the affinity between the nanofillers and selective layer, which results in an ideal selectivity. In addition, this modified membrane shows a high heavy metal ion (Cu2+, and Ni2+ and Pb2+) rejection (> 96%) in FO mode. Our finding offers a simple and efficient method to enhance the FO performance of membrane by designing the selective layer for treating heavy metal wastewater.
1. Introduction Wastewater with heavy metal ions are produced by diverse industrial processes, such as mining, machinery and electronics, which is harmful to human health and ecological environment [1,2]. Therefore, it is necessary to remove heavy metal ions from the wastewater before discharging. Traditional pressure driven membrane technologies, such as ultrafiltration and nanofiltration have been increasingly developed as a promising approach for heavy metal ions removal [3–6]. However, these membrane technologies own some drawbacks, such as fouling problem and high energy consumption. In the recent years, forward osmosis (FO) technology receives lots of attention for water treatment [7,8], including heavy metal ion removal [9,10], since its low polluted tendency, reduced energy cost and environmental friendly. FO membrane currently mainly studied is thin film composite (TFC) membrane, which is prepared by forming a dense m-Phenylenediamine/1,3,5-benzenetricarboxylic acid chloride (MPD/ TMC) cross-linked layer on a substrate [11,12]. Although the dense selective layer endows the membrane with an excellent solute rejection, the cross-linked networks limits the penetration of water molecule, thus
⁎
reduces the separation efficiency [13,14]. Hence, FO membranes with nanofiltration-like selective layer have been developed in some application areas, such as treating wastewater containing heavy metal ions and organic contaminant. However, there is needing further exploration before the process is effective by these membranes. Introduction of nanofillers in the selective layer, naming as thin film nanocomposite (TFN) [15,16], has been widely used to improve the separation efficiency of membrane. Metal organic frameworks (MOFs), a class of porous materials, has been explored for preparation of TFN membrane [17,18]. Compared with other nanofillers, such as silica, TiO2 and ZnO, the porous MOFs provide additional channels (pore volume of MOFs) for the passage of water molecules, reduce the mass transfer resistance, and thus increase the flux. Among MOF materials, zeolitic imidazolate framework-8 (ZIF-8) seems to be suitable for preparing TFN membrane due to its excellent water stability and welldefined sub-nanometer pores [19]. However, there are two challenges should be addressed to fabricate TNF membranes with ZIF-8, namely (1) dispersing ZIF-8 in polyamide polymer matrix uniformly without agglomeration; (2) improving the compatibility and interactions of ZIF8 with polymer matrix. Recently, polydopamine (PDA) has been used to
Corresponding author. E-mail address: [email protected] (C. He).
https://doi.org/10.1016/j.jhazmat.2018.12.096 Received 9 September 2018; Received in revised form 9 December 2018; Accepted 24 December 2018 Available online 26 December 2018 0304-3894/ © 2018 Elsevier B.V. All rights reserved.
Journal of Hazardous Materials 367 (2019) 339–347
M. Qiu, C. He
Fig. 1. Fabrication process of the ZIF-8@PDA modified TFN membrane.
2.2. Synthesis of nanofillers
modify ZIF-8 to improve the affinity between nanofillers and the membrane matrix for the gas separation [20,21]. The hydrophilic PDA layer can strongly adhere on ZIF-8 surface to enhance the dispersion of ZIF-8 in water and its porous structure would not block the pores of ZIF8 [22,23]. In addition, PDA layer can react with polymer that containing amine groups [24], which provide a feasible way to introduce this nanofillers in the aqueous solution during the interfacial polymerization. Nevertheless, there is few studies focusing on the preparation of FO membrane with a NF-like and nano-functional selective layer for removal of heavy metal ions. Here, a novel TFN FO membrane was developed for efficiency heavy metal ions removal based on a ZIF-8@PDA incorporated selective layer. As illustrated in Fig. 1, the substrate surface was firstly contacted with poly(ethyleneimine) (PEI) aqueous solution containing nanofillers with different concentration, followed by contacting with 1,3,5-benzenetricarboxylic acid chloride (TMC) to take place the interfacial polymerization. Finally, PEI was introduced on the membrane surface to produce a positively charged surface. The introduction of ZIF-8@PDA enhances the water permeability without sacrificing the selectivity. Moreover, the modified membrane shows high separation efficiency for removing heavy metal ions. This study may offer a convenient route to improve FO performance by designing the selective layer of membrane.
ZIF-8 was synthesized by following steps: 0.05 mol/L Zn (NO3)2·6H2O methanol solution (50.0 mL) was poured into 0.4 mol/L 2methylimidazole methanol solution (50.0 mL) with constant stirring (200 rpm). After stirring for 60 min, the crude product was obtained by centrifugation (10,000 rpm, 15 min). Then the nanofillers were purified with dispersion-centrifugation in methanol three circles, dried at 40 °C for 12 h. The dopamine modified ZIF-8 (ZIF-8@PDA) was prepared by following steps [21]: ZIF-8 (0.1 g) were dispersed in a mixed solution of methanol (50.0 mL) and Tris−HCl buffer solution (0.1 M, pH = 8.5, 50.0 mL). After adding the dopamine (0.4 g), the mixture solution was stirred at 200 rpm for 1 h. The crude product was obtained by centrifugation (10,000 rpm, 15 min). Then the nanofillers were purified with dispersion-centrifugation in methanol three circles, dried at 40 °C for 12 h. 2.3. Membrane preparation Firstly, the dense side of substrate was treated in PEI aqueous solution (2.0 wt%) containing different ZIF-8@PDA weight rates for 3 min. Secondly, the dense side of membrane was contacted with TMC hexane solution (0.1 wt%) for 1 min. After removing the unreacted TMC by washing with hexane, the dense skin side was contacted with PEI aqueous solution (0.1 wt%) for 1 min. Finally, the membrane was taken a 60 °C heat treatment for 10 min. The prepared TFN membranes with different ZIF-8@PDA loading (0, 0.025, 0.05, 0.1 wt%) are abbreviated into M0, M1, M2 and M3 membrane, respectively, and the membrane with pristine ZIF-8 loading (0.1 wt%) is denoted as M4.
2. Experimental section 2.1. Materials Polyethersulfone ultrafiltration membrane was obtained from RisingSun membrane technology CO., Ltd. The molecular weight cut off of this membrane is 50,000 Da and water flux is 260.0 LMH (3.5 bar). Hexane, methanol, MgCl2, CuSO4.5H2O, Pb(NO3)2, NiCl2 and Zn (NO3)2.6H2O were obtained from Shanghai chemical reagents medicine group Co., Ltd. Polyethylenimine (PEI, Mw = 10,000) was provided by Shanghai Gobekie New Materials Technology Co., Ltd. Trimesoylchloride (TMC), 2-Methylimidazole, dopamine and tris(hydroxymethyl)aminomethane were obtained from Sigma-Aldrich. Deionized (DI) water was used during the whole process of the experiments.
2.4. Characterizations of nanofillers and membranes Nanofillers were characterized by Fourier transform infrared spectroscopy (FTIR), X-ray diffraction (XRD), particle size distribution, field emission scanning electron microscope (FESEM) and transmission electron microscopy (TEM). Equipment information and measurement procedures were described in the Supporting Information. Attenuated total reflectance Fourier transform infrared spectroscopy 340
Journal of Hazardous Materials 367 (2019) 339–347
M. Qiu, C. He
Fig. 2. (a) FTIR spectra, (b) XRD patterns, and (c) TGA plots of nanofillers; (d) Photo images of nanofillers aqueous suspension at 0 h and 12 h standing.
(DDS-11 A, Leichi, China). PEG concentrations were determined by measuring absorbance at 535 nm after iodine complexation with a UV–Vis spectrophotometer (UV 1800, Shimadzu, Japan). The pore size r (nm) was determined based on the Mw value as shown in Eq. (3).
(ATR-FTIR) of membrane was recorded with a Nicolet 8700 instrument with the wavenumber ranging from 4000 cm−1 to 400 cm−1. X-ray photoelectron spectroscopy (XPS) of membrane surface was collected by a Thermo ESCALAB 250 instrument with a monochromatic Al Kα source (15 kV, 5 mA). The electrical property of membrane surface was collected using an electrokinetic analyzer (SurPASS 3, Anton Paar, Austria) with KCl solution (1.0 mM) as the electrolyte solution. The water contact angle of membrane was performed on a Dataphysics OCA40 micro instrument. The value was recorded every 30 s during a time of 180 s. The membrane intrinsic transport properties (water permeability, solute rejection and salt permeability) were recorded by a self-made cross-flow reverse osmosis device. The membrane with filtration area (S) of 12.5 cm2 was firstly operated at 5.0 bar pressure for 15 min to compact the membrane. Then the pressure (ΔP) was adjusted to 2.0 bar, and collected the passed water (J) every 5 min (t). The water permeability (A, L m−2 h-1 bar-1) was calculated by the Eq. (1).
A=
r = 16.73 × 10−12 × (MWCO)0.557
The salt permeability (B, L m−2 h-1) was determined based on the A and R value as shown in Eq. (4).
1−R B = R A (Δp − Δπ )
(1)
Cf − Cp Cf
× 100%
ΔV 1 Δt Am
(5)
where ΔV (L) is the increased volume of draw solution that passing through the membrane with area Am (m2) during a period of time Δt (h). Reverse salt flux (Js, gMH) was calculated using Eq. (6).
10.0 mM of MgCl2 solution was used as the feed solution to evaluate the rejection performance of the membrane. PEG molecules with different molecular weights (200, 400, 600 and 1000 Da) were used to measure the molecular weight cut-off (MWCO) of the membrane. The MWCO was taken the value of PEG rejection with 90.0%. The feed and permeate concentrations (Cf and Cp, respectively) of salt or PEG were recorded to calculate the solute rejection (R) by the Eq. (2).
R=
(4)
where Δπ is the osmosis pressure difference. The FO performance of as-prepared membrane was measured by a self-made device with an 8.5 cm/s crossflow velocity. DI water and MgCl2 solution (1.0 M) were chosen as the feed and draw solution, respectively. Water flux (Jw, LMH) was calculated using Eq. (5).
Jw =
J S⋅t⋅ΔP
(3)
Js =
Δ (Ct Vt ) ΔtAm
(6) −1
where Ct (g L ) is the MgCl2 concentration at the t time (h) of feed solution, Vt is the decreased volume of feed solution during the t time. The heavy metal ion rejection of prepared membrane was investigated under FO mode. 1.0 L of heavy metal solution (Cu2+, Ni2+, Pb2+) with different concentration was chosen as the feed solution, and the solution pH was adjusted to 6.5 by NaOH (0.1 M), while 1.0 L of
(2)
The MgCl2 concentration was detected using a conductivity meter 341
Journal of Hazardous Materials 367 (2019) 339–347
M. Qiu, C. He
Fig. 3. (a) ATR-FTIR spectra of prepared membranes; (b) XPS survey spectra and (c) Zn 2p core level spectra of M0 and M1 membrane; (d) N 1s core level spectra of M1 membrane.
diameter of about 75.0 nm (Fig. S1(c)). FTIR spectrometry was used to evaluate the synthesis of ZIF-8@PDA. The FTIR spectrum of ZIF-8 is in agreement with the previous literature [25]. For the ZIF-8@PDA, new peaks at approximately 3400–3500, 1630, 1510 and 1270 cm−1 are associated with the OH/NH, indole, indoline and phenolic groups of PDA, respectively [24,26], indicating the successfully modification of PDA on the surface of nanofillers. As shown in Fig. S2 (a), the zeta potential of ZIF-8 varies from positive to negative after modifying, further indicating the existence of PDA on the surface of ZIF-8. TGA results of ZIF-8, PDA and ZIF-8@PDA are shown in Fig. 2 (b). TGA plot shows that no well-defined decomposition temperature can be found for PDA. And PDA has multi-step degradation behavior due to its heterogeneous structure and the presence of different chemical groups [27]. For the ZIF-8@PDA, in addition to the decomposition of pure ZIF-8 phase from 450 °C, the multi-stage over the temperature from 200 to 450 °C should be assigned to the decomposition of PDA. As shown in Fig. 2 (c), the diffraction peaks of ZIF-8 are all found in the XRD patterns of ZIF-8@PDA, suggesting that the modification does not affect the crystal structure of ZIF-8. The dispersion ability of ZIF-8@PDA in water is evaluated and the photo images are shown in Fig. 2(d). The dried ZIF-8 would precipitate quickly when dispersed in water due to its intrinsic hydrophobicity and the formation of strong covalent Zn-methylimidazole-Zn links between different nanoparticles [28]. In contrast, ZIF-8@PDA can be dispersed stably in water for long time due to the improved hydrophilicity of ZIF-8 (Fig.S2 (b)). Although the ZIF-8@ PDA has aggregation after 12 h standing (Fig. S3), it still has a better dispersion stability than the pristine ZIF-8. The good dispersibility would beneficial to the uniformly loading of nanofillers on the
MgCl2 solution with different concentration was chosen as the draw solution. The heavy metal ion rejection Rm was calculated using Eq. (7).
C × Vd ⎞ Rm = ⎜⎛1 − d ⎟ × 100% Cf × Vp ⎠ ⎝
(7)
where Cd (ppm) and Vd (L) are the heavy metal ion concentration and volume of draw solution at last, respectively. Cf (ppm) and Vp (L) are the heavy metal ion concentration and decreased volume of feed solution. The heavy metal concentration was collected by an inductively coupled plasma emission spectrometer (ICP, Prodigy, Leeman, US). The absorption experiment was used to illustrate the chelation reaction between heavy metal ions and PEI. 0.1 g of PES and prepared FO membranes were immersed into a 300.0 ppm of heavy metal solution, respectively. The solution was placed in a water bath oscillator for 24 h. In addition, the membrane was conducted with an X-ray energy dispersive spectrometer (EDS, Oxford) to determine the element on the membrane surface. 3. Results and discussion 3.1. Characterization of nanofillers In this study, ZIF-8@PDA was obtained from ZIF-8 by surface coating PDA. The obvious polyhedral shape of ZIF-8 can be seen from the TEM images (Fig. S1(a)). After PDA coating, the structure of nanoparticles become more random. The average diameter of the ZIF-8 is measured to be about 65.0 nm, while the ZIF-8@PDA has a mean 342
Journal of Hazardous Materials 367 (2019) 339–347
M. Qiu, C. He
Fig. 4. The FESEM surface images and cross-sectional of (a) M0, (b) M1, (c) M2 and (d) M3 membrane. The “1” refers to the surface images with 10.0 K magnification; “2” refers to the surface images with 50.0 K magnification; “3” refers to the cross-sectional images with 50.0 K magnification. The existence of nanoparticles from the cross-sectional images was marked with red arrows (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article).
Fig. 5. AFM images of prepared membranes: (a) M0, (b) M1, (c) M2, and (d) M3.
343
Journal of Hazardous Materials 367 (2019) 339–347
M. Qiu, C. He
Table 1 The root mean square roughness (Rrms) and average roughness (Ra) of prepared membranes. Membrane
Rrms (nm)
Ra (nm)
M0 M1 M2 M3
13.0 15.6 16.4 21.8
10.9 12.5 13.5 17.8
membrane surface. 3.2. Characterization of TFN membranes Fig. 3(a) presents the ATR-FTIR spectra of prepared membranes. Compared with substrate, peak at 1660 cm−1 in the spectra of M0 and M1 membrane corresponds to C]O stretching vibration of amide group (Fig. 3(a)), evincing the formation of polyamide layer through the interfacial polymerization [13]. The broad peak at about 3450 cm−1 is ascribed to the unreacted amine group of PEI, which endows the membranes with a positively charged surface. Meanwhile, the peak intensity at 3450 cm−1 for M1 is apparently higher than that of M0, which can be attributed to the phenolic hydroxyl and amine group of PDA. Fig. 3(b) illustrates the XPS survey spectra of M0 and M1 membrane. Compared with M0 membrane, peaks of Zn (Zn 2p3 at 1021 eV and Zn 2p1 at 1044 eV) appear in the spectrum of M1 membrane (Fig. 3(c)) [29], which further confirms the existence of nanofillers. The high-solution N 1 s XPS spectrum of M1 membrane is shown in Fig. 3(d). Three peaks with binding energy at 398.6, 399.5 and 401.4 eV are associated with C-NH2, NeC = O, and C-N+H3 species [30], respectively, demonstrating the membrane with a positively charged surface. The surface and cross-sectional morphologies of prepared membranes were observed by FESEM. As Fig. 4 (a1-d1) shows, the M0 membrane exhibits a relatively smooth surface, while the M1, M2 and M3 membranes show a lot of grain-like structure on the membrane surface. These grain-like structure should be attributed to the loading of ZIF-8@PDA on the membrane surface. The number of grain-like structure increases with the increasing ZIF-8/PDA loading. In addition, the TFN membranes exhibit irregular wrinkle-shaped structure, especially in M3 membrane, which making the membrane surface rougher (Fig. 4 (a2-d2)). The ZIF-8@PDA provide additional amino group or phenolic hydroxyl group to react with acid chloride group, thus enhancing the cross-linking rate of polyamide layer. Moreover, when the ZIF-8@PDA concentration increases to 0.1 wt%, the aggregation of
Fig. 7. MWCOs of prepared membranes. Table 2 Transport property of as-prepared membranes. Membrane
Water permeabilitya A (L/m2 h bar)
MgCl2 rejectionb R (%)
MgCl2 permeability B (L/m2 h)
B/A (KPa)
M0 M1 M2 M3 M4
2.42 4.88 6.03 8.84 9.46
97.4 97.8 96.9 92.3 80.1
0.09 0.16 0.27 1.10 3.32
3.8 ± 0.5 3.3 ± 0.4 4.5 ± 0.8 12.5 ± 1.6 35.1 ± 3.8
a b
± ± ± ± ±
0.12 0.16 0.15 0.18 0.21
± ± ± ± ±
1.8 0.9 1.3 2.5 4.2
± ± ± ± ±
0.03 0.02 0.06 0.12 0.15
DI water was used as feed solution with 2.0 bar pressure. 10.0 mM of MgCl2 as feed solution with 2.0 bar pressure.
nanoparticles further affect the surface morphology. The cross-sectional SEM images are presented in Fig. 4 (a3-d3). It can be found that nanoparticles are dispersed on the membrane surface evenly for M1 and M2 membrane, while ZIF-8@PDA aggregates are found on the membrane surface for M3 membrane. The membrane surface roughness of prepared membranes was performed by AFM. As shown in Fig. 5, the grain-like structure and the ZIF-8@PDA aggregate on the membrane surface are also well-reflected in AFM images. Table 1 shows the extracted data for root mean square roughness (Rrms) and average roughness (Ra) of prepared membranes. The M0 membrane has root mean square roughness (Rrms) and average
Fig. 6. (a) Measured water contact angles and (b) zeta potential data of prepared membranes. 344
Journal of Hazardous Materials 367 (2019) 339–347
M. Qiu, C. He
Table 3 Performance comparison between the FO membrane with a NF-like selective layer synthesized in this work and other works. Membrane
A (L/m2 h bar)
MgCl2 Rejection (%)
B/A (KPa)
Operating conditions
Ref.
M2
5.95
95.8
3.6
This work
6-PAH/PSS
7.3
80
15
PAI/PEI
4.58
92
–
PVA
1.93
93.5
32
Feed: 5.0 mM MgCl2 Pressure: 1.0 bar Feed: 5.0 mM MgCl2 Pressure: 1.0 bar Feed: 500.0 ppm MgCl2 Pressure: 1.0 bar Feed: 5.0 mM MgCl2 Pressure: 5.0 bar
[37] [38] [39]
Fig. 8. (a) Water flux (Jw), reverse salt flux (Js) and (b) Js/Jw ratios of as-prepared membranes (DI water and 1.0 M MgCl2 were used as feed and draw solution).
molecules by hydrogen bond interactions. In order to reject the heavy metal ions, the nanofiltration-like FO membrane should possess a positively charged surface. The electrical property of membrane surface was determined with zeta potential and the data is shown in Fig. 6(b). All the membranes are found to have a positively charged surface at pH = 6.5, which is attributed to the unreacted primary amine of PEI [24,31,32]. After introduction of nanoparticles, the zeta potential values show a slightly decrease due to the existence of negatively charged nanoparticles (Fig. S2(a)). Overall, these prepared membranes would be suitable for the heavy metal ion removal under FO mode. The molecular weight cut-offs (MWCOs) of the prepared membranes were evaluated by the rejection of polyethylene glycol (PEG) with different molecular weight (200, 400, 600, and 1000 Da). The MWCOs of membranes are taken as the molecular weight of the PEG molecule with 90.0% rejection. As shown in the Fig.7, the MWCOs are 284, 317, 338 and 435 Da for M0, M1, M2 and M3 membrane, respectively. After introduction of nanoparticles, the MWCOs increases from 284 to 435 Da, suggesting that the selective layer becomes loose. This is because the nanoparticles may prevent polymer chains from packing closely together. Interestingly, the increment in MWCOs of M3 membrane are bigger than that of M1 and M2 membrane, which is caused by the aggregation of nanoparticles on the membrane surface. Table 2 shows the intrinsic transport properties of prepared membranes. Note that all TFN membranes have higher water permeability than that of M0 membrane, and the values increase with the increasing loading nanoparticles. The improvement of water permeability should be attributed to the following main factors. Firstly, the good compatibility between ZIF-8@PDA and polyamide layer avoids the formation of interface defects. The ideal generated interface voids between entrapped ZIF-8@PDA nanoparticles and polyamide potentially constitute continuous channels for transporting water molecules [29]. Meanwhile, the porous structure of nanoparticles also provides additional paths for water molecules for water molecules transporting [33]. Secondly, the PDA layer can absorb water molecules due to the strong hydrogen
Fig. 9. Measured water flux and rejection of three heavy metal ions in FO mode (1000 ppm heavy metal salt solution and 1.0 M MgCl2 were used as feed and draw solution, pH = 6.5, 25℃).
roughness (Ra) of 13.0 nm and Ra (average roughness) of 10.9 nm. After the introduction of nanoparticles, the surface morphology of membranes become much rougher with a Rrms value ranging from 15.6 nm to 21.8 nm and a Ra value ranging from 12.5 nm to 17.8 nm. Although the improved roughness of membrane surface may potentially increase the fouling propensity, the enhanced surface area is beneficial for a higher water flux. The hydrophilicity of prepared membranes was measured by the water contact angle (WCAs) and the results are shown in Fig. 6(a). The WCAs of prepared membranes at 0 s are almost same, suggesting that all the membrane surfaces are covered by the PEI layer. Time-dependent of WCAs measurement demonstrate that the TFN membranes have a better wettability than the pristine TFC membrane (Table S1). This result can be attributed to the PDA layer which adsorb the water 345
Journal of Hazardous Materials 367 (2019) 339–347
M. Qiu, C. He
Fig. 10. Schematic image of permeation and rejection mechanism of the prepared FO membranes.
increases to 0.1 wt%, the water flux of M3 membrane begins to decrease. This is caused by the higher salt reverse flux (Fig. 8(a)), which results in a severe ICP and a reduced water flux. In addition, M3 membrane show a higher Js/Jw ratio than that of M1 and M2 membrane (Fig. 8(b)), suggesting that M3 membrane has a poor separation efficiency. Therefore, this result indicates that M2 membrane is suitable for potential application. Fig. 9 shows the result of using M2 membrane to treat the heavy metal wastewater (Cu2+, Ni2+, Pb2+). The water flux of M2 membrane has a decrease when using 1000 ppm heavy metal salt solution as the feed solution. This is because the heavy metal solution increases the feed osmotic pressure, which decrease the driving force across the membrane, hence results in the reduced water flux. Moreover, M2 membrane shows a high rejection to the heavy metal ions in FO mode (99.1% for Cu2+, 98.3% for Ni2+ and 97.7% for Pb2+). Meanwhile, the effect of draw solution and feed solution concentration on the water flux and ions rejection of M2 membrane are presented in Fig. S4. The water flux increases with the draw solution concentration since the increased driving force. In addition, the Cu2+ rejection shows a slightly increasing trend. This is because the Cu2+ concentration across the membrane almost has no changing while the water flux increases, hence reaches a higher rejection. The water flux and Cu2+ rejection all decrease with the feed solution concentration. The Cu2+ flux would increase proportional with the feed solution concentration, while the Cu2+ rejection only has a slightly decrease. This is because the increased feed solution concentration reduces the driving force across the membrane for water and ions transporting. The rejection mechanism of the M2 membrane is presented in Fig. 10. Firstly, the M2 membrane with a MWCO of 338 Da (0.43 nm) can reject the heavy metal ions that have a hydrated radius around 0.4 nm (Table S2) by the size exclusion effect. Besides, the rejection shows a trend of Cu2+ > Ni2+ > Pb2+, which is consistent with the hydrated radius of these ions. Secondly, M2 membrane has a positive charged surface (Fig. 6(b)), which shows an electrostatic repulsion to the heavy metal ions. Thirdly, the residual amino group of PEI can adsorb the metal ions by the chelation reaction (Fig. S5), which is also beneficial to get a high rejection. However, only 6.2% of Cu2+, 6.9% of
bonds and facilitate water transport through the mentioned additional channels [34] (Fig. S2(b) and Table S1). Thirdly, the rougher surface is also beneficial for a higher water flux since an enlargement of the eff ;ective membrane surface area [35] (Fig.5 and Table 1). Last but not least, the pore size of the prepared membranes increases with the increasing nanoparticles concentration, ranged from 0.39 to 0.48 nm (Fig. 6), leading to an increment in permeation of water molecules. The MgCl2 rejection has a decrease with increasing nanoparticles concentration. The salt rejection mainly depends on the Donnan exclusion eff ;ect and size exclusion eff ;ect [36]. The Donnan exclusion eff ;ect is mostly responsible for the repulsion of divalent cations. The strong electrostatic repulsion between Mg2+ and the positively charged surface of prepared membranes, leading to a high repulsion of Mg2+. After introducing of the nanoparticles, the pore size of membranes increases, which causes the decrease of rejection. The B/A value was used to assess the separation efficiency of the FO membrane. Compared with the M1 and M2 membrane, M3 membrane shows a much higher B/A value than that of M0 membrane, which is caused by the decrease of salt rejection. Table 3 shows the water permeability and MgCl2 rejection of the membrane prepared in this work and in other works. All the membranes have a nanofiltration-like selective layer. In order to unify the operation condition, we have added the filtration experiment of M2 membrane using 5.0 mM of MgCl2 as the feed solution under 1.0 bar pressure. Our membrane shows a higher water permeability and salt rejection than other membranes. Although the membrane prepared by layer by layer method has the highest water permeability, it shows a low salt rejection, which would loss the draw solute during FO process, hence causes the serious ICP problem. This result suggests that incorporation of ZIF@PDA in selective layer, the water permeability of membrane can be improved without sacrificing the selectivity. Fig. 8(a) shows the water flux (Jw) and salt reverse flux (Js) of prepared membranes in FO mode. It can be seen that M1 and M2 membrane show a higher water flux than that of M0 membrane, which is consistent with the result of water permeability (A). The improvement of water flux is attributed to the relatively loose nano-functional selective layer. However, when the concentration of nanoparticles 346
Journal of Hazardous Materials 367 (2019) 339–347
M. Qiu, C. He
Ni2+ and 8.1% of Pb2+ were adsorbed by M2 membrane after 24 h, respectively, indicating that the heavy metal ions removal mainly relies on rejection.
performance: a review, Chem. Rev. 113 (2013) 4980–5028. [15] N. Song, X. Gao, Z. Ma, X. Wang, et al., A review of graphene-based separation membrane: materials, characteristics, preparation and applications, Desalination 437 (2018) 59–72. [16] W.J. Lau, S. Gray, T. Matsuura, D. Emadzadeh, et al., A review on polyamide thin film nanocomposite (TFN) membranes: history, applications, challenges and approaches, Water Res. 80 (2015) 306–324. [17] A. Zirehpour, A. Rahimpour, M. Ulbricht, Nano-sized metal organic framework to improve the structural properties and desalination performance of thin film composite forward osmosis membrane, J. Membr. Sci. 531 (2017) 59–67. [18] D. Ma, S.B. Peh, G. Han, S.B. Chen, Thin-film nanocomposite (TFN) membranes incorporated with super-hydrophilic metal–organic framework (MOF) UiO-66: toward enhancement of water flux and salt rejection, ACS Appl. Mater.Inter. 9 (2017) 7523–7534. [19] J. Duan, Y. Pan, F. Pacheco, E. Litwiller, et al., High-performance polyamide thinfilm-nanocomposite reverse osmosis membranes containing hydrophobic zeolitic imidazolate framework-8, J. Membr. Sci. 476 (2015) 303–310. [20] L. Dong, M. Chen, X. Wu, D. Shi, et al., Multi-functional polydopamine coating: simultaneous enhancement of interfacial adhesion and CO2 separation performance of mixed matrix membranes, New J. Chem. 40 (2016) 9148–9159. [21] Z. Wang, D. Wang, S. Zhang, L. Hu, et al., Interfacial design of mixed matrix membranes for improved gas separation performance, Adv. Mater. 28 (2016) 3399–3405. [22] Y. Liang, J. Wei, Y.X. Hu, X.F. Chen, et al., Metal-polydopamine frameworks and their transformation to hollow metal/N-doped carbon particles, Nanoscale 9 (2017) 5323–5328. [23] J.-C. Yang, Y. Chen, Y.-H. Li, X.-B. Yin, Magnetic resonance imaging-guided multidrug chemotherapy and photothermal synergistic therapy with pH and NIRStimulation release, ACS Appl. Mater. Inter. 9 (2017) 22278–22288. [24] F.Y. Zhao, Y.L. Ji, X.D. Weng, Y.F. Mi, et al., High-flux positively charged nanocomposite nanofiltration membranes filled with poly(dopamine) modified multiwall carbon nanotubes, ACS Appl. Mater. Inter. 8 (2016) 6693–6700. [25] J. Im, N. Yim, J. Kim, T. Vogt, et al., High-pressure chemistry of a zeolitic imidazolate framework compound in the presence of different fluids, J. Am. Chem. Soc. 138 (2016) 11477–11480. [26] Y. Wang, C. Hou, Y. Zhang, F. He, et al., Preparation of graphene nano-sheet bonded PDA/MOF microcapsules with immobilized glucose oxidase as a mimetic multienzyme system for electrochemical sensing of glucose, J. Mater. Chem. B Mater. Biol. Med. 4 (2016) 3695–3702. [27] S. Xiong, Y. Wang, J. Yu, L. Chen, et al., Polydopamine particles for next-generation multifunctional biocomposites, J. Mater. Chem. A Mater. Energy Sustain. 2 (2014) 7578–7587. [28] J. Cravillon, S. Münzer, S.-J. Lohmeier, A. Feldhoff, et al., Rapid room-temperature synthesis and characterization of nanocrystals of a prototypical zeolitic imidazolate framework, Chem. Mater. 21 (2009) 1410–1412. [29] J. Zhu, L. Qin, A. Uliana, J. Hou, et al., Elevated performance of thin film nanocomposite membranes enabled by modified hydrophilic MOFs for nanofiltration, ACS Appl. Mater. Inter. 9 (2017) 1975–1986. [30] L. Xu, J. Xu, B. Shan, X. Wang, et al., Novel thin-film composite membranes via manipulating the synergistic interaction of dopamine and m-phenylenediamine for highly efficient forward osmosis desalination, J. Mater. Chem. A Mater. Energy Sustain. 5 (2017) 7920–7932. [31] T. Ma, Y. Su, Y. Li, R. Zhang, et al., Fabrication of electro-neutral nanofiltration membranes at neutral pH with antifouling surface via interfacial polymerization from a novel zwitterionic amine monomer, J. Membr. Sci. 503 (2016) 101–109. [32] D. Wu, S. Yu, D. Lawless, X. Feng, Thin film composite nanofiltration membranes fabricated from polymeric amine polyethylenimine imbedded with monomeric amine piperazine for enhanced salt separations, React. Funct. Polym. 86 (2015) 168–183. [33] H. Sun, B. Tang, P. Wu, Hydrophilic hollow zeolitic imidazolate framework-8 modified ultrafiltration membranes with significantly enhanced water separation properties, J. Membr. Sci. 551 (2018) 283–293. [34] L. Bai, Y. Liu, N. Bossa, A. Ding, et al., Incorporation of cellulose nanocrystals (CNCs) into the polyamide layer of thin-film composite (TFC) nanofiltration membranes for enhanced separation performance and antifouling properties, Environ. Sci. Technol. 52 (2018) 11178–11187. [35] M. Qiu, C. He, Novel zwitterion-silver nanocomposite modified thin-film composite forward osmosis membrane with simultaneous improved water flux and biofouling resistance property, Appl. Surf. Sci. 455 (2018) 492–501. [36] Y. Chen, C. He, High salt permeation nanofiltration membranes based on NMGassisted polydopamine coating for dye/salt fractionation, Desalination 413 (2017) 29–39. [37] Q. Saren, C.Q. Qiu, C.Y. Tang, Synthesis and characterization of novel forward osmosis membranes based on layer-by-Layer assembly, Environ. Sci. Technol. 45 (2011) 5201–5208. [38] C. Qiu, L. Setiawan, R. Wang, C.Y. Tang, et al., High performance flat sheet forward osmosis membrane with an NF-like selective layer on a woven fabric embedded substrate, Desalination 287 (2012) 266–270. [39] D. Qin, Z. Liu, Z. Liu, H. Bai, et al., Superior antifouling capability of hydrogel forward osmosis membrane for treating wastewaters with high concentration of organic foulants, Environ. Sci. Technol. 52 (2018) 1421–1428.
4. Conclusions In summary, a novel TFN FO membrane had been developed via in situ interfacial polymerization for treating heavy metal wastewater. An optimized selective layer was designed by incorporating porous PDA/ ZIF-8 into a PEI/TMC crosslinked network to improve the separation efficiency. The introduction of nanoparticles endows the selective layer with good wettability and loose structure. More importantly, the modified PDA layer improve the affinity between the nanofillers and polyamide matrix, hence maintaining the selectivity of membrane. Furthermore, this membrane exhibits high rejection to heavy metal ions (> 96%) in FO mode. The outstanding performance of these membranes highlight its potential application of TFN FO membranes for heavy metal ion removal. Acknowledgments This work is supported by grants from the Innovation Funds for PhD Students of Donghua University (CUSF-DH-D-2017031) and the National Science Foundation of China (No.51873034). Appendix A. Supplementary data Supplementary material related to this article can be found, in the online version, at doi:https://doi.org/10.1016/j.jhazmat.2018.12.096. References [1] M.A. Alaei Shahmirzadi, S.S. Hosseini, J. Luo, I. Ortiz, Significance, evolution and recent advances in adsorption technology, materials and processes for desalination, water softening and salt removal, J. Environ. Manage. 215 (2018) 324–344. [2] Y. Lv, Y. Du, Z.-X. Chen, W.-Z. Qiu, et al., Nanocomposite membranes of polydopamine/electropositive nanoparticles/polyethyleneimine for nanofiltration, J. Membr. Sci. 545 (2018) 99–106. [3] L. Pino, C. Vargas, A. Schwarz, R. Borquez, Influence of operating conditions on the removal of metals and sulfate from copper acid mine drainage by nanofiltration, Chem. Eng. J. 345 (2018) 114–125. [4] J.E. Efome, D. Rana, T. Matsuura, C.Q. Lan, Insight studies on metal-organic framework nanofibrous membrane adsorption and activation for heavy metal ions removal from aqueous solution, ACS Appl. Mater.Inter. 10 (2018) 18619–18629. [5] M. Chen, K. Shafer-Peltier, S.J. Randtke, E. Peltier, Competitive association of cations with poly(sodium 4-styrenesulfonate) (PSS) and heavy metal removal from water by PSS-assisted ultrafiltration, Chem. Eng. J. 344 (2018) 155–164. [6] P. Roy Choudhury, S. Majumdar, G.C. Sahoo, S. Saha, et al., High pressure ultrafiltration CuO/hydroxyethyl cellulose composite ceramic membrane for separation of Cr (VI) and Pb (II) from contaminated water, Chem. Eng. J. 336 (2018) 570–578. [7] M. Qiu, J. Wang, C. He, A stable and hydrophilic substrate for thin-film composite forward osmosis membrane revealed by in-situ cross-linked polymerization, Desalination 433 (2018) 1–9. [8] S. Adham, A. Hussain, J. Minier-Matar, A. Janson, et al., Membrane applications and opportunities for water management in the oil & gas industry, Desalination 440 (2018) 2–17. [9] Q. Chen, W. Xu, Q. Ge, Novel multicharge hydroacid complexes that effectively remove heavy metal ions from water in forward osmosis processes, Environ. Sci. Technol. 52 (2018) 4464–4471. [10] X. Zhao, C. Liu, Efficient removal of heavy metal ions based on the optimized dissolution-diffusion-flow forward osmosis process, Chem. Eng. J. 334 (2018) 1128–1134. [11] J.M. Gohil, P. Ray, A review on semi-aromatic polyamide TFC membranes prepared by interfacial polymerization: potential for water treatment and desalination, Sep. Purif. Technol. 181 (2017) 159–182. [12] Q. Shi, L. Ni, Y. Zhang, X. Feng, et al., Poly(p-phenylene terephthamide) embedded in a polysulfone as the substrate for improving compaction resistance and adhesion of a thin film composite polyamide membrane, J. Mater. Chem. A Mater. Energy Sustain. 5 (2017) 13610–13624. [13] L. Yang, Z. Wang, J. Zhang, Highly permeable zeolite imidazolate framework composite membranes fabricated via a chelation-assisted interfacial reaction, J. Mater. Chem. A Mater. Energy Sustain. 5 (2017) 15342–15355. [14] H. Vinh-Thang, S. Kaliaguine, Predictive models for mixed-matrix membrane
347