- Email: [email protected]
Fe3O4-CTAB composite for the removal of nitrate and phosphate from water
G Model
JIEC-2992; No. of Pages 10 Journal of Industrial and Engineering Chemistry xxx (2016) xxx–xxx
Contents lists available at ScienceDirect
Journal of Industrial and Engineering Chemistry journal homepage: www.elsevier.com/locate/jiec
Synthesis of rectorite/Fe3O4-CTAB composite for the removal of nitrate and phosphate from water Fei Wang a, Dan Liu b, Pengwu Zheng c, Xiaofei Ma a,* a
Chemistry Department, School of Science, Tianjin University, Tianjin 300072, China School of Physics and Materials Engineering, Dalian Nationalities University, Dalian 116600, China c School of Pharmacy, Jiangxi Science and Technology Normal University, Nanchang, Jiangxi 330013, China b
A R T I C L E I N F O
Article history: Received 24 August 2015 Received in revised form 5 March 2016 Accepted 14 July 2016 Available online xxx Keywords: Rectorite Modification Nitrate Phosphate Adsorption
A B S T R A C T
The rectorite/Fe3O4-CTAB composite (REC/Fe3O4-CTAB) was fabricated by introducing cetyl trimethyl ammonium bromide (CTAB) and Fe3O4 onto the layers of raw rectorite (REC). The layers of rectorite were intercalated or exfoliated, and CTAB enhanced the electrostatic attraction by quaternary ammonium cations towards nitrate and phosphate anions. The order of Fe3O4 and CTAB introduction had great effect on the adsorption. When Fe3O4 was loaded before CTAB on REC, the obtained composites exhibited better adsorption. The maximum adsorption capacities could reach 182.1 mg/g for NO3 and 174.5 mg/g for PO43. Besides, REC/Fe3O4-CTAB composite could be easily regenerated with NaOH solution. ß 2016 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights reserved.
Introduction Agricultural nitrate and phosphate pollution has become the most predominant source of water pollution in China. Excessive use of nitrogenous and phosphorous fertilizers causes eutrophication, which stimulates the overgrowth of algae and aquatic plants, leading to human health disorders, such as blue-baby syndrome, infant methemoglobinemia as well as cancer [1]. The present available methods for the removal of nitrate and phosphate include ion exchange [2–4], reverse osmosis [5], adsorption [6–13], catalytic [14,15] and biological denitrification [16]. Among all the methods studied, adsorption method was attached great importance due to its high efficiency, simple equipment required and good reliability [17]. Many natural materials can be good adsorbents for nitrate and phosphate in water, especially when they are modified, such as bamboo powder charcoal [18], Fe3O4/ZrO2/chitosan composites [19], mixed lanthanum/aluminum modified montmorillonite [20], Fe(III)-montmorillonites [21], positively charged kaolinites [22]. Rectorite (REC) is an interstratified natural clay mineral composed of a regular stacking of dioctahedral mica-like layer and dioctahedral smectite-like layer in a ratio of 1:1. As the smectite-like layers in REC are readily intercalated, REC possesses
* Corresponding author. Tel.: +86 22 27406144; fax: +86 22 27403475. E-mail address: [email protected] (X. Ma).
good adsorption capacity for cationic metals and dyes [23–26] by exchanging interlayer Na+ with either organic or inorganic cations. Previous work proved that modified REC is a good adsorbent for MB (methylene blue) and Pb2+, with maximum adsorption values of 277 mg/g and 180.8 mg/g, respectively [27]. However, REC exhibited weak adsorption for anionic contaminants such as nitrate, phosphate, sulfate and dichromate. Therefore, it is of great necessity to modify REC in order to meet the demand for the removal of nitrate and phosphate contaminants in water. Recently, many studies focused on tackling such similar problem, and quaternary ammonium salt is a very commonly used and advisable modifier [28]. By ion exchange between organic cations and Na+ in the interlayer of REC, the adsorption capacity of REC for anionic contaminants can be greatly enhanced. In addition, another disadvantage of REC adsorbent is its difficult withdrawal from water after the adsorption process. By introducing magnetic Fe3O4 into REC, REC composites can be easily separated from water with the magnetic field, thus avoiding secondary pollution [29]. In this work, magnetic Fe3O4 was introduced into the interlayer of REC for the facile separation from water. At the same time, Fe3O4 particles could improve the interlayer spaces of REC, which would be beneficial for intercalating quaternary ammonium salt into the enlarged interlayer of REC. The REC/Fe3O4 composite was then treated with CTAB to obtain REC/Fe3O4-CTAB composite. The modifications were expected to improve the adsorption capacities for nitrate (NO3) and phosphate (PO43). In this process, magnetic REC actually acted
http://dx.doi.org/10.1016/j.jiec.2016.07.017 1226-086X/ß 2016 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights reserved.
Please cite this article in press as: F. Wang, et al., J. Ind. Eng. Chem. (2016), http://dx.doi.org/10.1016/j.jiec.2016.07.017
G Model
JIEC-2992; No. of Pages 10 2
F. Wang et al. / Journal of Industrial and Engineering Chemistry xxx (2016) xxx–xxx
as the support of CTAB, while CTAB on magnetic REC was the adsorbent for nitrate and phosphate.
spectra scanner. The sample powders were dispersed in KBr and pressed into transparent sheets for testing. Thermogravimetric analysis (TGA)
Experimental Materials Sodium rectorite was purchased by Hubei Zhongxiang Rectorite Mine (Wuhan, China). All other reagents were of analytical grade and commercially available.
Thermal properties of REC, REC-CTAB, REC/Fe3O4, REC-CTAB/ Fe3O4 and REC/Fe3O4-CTAB were measured using a ZTY-ZP type thermal analyzer. Sample weights ranged from 10 to 15 mg. Samples were heated from ambient temperature to 800 8C at a heating rate of 10 8C/min in a nitrogen atmosphere. Vibrating sample magnetometer (VSM)
Preparation of REC/Fe3O4 composite REC/Fe3O4 composites were fabricated in a modified method of Ref. [29]. 1 g REC, 1.165 g FeCl36H2O and 0.6 g FeSO47H2O were added into 20 mL distilled water. 20 mL NH3H2O solution (8 M) was added dropwise to produce magnetic iron oxide under N2 atmosphere. The pH of the mixture was approximately 11–12. The suspension was held at 70 8C for 4 h and then washed with distilled water several times. The obtained REC/Fe3O4 composite was dried at 50 8C for 4 h.
The magnetic properties of REC/Fe3O4, REC-CTAB/Fe3O4, and REC/Fe3O4-CTAB were tested on a vibrating sample magnetometer (LDJ 9600-1, LDJ Electronics Inc., USA). Scanning electron microscope (SEM) Powders of REC, REC-CTAB, REC/Fe3O4, REC-CTAB/Fe3O4 and REC/Fe3O4-CTAB were immobilized in a sample holder, coated with gold, and viewed with an S-4800 scanning electron microscope.
Preparation of REC-CTAB composites Transmission electron microscope (TEM) REC-CTAB composites were prepared in a modified method of Ref. [30]. 4 g REC was dispersed in 200 mL 0.014 mol/L CTAB solution using ultrasonication for 10 min. The pH of the mixture was adjusted to 2, 7 and 13 with hydrochloric and sodium hydroxide solution. Then the mixture was shaken on an immersion oscillator registration at 100 rpm for 12 h at 60 8C. After washed for 6–8 times, the sample was dried at 50 8C for 4 h. The obtained RECCTAB composites were labeled as CTAB-REC-2, CTAB-REC-7 and CTAB-REC-13, respectively. Assembly of REC-CTAB/Fe3O4 composite 1 g as-obtained CTAB-REC-13, 1.165 g FeCl36H2O and 0.6 g FeSO47H2O were added into 20 mL distilled water, and 20 mL NH3H2O solution (8 M) was added dropwise to produce magnetic iron oxide under N2 atmosphere. The pH of the mixture was approximately 11–12. The suspension was held at 70 8C for 4 h and then washed with distilled water several times. The obtained RECCTAB/Fe3O4 composite was dried at 50 8C for 4 h.
REC, REC-CTAB, REC/Fe3O4, REC-CTAB/Fe3O4, REC/Fe3O4-CTAB powders were dispersed in alcohol. Sample suspensions were dropped into a copper grid, dried in air, and tested with a JEM2100F transmission electron microscope. Energy dispersive spectroscopic (EDS) mapping The composition of REC-CTAB/Fe3O4 and REC/Fe3O4-CTAB was measured with energy dispersive spectroscopic mapping with a JEM-2100F transmission electron microscope. Specific surface area The specific surface area of REC, REC-CTAB/Fe3O4 and REC/ Fe3O4-CTAB were measured with an Autosorb-1 specific surface area analyzer (Quantachrome Instruments, USA) using the BET method on the base of N2 adsorption data. The point of zero charge (pHZPC)
Assembly of REC/Fe3O4-CTAB composite 1 g as-obtained REC/Fe3O4 composite was dispersed in 50 mL 0.014 mol/L CTAB solution using ultrasonication for 10 min. The pH of the mixture was adjusted to 13 with sodium hydroxide solution. Then the mixture was shaken on an immersion oscillator registration at 100 rpm for 12 h at 60 8C. After washed for 6–8 times, the obtained REC/Fe3O4-CTAB composite was dried at 50 8C for 4 h.
The point of zero charge (pHZPC) of adsorbent REC/Fe3O4-CTAB was determined. 10 mg REC/Fe3O4-CTAB composite was added to 10 mL of 0.1 mol/L KNO3 solution at different initial pH from 3 to 10. The initial pH of solutions were adjusted with NaOH or HCl solution, and measured by pH meter (PHSJ-4A, China). Afterwards, the mixtures were shaken on a rotary shaker at 100 rpm for 12 h at 30 8C and the final pH values of solutions were measured at equilibrium [31].
X-ray diffraction (XRD)
Adsorption of nitrate (NO3)
X-ray diffraction patterns for REC, REC-CTAB, REC/Fe3O4, RECCTAB/Fe3O4 and REC/Fe3O4-CTAB were recorded in reflection mode in the angular range of 1–408 (2u), at room temperature by a D/ MAX-2500 operated at a CuKa wavelength of 1.542 A˚.
Adsorption experiments were conducted using glass bottles containing 10 mg adsorbents and 10 mL nitrate solution (NO3 100 mg/L) at pH 5. During the adsorption process, the glass bottles were placed on a slow-moving platform shaker and nitrate solutions were taken at intervals to analyze the effect of contact time on equilibrium adsorption capacity. The NO3 concentrations were tested using UV–vis spectrometry at 204 nm to determine the adsorbed amounts of NO3 at 30 8C [32].
Fourier transform infrared spectroscopy (FTIR) spectra FTIR analysis of REC, REC-CTAB, REC/Fe3O4, REC-CTAB/Fe3O4 and REC/Fe3O4-CTAB were performed on a BIO-RAD FTS3000 IR
Please cite this article in press as: F. Wang, et al., J. Ind. Eng. Chem. (2016), http://dx.doi.org/10.1016/j.jiec.2016.07.017
G Model
JIEC-2992; No. of Pages 10 F. Wang et al. / Journal of Industrial and Engineering Chemistry xxx (2016) xxx–xxx
3
(a)
In adsorption isotherm testing, the concentration of the adsorbent was 1 g/L and the NO3 concentrations varied from 80 to 500 mg/L. These experiments were carried out at 30 8C and all suspensions were shaken on a rotary shaker at 100 rpm for 2 h to reach the adsorption equilibrium. Adsorption of phosphate (PO43)
Intensity
CTAB-REC-13
Adsorption experiments were conducted using glass bottles containing 1 g/L adsorbents and 10 mL phosphate solution (PO43 100 mg/L) at pH 5 (mainly in the form of H2PO42). Solutions of ammonium molybdate-sulfuric acid and stannous chloride were added to show blue color. During the adsorption process, the glass bottles were placed on a slow-moving platform shaker and phosphate solution were taken at intervals to study the effect of contact time on equilibrium adsorption capacity. The PO43 concentrations were tested using UV–vis spectrometry at 650 nm to determine the adsorbed amounts of PO43 at 30 8C. In adsorption isotherm testing, the concentration of the adsorbent was 1 g/L and the PO43 concentrations varied from 80 to 500 mg/L. These experiments were carried out at 30 8C and all suspensions were shaken on a rotary shaker at 100 rpm for 2 h to reach the adsorption equilibrium.
CTAB-REC-7 CTAB-REC-2 R(001)
2
4
R(002)
REC
6
8
10
2 theta (degree) (b)
REC/Fe 3O4-CTAB The regeneration of REC/Fe3O4-CTAB
Intensity
For the readsorption study of NO3 and PO43–, 20 mg REC/ Fe3O4-CTAB was added to 20 mL NO3 (100 mg/L) and PO43 (100 mg/L) solution respectively, and placed on a rotary shaker at 100 rpm for 2 h. REC/Fe3O4-CTAB was taken out from the solution, and the concentrations of NO3 and PO43 in the solution were determined with a UV–vis spectrometer. REC/Fe3O4-CTAB was dried and immersed in 20 mL NaOH (1 mol/L) solution for 12 h to regenerate. The obtained REC/Fe3O4-CTAB was washed with distilled water for 4–6 times and readsorb at pH 5 (adjusted with 1 mol/L HCl solution).
REC-CTAB/Fe 3O4 R(001) REC/Fe 3O4 R(002) REC 2
4
6
8
10
2 theta (degree) Results and discussion
(c)
Characterization of REC/Fe3O4-CTAB composites
REC/Fe 3O4-CTAB REC-CTAB/Fe 3O4 intensity
As shown in Fig. 1(a), raw REC displayed strong R(0 0 1) and R(0 0 2) diffraction peaks at 2u = 4.108 and 8.118. According to the Bragg diffraction equation, 2dsinu = l, the distance between the REC layers (d0 0 1 and d0 0 2) can be approximately 2.16 and 1.09 nm. After being modified by CTAB, the R(0 0 1) patterns shifted to 3.128, 2.848 and 2.788 for CTAB-REC-2, CTAB-REC-7 and CTAB-REC-13, respectively, and their interlayer distance expanded to 2.83, 3.11 and 3.18 nm. Moreover, R(0 0 2) patterns shifted to the lower 2 theta values. These indicated that CTAB successfully intercalated the layers of REC, and cationic exchange reaction occurred between REC and CTAB. As CTAB-REC-13 expanded most obviously, REC modifications by CTAB were most effective at pH 13. On the one hand, REC showed better dispersion in basic solution, which facilitated better interaction between REC layers and CTAB. On the other hand, at pH 13 the surface of REC would become negatively charged, which would in turn favors the positively charged quaternary ammonium cations and enhance the cationic exchange between REC and CTAB [33]. Therefore, the synthesis of REC-CTAB/Fe3O4 and REC/Fe3O4-CTAB involved adjusting pH value to 13 using sodium hydroxide solution. As shown in Fig. 1(b), after being modified by iron oxide, the R(0 0 1) shifted to the lower values of 2 theta for REC/Fe3O4, RECCTAB/Fe3O4 and REC/Fe3O4-CTAB. It indicated that the cationic exchange occurred between Fe3+/Fe2+ and Na+ in the layers of REC, and the formation of Fe3O4 occurred at the surface of REC.
F(311) F(220)
REC/Fe 3O4
REC
10
20
30
40
2 theta (degree) Fig. 1. X-ray diffraction peaks of REC-CTAB prepared at the different pH values (a), and REC/Fe3O4, REC-CTAB/Fe3O4, REC/Fe3O4-CTAB (from 1 to 108) (b) and (from 1 to 408) (c).
However, the intensity of R(0 0 2) reflection was obviously weakened. This suggested that the introduction of Fe3O4 partially damaged the ordered structure of REC, and that the layers of REC were intercalated or exfoliated with different degrees. And CTAB could further intercalate or exfoliate the REC layer. However, Fe3O4 hardly improved interlayer distance if REC was firstly intercalated
Please cite this article in press as: F. Wang, et al., J. Ind. Eng. Chem. (2016), http://dx.doi.org/10.1016/j.jiec.2016.07.017
G Model
JIEC-2992; No. of Pages 10 F. Wang et al. / Journal of Industrial and Engineering Chemistry xxx (2016) xxx–xxx
Transmittance
REC/Fe3O4-CTAB 2852 696 2922 1401 3641 3136 REC-CTAB/Fe3O4 1021 3641 685 1401 3130 1 102 REC/Fe3O4 684 3642 1403 3126 1022 702 2926 2850 REC-CTAB 1479 3645 REC
3046
1058 703
1440 3644
Mass loss (%)
(a) 100
REC CTAB-REC-2 CTAB-REC-7 CTAB-REC-13
DTG
80
0
421
262 -2 200
400
600
800
600
800
Mass loss (%)
(b) 100
80
REC REC-CTAB/Fe3O4 REC/Fe3O4-CTAB
0
DTG
by CTAB. Therefore, the order of the introduction of Fe3O4 and CTAB had great effect on the intercalation of REC. As observed in Fig. 1(c), the characteristic diffraction peaks of REC-Fe3O4 at 30.2 and 35.7 are related to [2 2 0] and [3 1 1] crystal planes of face-centered cubic Fe3O4. These two peaks are also seen in REC-CTAB/Fe3O4 and REC/Fe3O4-CTAB, confirming that iron oxide has been grafted and attached on REC layers [34]. The FTIR spectra of REC, REC-CTAB, REC/Fe3O4, REC-CTAB/Fe3O4 and REC/Fe3O4-CTAB were shown in Fig. 2. In FTIR spectra of REC [35], the peak at 3644 cm1 was related to the hydroxyl stretching bonding of Si–O–H. The peak at 1021 cm1 was associated to the in-plane Si–O–Si stretching vibration, and the peaks at 818 and 703 cm1 were respectively attributed to Si–O–Al bending and Al– O– out of plane. These peaks appeared in the FTIR spectra of RECCTAB, REC/Fe3O4, REC-CTAB/Fe3O4 and REC/Fe3O4-CTAB. After REC was modified by CTAB, there appeared two new peaks at 2850 and 2926 cm1 for REC-CTAB. These new peaks were attributed to the symmetric and asymmetric CH2– stretching vibration, which could also be seen in REC-CTAB/Fe3O4 and REC/Fe3O4-CTAB [36]. However, CH2– stretching vibration was much weaker in RECCTAB/Fe3O4. This was related to the overlapping by Fe3O4, weakening the effect of CTAB. According to Fig. 3, the thermogravimetric (TG) curves of REC exhibited no weight loss under 500 8C. For REC-CTAB, the weightloss curve decreased continuously after 200 8C in Fig. 3(a). The weight loss at about 262 8C was attributed to thermal decomposition of the quaternary ammonium cations on the REC layers [30]. At 421 8C, the weight loss could be related to the pyrolysis of the long alkyl chains of CTAB [37], and the interaction between CTAB and REC layers improved the decomposed temperature of CTAB. Obviously, at pH 13 more CTAB were loaded in REC layers, in accordance with the highest weigh loss peak at 262 8C in Fig. 3(a). The thermogravimetric (TG) curves of REC-CTAB/Fe3O4 and REC/ Fe3O4-CTAB were shown in Fig. 3(b). As for REC-CTAB/Fe3O4, the loss of free water and the decomposition of iron hydroxide impurity could result in the weight loss below 200 8C and at 335 8C, respectively. In views of weight loss at 411 8C, a higher weight loss in REC/Fe3O4-CTAB was observed because more CTAB was attached on REC in REC/Fe3O4-CTAB than REC-CTAB/Fe3O4. As REC can disperse well in water, it is necessary to separate raw REC or modified REC in determining residual nitrate or phosphate or reclaiming REC. Fig. 4 displays the magnetic hysteresis curves of REC/Fe3O4, REC-CTAB/Fe3O4 and REC/Fe3O4-CTAB measured at
265
411 335
-2 200
400
Fig. 3. TG curves of REC-CTAB prepared at the different pH values (a), and RECCTAB/Fe3O4, REC/Fe3O4-CTAB (b).
300 K. REC/Fe3O4 and REC-CTAB/Fe3O4 composites possessed magnetic properties with saturation magnetization (18.28 and 17.55 emu/g, respectively) and exhibited superparamagnetic behavior with small coercivity in the applied magnetic fields
20 15
REC-Fe3O4 REC-CTAB/Fe3O4
magnetization (emu/g)
4
10
REC/Fe3O4-CTAB
5 0 -5 -10 -15 -20
1021 4000
3500
3000
2500
2000
1500
1000
500
Wavenumber (cm-1) Fig. 2. FTIR spectra of REC, REC-CTAB, REC/Fe3O4, REC-CTAB/Fe3O4 and REC/Fe3O4CTAB.
-10000
-5000
0
5000
10000
Field (Oe) Fig. 4. Hysteresis loops of REC/Fe3O4, REC-CTAB/Fe3O4, REC/Fe3O4-CTAB measure at 300 K.
Please cite this article in press as: F. Wang, et al., J. Ind. Eng. Chem. (2016), http://dx.doi.org/10.1016/j.jiec.2016.07.017
G Model
JIEC-2992; No. of Pages 10 F. Wang et al. / Journal of Industrial and Engineering Chemistry xxx (2016) xxx–xxx
[38]. REC/Fe3O4-CTAB exhibited a smaller saturation magnetization value (8.49 emu/g) than REC/Fe3O4. The magnetism of Fe3O4 was reduced in the process of CTAB introduction. Nevertheless, the saturation magnetization values of them were enough to separate them from aqueous solution with a strong magnet. SEM images were shown in Fig. 5. The aggregated layers of REC (Fig. 5(a)) were identified. After being modified by CTAB, the layers of REC were intercalated, and became obvious, as marked with red circles in Fig. 5(b and c). In Fig. 5 (d, e and f), Fe3O4 particles were introduced and grafted onto the surface of the layers in REC [39]. As shown in Fig. 6, REC possessed platelets with a thickness of 30–100 nm and the layers were transparent. After modification with quaternary ammonium salt, the CTAB molecules were easily visualized on the surface of REC layers via contrast differences between REC and REC-CTAB as some scattered dots could be seen in Fig. 6(b). With the introduction of iron oxide, Fe3O4 particles could be easily observed with a diameter of 40–90 nm in Fig. 6(c and d). After modification, some REC stacks were still consisted of
5
several layers, while some REC layers were exfoliated, and the REC was thinner [40]. In order to determine the composition of modified REC, energy dispersive spectroscopic (EDS) mapping was carried out [41]. Fig. 7 confirmed that the quaternary ammonium groups and Fe3O4 particles were successfully introduced onto the layers of REC. Besides, CTAB and Fe3O4 particles could be homogeneously dispersed on REC surface. The BET surface area of REC was 9.94 m2/g [42]. Nevertheless, the BET surface areas of REC/Fe3O4 and REC/Fe3O4-CTAB were 406.05 and 434.42 m2/g, respectively, much higher than that of REC. The increased surface area of REC/Fe3O4 composite was attributed to the intercalation of REC layers and magnetic iron oxide particles. When CTAB was introduced onto the REC interlayer afterward, the basal spacing of REC was expanded further, implying that CTAB was definitely grafted onto the layers of REC. The point of zero charge (pHZPC) of REC/Fe3O4-CTAB composite was shown in Fig. 8. It could be noted that pHZPC, namely the pH at
Fig. 5. SEM images of REC (a), REC-CTAB (b and c) , REC/Fe3O4 (d), REC-CTAB/Fe3O4 (e) and REC/Fe3O4-CTAB (f). (For interpretation of the references to color in the text, the reader is referred to the web version of this article.)
Please cite this article in press as: F. Wang, et al., J. Ind. Eng. Chem. (2016), http://dx.doi.org/10.1016/j.jiec.2016.07.017
G Model
JIEC-2992; No. of Pages 10 6
F. Wang et al. / Journal of Industrial and Engineering Chemistry xxx (2016) xxx–xxx
Fig. 6. TEM micrographs for REC (a), REC-CTAB (b), REC-CTAB/Fe3O4 (c) and REC/Fe3O4-CTAB (d).
Fig. 7. Energy dispersive X-ray analysis and dot mapping on REC-CTAB/Fe3O4 (a) and REC/Fe3O4-CTAB (b).
Please cite this article in press as: F. Wang, et al., J. Ind. Eng. Chem. (2016), http://dx.doi.org/10.1016/j.jiec.2016.07.017
G Model
JIEC-2992; No. of Pages 10 F. Wang et al. / Journal of Industrial and Engineering Chemistry xxx (2016) xxx–xxx
11
7
(a) pH (initial) pH (Final)
10
80
REC/Fe 3O4-CTAB
9 60
REC-CTAB
qt (mg/g)
pH (Final)
8 7 6 5
40
REC-CTAB/Fe 3O4 REC
20
4 0
3 2 2
3
4
5
6
7
8
9
10
0
11
20
40
60
80
100
120
t (min)
pH (Initial) Fig. 8. Determination of the pH of point of zero charge (pHZPC).
60
which the net surface charge on REC/Fe3O4-CTAB composite was zero (DpH 0) was associated with a pH value around 5.4. It means that at pH values less that 5.4, the REC/Fe3O4-CTAB surface had a net positive charge, while at pH above 5.4, the surface had a net negative charge. Therefore, the acidic pH could enhance the adsorption of NO3 and PO43 onto REC/ Fe3O4-CTAB surface [31]. However, if the solution was excessively acidic, molecular form of H3PO4 would prevail, which would damage the adsorption for phosphate.
(b)
REC/Fe 3O4-CTAB REC-CTAB
50
qt (mg/g)
40
REC-CTAB/Fe 3O4
30
REC 20
10
Adsorption Kinetic of NO3 and PO43 Fig. 9 reveals the effect of contact time on adsorption of NO3 and PO43 by REC, REC-CTAB, REC-CTAB/Fe3O4 and REC/Fe3O4CTAB. For about 2 h, the adsorptions could reach the equilibrium. In Fig. 9, CTAB possessed a higher adsorption capacity for NO3 and PO43 than raw REC because the grafted positively-charged CTAB generated electrostatic attraction towards anions. It was obvious that REC/Fe3O4-CTAB had absolute predominance whether in the nitrate adsorption or phosphate adsorption. However, the adsorption capacity for NO3 and PO43 of REC-CTAB/Fe3O4 was far less than that of REC/Fe3O4-CTAB, even though they went thorough same procedures in different order. The reason could be attributed to the chemical site occupation of Fe3O4 on the adsorption sites of REC, damaging the adsorption performance of quaternary ammonium cations on the layers of REC. Besides, as Fe3O4 particles contributed little to the adsorption for anions, the introduction of Fe3O4 increased the weight of adsorbents and decreased the adsorption capacity of 1 g adsorbents when iron oxide was grafted after CTAB. However, when Fe3O4 was firstly introduced before CTAB, it would enhance the interlayer spaces of REC, thus allowing more CTAB molecules enter REC layers and facilitating higher adsorption capacity for anions [36]. That could be verified in Fig. 3(b) that REC/ Fe3O4-CTAB experienced a higher weight loss peak at 411 8C than REC-CTAB/Fe3O4. But the overall modified REC
0 0
20
40
60
80
100
120
t (min) Fig. 9. (a) Effect of contact time on adsorption of NO3 by REC, REC-CTAB, REC-CTAB/ Fe3O4 and REC/Fe3O4-CTAB at 30 8C. Initial concentration: NO3 100 mg/L; adsorbents 1 g/L. (b) Effect of contact time on adsorption of PO43 by REC, RECCTAB, REC-CTAB/Fe3O4 and REC/Fe3O4-CTAB at 30 8C. Initial concentration: PO43 100 mg/L; adsorbents 1 g/L.
displayed larger adsorption ability for NO3 and PO43 than raw REC. The adsorption process can be measured using pseudo-secondorder kinetic model. According to this model, the amount of NO3 and PO43 adsorbed at any time, t (min), is defined as [43]: t 1 t ¼ þ qt kqe af;2 qe
(1)
where k (mg/g min1) is the second order rate constant, and qt (mg/g) and qe (mg/g) represent the amount of NO3 and PO43 adsorbed at any time (min) and equilibrium, respectively. Kinetic constants were calculated from the experimental data in Fig. 9, and listed in Tables 1 and 2. The adsorption process of
Table 1 Kinetic constants for NO3 adsorption at 30 8C. Pseudo-second-order model
REC REC-CTAB REC-CTAB/Fe3O4 REC/Fe3O4-CTAB
Intra-particle diffusion model
qe (mg/g)
k (g/mg min)
R
kid (mg/g min1/2)
C
R
19.2 58.2 34.3 87.0
0.015 0.0038 0.0019 0.0012
0.9994 0.9998 0.9984 0.9944
0.65 2.52 2.31 3.02
12.36 32.13 6.24 39.85
0.9354 0.9195 0.9452 0.9444
Please cite this article in press as: F. Wang, et al., J. Ind. Eng. Chem. (2016), http://dx.doi.org/10.1016/j.jiec.2016.07.017
G Model
JIEC-2992; No. of Pages 10 F. Wang et al. / Journal of Industrial and Engineering Chemistry xxx (2016) xxx–xxx
8
Table 2 Kinetic constants for PO43 adsorption at 30 8C. Pseudo-second-order model
REC REC-CTAB REC-CTAB/Fe3O4 REC/Fe3O4-CTAB
Intra-particle diffusion model
qe (mg/g)
k (g/mg min)
R
kid (mg/g min1/2)
C
R
25.3 51 38.1 59.3
0.0047 0.0044 0.0043 0.0019
0.9996 0.9997 0.9993 0.9999
1.46 2.08 1.94 3.63
9.39 29.17 17.71 20.11
0.9363 0.9250 0.9219 0.9233
NO3 and PO43 matched the pseudo second-order model well with a high correlation coefficient (R > 0.992), and qe values of REC-CTAB, REC-CTAB/Fe3O4 and REC/Fe3O4-CTAB were 58.2, 34.3, 87.0 mg/g for NO3 and 51.0, 38.1, 59.3 mg/g for PO43, respectively. The intra-particle diffusion model was also widely applied to measure the diffusion mechanism of adsorption process. It is defined as [25]:
(a)
180
REC/Fe 3O4-CTAB 160 140
qe (mg/g)
qt ¼ K id t 1=2 þ C
200
(2)
120
REC-CTAB
100 80
REC-CTAB/Fe 3O4
where Kid (mg/g min1/2) is the intra-particle diffusion rate constant and C is the intercept. As shown in Tables 1 and 2, the pseudo-second-order kinetic model fitted the adsorption process of NO3 and PO43 better than the intra-particle diffusion model, as the pseudo-second-order kinetic model exhibited higher R values.
60
REC 40 20 0
100
200
300
400
500
Ce (mg/L)
Adsorption isotherms of NO3 and PO43 Adsorption isotherms of NO3 and PO43 by REC, REC-CTAB, REC-CTAB/Fe3O4 and REC/Fe3O4-CTAB were shown in Fig. 10, and investigated with the Langmuir model, expressed as [43]:
160
ce 1 ce ¼ þ qe bQ max Q max
120
(b) REC/Fe 3O4-CTAB
140
qe (mg/g)
(3)
where ce (mg/L) is the equilibrium concentration in the solution; qe (mg/g) is the equilibrium adsorption capacity. Qmax (mg/g) is the maximum adsorption capacity; b (L/mg) is the Langmuir adsorption equilibrium constant. The adsorption isotherm process of NO3 and PO43 by REC, REC-CTAB, REC-CTAB/Fe3O4 and REC/Fe3O4-CTAB (Fig. 10) fitted the Langmuir isotherm model well, and the maximum monolayer adsorption capacities (Qmax) for REC-CTAB, REC-CTAB/Fe3O4 and REC/Fe3O4-CTAB were 134.4, 126.7, 182.1 mg/g for NO3 and 152.9, 117.0, 174.5 mg/g for PO43, respectively, as listed in Tables 3 and 4. Above all, it was easily observed that REC/Fe3O4-CTAB was an excellent adsorbent with predominant maximum adsorption capacity for both NO3 (182.1 mg/g) and PO43 (174.5 mg/g). The Qmax values of REC/Fe3O4-CTAB for nitrate (NO3) were competitive with other adsorbents (Table 5) such as polyacrylonitrilealumina nanoparticle mixed matrix membrane (15 mg/g) [43], modified pine sawdust (32.8 mg/g) [44], quaternized chitosanmelamine-glutaraldehyde resin (97.5 mg/g) [45], Zr(IV) loaded
REC-CTAB
100 80
REC-CTAB/Fe 3O4
60
REC 40 20 0
100
200
300
400
500
Ce (mg/L) Fig. 10. (a) Adsorption isotherms for NO3 adsorption by REC, REC-CTAB, RECCTAB/Fe3O4 and REC/Fe3O4-CTAB at 30 8C. Adsorbents 1 g/L. (a) Adsorption isotherms for PO43 adsorption by REC, REC-CTAB, REC-CTAB/Fe3O4 and REC/ Fe3O4-CTAB at 30 8C. Adsorbents 1 g/L.
Table 3 Isothermal constants for NO3 adsorption at 30 8C. Langmuir model
REC REC-CTAB REC-CTAB/Fe3O4 REC/Fe3O4-CTAB
Freundlich model
Qmax (mg/g)
b (L/mg)
R
1/n
KF (mg11/n L1/n g1)
R
82.4 134.4 126.7 182.1
0.0034 0.014 0.0046 0.0422
0.9956 0.9956 0.9955 0.9999
0.62 0.37 0.60 0.29
1.21 13.4 2.35 34.8
0.9981 0.9927 0.9860 0.9802
Please cite this article in press as: F. Wang, et al., J. Ind. Eng. Chem. (2016), http://dx.doi.org/10.1016/j.jiec.2016.07.017
G Model
JIEC-2992; No. of Pages 10 F. Wang et al. / Journal of Industrial and Engineering Chemistry xxx (2016) xxx–xxx
9
Table 4 Isothermal constants for PO43 adsorption at 30 8C. Langmuir model
REC REC-CTAB REC-CTAB/Fe3O4 REC/Fe3O4-CTAB
Freundlich model
Qmax (mg/g)
b (L/mg)
R
1/n
KF (mg11/n L1/n g1)
R
61.6 152.9 117 174.5
0.0117 0.0102 0.0077 0.0226
0.9992 0.9936 0.9928 0.9992
0.36 0.49 0.52 0.37
5.99 7.18 4.30 19.5
0.9822 0.9694 0.9714 0.9705
Table 5 The comparison of maximum adsorption capacity for NO3 and PO43. Adsorbents
84.5
Adsorption capacities (mg g1)
NO3- adsorption
88.3
PO43-adsorption 79.3
80
74.2
72.6
S
65.0
15 mg/g [43] 32.8 mg/g [44] 97.5 mg/g [45] 128.4 mg/g [46] 182.1 mg/g [this work] 11.9 mg/g [47] 35.4 mg/g [48] 60.6 mg/g [49] 112.5 mg/g [45]
efficiency (%)
NO3 adsorption Polyacrylonitrile-alumina nanoparticle mixed matrix membrane Modified pine sawdust Quaternized chitosan-melamine- glutaraldehyde resin Zr(IV) loaded cross-linked chitosan beads REC/Fe3O4-CTAB composite PO43S adsorption Mg–Al hydrotalcite-loaded kaolin clay Fe–Ti bimetal oxide Modified chitosan beads Quaternized chitosan-melamine-glutaraldehyde resin Zr(IV) loaded cross-linked chitosan beads REC/Fe3O4-CTAB composite
60
40
20
149.4 mg/g [46] 174.5 mg/g [this work]
0 1
2
3
cycle time cross-linked chitosan beads (128.4 mg/g) [46]. Also, the adsorption capacities of REC/Fe3O4-CTAB for phosphate (PO43) were higher than other adsorbents, including Mg–Al hydrotalcite-loaded kaolin clay (11.9 mg/g) [47], Fe–Ti bimetal oxide (35.4 mg/g) [48], modified chitosan beads (60.6 mg/g) [49], quaternized chitosanmelamine-glutaraldehyde resin (112.5 mg/g) [45], and Zr(IV) loaded cross-linked chitosan beads (149.4 mg/g) [46]. The Freundlich model was also widely used to describe adsorption isotherm, and expressed as [50]: 1 log qe ¼ logK F þ logC e n
(4)
where 1/n is the Freundlich adsorption intensity parameter and KF (mg11/n L1/n g1) is the Freundlich constant. As shown in Tables 3 and 4, as for the adsorption process of NO3 and PO43, the
Fig. 11. Three consecutive adsorption cycles of REC/Fe3O4-CTAB for NO3 and PO43.
Langmuir model exhibited better fitness than the Freundlich model. The 1/n values were less than 1, so the isotherm adsorption was heterogeneous [49]. The regeneration of REC/Fe3O4-CTAB The adsorption cycles of REC/Fe3O4-CTAB were shown in Fig. 11. The adsorption efficiency (%) meant the ratio of the weight of adsorbed NO3 and PO43 at any time and that at the first time. At a higher pH, the surface of adsorbents could carry more negative charges and therefore would more significantly repulse the negatively charged ions in solution, such as NO3 and PO43.
Fig. 12. Mechanism of NO3 and PO43 removal.
Please cite this article in press as: F. Wang, et al., J. Ind. Eng. Chem. (2016), http://dx.doi.org/10.1016/j.jiec.2016.07.017
G Model
JIEC-2992; No. of Pages 10 10
F. Wang et al. / Journal of Industrial and Engineering Chemistry xxx (2016) xxx–xxx
The adsorption efficiency of NO3 and PO43 reduced with the increasing of adsorption cycles. The adsorption efficiency of NO3 and PO43 on REC/Fe3O4-CTAB decreased to 65.0 and 72.6% at the third time. After several circles, REC/Fe3O4-CTAB still had good adsorption capacity even though the adsorption efficiency decreased much. Mechanism of nitrate and phosphate removal The mechanism of nitrate and phosphate removal by REC/ Fe3O4-CTAB composite was shown in Fig. 12. The fabrication of REC/Fe3O4-CTAB composite involved two procedures: the intercalation of REC by magnetic iron oxide particles and cationic exchange between REC and cationic surfactant (i.e. CTAB). The introduction of Fe3O4 particles expanded the interlayer spacing of REC and therefore permitted more CATB molecules to enter REC layers and contributed to a higher adsorption capacity for nitrate and phosphate anions. The grafted positively-charged CTAB molecules in REC layers could cause electrostatic attraction towards anions, such as NO3 and PO43 [50]. Conclusion The raw REC was successfully modified by introducing Fe3O4 particles and CTAB cations. During the modifications, iron oxide first formed in the layers of REC, and then CTAB easily intercalated the expanded REC interlayer. CTAB on magnetic REC could remove nitrate and phosphate. Besides, it is convenient to reclaim adsorbents from the agricultural wastewater with the existence of magnetic Fe3O4 particle on the layers of REC. The adsorption process of nitrate and phosphate matched the pseudo secondorder model and Langmuir isotherm model well. The maximum adsorption values were 182.1 mg/g and 174.5 mg/g for NO3 and PO43, respectively, which exhibited potential value in removal of nitrate and phosphate wastes in water. REC/Fe3O4-CTAB exhibited facile regeneration and good adsorption capacity in the adsorption cycles. Acknowledgements This research was supported by the Science and Technology Project of Jiangxi Provincial Office of Education (KJLD12082); the Fundamental Research Funds for the Central Universities (DC201502080402), the Scientific Research Fund of Liaoning
Provincial Education Department (L2013505), and the National Nature Science Foundation of China (51462009). References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50]
P. Loganathan, S. Vigneswaran, J. Kandasamy, J. Environ. Manag. 131 (2013) 363. H. Song, Y. Zhou, A. Li, S. Mueller, Desalination 296 (2012) 53. O. Primo, M.J. Rivero, A.M. Urtiaga, I. Ortiz, J. Hazard. Mater. 164 (2009) 389. A. Pintar, J. Batista, J. Levec, Chem. Eng. Sci. 56 (2001) 1551. J.H. Ahn, K.H. Choo, H.S. Park, J. Membr. Sci. 310 (2008) 296. H.H. Hwang, D.G. Kim, H.S. Shin, J. Hazard. Mater. 185 (2011) 1513. M.Z. Kassaee, E. Motamedi, A. Mikhak, R. Rahnemaie, Chem. Eng. J. 166 (2011) 490. L. Jiang, L. Lu, W. Zhang, Q. Du, B. Pan, L. Yang, Q. Zhang, Water Res. 45 (2011) 2191. Y.H. Liou, S.L. Lo, W.H. Kuan, C.J. Lin, S.C. Weng, Water Res. 40 (2006) 2485. J.S. Shengnan Yi, H. He, C. Long, A. Li, Chem. Eng. J. 230 (2013) 166. M. Kumar, S. Chakraborty, J. Hazard. Mater. 135 (2006) 112. S.M. Hosseini, B. Ataie-Ashtiani, M. Kholghi, Desalination 276 (2011) 214. Y.A. Tielong Li, Z. Jin, M. Dong, H. Xia, X. Wang, Bioresour. Technol. 101 (2010) 9825. P.T. Chiueh, Y.H. Lee, C.Y. Su, S.L. Lo, J. Hazard. Mater. 192 (2011) 837. X. Xing, B.Y. Gao, Q.Q. Zhong, Q.Y. Yue, Q. Li, J. Hazard. Mater. 186 (2011) 206. A. Bhatnagar, M. Sillanpaa, Chem. Eng. J. 168 (2011) 493. A. Maleki, B. Hayati, M. Naghizadeh, S.W. Joo, J. Ind. Eng. Chem. 28 (2015) 211. K. Mizuta, T. Matsumoto, Y. Hatate, K. Nishihara, T. Nakanishi, Bioresour. Technol. 95 (2004) 255. H.L. Jiang, P.H. Chen, S.L. Luo, Appl. Clay Sci. 284 (2013) 942. S.L. Tian, P.X. Jiang, P. Ning, Chem. Eng. J. 151 (2009) 141. L. Borgnino, M.J. Arena, C.P. Pauli, Colloid Surf. A 341 (2009) 46. J. Matusik, T. Bajda, J. Colloid Interface Sci. 398 (2013) 74. L.L. Yang, X.Y. Ma, N.N. Guo, Carbohydr. Polym. 90 (2012) 853. L.X. Zeng, Y.F. Chen, Q.Y. Zhang, Carbohydr. Polym. 130 (2015) 333. Y.J. Lu, P.R. Chang, P.W. Zheng, Appl. Clay Sci. 107 (2015) 21. G.K. Zhang, Y. Liu, X. Yang, Appl. Clay Sci. 29 (2005) 15. F. Wang, P.R. Chang, P.W. Zheng, X.F. Ma, Appl. Surf. Sci. 349 (2015) 251. H.L. Hong, W.T. Jiang, X.L. Zhang, Appl. Clay Sci. 42 (2008) 292. Y.F. Xie, D.Y. Qian, D.L. Wu, X.F. Ma, Chem. Eng. J. 168 (2011) 959. J.H. Wang, X. Zhang, B. Zhang, Desalination 259 (2010) 22. S.S. Mehdi, A. Khataee, A. Hassani, S. Karaca, C.R. Chim. 18 (2015) 204. Y.J. Lu, P.R. Lu, P.W. Zheng, X.F. Ma, Chem. Eng. J. 255 (2014) 49. W.B. Wang, A.Q. Wang, Carbohydr. Polym. 77 (2009) 891. L.X. Zheng, M.J. Xie, Q.Y. Zhang, Carbohydr. Polym. 123 (2015) 89. G.C. Han, Y. Han, X.Y. Wang, Appl. Surf. Sci. 317 (2014) 35. R.F. Xu, S.J. Xin, X. Zhou, Int. J. Pharm. 438 (2012) 258. D. Cheng, Y.B. Wen, L.J. Wang, Carbohydr. Polym. 123 (2015) 157. P.R. Chang, P.W. Zheng, B.X. Liu, J. Hazard. Mater. 186 (2011) 2144. J.B. Liu, J. Hao, F. Cheng, J. Environ. Eng. 141 (2015) 04014055. D.L. Wu, P.W. Zheng, P.R. Chang, X.F. Ma, Chem. Eng. J. 174 (2011) 489. W. Weng, V.G. Pol, K. Amine, Adv. Mater. 25 (2013) 1608. R. Mukherjee, S. De, J. Membr. Sci. 466 (2014) 281. S. Kilpimaa, H. Runtti, T. Kangas, U. Lassi, T. Kuokkanen, J. Ind. Eng. Chem. 21 (2015) 1354. A. Keranen, T. Leiviska, O. Hormi, J. Tanskanen, J. Environ. Manage. 147 (2015) 46. A. Sowmya, S. Meenakshi, Int. J. Biol. Macromol. 64 (2014) 224. A. Sowmya, S. Meenakshi, Int. J. Biol. Macromol. 69 (2014) 336. L. Deng, Z. Shi, J. Alloys Compd. 637 (2015) 188. J.B. Lu, D.F. Liu, J. Hao, Chem. Eng. Res. Des. 93 (2015) 652. X. Liu, L.F. Zhang, Powder Technol. 277 (2015) 112. S.M. Prabhu, S. Meenakshi, Powder Technol. 268 (2014) 306.
Please cite this article in press as: F. Wang, et al., J. Ind. Eng. Chem. (2016), http://dx.doi.org/10.1016/j.jiec.2016.07.017
JIEC-2992; No. of Pages 10 Journal of Industrial and Engineering Chemistry xxx (2016) xxx–xxx
Contents lists available at ScienceDirect
Journal of Industrial and Engineering Chemistry journal homepage: www.elsevier.com/locate/jiec
Synthesis of rectorite/Fe3O4-CTAB composite for the removal of nitrate and phosphate from water Fei Wang a, Dan Liu b, Pengwu Zheng c, Xiaofei Ma a,* a
Chemistry Department, School of Science, Tianjin University, Tianjin 300072, China School of Physics and Materials Engineering, Dalian Nationalities University, Dalian 116600, China c School of Pharmacy, Jiangxi Science and Technology Normal University, Nanchang, Jiangxi 330013, China b
A R T I C L E I N F O
Article history: Received 24 August 2015 Received in revised form 5 March 2016 Accepted 14 July 2016 Available online xxx Keywords: Rectorite Modification Nitrate Phosphate Adsorption
A B S T R A C T
The rectorite/Fe3O4-CTAB composite (REC/Fe3O4-CTAB) was fabricated by introducing cetyl trimethyl ammonium bromide (CTAB) and Fe3O4 onto the layers of raw rectorite (REC). The layers of rectorite were intercalated or exfoliated, and CTAB enhanced the electrostatic attraction by quaternary ammonium cations towards nitrate and phosphate anions. The order of Fe3O4 and CTAB introduction had great effect on the adsorption. When Fe3O4 was loaded before CTAB on REC, the obtained composites exhibited better adsorption. The maximum adsorption capacities could reach 182.1 mg/g for NO3 and 174.5 mg/g for PO43. Besides, REC/Fe3O4-CTAB composite could be easily regenerated with NaOH solution. ß 2016 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights reserved.
Introduction Agricultural nitrate and phosphate pollution has become the most predominant source of water pollution in China. Excessive use of nitrogenous and phosphorous fertilizers causes eutrophication, which stimulates the overgrowth of algae and aquatic plants, leading to human health disorders, such as blue-baby syndrome, infant methemoglobinemia as well as cancer [1]. The present available methods for the removal of nitrate and phosphate include ion exchange [2–4], reverse osmosis [5], adsorption [6–13], catalytic [14,15] and biological denitrification [16]. Among all the methods studied, adsorption method was attached great importance due to its high efficiency, simple equipment required and good reliability [17]. Many natural materials can be good adsorbents for nitrate and phosphate in water, especially when they are modified, such as bamboo powder charcoal [18], Fe3O4/ZrO2/chitosan composites [19], mixed lanthanum/aluminum modified montmorillonite [20], Fe(III)-montmorillonites [21], positively charged kaolinites [22]. Rectorite (REC) is an interstratified natural clay mineral composed of a regular stacking of dioctahedral mica-like layer and dioctahedral smectite-like layer in a ratio of 1:1. As the smectite-like layers in REC are readily intercalated, REC possesses
* Corresponding author. Tel.: +86 22 27406144; fax: +86 22 27403475. E-mail address: [email protected] (X. Ma).
good adsorption capacity for cationic metals and dyes [23–26] by exchanging interlayer Na+ with either organic or inorganic cations. Previous work proved that modified REC is a good adsorbent for MB (methylene blue) and Pb2+, with maximum adsorption values of 277 mg/g and 180.8 mg/g, respectively [27]. However, REC exhibited weak adsorption for anionic contaminants such as nitrate, phosphate, sulfate and dichromate. Therefore, it is of great necessity to modify REC in order to meet the demand for the removal of nitrate and phosphate contaminants in water. Recently, many studies focused on tackling such similar problem, and quaternary ammonium salt is a very commonly used and advisable modifier [28]. By ion exchange between organic cations and Na+ in the interlayer of REC, the adsorption capacity of REC for anionic contaminants can be greatly enhanced. In addition, another disadvantage of REC adsorbent is its difficult withdrawal from water after the adsorption process. By introducing magnetic Fe3O4 into REC, REC composites can be easily separated from water with the magnetic field, thus avoiding secondary pollution [29]. In this work, magnetic Fe3O4 was introduced into the interlayer of REC for the facile separation from water. At the same time, Fe3O4 particles could improve the interlayer spaces of REC, which would be beneficial for intercalating quaternary ammonium salt into the enlarged interlayer of REC. The REC/Fe3O4 composite was then treated with CTAB to obtain REC/Fe3O4-CTAB composite. The modifications were expected to improve the adsorption capacities for nitrate (NO3) and phosphate (PO43). In this process, magnetic REC actually acted
http://dx.doi.org/10.1016/j.jiec.2016.07.017 1226-086X/ß 2016 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights reserved.
Please cite this article in press as: F. Wang, et al., J. Ind. Eng. Chem. (2016), http://dx.doi.org/10.1016/j.jiec.2016.07.017
G Model
JIEC-2992; No. of Pages 10 2
F. Wang et al. / Journal of Industrial and Engineering Chemistry xxx (2016) xxx–xxx
as the support of CTAB, while CTAB on magnetic REC was the adsorbent for nitrate and phosphate.
spectra scanner. The sample powders were dispersed in KBr and pressed into transparent sheets for testing. Thermogravimetric analysis (TGA)
Experimental Materials Sodium rectorite was purchased by Hubei Zhongxiang Rectorite Mine (Wuhan, China). All other reagents were of analytical grade and commercially available.
Thermal properties of REC, REC-CTAB, REC/Fe3O4, REC-CTAB/ Fe3O4 and REC/Fe3O4-CTAB were measured using a ZTY-ZP type thermal analyzer. Sample weights ranged from 10 to 15 mg. Samples were heated from ambient temperature to 800 8C at a heating rate of 10 8C/min in a nitrogen atmosphere. Vibrating sample magnetometer (VSM)
Preparation of REC/Fe3O4 composite REC/Fe3O4 composites were fabricated in a modified method of Ref. [29]. 1 g REC, 1.165 g FeCl36H2O and 0.6 g FeSO47H2O were added into 20 mL distilled water. 20 mL NH3H2O solution (8 M) was added dropwise to produce magnetic iron oxide under N2 atmosphere. The pH of the mixture was approximately 11–12. The suspension was held at 70 8C for 4 h and then washed with distilled water several times. The obtained REC/Fe3O4 composite was dried at 50 8C for 4 h.
The magnetic properties of REC/Fe3O4, REC-CTAB/Fe3O4, and REC/Fe3O4-CTAB were tested on a vibrating sample magnetometer (LDJ 9600-1, LDJ Electronics Inc., USA). Scanning electron microscope (SEM) Powders of REC, REC-CTAB, REC/Fe3O4, REC-CTAB/Fe3O4 and REC/Fe3O4-CTAB were immobilized in a sample holder, coated with gold, and viewed with an S-4800 scanning electron microscope.
Preparation of REC-CTAB composites Transmission electron microscope (TEM) REC-CTAB composites were prepared in a modified method of Ref. [30]. 4 g REC was dispersed in 200 mL 0.014 mol/L CTAB solution using ultrasonication for 10 min. The pH of the mixture was adjusted to 2, 7 and 13 with hydrochloric and sodium hydroxide solution. Then the mixture was shaken on an immersion oscillator registration at 100 rpm for 12 h at 60 8C. After washed for 6–8 times, the sample was dried at 50 8C for 4 h. The obtained RECCTAB composites were labeled as CTAB-REC-2, CTAB-REC-7 and CTAB-REC-13, respectively. Assembly of REC-CTAB/Fe3O4 composite 1 g as-obtained CTAB-REC-13, 1.165 g FeCl36H2O and 0.6 g FeSO47H2O were added into 20 mL distilled water, and 20 mL NH3H2O solution (8 M) was added dropwise to produce magnetic iron oxide under N2 atmosphere. The pH of the mixture was approximately 11–12. The suspension was held at 70 8C for 4 h and then washed with distilled water several times. The obtained RECCTAB/Fe3O4 composite was dried at 50 8C for 4 h.
REC, REC-CTAB, REC/Fe3O4, REC-CTAB/Fe3O4, REC/Fe3O4-CTAB powders were dispersed in alcohol. Sample suspensions were dropped into a copper grid, dried in air, and tested with a JEM2100F transmission electron microscope. Energy dispersive spectroscopic (EDS) mapping The composition of REC-CTAB/Fe3O4 and REC/Fe3O4-CTAB was measured with energy dispersive spectroscopic mapping with a JEM-2100F transmission electron microscope. Specific surface area The specific surface area of REC, REC-CTAB/Fe3O4 and REC/ Fe3O4-CTAB were measured with an Autosorb-1 specific surface area analyzer (Quantachrome Instruments, USA) using the BET method on the base of N2 adsorption data. The point of zero charge (pHZPC)
Assembly of REC/Fe3O4-CTAB composite 1 g as-obtained REC/Fe3O4 composite was dispersed in 50 mL 0.014 mol/L CTAB solution using ultrasonication for 10 min. The pH of the mixture was adjusted to 13 with sodium hydroxide solution. Then the mixture was shaken on an immersion oscillator registration at 100 rpm for 12 h at 60 8C. After washed for 6–8 times, the obtained REC/Fe3O4-CTAB composite was dried at 50 8C for 4 h.
The point of zero charge (pHZPC) of adsorbent REC/Fe3O4-CTAB was determined. 10 mg REC/Fe3O4-CTAB composite was added to 10 mL of 0.1 mol/L KNO3 solution at different initial pH from 3 to 10. The initial pH of solutions were adjusted with NaOH or HCl solution, and measured by pH meter (PHSJ-4A, China). Afterwards, the mixtures were shaken on a rotary shaker at 100 rpm for 12 h at 30 8C and the final pH values of solutions were measured at equilibrium [31].
X-ray diffraction (XRD)
Adsorption of nitrate (NO3)
X-ray diffraction patterns for REC, REC-CTAB, REC/Fe3O4, RECCTAB/Fe3O4 and REC/Fe3O4-CTAB were recorded in reflection mode in the angular range of 1–408 (2u), at room temperature by a D/ MAX-2500 operated at a CuKa wavelength of 1.542 A˚.
Adsorption experiments were conducted using glass bottles containing 10 mg adsorbents and 10 mL nitrate solution (NO3 100 mg/L) at pH 5. During the adsorption process, the glass bottles were placed on a slow-moving platform shaker and nitrate solutions were taken at intervals to analyze the effect of contact time on equilibrium adsorption capacity. The NO3 concentrations were tested using UV–vis spectrometry at 204 nm to determine the adsorbed amounts of NO3 at 30 8C [32].
Fourier transform infrared spectroscopy (FTIR) spectra FTIR analysis of REC, REC-CTAB, REC/Fe3O4, REC-CTAB/Fe3O4 and REC/Fe3O4-CTAB were performed on a BIO-RAD FTS3000 IR
Please cite this article in press as: F. Wang, et al., J. Ind. Eng. Chem. (2016), http://dx.doi.org/10.1016/j.jiec.2016.07.017
G Model
JIEC-2992; No. of Pages 10 F. Wang et al. / Journal of Industrial and Engineering Chemistry xxx (2016) xxx–xxx
3
(a)
In adsorption isotherm testing, the concentration of the adsorbent was 1 g/L and the NO3 concentrations varied from 80 to 500 mg/L. These experiments were carried out at 30 8C and all suspensions were shaken on a rotary shaker at 100 rpm for 2 h to reach the adsorption equilibrium. Adsorption of phosphate (PO43)
Intensity
CTAB-REC-13
Adsorption experiments were conducted using glass bottles containing 1 g/L adsorbents and 10 mL phosphate solution (PO43 100 mg/L) at pH 5 (mainly in the form of H2PO42). Solutions of ammonium molybdate-sulfuric acid and stannous chloride were added to show blue color. During the adsorption process, the glass bottles were placed on a slow-moving platform shaker and phosphate solution were taken at intervals to study the effect of contact time on equilibrium adsorption capacity. The PO43 concentrations were tested using UV–vis spectrometry at 650 nm to determine the adsorbed amounts of PO43 at 30 8C. In adsorption isotherm testing, the concentration of the adsorbent was 1 g/L and the PO43 concentrations varied from 80 to 500 mg/L. These experiments were carried out at 30 8C and all suspensions were shaken on a rotary shaker at 100 rpm for 2 h to reach the adsorption equilibrium.
CTAB-REC-7 CTAB-REC-2 R(001)
2
4
R(002)
REC
6
8
10
2 theta (degree) (b)
REC/Fe 3O4-CTAB The regeneration of REC/Fe3O4-CTAB
Intensity
For the readsorption study of NO3 and PO43–, 20 mg REC/ Fe3O4-CTAB was added to 20 mL NO3 (100 mg/L) and PO43 (100 mg/L) solution respectively, and placed on a rotary shaker at 100 rpm for 2 h. REC/Fe3O4-CTAB was taken out from the solution, and the concentrations of NO3 and PO43 in the solution were determined with a UV–vis spectrometer. REC/Fe3O4-CTAB was dried and immersed in 20 mL NaOH (1 mol/L) solution for 12 h to regenerate. The obtained REC/Fe3O4-CTAB was washed with distilled water for 4–6 times and readsorb at pH 5 (adjusted with 1 mol/L HCl solution).
REC-CTAB/Fe 3O4 R(001) REC/Fe 3O4 R(002) REC 2
4
6
8
10
2 theta (degree) Results and discussion
(c)
Characterization of REC/Fe3O4-CTAB composites
REC/Fe 3O4-CTAB REC-CTAB/Fe 3O4 intensity
As shown in Fig. 1(a), raw REC displayed strong R(0 0 1) and R(0 0 2) diffraction peaks at 2u = 4.108 and 8.118. According to the Bragg diffraction equation, 2dsinu = l, the distance between the REC layers (d0 0 1 and d0 0 2) can be approximately 2.16 and 1.09 nm. After being modified by CTAB, the R(0 0 1) patterns shifted to 3.128, 2.848 and 2.788 for CTAB-REC-2, CTAB-REC-7 and CTAB-REC-13, respectively, and their interlayer distance expanded to 2.83, 3.11 and 3.18 nm. Moreover, R(0 0 2) patterns shifted to the lower 2 theta values. These indicated that CTAB successfully intercalated the layers of REC, and cationic exchange reaction occurred between REC and CTAB. As CTAB-REC-13 expanded most obviously, REC modifications by CTAB were most effective at pH 13. On the one hand, REC showed better dispersion in basic solution, which facilitated better interaction between REC layers and CTAB. On the other hand, at pH 13 the surface of REC would become negatively charged, which would in turn favors the positively charged quaternary ammonium cations and enhance the cationic exchange between REC and CTAB [33]. Therefore, the synthesis of REC-CTAB/Fe3O4 and REC/Fe3O4-CTAB involved adjusting pH value to 13 using sodium hydroxide solution. As shown in Fig. 1(b), after being modified by iron oxide, the R(0 0 1) shifted to the lower values of 2 theta for REC/Fe3O4, RECCTAB/Fe3O4 and REC/Fe3O4-CTAB. It indicated that the cationic exchange occurred between Fe3+/Fe2+ and Na+ in the layers of REC, and the formation of Fe3O4 occurred at the surface of REC.
F(311) F(220)
REC/Fe 3O4
REC
10
20
30
40
2 theta (degree) Fig. 1. X-ray diffraction peaks of REC-CTAB prepared at the different pH values (a), and REC/Fe3O4, REC-CTAB/Fe3O4, REC/Fe3O4-CTAB (from 1 to 108) (b) and (from 1 to 408) (c).
However, the intensity of R(0 0 2) reflection was obviously weakened. This suggested that the introduction of Fe3O4 partially damaged the ordered structure of REC, and that the layers of REC were intercalated or exfoliated with different degrees. And CTAB could further intercalate or exfoliate the REC layer. However, Fe3O4 hardly improved interlayer distance if REC was firstly intercalated
Please cite this article in press as: F. Wang, et al., J. Ind. Eng. Chem. (2016), http://dx.doi.org/10.1016/j.jiec.2016.07.017
G Model
JIEC-2992; No. of Pages 10 F. Wang et al. / Journal of Industrial and Engineering Chemistry xxx (2016) xxx–xxx
Transmittance
REC/Fe3O4-CTAB 2852 696 2922 1401 3641 3136 REC-CTAB/Fe3O4 1021 3641 685 1401 3130 1 102 REC/Fe3O4 684 3642 1403 3126 1022 702 2926 2850 REC-CTAB 1479 3645 REC
3046
1058 703
1440 3644
Mass loss (%)
(a) 100
REC CTAB-REC-2 CTAB-REC-7 CTAB-REC-13
DTG
80
0
421
262 -2 200
400
600
800
600
800
Mass loss (%)
(b) 100
80
REC REC-CTAB/Fe3O4 REC/Fe3O4-CTAB
0
DTG
by CTAB. Therefore, the order of the introduction of Fe3O4 and CTAB had great effect on the intercalation of REC. As observed in Fig. 1(c), the characteristic diffraction peaks of REC-Fe3O4 at 30.2 and 35.7 are related to [2 2 0] and [3 1 1] crystal planes of face-centered cubic Fe3O4. These two peaks are also seen in REC-CTAB/Fe3O4 and REC/Fe3O4-CTAB, confirming that iron oxide has been grafted and attached on REC layers [34]. The FTIR spectra of REC, REC-CTAB, REC/Fe3O4, REC-CTAB/Fe3O4 and REC/Fe3O4-CTAB were shown in Fig. 2. In FTIR spectra of REC [35], the peak at 3644 cm1 was related to the hydroxyl stretching bonding of Si–O–H. The peak at 1021 cm1 was associated to the in-plane Si–O–Si stretching vibration, and the peaks at 818 and 703 cm1 were respectively attributed to Si–O–Al bending and Al– O– out of plane. These peaks appeared in the FTIR spectra of RECCTAB, REC/Fe3O4, REC-CTAB/Fe3O4 and REC/Fe3O4-CTAB. After REC was modified by CTAB, there appeared two new peaks at 2850 and 2926 cm1 for REC-CTAB. These new peaks were attributed to the symmetric and asymmetric CH2– stretching vibration, which could also be seen in REC-CTAB/Fe3O4 and REC/Fe3O4-CTAB [36]. However, CH2– stretching vibration was much weaker in RECCTAB/Fe3O4. This was related to the overlapping by Fe3O4, weakening the effect of CTAB. According to Fig. 3, the thermogravimetric (TG) curves of REC exhibited no weight loss under 500 8C. For REC-CTAB, the weightloss curve decreased continuously after 200 8C in Fig. 3(a). The weight loss at about 262 8C was attributed to thermal decomposition of the quaternary ammonium cations on the REC layers [30]. At 421 8C, the weight loss could be related to the pyrolysis of the long alkyl chains of CTAB [37], and the interaction between CTAB and REC layers improved the decomposed temperature of CTAB. Obviously, at pH 13 more CTAB were loaded in REC layers, in accordance with the highest weigh loss peak at 262 8C in Fig. 3(a). The thermogravimetric (TG) curves of REC-CTAB/Fe3O4 and REC/ Fe3O4-CTAB were shown in Fig. 3(b). As for REC-CTAB/Fe3O4, the loss of free water and the decomposition of iron hydroxide impurity could result in the weight loss below 200 8C and at 335 8C, respectively. In views of weight loss at 411 8C, a higher weight loss in REC/Fe3O4-CTAB was observed because more CTAB was attached on REC in REC/Fe3O4-CTAB than REC-CTAB/Fe3O4. As REC can disperse well in water, it is necessary to separate raw REC or modified REC in determining residual nitrate or phosphate or reclaiming REC. Fig. 4 displays the magnetic hysteresis curves of REC/Fe3O4, REC-CTAB/Fe3O4 and REC/Fe3O4-CTAB measured at
265
411 335
-2 200
400
Fig. 3. TG curves of REC-CTAB prepared at the different pH values (a), and RECCTAB/Fe3O4, REC/Fe3O4-CTAB (b).
300 K. REC/Fe3O4 and REC-CTAB/Fe3O4 composites possessed magnetic properties with saturation magnetization (18.28 and 17.55 emu/g, respectively) and exhibited superparamagnetic behavior with small coercivity in the applied magnetic fields
20 15
REC-Fe3O4 REC-CTAB/Fe3O4
magnetization (emu/g)
4
10
REC/Fe3O4-CTAB
5 0 -5 -10 -15 -20
1021 4000
3500
3000
2500
2000
1500
1000
500
Wavenumber (cm-1) Fig. 2. FTIR spectra of REC, REC-CTAB, REC/Fe3O4, REC-CTAB/Fe3O4 and REC/Fe3O4CTAB.
-10000
-5000
0
5000
10000
Field (Oe) Fig. 4. Hysteresis loops of REC/Fe3O4, REC-CTAB/Fe3O4, REC/Fe3O4-CTAB measure at 300 K.
Please cite this article in press as: F. Wang, et al., J. Ind. Eng. Chem. (2016), http://dx.doi.org/10.1016/j.jiec.2016.07.017
G Model
JIEC-2992; No. of Pages 10 F. Wang et al. / Journal of Industrial and Engineering Chemistry xxx (2016) xxx–xxx
[38]. REC/Fe3O4-CTAB exhibited a smaller saturation magnetization value (8.49 emu/g) than REC/Fe3O4. The magnetism of Fe3O4 was reduced in the process of CTAB introduction. Nevertheless, the saturation magnetization values of them were enough to separate them from aqueous solution with a strong magnet. SEM images were shown in Fig. 5. The aggregated layers of REC (Fig. 5(a)) were identified. After being modified by CTAB, the layers of REC were intercalated, and became obvious, as marked with red circles in Fig. 5(b and c). In Fig. 5 (d, e and f), Fe3O4 particles were introduced and grafted onto the surface of the layers in REC [39]. As shown in Fig. 6, REC possessed platelets with a thickness of 30–100 nm and the layers were transparent. After modification with quaternary ammonium salt, the CTAB molecules were easily visualized on the surface of REC layers via contrast differences between REC and REC-CTAB as some scattered dots could be seen in Fig. 6(b). With the introduction of iron oxide, Fe3O4 particles could be easily observed with a diameter of 40–90 nm in Fig. 6(c and d). After modification, some REC stacks were still consisted of
5
several layers, while some REC layers were exfoliated, and the REC was thinner [40]. In order to determine the composition of modified REC, energy dispersive spectroscopic (EDS) mapping was carried out [41]. Fig. 7 confirmed that the quaternary ammonium groups and Fe3O4 particles were successfully introduced onto the layers of REC. Besides, CTAB and Fe3O4 particles could be homogeneously dispersed on REC surface. The BET surface area of REC was 9.94 m2/g [42]. Nevertheless, the BET surface areas of REC/Fe3O4 and REC/Fe3O4-CTAB were 406.05 and 434.42 m2/g, respectively, much higher than that of REC. The increased surface area of REC/Fe3O4 composite was attributed to the intercalation of REC layers and magnetic iron oxide particles. When CTAB was introduced onto the REC interlayer afterward, the basal spacing of REC was expanded further, implying that CTAB was definitely grafted onto the layers of REC. The point of zero charge (pHZPC) of REC/Fe3O4-CTAB composite was shown in Fig. 8. It could be noted that pHZPC, namely the pH at
Fig. 5. SEM images of REC (a), REC-CTAB (b and c) , REC/Fe3O4 (d), REC-CTAB/Fe3O4 (e) and REC/Fe3O4-CTAB (f). (For interpretation of the references to color in the text, the reader is referred to the web version of this article.)
Please cite this article in press as: F. Wang, et al., J. Ind. Eng. Chem. (2016), http://dx.doi.org/10.1016/j.jiec.2016.07.017
G Model
JIEC-2992; No. of Pages 10 6
F. Wang et al. / Journal of Industrial and Engineering Chemistry xxx (2016) xxx–xxx
Fig. 6. TEM micrographs for REC (a), REC-CTAB (b), REC-CTAB/Fe3O4 (c) and REC/Fe3O4-CTAB (d).
Fig. 7. Energy dispersive X-ray analysis and dot mapping on REC-CTAB/Fe3O4 (a) and REC/Fe3O4-CTAB (b).
Please cite this article in press as: F. Wang, et al., J. Ind. Eng. Chem. (2016), http://dx.doi.org/10.1016/j.jiec.2016.07.017
G Model
JIEC-2992; No. of Pages 10 F. Wang et al. / Journal of Industrial and Engineering Chemistry xxx (2016) xxx–xxx
11
7
(a) pH (initial) pH (Final)
10
80
REC/Fe 3O4-CTAB
9 60
REC-CTAB
qt (mg/g)
pH (Final)
8 7 6 5
40
REC-CTAB/Fe 3O4 REC
20
4 0
3 2 2
3
4
5
6
7
8
9
10
0
11
20
40
60
80
100
120
t (min)
pH (Initial) Fig. 8. Determination of the pH of point of zero charge (pHZPC).
60
which the net surface charge on REC/Fe3O4-CTAB composite was zero (DpH 0) was associated with a pH value around 5.4. It means that at pH values less that 5.4, the REC/Fe3O4-CTAB surface had a net positive charge, while at pH above 5.4, the surface had a net negative charge. Therefore, the acidic pH could enhance the adsorption of NO3 and PO43 onto REC/ Fe3O4-CTAB surface [31]. However, if the solution was excessively acidic, molecular form of H3PO4 would prevail, which would damage the adsorption for phosphate.
(b)
REC/Fe 3O4-CTAB REC-CTAB
50
qt (mg/g)
40
REC-CTAB/Fe 3O4
30
REC 20
10
Adsorption Kinetic of NO3 and PO43 Fig. 9 reveals the effect of contact time on adsorption of NO3 and PO43 by REC, REC-CTAB, REC-CTAB/Fe3O4 and REC/Fe3O4CTAB. For about 2 h, the adsorptions could reach the equilibrium. In Fig. 9, CTAB possessed a higher adsorption capacity for NO3 and PO43 than raw REC because the grafted positively-charged CTAB generated electrostatic attraction towards anions. It was obvious that REC/Fe3O4-CTAB had absolute predominance whether in the nitrate adsorption or phosphate adsorption. However, the adsorption capacity for NO3 and PO43 of REC-CTAB/Fe3O4 was far less than that of REC/Fe3O4-CTAB, even though they went thorough same procedures in different order. The reason could be attributed to the chemical site occupation of Fe3O4 on the adsorption sites of REC, damaging the adsorption performance of quaternary ammonium cations on the layers of REC. Besides, as Fe3O4 particles contributed little to the adsorption for anions, the introduction of Fe3O4 increased the weight of adsorbents and decreased the adsorption capacity of 1 g adsorbents when iron oxide was grafted after CTAB. However, when Fe3O4 was firstly introduced before CTAB, it would enhance the interlayer spaces of REC, thus allowing more CTAB molecules enter REC layers and facilitating higher adsorption capacity for anions [36]. That could be verified in Fig. 3(b) that REC/ Fe3O4-CTAB experienced a higher weight loss peak at 411 8C than REC-CTAB/Fe3O4. But the overall modified REC
0 0
20
40
60
80
100
120
t (min) Fig. 9. (a) Effect of contact time on adsorption of NO3 by REC, REC-CTAB, REC-CTAB/ Fe3O4 and REC/Fe3O4-CTAB at 30 8C. Initial concentration: NO3 100 mg/L; adsorbents 1 g/L. (b) Effect of contact time on adsorption of PO43 by REC, RECCTAB, REC-CTAB/Fe3O4 and REC/Fe3O4-CTAB at 30 8C. Initial concentration: PO43 100 mg/L; adsorbents 1 g/L.
displayed larger adsorption ability for NO3 and PO43 than raw REC. The adsorption process can be measured using pseudo-secondorder kinetic model. According to this model, the amount of NO3 and PO43 adsorbed at any time, t (min), is defined as [43]: t 1 t ¼ þ qt kqe af;2 qe
(1)
where k (mg/g min1) is the second order rate constant, and qt (mg/g) and qe (mg/g) represent the amount of NO3 and PO43 adsorbed at any time (min) and equilibrium, respectively. Kinetic constants were calculated from the experimental data in Fig. 9, and listed in Tables 1 and 2. The adsorption process of
Table 1 Kinetic constants for NO3 adsorption at 30 8C. Pseudo-second-order model
REC REC-CTAB REC-CTAB/Fe3O4 REC/Fe3O4-CTAB
Intra-particle diffusion model
qe (mg/g)
k (g/mg min)
R
kid (mg/g min1/2)
C
R
19.2 58.2 34.3 87.0
0.015 0.0038 0.0019 0.0012
0.9994 0.9998 0.9984 0.9944
0.65 2.52 2.31 3.02
12.36 32.13 6.24 39.85
0.9354 0.9195 0.9452 0.9444
Please cite this article in press as: F. Wang, et al., J. Ind. Eng. Chem. (2016), http://dx.doi.org/10.1016/j.jiec.2016.07.017
G Model
JIEC-2992; No. of Pages 10 F. Wang et al. / Journal of Industrial and Engineering Chemistry xxx (2016) xxx–xxx
8
Table 2 Kinetic constants for PO43 adsorption at 30 8C. Pseudo-second-order model
REC REC-CTAB REC-CTAB/Fe3O4 REC/Fe3O4-CTAB
Intra-particle diffusion model
qe (mg/g)
k (g/mg min)
R
kid (mg/g min1/2)
C
R
25.3 51 38.1 59.3
0.0047 0.0044 0.0043 0.0019
0.9996 0.9997 0.9993 0.9999
1.46 2.08 1.94 3.63
9.39 29.17 17.71 20.11
0.9363 0.9250 0.9219 0.9233
NO3 and PO43 matched the pseudo second-order model well with a high correlation coefficient (R > 0.992), and qe values of REC-CTAB, REC-CTAB/Fe3O4 and REC/Fe3O4-CTAB were 58.2, 34.3, 87.0 mg/g for NO3 and 51.0, 38.1, 59.3 mg/g for PO43, respectively. The intra-particle diffusion model was also widely applied to measure the diffusion mechanism of adsorption process. It is defined as [25]:
(a)
180
REC/Fe 3O4-CTAB 160 140
qe (mg/g)
qt ¼ K id t 1=2 þ C
200
(2)
120
REC-CTAB
100 80
REC-CTAB/Fe 3O4
where Kid (mg/g min1/2) is the intra-particle diffusion rate constant and C is the intercept. As shown in Tables 1 and 2, the pseudo-second-order kinetic model fitted the adsorption process of NO3 and PO43 better than the intra-particle diffusion model, as the pseudo-second-order kinetic model exhibited higher R values.
60
REC 40 20 0
100
200
300
400
500
Ce (mg/L)
Adsorption isotherms of NO3 and PO43 Adsorption isotherms of NO3 and PO43 by REC, REC-CTAB, REC-CTAB/Fe3O4 and REC/Fe3O4-CTAB were shown in Fig. 10, and investigated with the Langmuir model, expressed as [43]:
160
ce 1 ce ¼ þ qe bQ max Q max
120
(b) REC/Fe 3O4-CTAB
140
qe (mg/g)
(3)
where ce (mg/L) is the equilibrium concentration in the solution; qe (mg/g) is the equilibrium adsorption capacity. Qmax (mg/g) is the maximum adsorption capacity; b (L/mg) is the Langmuir adsorption equilibrium constant. The adsorption isotherm process of NO3 and PO43 by REC, REC-CTAB, REC-CTAB/Fe3O4 and REC/Fe3O4-CTAB (Fig. 10) fitted the Langmuir isotherm model well, and the maximum monolayer adsorption capacities (Qmax) for REC-CTAB, REC-CTAB/Fe3O4 and REC/Fe3O4-CTAB were 134.4, 126.7, 182.1 mg/g for NO3 and 152.9, 117.0, 174.5 mg/g for PO43, respectively, as listed in Tables 3 and 4. Above all, it was easily observed that REC/Fe3O4-CTAB was an excellent adsorbent with predominant maximum adsorption capacity for both NO3 (182.1 mg/g) and PO43 (174.5 mg/g). The Qmax values of REC/Fe3O4-CTAB for nitrate (NO3) were competitive with other adsorbents (Table 5) such as polyacrylonitrilealumina nanoparticle mixed matrix membrane (15 mg/g) [43], modified pine sawdust (32.8 mg/g) [44], quaternized chitosanmelamine-glutaraldehyde resin (97.5 mg/g) [45], Zr(IV) loaded
REC-CTAB
100 80
REC-CTAB/Fe 3O4
60
REC 40 20 0
100
200
300
400
500
Ce (mg/L) Fig. 10. (a) Adsorption isotherms for NO3 adsorption by REC, REC-CTAB, RECCTAB/Fe3O4 and REC/Fe3O4-CTAB at 30 8C. Adsorbents 1 g/L. (a) Adsorption isotherms for PO43 adsorption by REC, REC-CTAB, REC-CTAB/Fe3O4 and REC/ Fe3O4-CTAB at 30 8C. Adsorbents 1 g/L.
Table 3 Isothermal constants for NO3 adsorption at 30 8C. Langmuir model
REC REC-CTAB REC-CTAB/Fe3O4 REC/Fe3O4-CTAB
Freundlich model
Qmax (mg/g)
b (L/mg)
R
1/n
KF (mg11/n L1/n g1)
R
82.4 134.4 126.7 182.1
0.0034 0.014 0.0046 0.0422
0.9956 0.9956 0.9955 0.9999
0.62 0.37 0.60 0.29
1.21 13.4 2.35 34.8
0.9981 0.9927 0.9860 0.9802
Please cite this article in press as: F. Wang, et al., J. Ind. Eng. Chem. (2016), http://dx.doi.org/10.1016/j.jiec.2016.07.017
G Model
JIEC-2992; No. of Pages 10 F. Wang et al. / Journal of Industrial and Engineering Chemistry xxx (2016) xxx–xxx
9
Table 4 Isothermal constants for PO43 adsorption at 30 8C. Langmuir model
REC REC-CTAB REC-CTAB/Fe3O4 REC/Fe3O4-CTAB
Freundlich model
Qmax (mg/g)
b (L/mg)
R
1/n
KF (mg11/n L1/n g1)
R
61.6 152.9 117 174.5
0.0117 0.0102 0.0077 0.0226
0.9992 0.9936 0.9928 0.9992
0.36 0.49 0.52 0.37
5.99 7.18 4.30 19.5
0.9822 0.9694 0.9714 0.9705
Table 5 The comparison of maximum adsorption capacity for NO3 and PO43. Adsorbents
84.5
Adsorption capacities (mg g1)
NO3- adsorption
88.3
PO43-adsorption 79.3
80
74.2
72.6
S
65.0
15 mg/g [43] 32.8 mg/g [44] 97.5 mg/g [45] 128.4 mg/g [46] 182.1 mg/g [this work] 11.9 mg/g [47] 35.4 mg/g [48] 60.6 mg/g [49] 112.5 mg/g [45]
efficiency (%)
NO3 adsorption Polyacrylonitrile-alumina nanoparticle mixed matrix membrane Modified pine sawdust Quaternized chitosan-melamine- glutaraldehyde resin Zr(IV) loaded cross-linked chitosan beads REC/Fe3O4-CTAB composite PO43S adsorption Mg–Al hydrotalcite-loaded kaolin clay Fe–Ti bimetal oxide Modified chitosan beads Quaternized chitosan-melamine-glutaraldehyde resin Zr(IV) loaded cross-linked chitosan beads REC/Fe3O4-CTAB composite
60
40
20
149.4 mg/g [46] 174.5 mg/g [this work]
0 1
2
3
cycle time cross-linked chitosan beads (128.4 mg/g) [46]. Also, the adsorption capacities of REC/Fe3O4-CTAB for phosphate (PO43) were higher than other adsorbents, including Mg–Al hydrotalcite-loaded kaolin clay (11.9 mg/g) [47], Fe–Ti bimetal oxide (35.4 mg/g) [48], modified chitosan beads (60.6 mg/g) [49], quaternized chitosanmelamine-glutaraldehyde resin (112.5 mg/g) [45], and Zr(IV) loaded cross-linked chitosan beads (149.4 mg/g) [46]. The Freundlich model was also widely used to describe adsorption isotherm, and expressed as [50]: 1 log qe ¼ logK F þ logC e n
(4)
where 1/n is the Freundlich adsorption intensity parameter and KF (mg11/n L1/n g1) is the Freundlich constant. As shown in Tables 3 and 4, as for the adsorption process of NO3 and PO43, the
Fig. 11. Three consecutive adsorption cycles of REC/Fe3O4-CTAB for NO3 and PO43.
Langmuir model exhibited better fitness than the Freundlich model. The 1/n values were less than 1, so the isotherm adsorption was heterogeneous [49]. The regeneration of REC/Fe3O4-CTAB The adsorption cycles of REC/Fe3O4-CTAB were shown in Fig. 11. The adsorption efficiency (%) meant the ratio of the weight of adsorbed NO3 and PO43 at any time and that at the first time. At a higher pH, the surface of adsorbents could carry more negative charges and therefore would more significantly repulse the negatively charged ions in solution, such as NO3 and PO43.
Fig. 12. Mechanism of NO3 and PO43 removal.
Please cite this article in press as: F. Wang, et al., J. Ind. Eng. Chem. (2016), http://dx.doi.org/10.1016/j.jiec.2016.07.017
G Model
JIEC-2992; No. of Pages 10 10
F. Wang et al. / Journal of Industrial and Engineering Chemistry xxx (2016) xxx–xxx
The adsorption efficiency of NO3 and PO43 reduced with the increasing of adsorption cycles. The adsorption efficiency of NO3 and PO43 on REC/Fe3O4-CTAB decreased to 65.0 and 72.6% at the third time. After several circles, REC/Fe3O4-CTAB still had good adsorption capacity even though the adsorption efficiency decreased much. Mechanism of nitrate and phosphate removal The mechanism of nitrate and phosphate removal by REC/ Fe3O4-CTAB composite was shown in Fig. 12. The fabrication of REC/Fe3O4-CTAB composite involved two procedures: the intercalation of REC by magnetic iron oxide particles and cationic exchange between REC and cationic surfactant (i.e. CTAB). The introduction of Fe3O4 particles expanded the interlayer spacing of REC and therefore permitted more CATB molecules to enter REC layers and contributed to a higher adsorption capacity for nitrate and phosphate anions. The grafted positively-charged CTAB molecules in REC layers could cause electrostatic attraction towards anions, such as NO3 and PO43 [50]. Conclusion The raw REC was successfully modified by introducing Fe3O4 particles and CTAB cations. During the modifications, iron oxide first formed in the layers of REC, and then CTAB easily intercalated the expanded REC interlayer. CTAB on magnetic REC could remove nitrate and phosphate. Besides, it is convenient to reclaim adsorbents from the agricultural wastewater with the existence of magnetic Fe3O4 particle on the layers of REC. The adsorption process of nitrate and phosphate matched the pseudo secondorder model and Langmuir isotherm model well. The maximum adsorption values were 182.1 mg/g and 174.5 mg/g for NO3 and PO43, respectively, which exhibited potential value in removal of nitrate and phosphate wastes in water. REC/Fe3O4-CTAB exhibited facile regeneration and good adsorption capacity in the adsorption cycles. Acknowledgements This research was supported by the Science and Technology Project of Jiangxi Provincial Office of Education (KJLD12082); the Fundamental Research Funds for the Central Universities (DC201502080402), the Scientific Research Fund of Liaoning
Provincial Education Department (L2013505), and the National Nature Science Foundation of China (51462009). References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50]
P. Loganathan, S. Vigneswaran, J. Kandasamy, J. Environ. Manag. 131 (2013) 363. H. Song, Y. Zhou, A. Li, S. Mueller, Desalination 296 (2012) 53. O. Primo, M.J. Rivero, A.M. Urtiaga, I. Ortiz, J. Hazard. Mater. 164 (2009) 389. A. Pintar, J. Batista, J. Levec, Chem. Eng. Sci. 56 (2001) 1551. J.H. Ahn, K.H. Choo, H.S. Park, J. Membr. Sci. 310 (2008) 296. H.H. Hwang, D.G. Kim, H.S. Shin, J. Hazard. Mater. 185 (2011) 1513. M.Z. Kassaee, E. Motamedi, A. Mikhak, R. Rahnemaie, Chem. Eng. J. 166 (2011) 490. L. Jiang, L. Lu, W. Zhang, Q. Du, B. Pan, L. Yang, Q. Zhang, Water Res. 45 (2011) 2191. Y.H. Liou, S.L. Lo, W.H. Kuan, C.J. Lin, S.C. Weng, Water Res. 40 (2006) 2485. J.S. Shengnan Yi, H. He, C. Long, A. Li, Chem. Eng. J. 230 (2013) 166. M. Kumar, S. Chakraborty, J. Hazard. Mater. 135 (2006) 112. S.M. Hosseini, B. Ataie-Ashtiani, M. Kholghi, Desalination 276 (2011) 214. Y.A. Tielong Li, Z. Jin, M. Dong, H. Xia, X. Wang, Bioresour. Technol. 101 (2010) 9825. P.T. Chiueh, Y.H. Lee, C.Y. Su, S.L. Lo, J. Hazard. Mater. 192 (2011) 837. X. Xing, B.Y. Gao, Q.Q. Zhong, Q.Y. Yue, Q. Li, J. Hazard. Mater. 186 (2011) 206. A. Bhatnagar, M. Sillanpaa, Chem. Eng. J. 168 (2011) 493. A. Maleki, B. Hayati, M. Naghizadeh, S.W. Joo, J. Ind. Eng. Chem. 28 (2015) 211. K. Mizuta, T. Matsumoto, Y. Hatate, K. Nishihara, T. Nakanishi, Bioresour. Technol. 95 (2004) 255. H.L. Jiang, P.H. Chen, S.L. Luo, Appl. Clay Sci. 284 (2013) 942. S.L. Tian, P.X. Jiang, P. Ning, Chem. Eng. J. 151 (2009) 141. L. Borgnino, M.J. Arena, C.P. Pauli, Colloid Surf. A 341 (2009) 46. J. Matusik, T. Bajda, J. Colloid Interface Sci. 398 (2013) 74. L.L. Yang, X.Y. Ma, N.N. Guo, Carbohydr. Polym. 90 (2012) 853. L.X. Zeng, Y.F. Chen, Q.Y. Zhang, Carbohydr. Polym. 130 (2015) 333. Y.J. Lu, P.R. Chang, P.W. Zheng, Appl. Clay Sci. 107 (2015) 21. G.K. Zhang, Y. Liu, X. Yang, Appl. Clay Sci. 29 (2005) 15. F. Wang, P.R. Chang, P.W. Zheng, X.F. Ma, Appl. Surf. Sci. 349 (2015) 251. H.L. Hong, W.T. Jiang, X.L. Zhang, Appl. Clay Sci. 42 (2008) 292. Y.F. Xie, D.Y. Qian, D.L. Wu, X.F. Ma, Chem. Eng. J. 168 (2011) 959. J.H. Wang, X. Zhang, B. Zhang, Desalination 259 (2010) 22. S.S. Mehdi, A. Khataee, A. Hassani, S. Karaca, C.R. Chim. 18 (2015) 204. Y.J. Lu, P.R. Lu, P.W. Zheng, X.F. Ma, Chem. Eng. J. 255 (2014) 49. W.B. Wang, A.Q. Wang, Carbohydr. Polym. 77 (2009) 891. L.X. Zheng, M.J. Xie, Q.Y. Zhang, Carbohydr. Polym. 123 (2015) 89. G.C. Han, Y. Han, X.Y. Wang, Appl. Surf. Sci. 317 (2014) 35. R.F. Xu, S.J. Xin, X. Zhou, Int. J. Pharm. 438 (2012) 258. D. Cheng, Y.B. Wen, L.J. Wang, Carbohydr. Polym. 123 (2015) 157. P.R. Chang, P.W. Zheng, B.X. Liu, J. Hazard. Mater. 186 (2011) 2144. J.B. Liu, J. Hao, F. Cheng, J. Environ. Eng. 141 (2015) 04014055. D.L. Wu, P.W. Zheng, P.R. Chang, X.F. Ma, Chem. Eng. J. 174 (2011) 489. W. Weng, V.G. Pol, K. Amine, Adv. Mater. 25 (2013) 1608. R. Mukherjee, S. De, J. Membr. Sci. 466 (2014) 281. S. Kilpimaa, H. Runtti, T. Kangas, U. Lassi, T. Kuokkanen, J. Ind. Eng. Chem. 21 (2015) 1354. A. Keranen, T. Leiviska, O. Hormi, J. Tanskanen, J. Environ. Manage. 147 (2015) 46. A. Sowmya, S. Meenakshi, Int. J. Biol. Macromol. 64 (2014) 224. A. Sowmya, S. Meenakshi, Int. J. Biol. Macromol. 69 (2014) 336. L. Deng, Z. Shi, J. Alloys Compd. 637 (2015) 188. J.B. Lu, D.F. Liu, J. Hao, Chem. Eng. Res. Des. 93 (2015) 652. X. Liu, L.F. Zhang, Powder Technol. 277 (2015) 112. S.M. Prabhu, S. Meenakshi, Powder Technol. 268 (2014) 306.
Please cite this article in press as: F. Wang, et al., J. Ind. Eng. Chem. (2016), http://dx.doi.org/10.1016/j.jiec.2016.07.017