w a t e r r e s e a r c h 6 9 ( 2 0 1 5 ) 2 5 2 e2 6 0
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.elsevier.com/locate/watres
Removal of Hg(II) by poly(1-vinylimidazole)-grafted Fe3O4@SiO2 magnetic nanoparticles Chao Shan, Zhiyao Ma, Meiping Tong*, Jinren Ni The Key Laboratory of Water and Sediment Sciences, Ministry of Education, College of Environmental Sciences and Engineering, Peking University, Beijing 100871, PR China
article info
abstract
Article history:
Fe3O4@SiO2 magnetic nanoparticles modified by grafting poly(1-vinylimidazole) oligomer
Received 5 September 2014
(FSPV) was fabricated as a novel adsorbent to remove Hg(II) from water. Fourier transform
Received in revised form
infra-red spectroscopy confirmed the successful grafting of oligomer, and thermogravi-
13 November 2014
metric analysis showed FSPV had a high grafting yield with organic content of 22.8%.
Accepted 18 November 2014
Transmission electron microscopy image displayed that FSPV particles were polymer-
Available online 26 November 2014
coated spheres with size of 10e20 nm. With saturation magnetization of 44.7 emu/g, FSPV particles could be easily separated from water with a simple magnetic process in
Keywords:
5 min. The Hg(II) adsorption capacity of FSPV was found to be 346 mg/g at pH 7 and 25 C in
Mercury
10 mM NaCl. Moreover, the removal of Hg(II) by FSPV was not obviously affected by solu-
Adsorption
tion pH (from 4 to 10) or humic acid (up to 8 mg/L as TOC). The presence of seven common
Magnetite
2 (up to 100 mM ionic strength) ions including Naþ, Kþ, Ca2þ, Mg2þ, Cl, NO 3 , and SO4
Imidazole
slightly increased the adsorption of Hg(II) by FSPV. X-ray photoelectron spectroscopy
Silica
analysis revealed that the N atom of the imidazole ring was responsible for the bonding
Magnetic nanoparticle
with Hg(II), whereas the bonding of Hg with N did not result in cleavage of HgeCl bond in HgCl2 and HgClOH. The regeneration of Hg(II)-loaded FSPV could be achieved with 0.5 M HCl rapidly in 10 min, and the removal of Hg(II) maintained above 94% in five consecutive adsorptionedesorption cycles. Therefore, FSPV could serve as a promising adsorbent for Hg(II) removal from water. © 2014 Elsevier Ltd. All rights reserved.
1.
Introduction
As one of the most toxic metals, mercury (Hg) has attracted increasing attention due to its persistence and bioaccumulation in the food chain (Langford and Ferner, 1999). Inappropriate discharges from anthropogenic activities, such as Au and Hg mining, smelting, electronics and battery manufacturing, and chemical industry would lead to the release of Hg into aquatic systems (Boening, 2000; Wang et al.,
* Corresponding author. Tel.: þ86 10 62756491; fax: þ86 10 62756526. E-mail address: [email protected] (M. Tong). http://dx.doi.org/10.1016/j.watres.2014.11.030 0043-1354/© 2014 Elsevier Ltd. All rights reserved.
2004; Jiang et al., 2006). Hg contamination in aquatic systems is extremely harmful to both public health and aquatic life. Thus, serious concerns have arisen in the vicinity of Hg contaminated sites, especially in Asia (Li et al., 2009). Due to its solubility and stability, Hg(II) is the common inorganic form of mercury in both the anthropogenic effluents and the aquatic environment, which could be converted into more toxic organic forms via biological methylation (Loux, 1998; Wang et al., 2004). Therefore, it is very imperative to develop efficient technologies to remove Hg(II) from water.
w a t e r r e s e a r c h 6 9 ( 2 0 1 5 ) 2 5 2 e2 6 0
Adsorption has been regarded as one of the most promising technologies for Hg(II) removal (Miretzky and Fernandez Cirelli, 2009). Among the numerous types of adsorbents, magnetic nanoparticles (MNPs) have drawn extensive attention due to the convenient separation from water with a simple magnetic process (Tang and Lo, 2013). According to the hard-soft acid-base (HSAB) theory, soft bases would be favorable ligands for immobilization of soft acid Hg(II). Thereby, MNPs (Fe3O4 or g-Fe2O3) have been previously modified with organic compounds (mainly containing nitrogen or sulfur groups) for the removal of Hg(II) from water (Figueira et al., 2011; Song et al., 2011; Mandel et al., 2012; Pan et al., 2012; Zhang et al., 2012). Although Hg(II) could be effectively removed by using these adsorbents, yet it was very difficult to regenerate and reuse these adsorbents. First, the adsorbents without silica shell are unlikely to be regenerated in acid solutions due to the dissolution of Fe. Moreover, for the adsorbents containing sulfur ligands, the tight bonds between reduced sulfur atom and Hg would also inhibit the desorption of Hg(II) from the adsorbents. For example, Hakami et al. (2012) found that the desorption of Hg(II) from thiolfunctionalized MNPs was only around 10% and 70% by using 1 M and 3 M HCl, respectively. Taking desorption into consideration, the nitrogen-containing ligands therefore would be more favorable than those containing sulfur in longterm application. Imidazole is a nitrogen-substituted aromatic heterocycle compound with strong affinity for Hg(II) (Brooks and Davidson, 1960). Furthermore, the adsorption capacity would be considerably enhanced via polymerization of imidazolebased monomers. Molina et al. (2001) have found that Hg(II) could be effectively removed from acid aqueous solutions by using poly(1-vinylimidazole) hydrogel particles. By grafting the poly(1-vinylimidazole) oligomer onto MNPs, Takafuji et al. (2004) showed that the fabricated adsorbent could adsorb Cu2þ, Ni2þ, and Co2þ. However, the removal of Hg(II) has not been investigated in this previous study. Moreover, the oligomer was grafted onto the bare g-Fe2O3, which would disable the regeneration and reuse of the adsorbent due to the dissolution of Fe in acid solutions. Accordingly, the regeneration and reuse of the adsorbent, which is of crucial importance for magnetic adsorbents, has not been concerned in this previous study. Thus, to guarantee their long-term application, it is imperative to improve the MNPs by resolving the issue of Fe dissolution during the regeneration process. By employing SiO2 as a protective shell, Wang et al. (2010) found that the dissolution of Fe from the MNPs could be greatly inhibited. Specifically, the authors showed that the leaching of Fe from the amino-functionalized Fe3O4@SiO2 nanoparticles was less than 1% in 1 M HCl for 12 h. Due to the stable chemical nature of the silica shell, the regeneration of SiO2 modified MNPs thus could be achieved by desorption of heavy metal from the absorbents in acid solutions. Clearly, coating the bare Fe3O4 first by SiO2 and then grafting poly(1vinylimidazole) oligomer onto Fe3O4@SiO2 MNPs would ensure regeneration and reuse of the fabricated Hg adsorbent. Therefore, the objective of our study is to synthesize a novel adsorbent by grafting poly(1-vinylimidazole) oligomer onto Fe3O4@SiO2 MNPs and to investigate its application for Hg(II) removal from water. In addition to the isotherm and kinetic
253
studies, the effects of solution pH, coexisting ions, and humic acid on the removal of Hg(II) were also investigated. Based on the spectroscopy studies, the adsorption mechanism was explored and elucidated. Moreover, the regeneration and reuse of the adsorbent was evaluated.
2.
Materials and methods
2.1.
Materials
Hg(II) stock solutions were synthesized by dissolving HgCl2 (GR, Xiya Reagent, Chengdu, Sichuan, China) with ultrapure water (resistivity > 18.2 MU cm) prepared with gradient water purification system (Milli-Q, Millipore, Billerica, MA, USA). 1vinylimidazole (1-VIm, 99%, Aldrich, St. Louis, MO, USA), triethoxyvinylsilane (TEOVS, 97%, Aladdin, Shanghai, China), and 2,20 -azobis(2-methylpropionitrile) (AIBN, 98%, Aladdin) were employed as the monomer, the terminal, and the initiator, respectively. Note that AIBN was purified by recrystallization from 95% ethanol (v/v) to yield needle-like white crystal prior to use. Humic acid sodium salt was from ACROS Organics (Geel, Belgium). HCl and NaOH were guaranteed reagents (GR) from Beijing Chemical Works. Absolute methanol (GR, Beijing Tongguang Fine Chemicals) was used as received. All the other chemicals used in this study were analytical reagents (AR) from Sinopharm Chemical Reagent (Shanghai, China) and Xilong Chemical Co., Ltd (Shantou, Guangdong, China). Standard solutions were purchased from the National Institute of Metrology of China.
2.2.
Adsorbent preparation
2.2.1.
Preparation of Fe3O4 magnetic nanoparticles
Fe3O4 magnetic nanoparticles were prepared via aqueous coprecipitation in a 500-mL flask. Throughout the preparation of the Fe3O4 magnetic nanoparticles, nitrogen was continually bubbled through 200 mL of solution (0.2 M FeCl3 þ 0.1 M FeSO4) to expel oxygen under vigorous mechanical agitation. Black precipitate was formed upon the slow addition of ammonium hydroxide (NH3$H2O, 25e28% wt.) at room temperature until pH reached 10. Following that, the suspension was heated up to 80 C, maintained at this temperature for 25 min, and was then cooled to room temperature in cold water bath. The resultant precipitate was separated with an external magnet, and then washed repeatedly with deionized water until the supernatant was neutral (pH 7). The magnetic nanoparticles were re-dispersed with 200 mL of deionized water by ultrasonication for 5 min.
2.2.2.
SiO2 coating of the Fe3O4 magnetic nanoparticles
SiO2 was coated onto the Fe3O4 magnetic nanoparticles according to a cost-effective method derived from previous studies (Hu et al., 2010; Wang et al., 2010). The above suspension of Fe3O4 magnetic nanoparticles was mixed with 50 mL of Na2SiO3 solution (0.35 M) and the mixed suspension was ultrasonicated for 2 min. With vigorous mechanical agitation and nitrogen flow, the suspension was heated to 80 C, and then 3 M HCl solution was added dropwisely within 20 min until the pH reached 6. The mixture was stirred for
254
w a t e r r e s e a r c h 6 9 ( 2 0 1 5 ) 2 5 2 e2 6 0
2 h at 80 C, during which the pH was constantly adjusted to 6 using 3 M HCl. Finally, the suspension was cooled to room temperature, and the coated magnetic nanoparticles were washed repeatedly with the help of magnetic separation, dried at 60 C overnight, and pulverized to obtain black powder (referred to as Fe3O4@SiO2).
2.2.3.
Polymerization of 1-VIm
The 1-VIm oligomer was obtained via radical polymerization derived from Takafuji et al. (2004). Briefly, 5 mL of 1-VIm, 0.6 mL of TEOVS, and 112 mg of the recrystallized AIBN were dissolved in 80 mL of absolute methanol. The solution was stirred at reflux temperature for 36 h, during which ultrapure nitrogen (99.999%) was bubbled through the solution. The product (referred to as pVIm) was separated by gel permeation chromatography (GPC) and analyzed with electro-spray ionization Fourier transform mass spectrometry (ESI-FTMS, APEX IV, Bruker Daltonics Inc., Billerica, MA, USA).
2.2.4.
Grafting
The pVIm oligomer was grafted onto the surface of Fe3O4@SiO2 through silanation in 200 mL of mixed solution of methanol and toluene (30:70, v/v) at reflux temperature for 24 h. After cooled to room temperature, the grafted magnetic particles were collected with a magnet and washed with organic solvents, and the washing details were provided in Table S1. Finally, the washed magnetic material was dried at 60 C overnight, and pulverized to obtain black powder (referred to as FSPV).
2.3.
Adsorbent characterization
The adsorbent constituents were identified by Fourier transform infra-red spectroscopy (FTIR, Tensor 27, Bruker, Germany), in which the solid samples were measured in transmission mode through KBr pellets and the liquid samples with attenuated total reflection accessory (ATR, MIRacle, Pike Technologies, Madison, WI, USA). The elemental contents and the chemical states of the prepared materials were characterized with X-ray photoelectron spectroscopy (XPS, Axis Ultra, Kratos Analytical Ltd., Manchester, UK) with a monochromatic Al Ka radiation source (225 W, 15 mA, 15 kV), and the organic contents were quantified by thermogravimetric analysis (TGA, Q50, TA Instruments, New Castle, DE, USA). The morphology and size of the nanoparticles were characterized by transmission electron microscope (TEM, Tecnai G2 T20, FEI, Hillsboro, OR, USA). Hysteresis loops were obtained with an alternating gradient magnetometer (AGM, 2900-04C, Princeton Measurements Corporation, Princeton, NJ, USA) to characterize the magnetic properties of the materials. Zetasizer Nano ZS90 (Malvern Instruments Ltd, Worcestershire, UK) was employed to measure the zeta potentials of the prepared adsorbent at various pH from 4 to 11.
2.4.
Hg(II) solution contained in a 100-mL Erlenmeyer flask under ultrasonic wave for 2 min. The flasks were sealed and shaken at 180 rpm under 25 C in a thermostatic orbit incubator for sufficient time to reach equilibrium. Unless stated otherwise, the initial pH of Hg(II) solution was adjusted to 7.0 ± 0.1 with NaOH, and 10 mM of NaCl served as the background electrolyte to represent the typical ionic strength in the prepared Hg(II) solutions. After the adsorption reached equilibrium, the suspension was separated with a magnet, and the supernatant was analyzed for Hg(II) concentration with atomic fluorescence spectroscopy (AFS9130, Titan Instruments, Beijing, China) and inductively coupled plasma optical emission spectroscopy (ICP-OES, Prodigy, Teledyne Leeman Labs, Hudson, NH, USA) at low concentrations (below 1 mg/L) and at high concentrations (above 1 mg/L), respectively. The amount of adsorbed Hg was calculated by conducting mass balance between initial and final Hg concentrations. All batch adsorption experiments were performed in triplicate. The adsorption isotherms of Fe3O4, Fe3O4@SiO2, and FSPV were first acquired by varying the initial Hg(II) concentration in appropriate ranges for comparison among the three materials, whereas, the modified adsorbent FSPV was solely investigated for all the rest adsorption experiments. In the adsorption kinetics study, 15 mg of FSPV were dispersed into 150 mL of 1 mg/ L Hg(II) solution, and a 1-mL aliquot was sampled and filtered through 0.22 mm polyethersulfone membrane at a series of time intervals. Effect of solution pH was investigated by adjusting the initial solution pH from 2 to 12 with HCl or NaOH, and meanwhile the release of Fe in the treated solution was determined. To investigate the effects of common cations 2 (Naþ, Kþ, Ca2þ, and Mg2þ) and anions (Cl, NO 3 , and SO4 ) on Hg(II) removal, adsorption experiments were performed in six salts solutions including NaCl, Na2SO4, NaNO3, KNO3, Ca(NO3)2, and Mg(NO3)2 with ionic strength from 0 to 100 mM. Humic acid stock solution was adjusted to pH 7 with 1 M HCl and then filtered through 0.22 mm polyethersulfone membrane prior to use. Total organic carbon (TOC) content of the filtered humic acid stock solution was measured by a combustion-type TOC meter (TOC-V CPN, Shimadzu, Kyoto, Japan).
2.5.
Adsorbent regeneration and reuse
Five consecutive adsorptionedesorption cycles were performed in triplicate to test the reusability of FSPV. Each adsorption process was carried out as in batch experiment for 12 h. Then, the adsorbent was magnetically separated from the solution and then washed gently with ultrapure water to remove the adhered solution. The collected adsorbent was then dispersed into 20 mL of 0.5 M HCl and shaken at 180 rpm shortly for 10 min to regenerate the adsorbent. Prior to the next cycle, the regenerated adsorbent was washed repeatedly with ultrapure water until the effluent was neutral (pH 7).
Adsorption experiments
All the containers and sample vials used in the adsorption experiments were glassware instead of plastic to minimize adsorption of Hg on the containers, and the glassware vessels were pretreated by soaking in 2 M HNO3 overnight prior to use. Generally, 5 mg of adsorbent were dispersed into 50 mL of
3.
Results and discussion
3.1.
Adsorbent characterization
The mass spectrum of pVIm (Fig. S1) showed that the molecular weight of pVIm distributed approximately from 600 to
w a t e r r e s e a r c h 6 9 ( 2 0 1 5 ) 2 5 2 e2 6 0
2300. Thus, the degree of polymerization of pVIm could be estimated to be between 5 and 16. This observation demonstrated that the 1-VIm has been successfully telomerized into a chain of the desired length. Comparison of the FTIR spectrum of FSPV with those of 1-VIm, Fe3O4, and Fe3O4@SiO2 (Fig. S2) clearly indicated that the pVIm oligomer has been successfully grafted onto the surface of Fe3O4@SiO2 core. To estimate the grafting yield of pVIm on FSPV, thermogravimetric analysis of Fe3O4, Fe3O4@SiO2, and FSPV was conducted and the results were presented in Fig. S3. Negligible weight loss was obtained for either Fe3O4 or Fe3O4@SiO2, whereas, two obvious weight loss stages were observed in the TGA curve of FSPV. The first stage in the temperature region from 30 to 200 C could be attributed to the loss of water in the sample, while the second stage in the temperature range from 230 to 700 C was due to the decomposition of the organic components (Lai et al., 2012). The organic content of FSPV was found to be 22.8%, which indicated the high grafting yield of pVIm. Surface element survey of FSPV by XPS (Table S2) showed that Si and N contents were 11.3 and 15.9% wt., respectively, which demonstrated the effective SiO2 coating and pVIm grafting on FSPV. In contrast, the mass concentration of Fe element on the surface of FSPV was found to be only 5.3%, which was significantly lower than the theoretical concentration in bulk Fe3O4 (72.4%). This observation implied that the Fe3O4 has been firmly coated by the pVIm-SiO2 shell. Thereby, the detection of Fe element was inhibited due to the limited probing depth of XPS. TEM images (Fig. 1) displayed that the particles of Fe3O4, Fe3O4@SiO2, and FSPV were all spherical in shape with size of 10e20 nm. The morphologies of Fe3O4 and Fe3O4@SiO2 were similar, whereas, it can be obviously observed that the FSPV particles were wrapped by a thin layer of translucent polymer coat. The hysteresis loops of Fe3O4, Fe3O4@SiO2, and FSPV (Fig. S4) indicated that all three materials were superparamagnetic, and the saturation magnetizations of Fe3O4, Fe3O4@SiO2, and FSPV were 67.5, 55.9, and 44.7 emu/g, respectively. Thus, the FSPV particles could be easily separated from water with a magnet in short time of 5 min (Fig. S5). The magnetic properties of FSPV could ensure its convenient and rapid separation from water with a simple magnetic
255
Fig. 2 e Hg(II) adsorption isotherms of Fe3O4, Fe3O4@SiO2, and FSPV at 25 C. Initial solution pH 7.0 ± 0.1, background electrolyte 10 mM NaCl, contact time 36 h.
process after application. The zeta potentials of FSPV gradually decreased from þ43.9 mV to 31.5 mV with increasing pH from 4 to 11 (Fig. S6), and the isoelectric point of FSPV was estimated at around pH 7.4.
3.2.
Adsorption isotherms
Hg(II) adsorption isotherms of Fe3O4, Fe3O4@SiO2, and FSPV were investigated and the results were shown in Fig. 2. The three isotherms were all fitted with both Langmuir and Freundlich models, respectively. The detailed information of the fitting models was provided in Supplementary Information (Text S1). The fitted parameters for all three isotherms with both models were summarized in Table S3. The experimental data of both Fe3O4 and FSPV were fitted well with Langmuir model (r2 ¼ 0.974 and 0.988, respectively), whereas Freundlich model could better describe the trend of isotherm of Fe3O4@SiO2 (r2 ¼ 0.993). The adsorption capacity of Fe3O4@SiO2 (0.34 mg/g) was smaller than that of Fe3O4 (1.14 mg/g). The decrease in adsorption capacity could be attributed to the relatively small affinity of the SiO2 shell for HgeCl complex (Bonnissel-Gissinger et al., 1999), which
Fig. 1 e TEM images of Fe3O4 (a), Fe3O4@SiO2 (b), and FSPV (c).
256
w a t e r r e s e a r c h 6 9 ( 2 0 1 5 ) 2 5 2 e2 6 0
Table 1 e Comparison of Hg(II) adsorption capacity of FSPV with those of other functionalized magnetic adsorbents. Adsorbent humic acid coated Fe3O4 dithiocarbamate functionalized magnetite particles thiol modified Fe3O4@SiO2 thiol-functionalized mesoporous silica coated Fe3O4 thiol-functionalized superparamagnetic carbon nanotubes thiol-functionalized magnetite/graphene oxide hybrid polymer-brush-grafted magnetic nanoparticles Rhodamine hydrazide modified Fe3O4 naphthalimide functionalized Fe3O4@SiO2 CoFe2O4-reduced graphene oxide poly(1-vinylimidazole)-grafted Fe3O4@SiO2 (FSPV) a b
Temperature ( C)
Adsorption capacity (mg/g)
Reference
20 22 r.t.a 22.5 25
6.0 7.0 4.0 6.0 6.5
pH
98 206 79 208 65.6
Liu et al. (2008) Figueira et al. (2011) Mandel et al. (2012) Hakami et al. (2012) Zhang et al. (2012)
n.a.b r.t.
n.a. neutral pH
290 59.4
Bao et al. (2013) Farrukh et al. (2013)
n.a. 25 25 25
n.a. n.a. 4.6 7.0
7.5 32 158 346
Wang et al. (2013) Zhu et al. (2013) Zhang et al. (2014) present study
Room temperature. Not available.
prevented the immobilization of Hg(II) on the surface of Fe3O4. In contrast, grafting of pVIm on Fe3O4@SiO2 significantly increased Hg(II) adsorption. Specifically, Hg(II) adsorption capacity of FSPV remarkably reached 346 mg/g, which was two orders of magnitude larger than that of Fe3O4, and three orders of magnitude larger than that of Fe3O4@SiO2. Clearly, the high grafting yield of pVIm onto FSPV could provide a large number of extra sites for Hg(II) adsorption, leading to the significant enhancement of adsorption capacity. Comparing to the large adsorption capacity of FSPV (346 mg/g), the trivial decrease in adsorption capacity (0.8 mg/g) induced by the SiO2 shell obviously could be negligible. Moreover, the adsorption capacity of FSPV was greater than those of other functionalized magnetic adsorbents reported previously (Table 1). The above results suggested that the FSPV fabricated in our study contained a large adsorption capacity towards Hg(II) and could serve as a promising adsorbent for Hg(II) removal from water.
3.3.
Adsorption kinetics
The adsorption kinetics of Hg(II) by FSPV was investigated and the results were depicted in Fig. 3. As can be seen, over 90% of Hg(II) was removed by FSPV within 2.5 h, and the adsorption reached equilibrium in 7 h. Similar trend of Hg(II) adsorption kinetics has also been observed by previous studies (Arkas and Tsiourvas, 2009; Farrukh et al., 2013). The kinetics experimental data were fitted with both Lagergren pseudo-first order kinetic model and the pseudo-second order kinetic model (Ho and McKay, 1999), respectively. The detailed information of the fitting models can be found in Supplementary Information (Text S2). As shown in Table S4, the calculated qe derived from the pseudo-second order kinetic model (9.35 mg/g) was closer to the experimental value (9.50 mg/g) relative to that from the pseudo-first order kinetic model (8.83 mg/g). Furthermore, the coefficient of determination (r2) of pseudo-second order kinetic model (0.964) was larger than that of the pseudo-first order model (0.943). Therefore, the adsorption kinetics of Hg(II) by FSPV could be better described with the pseudosecond order kinetic model. Moreover, the calculated rate constant (k2) of Hg(II) adsorption by FSPV was greater than those of other functionalized magnetic adsorbents (Table S5).
These observations indicated that the FSPV fabricated in our study had a good kinetics performance of Hg(II) adsorption.
3.4.
Effect of solution pH
The influence of initial solution pH (from 2 to 12) on the uptake of Hg(II) by FSPV was investigated and the results were displayed in Fig. 4. According to the thermodynamic calculation results with Visual MINTEQ software (Fig. S7), the Hg(II) species were all soluble without precipitation over the whole pH range under the experimental conditions. As can be seen from Fig. 4, the removal percentage of Hg(II) was above 94% over a broad pH range from 4 to 10. Only under extreme acidic (i.e. at pH 2 and 3) or extreme basic (i.e. at pH 11 and 12) circumstance, the removal percentage of Hg(II) dropped below 80%. The decrease in removal of Hg(II) at low pH has also been reported in previous studies, which employed other pVImcontaining adsorbents (Molina et al., 2001; Sun et al., 2013). C¸am et al. (2013) reported that the protonation constant (log KH) of the ligand 1-Vim was 3.61 at 25 C determined by potentiometric titration. The imidazole groups thus would be highly protonated at pH lower than 3. The protons would increasingly
Fig. 3 e Adsorption kinetics of Hg(II) by FSPV at 25 C. Initial solution pH 7.0, background electrolyte 10 mM NaCl, adsorbent dosage 0.1 g/L.
w a t e r r e s e a r c h 6 9 ( 2 0 1 5 ) 2 5 2 e2 6 0
Fig. 4 e Effect of solution pH on Hg(II) removal by FSPV at 25 C. Initial Hg(II) concentration 1.0 mg/L, background electrolyte 10 mM NaCl, adsorbent dosage 0.1 g/L, contact time 24 h.
compete with Hg(II) cations to be bonded by the imidazole groups with the decrease of pH (below 3), resulting in the decrease in the uptake of Hg(II) by FSPV under extreme acidic conditions (pH 2 and 3). The inhibition of Hg(II) removal in the high pH (>10) region could be attributed to the ligand competition of OH with imidazole for Hg. The affinity between Hg(II) and OH is very high, as can be seen from the large stability constant (log b2 ¼ 21.7). Consequently, under extreme basic conditions (pH 11 and 12), the mobility of Hg was significantly enhanced by forming soluble Hg(OH)2 species in the presence of large amounts of OH (Fig. S7), leading to the decrease in the adsorption of Hg(II) under these solution conditions. The concentrations of Fe released into solution after adsorption of Hg(II) at different solution pH were simultaneously determined and the results were shown in Table S6. The release of Fe from FSPV particles into the final solution was not detected at pH from 3 to 12. Only under very acidic circumstance (i.e. at pH 2), a small amount of released Fe (0.92 mg/L) was detected. The observations indicated that the SiO2 shell could effectively protect the Fe3O4 core from dissolution at low solution pH, and FSPV was a safe and stable adsorbent under the experimental conditions (except pH 2). Thus, considering both the material stability and the removal performance, the FSPV synthesized in our study is suitable for removal of Hg(II) from water over a wide range of pH from 4 to 10.
3.5.
Effect of coexisting ions
The effects of four types of cations (Naþ, Kþ, Ca2þ, and Mg2þ) 2 and three types of anions (Cl, NO 3 , and SO4 ) on the adsorption of Hg(II) by FSPV were investigated. The results (Fig. 5) demonstrated that all seven types of ions including four cations and three anions slightly increased the adsorption of Hg(II) onto FSPV. Meanwhile, enhancement of adsorption increased with increasing ionic strength up to 100 mM regardless of all seven examined ion types. It has been commonly reported that cations including Naþ, Kþ, Ca2þ, 2 Mg2þ, and anions such as NO 3 or SO4 have no obvious effect on Hg(II) adsorption (Wang et al., 2009a, 2009b; Carro et al.,
257
Fig. 5 e Effect of coexisting ions on Hg(II) removal by FSPV at 25 C. Initial Hg(II) concentration 1.0 mg/L, initial solution pH 7.0 ± 0.1, adsorbent dosage 0.1 g/L, contact time 24 h.
2010; Zhang et al., 2010; Cui et al., 2012; Zhou et al., 2014). Due to the strong complexation ability of Cl with Hg2þ to 2 form HgCl 3 and HgCl4 , many studies yet found that increasing the concentration of chloride ion (Cl) significantly inhibited the uptake of Hg(II) (Herrero et al., 2005; Wang et al., 2009a, 2009b; Carro et al., 2010; Zhang et al., 2010; Cui et al., 2012). In contrast, the removal efficiency of Hg(II) by FSPV fabricated in our study yet slightly increased with the copresence of four cations (Naþ, Kþ, Ca2þ, Mg2þ) and three anions 2 (Cl, NO 3 , and SO4 ). Since the enhanced removal of Hg(II) by FSPV with increasing ionic strength was obtained for all seven examined coexisting ions (cations and anions), we presume that the mechanism controlling the increased uptake of Hg(II) at higher ionic strength possibly could be attributed to the change in conformation of the grafted pVIm polymer. Wang et al. (2011) have thoroughly elucidated the conformational behavior of grafted weak polyelectrolyte chains under different ionic strength conditions. They found that the uncharged weak polyelectrolyte chains were in collapse state at low ionic strength, whereas, with increased ionic strength, the chains would swell to a relative extended conformation as the anions were adsorbed on the chains via nonelectrostatic forces. Similarly, we assume that the increase of solution ionic strength might also lead to the transformation of conformation of pVIm grafted on FSPV from tightly collapsed state to loosely extended form. Thus, more adsorption sites on the grafted pVIm are exposed to Hg(II) at elevated ionic strength, which would contribute to the increase of Hg(II) removal. Although the detailed mechanism driving to the enhanced Hg(II) uptake at high ionic strength still requires further exploration, the above observations yet clearly demonstrated that Hg(II) could be effectively removed by FSPV in the presence of commonly coexisting cations (Naþ, Kþ, Ca2þ, or Mg2þ) 2 and anions (Cl, NO 3 , or SO4 ). Furthermore, unlike the trend observed in most previous studies, the adsorption of Hg(II) on FSPV would be more favorable at higher ionic strength up to 100 mM, regardless of cation/anion types (including chloride ion).
258
3.6.
w a t e r r e s e a r c h 6 9 ( 2 0 1 5 ) 2 5 2 e2 6 0
Effect of humic acid
The effect of humic acid, which is a ubiquitous and representative natural organic matter (NOM) in the aquatic environment, on the adsorption of Hg(II) by FSPV was explored, and the results were listed in Table S7 in Supplementary Information. Briefly, the removal of Hg(II) remained almost constant (from 95.5% to 97.9%) in the presence of humic acid (from 0 to 8 mg/L as TOC). Obviously, the presence of humic acid had negligible influence on the removal of Hg(II) by FSPV. By serving as a competitive ligand to complex mercury, previous studies have shown that humic acid would increase the mobility of Hg(II) and thus could substantially inhibit the uptake of Hg(II) (Wang et al., 2009a, 2009b; do Nascimento and Masini, 2014). Since humic acid had high affinity for Hg(II) with large stability constant (log b) from around 11 to 19 (Gismera et al., 2007), humic acid-containing adsorbents therefore have also been extensively developed for Hg(II) removal from water (Liu et al., 2008; Zhang et al., 2010). However, our study showed that adsorption of Hg(II) by FSPV was not obviously affected by the presence of humic acid. It is reasonable to deduce that pVIm possibly has stronger affinity with Hg(II) than humic acid, or alternatively, it also has high affinity for Hg(II)-humic acid complex besides Hg(II). Under the latter circumstance, high removal of Hg(II) with the presence of humic acid by FSPV still could be achieved even if Hg(II) was complexed with humic acid. The above observation demonstrated that the synthesized FSPV could serve as an excellent adsorbent for removal of Hg(II) from water containing humic acid with concentration up to 8 mg/L as TOC.
3.7.
Adsorption mechanism
The mechanism of adsorption of Hg(II) by FSPV was investigated with XPS analysis of FSPV before and after adsorption of Hg(II), and the results were shown in Fig. S8, Table S2, and Table S8 in Supplementary Information. Comparison the survey spectra of FSPV versus that of FSPV-Hg in Fig. S8a showed that new peaks of Hg 4f, Cl 2p, Cl 2s, Hg 4d, and Hg 4p appeared in the spectrum of FPSV-Hg, which confirmed that Hg(II) was adsorbed onto the surface of FSPV. As can be seen from Fig. S8b, the N 1 s spectra could be divided into two peaks at 400.1 and 398.2 eV, which were assigned to the amine and imine nitrogen of the imidazole ring, respectively (Ang et al., 2000; Sun et al., 2013). After the adsorption of Hg(II), the maximum position of the peak at 400.1 eV slightly shifted by 0.2 eV (from 400.15 to 399.95 eV). Moreover, the peak at 400.1 eV was strengthened, while the peak at 398.2 eV was weakened. The above observations revealed that the nitrogen atom of the imidazole ring on the FSPV was responsible for the bonding with Hg(II), which was in agreement with previous study (Sun et al., 2013). The fitting results of Hg 4f (Fig. S8c) and Cl 2p (Fig. S8d) peaks indicated that the valence states of Hg and Cl remained þ2 and 1, respectively (Kadono et al., 1994; Devi et al., 2006). This observation confirmed that no redox reaction was involved during the adsorption of Hg(II) by FSPV. Element quantification results (Table S2) showed that the atomic ratio of Cl/Hg was 1.79, which implied that the adsorbed Hg(II) species were mainly HgCl2 and HgClOH. This observation displayed that the bonding of Hg with N did not
result in cleavage of HgeCl bond in HgCl2 and HgClOH. Clearly, the adsorption of Hg(II) by FSPV was not a ligand exchange process, and thus the speciation of Hg(II) would not change after the adsorption. Due to the absence of competition between the ligands, the presence of Cl could not inhibit the removal of Hg(II). Moreover, unlike other metals such as Naþ, Kþ, Ca2þ, and Mg2þ, 92% of the total Hg(II) was in the form of electroneutral species such as HgCl2 and HgClOH (Fig. S7), and thus could be adsorbed by FSPV via nonelectrostatic forces. Thereby, even in the copresence of other metals, FSPV fabricated in our study had high selectivity for Hg(II) adsorption and thus the uptake of Hg(II) was not inhibited.
3.8.
Adsorbent regeneration and reuse
The regeneration and reuse of FSPV were evaluated in five consecutive adsorption/desorption cycles, and the corresponding results were presented in Fig. 6. Preliminary experiment showed that the desorption of Hg(II) from FSPV with HCl was very fast, and equilibrium could be obtained in less than 10 min (data not shown). Thereby, the contact time of desorption process was set to be 10 min. High and stable adsorption efficiency was achieved during the regeneration and reuse processes. Specifically, the adsorption efficiency remained from 93.7% to 96.7% throughout the five consecutive cycles. Moreover, desorption efficiency was also around 90% throughout the five cycles except for the first cycle. The observation confirmed that the Hg(II)-loaded FSPV could be effectively regenerated with 0.5 M HCl in very short time (10 min), and the FSPV could be reused for efficient removal of Hg(II) from water in consecutive cycles. Meanwhile, Fe release was not detected in the treated water. During desorption with HCl, negligible release of Fe was observed. Moreover, the release concentration obviously decreased with the number of cycles (Table S9). A small proportion of Fe3O4 might not be firmly coated by SiO2. After the dissolution of this proportion of Fe3O4, FSPV tended to be stable and thus the decreased
Fig. 6 e Hg(II) adsorption and desorption efficiencies with FSPV in five consecutive cycles at 25 C. Adsorption conditions: initial Hg(II) concentration 1.0 mg/L, initial solution pH 7.0 ± 0.1, background electrolyte 10 mM NaCl, adsorbent dosage 0.1 g/L, contact time 12 h. Desorption conditions: 0.5 M HCl, dosage 4 mL/mg, contact time 10 min.
w a t e r r e s e a r c h 6 9 ( 2 0 1 5 ) 2 5 2 e2 6 0
concentration of Fe was observed with increasing desorption cycles. It is reasonable to deduce that FSPV would be increasingly stable if further cycles were conducted. It should be noted that the total Fe loss was 1.78% throughout the five cycles, which was insignificant for the adsorbent mass. Hence, SiO2 coating on FSPV successfully inhibited the dissolution of Fe during acid regeneration process. The above observations clearly showed that after Hg(II) adsorption, FSPV could be effectively regenerated with a rapid and simple process. Clearly, FSPV fabricated in our study had good potential for repeated utilization for removal of Hg(II) from water.
4.
Conclusions
For the first time, FSPV was synthesized by grafting poly(1vinylimidazole) oligomer onto Fe3O4@SiO2 particles as an efficient adsorbent to remove Hg(II) from water. Containing a magnetic core, the FSPV particles could be separated from water with a simple magnetic process. The Hg(II) adsorption capacity of FSPV was 346 mg/g, which was two orders of magnitude larger than that of Fe3O4, and three orders of magnitude larger than that of Fe3O4@SiO2. The removal of Hg(II) slightly increased with the increase of ionic strength up to 100 mM regardless of cation (Naþ, Kþ, 2 Ca2þ, or Mg2þ) or anion (Cl, NO 3 , or SO4 ) types. Unlike the trend observed in previous studies, the presence of Cl also promoted the uptake of Hg(II) by FSPV. The nitrogen atom of the imidazole ring was responsible for the bonding with Hg(II). The adsorption of Hg(II) by FSPV was not a ligand exchange process. Thus, the removal of Hg(II) could not be inhibited even in the presence of other strong ligands such as Cl. FSPV could be effectively regenerated after Hg(II) adsorption with a rapid and simple process, and exhibited good potential for repeated utilization for removal of Hg(II) from water.
Acknowledgments This work was supported by the National Natural Science Foundation of China under Grant No. 21377006 and the program for New Century Excellent Talents in University under Grant No. NCET-13-0010.
Appendix A. Supplementary data Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.watres.2014.11.030.
references
Ang, A.K.S., Kang, E.T., Neoh, K.G., Tan, K.L., Cui, C.Q., Lim, T.B., 2000. Low-temperature graft copolymerization of 1-vinyl imidazole on polyimide films with simultaneous lamination to
259
copper foils - effect of crosslinking agents. Polymer 41 (2), 489e498. Arkas, M., Tsiourvas, D., 2009. Organic/inorganic hybrid nanospheres based on hyperbranched poly(ethylene imine) encapsulated into silica for the sorption of toxic metal ions and polycyclic aromatic hydrocarbons from water. J. Hazard. Mater. 170 (1), 35e42. Bao, J., Fu, Y., Bao, Z., 2013. Thiol-functionalized magnetite/ graphene oxide hybrid as a reusable adsorbent for Hg2þ removal. Nanoscale Res. Lett. 8, 486. Boening, D.W., 2000. Ecological effects, transport, and fate of mercury: a general review. Chemosphere 40 (12), 1335e1351. Bonnissel-Gissinger, P., Alnot, M., Lickes, J.P., Ehrhardt, J.J., Behra, P., 1999. Modeling the adsorption of mercury(II) on (hydr)oxides II: a-FeOOH (goethite) and amorphous silica. J. Colloid Interface Sci. 215 (2), 313e322. Brooks, P., Davidson, N., 1960. Mercury(II) complexes of imidazole and histidine. J. Am. Chem. Soc. 82 (9), 2118e2123. _ C¸am, T., Osman, B., Kara, A., Demirbel, E., Bes‚irli, N., Irez, G., 2013. Complexation of Zn2þ with vinyl imidazole containing IMAC adsorbent: potentiometric, kinetic, and thermodynamic studies. Hacet. J. Biol. Chem. 41 (3), 207e219. Carro, L., Anagnostopoulos, V., Lodeiro, P., Barriada, J.L., Herrero, R., Sastre de Vicente, M.E., 2010. A dynamic proof of mercury elimination from solution through a combined sorption-reduction process. Bioresour. Technol. 101 (23), 8969e8974. Cui, H., Qian, Y., Li, Q., Zhang, Q., Zhai, J., 2012. Adsorption of aqueous Hg(II) by a polyaniline/attapulgite composite. Chem. Eng. J. 211, 216e223. Devi, P.S.R., Kumar, S., Verma, R., Sudersanan, M., 2006. Sorption of mercury on chemically synthesized polyaniline. J. Radioanal. Nucl. Chem. 269 (1), 217e222. do Nascimento, F.H., Masini, J.C., 2014. Influence of humic acid on adsorption of Hg(II) by vermiculite. J. Environ. Manag. 143, 1e7. Farrukh, A., Akram, A., Ghaffar, A., Hanif, S., Hamid, A., Duran, H., Yameen, B., 2013. Design of polymer-brush-grafted magnetic nanoparticles for highly efficient water remediation. ACS Appl. Mater. Interfaces 5 (9), 3784e3793. Figueira, P., Lopes, C.B., Daniel-da-Silva, A.L., Pereira, E., Duarte, A.C., Trindade, T., 2011. Removal of mercury (II) by dithiocarbamate surface functionalized magnetite particles: application to synthetic and natural spiked waters. Water Res. 45 (17), 5773e5784. Gismera, M.J., Procopio, J.R., Sevilla, M.T., 2007. Characterization of mercury e humic acids interaction by potentiometric titration with a modified carbon paste mercury sensor. Electroanalysis 19 (10), 1055e1061. Hakami, O., Zhang, Y., Banks, C.J., 2012. Thiol-functionalised mesoporous silica-coated magnetite nanoparticles for high efficiency removal and recovery of Hg from water. Water Res. 46 (12), 3913e3922. Herrero, R., Lodeiro, P., Rey-Castro, C., Vilarino, T., de Vicente, M.E.S., 2005. Removal of inorganic mercury from aqueous solutions by biomass of the marine macroalga Cystoseira baccata. Water Res. 39 (14), 3199e3210. Ho, Y.S., McKay, G., 1999. Pseudo-second order model for sorption processes. Process Biochem. 34 (5), 451e465. Hu, H., Wang, Z., Pan, L., 2010. Synthesis of monodisperse Fe3O4@silica core-shell microspheres and their application for removal of heavy metal ions from water. J. Alloys Compd. 492 (1e2), 656e661. Jiang, G.B., Shi, J.B., Feng, X.B., 2006. Mercury pollution in China. Environ. Sci. Technol. 40 (12), 3672e3678. Kadono, K., Kinugawa, K., Tanaka, H., 1994. X-ray photoelectronspectroscopy of ZnCl2 based glasses. Phys. Chem. Glasses 35 (2), 59e64.
260
w a t e r r e s e a r c h 6 9 ( 2 0 1 5 ) 2 5 2 e2 6 0
Lai, B., Yeh, C., Chen, D., 2012. Surface modification of iron oxide nanoparticles with polyarginine as a highly positively charged magnetic nano-adsorbent for fast and effective recovery of acid proteins. Process Biochem. 47 (5), 799e805. Langford, N.J., Ferner, R.E., 1999. Toxicity of mercury. J. Hum. Hypertens. 13 (10), 651e656. Li, P., Feng, X.B., Qiu, G.L., Shang, L.H., Li, Z.G., 2009. Mercury pollution in Asia: a review of the contaminated sites. J. Hazard. Mater. 168 (2e3), 591e601. Liu, J., Zhao, Z., Jiang, G., 2008. Coating Fe3O4 magnetic nanoparticles with humic acid for high efficient removal of heavy metals in water. Environ. Sci. Technol. 42 (18), 6949e6954. Loux, N.T., 1998. An assessment of mercury-species-dependent binding with natural organic carbon. Chem. Speciat. Bioavailab. 10 (4), 127e136. Mandel, K., Hutter, F., Gellermann, C., Sextl, G., 2012. Modified superparamagnetic nanocomposite microparticles for highly selective Hg-II or Cu-II separation and recovery from aqueous solutions. ACS Appl. Mater. Interfaces 4 (10), 5633e5642. Miretzky, P., Fernandez Cirelli, A., 2009. Hg(II) removal from water by chitosan and chitosan derivatives: a review. J. Hazard. Mater. 167 (1e3), 10e23. Molina, M.J., Gomez-Anton, M.R., Rivas, B.L., Maturana, H.A., Pierola, I.F., 2001. Removal of Hg(II) from acid aqueous solutions by poly(N-vinylimidazole) hydrogel. J. Appl. Polym. Sci. 79 (8), 1467e1475. Pan, S., Shen, H., Xu, Q., Luo, J., Hu, M., 2012. Surface mercapto engineered magnetic Fe3O4 nanoadsorbent for the removal of mercury from aqueous solutions. J. Colloid Interface Sci. 365 (1), 204e212. Song, J., Kong, H., Jang, J., 2011. Adsorption of heavy metal ions from aqueous solution by polyrhodanine-encapsulated magnetic nanoparticles. J. Colloid Interface Sci. 359 (2), 505e511. Sun, J., Chen, Z., Ge, M., Xu, L., Zhai, M., 2013. Selective adsorption of Hg(II) by g-radiation synthesized silica-graft-vinyl imidazole adsorbent. J. Hazard. Mater. 244, 94e101. Takafuji, M., Ide, S., Ihara, H., Xu, Z.H., 2004. Preparation of poly(1-vinylimidazole)-grafted magnetic nanoparticles and their application for removal of metal ions. Chem. Mater. 16 (10), 1977e1983.
Tang, S.C.N., Lo, I.M.C., 2013. Magnetic nanoparticles: essential factors for sustainable environmental applications. Water Res. 47 (8), 2613e2632. Wang, J., Deng, B., Chen, H., Wang, X., Zheng, J., 2009a. Removal of aqueous Hg(II) by polyaniline: sorption characteristics and mechanisms. Environ. Sci. Technol. 43 (14), 5223e5228. Wang, J., Deng, B., Wang, X., Zheng, J., 2009b. Adsorption of aqueous Hg(II) by sulfur-impregnated activated carbon. Environ. Eng. Sci. 26 (12), 1693e1699. Wang, J., Zheng, S., Shao, Y., Liu, J., Xu, Z., Zhu, D., 2010. Aminofunctionalized Fe3O4@SiO2 core-shell magnetic nanomaterial as a novel adsorbent for aqueous heavy metals removal. J. Colloid Interface Sci. 349 (1), 293e299. Wang, Q.R., Kim, D., Dionysiou, D.D., Sorial, G.A., Timberlake, D., 2004. Sources and remediation for mercury contamination in aquatic systems - a literature review. Environ. Pollut. 131 (2), 323e336. Wang, X., Liu, G., Zhang, G., 2011. Conformational behavior of grafted weak polyelectrolyte chains: effects of counterion condensation and nonelectrostatic anion adsorption. Langmuir 27 (16), 9895e9901. Wang, Z., Wu, D., Wu, G., Yang, N., Wu, A., 2013. Modifying Fe3O4 microspheres with rhodamine hydrazide for selective detection and removal of Hg2þ ion in water. J. Hazard. Mater. 244, 621e627. Zhang, C., Sui, J., Li, J., Tang, Y., Cai, W., 2012. Efficient removal of heavy metal ions by thiol-functionalized superparamagnetic carbon nanotubes. Chem. Eng. J. 210, 45e52. Zhang, Y., Li, Q., Sun, L., Tang, R., Zhai, J., 2010. High efficient removal of mercury from aqueous solution by polyaniline/humic acid nanocomposite. J. Hazard. Mater. 175 (1e3), 404e409. Zhang, Y.K., Yan, L.G., Xu, W.Y., Guo, X.Y., Cui, L.M., Gao, L., Wei, Q., Du, B., 2014. Adsorption of Pb(II) and Hg(II) from aqueous solution using magnetic CoFe2O4-reduced graphene oxide. J. Mol. Liq. 191, 177e182. Zhou, Y., Zhang, J., Luo, X., Lin, X., 2014. Adsorption of Hg(II) in aqueous solutions using mercapto-functionalized alkali lignin. J. Appl. Polym. Sci. 131 (18). Zhu, B., Zhao, J., Yu, H., Yan, L., Wei, Q., Du, B., 2013. Naphthalimide-functionalized Fe3O4@SiO2 core/shell nanoparticles for selective and sensitive adsorption and detection of Hg2þ. Chem. Eng. J. 219, 411e418.
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.elsevier.com/locate/watres
Removal of Hg(II) by poly(1-vinylimidazole)-grafted Fe3O4@SiO2 magnetic nanoparticles Chao Shan, Zhiyao Ma, Meiping Tong*, Jinren Ni The Key Laboratory of Water and Sediment Sciences, Ministry of Education, College of Environmental Sciences and Engineering, Peking University, Beijing 100871, PR China
article info
abstract
Article history:
Fe3O4@SiO2 magnetic nanoparticles modified by grafting poly(1-vinylimidazole) oligomer
Received 5 September 2014
(FSPV) was fabricated as a novel adsorbent to remove Hg(II) from water. Fourier transform
Received in revised form
infra-red spectroscopy confirmed the successful grafting of oligomer, and thermogravi-
13 November 2014
metric analysis showed FSPV had a high grafting yield with organic content of 22.8%.
Accepted 18 November 2014
Transmission electron microscopy image displayed that FSPV particles were polymer-
Available online 26 November 2014
coated spheres with size of 10e20 nm. With saturation magnetization of 44.7 emu/g, FSPV particles could be easily separated from water with a simple magnetic process in
Keywords:
5 min. The Hg(II) adsorption capacity of FSPV was found to be 346 mg/g at pH 7 and 25 C in
Mercury
10 mM NaCl. Moreover, the removal of Hg(II) by FSPV was not obviously affected by solu-
Adsorption
tion pH (from 4 to 10) or humic acid (up to 8 mg/L as TOC). The presence of seven common
Magnetite
2 (up to 100 mM ionic strength) ions including Naþ, Kþ, Ca2þ, Mg2þ, Cl, NO 3 , and SO4
Imidazole
slightly increased the adsorption of Hg(II) by FSPV. X-ray photoelectron spectroscopy
Silica
analysis revealed that the N atom of the imidazole ring was responsible for the bonding
Magnetic nanoparticle
with Hg(II), whereas the bonding of Hg with N did not result in cleavage of HgeCl bond in HgCl2 and HgClOH. The regeneration of Hg(II)-loaded FSPV could be achieved with 0.5 M HCl rapidly in 10 min, and the removal of Hg(II) maintained above 94% in five consecutive adsorptionedesorption cycles. Therefore, FSPV could serve as a promising adsorbent for Hg(II) removal from water. © 2014 Elsevier Ltd. All rights reserved.
1.
Introduction
As one of the most toxic metals, mercury (Hg) has attracted increasing attention due to its persistence and bioaccumulation in the food chain (Langford and Ferner, 1999). Inappropriate discharges from anthropogenic activities, such as Au and Hg mining, smelting, electronics and battery manufacturing, and chemical industry would lead to the release of Hg into aquatic systems (Boening, 2000; Wang et al.,
* Corresponding author. Tel.: þ86 10 62756491; fax: þ86 10 62756526. E-mail address: [email protected] (M. Tong). http://dx.doi.org/10.1016/j.watres.2014.11.030 0043-1354/© 2014 Elsevier Ltd. All rights reserved.
2004; Jiang et al., 2006). Hg contamination in aquatic systems is extremely harmful to both public health and aquatic life. Thus, serious concerns have arisen in the vicinity of Hg contaminated sites, especially in Asia (Li et al., 2009). Due to its solubility and stability, Hg(II) is the common inorganic form of mercury in both the anthropogenic effluents and the aquatic environment, which could be converted into more toxic organic forms via biological methylation (Loux, 1998; Wang et al., 2004). Therefore, it is very imperative to develop efficient technologies to remove Hg(II) from water.
w a t e r r e s e a r c h 6 9 ( 2 0 1 5 ) 2 5 2 e2 6 0
Adsorption has been regarded as one of the most promising technologies for Hg(II) removal (Miretzky and Fernandez Cirelli, 2009). Among the numerous types of adsorbents, magnetic nanoparticles (MNPs) have drawn extensive attention due to the convenient separation from water with a simple magnetic process (Tang and Lo, 2013). According to the hard-soft acid-base (HSAB) theory, soft bases would be favorable ligands for immobilization of soft acid Hg(II). Thereby, MNPs (Fe3O4 or g-Fe2O3) have been previously modified with organic compounds (mainly containing nitrogen or sulfur groups) for the removal of Hg(II) from water (Figueira et al., 2011; Song et al., 2011; Mandel et al., 2012; Pan et al., 2012; Zhang et al., 2012). Although Hg(II) could be effectively removed by using these adsorbents, yet it was very difficult to regenerate and reuse these adsorbents. First, the adsorbents without silica shell are unlikely to be regenerated in acid solutions due to the dissolution of Fe. Moreover, for the adsorbents containing sulfur ligands, the tight bonds between reduced sulfur atom and Hg would also inhibit the desorption of Hg(II) from the adsorbents. For example, Hakami et al. (2012) found that the desorption of Hg(II) from thiolfunctionalized MNPs was only around 10% and 70% by using 1 M and 3 M HCl, respectively. Taking desorption into consideration, the nitrogen-containing ligands therefore would be more favorable than those containing sulfur in longterm application. Imidazole is a nitrogen-substituted aromatic heterocycle compound with strong affinity for Hg(II) (Brooks and Davidson, 1960). Furthermore, the adsorption capacity would be considerably enhanced via polymerization of imidazolebased monomers. Molina et al. (2001) have found that Hg(II) could be effectively removed from acid aqueous solutions by using poly(1-vinylimidazole) hydrogel particles. By grafting the poly(1-vinylimidazole) oligomer onto MNPs, Takafuji et al. (2004) showed that the fabricated adsorbent could adsorb Cu2þ, Ni2þ, and Co2þ. However, the removal of Hg(II) has not been investigated in this previous study. Moreover, the oligomer was grafted onto the bare g-Fe2O3, which would disable the regeneration and reuse of the adsorbent due to the dissolution of Fe in acid solutions. Accordingly, the regeneration and reuse of the adsorbent, which is of crucial importance for magnetic adsorbents, has not been concerned in this previous study. Thus, to guarantee their long-term application, it is imperative to improve the MNPs by resolving the issue of Fe dissolution during the regeneration process. By employing SiO2 as a protective shell, Wang et al. (2010) found that the dissolution of Fe from the MNPs could be greatly inhibited. Specifically, the authors showed that the leaching of Fe from the amino-functionalized Fe3O4@SiO2 nanoparticles was less than 1% in 1 M HCl for 12 h. Due to the stable chemical nature of the silica shell, the regeneration of SiO2 modified MNPs thus could be achieved by desorption of heavy metal from the absorbents in acid solutions. Clearly, coating the bare Fe3O4 first by SiO2 and then grafting poly(1vinylimidazole) oligomer onto Fe3O4@SiO2 MNPs would ensure regeneration and reuse of the fabricated Hg adsorbent. Therefore, the objective of our study is to synthesize a novel adsorbent by grafting poly(1-vinylimidazole) oligomer onto Fe3O4@SiO2 MNPs and to investigate its application for Hg(II) removal from water. In addition to the isotherm and kinetic
253
studies, the effects of solution pH, coexisting ions, and humic acid on the removal of Hg(II) were also investigated. Based on the spectroscopy studies, the adsorption mechanism was explored and elucidated. Moreover, the regeneration and reuse of the adsorbent was evaluated.
2.
Materials and methods
2.1.
Materials
Hg(II) stock solutions were synthesized by dissolving HgCl2 (GR, Xiya Reagent, Chengdu, Sichuan, China) with ultrapure water (resistivity > 18.2 MU cm) prepared with gradient water purification system (Milli-Q, Millipore, Billerica, MA, USA). 1vinylimidazole (1-VIm, 99%, Aldrich, St. Louis, MO, USA), triethoxyvinylsilane (TEOVS, 97%, Aladdin, Shanghai, China), and 2,20 -azobis(2-methylpropionitrile) (AIBN, 98%, Aladdin) were employed as the monomer, the terminal, and the initiator, respectively. Note that AIBN was purified by recrystallization from 95% ethanol (v/v) to yield needle-like white crystal prior to use. Humic acid sodium salt was from ACROS Organics (Geel, Belgium). HCl and NaOH were guaranteed reagents (GR) from Beijing Chemical Works. Absolute methanol (GR, Beijing Tongguang Fine Chemicals) was used as received. All the other chemicals used in this study were analytical reagents (AR) from Sinopharm Chemical Reagent (Shanghai, China) and Xilong Chemical Co., Ltd (Shantou, Guangdong, China). Standard solutions were purchased from the National Institute of Metrology of China.
2.2.
Adsorbent preparation
2.2.1.
Preparation of Fe3O4 magnetic nanoparticles
Fe3O4 magnetic nanoparticles were prepared via aqueous coprecipitation in a 500-mL flask. Throughout the preparation of the Fe3O4 magnetic nanoparticles, nitrogen was continually bubbled through 200 mL of solution (0.2 M FeCl3 þ 0.1 M FeSO4) to expel oxygen under vigorous mechanical agitation. Black precipitate was formed upon the slow addition of ammonium hydroxide (NH3$H2O, 25e28% wt.) at room temperature until pH reached 10. Following that, the suspension was heated up to 80 C, maintained at this temperature for 25 min, and was then cooled to room temperature in cold water bath. The resultant precipitate was separated with an external magnet, and then washed repeatedly with deionized water until the supernatant was neutral (pH 7). The magnetic nanoparticles were re-dispersed with 200 mL of deionized water by ultrasonication for 5 min.
2.2.2.
SiO2 coating of the Fe3O4 magnetic nanoparticles
SiO2 was coated onto the Fe3O4 magnetic nanoparticles according to a cost-effective method derived from previous studies (Hu et al., 2010; Wang et al., 2010). The above suspension of Fe3O4 magnetic nanoparticles was mixed with 50 mL of Na2SiO3 solution (0.35 M) and the mixed suspension was ultrasonicated for 2 min. With vigorous mechanical agitation and nitrogen flow, the suspension was heated to 80 C, and then 3 M HCl solution was added dropwisely within 20 min until the pH reached 6. The mixture was stirred for
254
w a t e r r e s e a r c h 6 9 ( 2 0 1 5 ) 2 5 2 e2 6 0
2 h at 80 C, during which the pH was constantly adjusted to 6 using 3 M HCl. Finally, the suspension was cooled to room temperature, and the coated magnetic nanoparticles were washed repeatedly with the help of magnetic separation, dried at 60 C overnight, and pulverized to obtain black powder (referred to as Fe3O4@SiO2).
2.2.3.
Polymerization of 1-VIm
The 1-VIm oligomer was obtained via radical polymerization derived from Takafuji et al. (2004). Briefly, 5 mL of 1-VIm, 0.6 mL of TEOVS, and 112 mg of the recrystallized AIBN were dissolved in 80 mL of absolute methanol. The solution was stirred at reflux temperature for 36 h, during which ultrapure nitrogen (99.999%) was bubbled through the solution. The product (referred to as pVIm) was separated by gel permeation chromatography (GPC) and analyzed with electro-spray ionization Fourier transform mass spectrometry (ESI-FTMS, APEX IV, Bruker Daltonics Inc., Billerica, MA, USA).
2.2.4.
Grafting
The pVIm oligomer was grafted onto the surface of Fe3O4@SiO2 through silanation in 200 mL of mixed solution of methanol and toluene (30:70, v/v) at reflux temperature for 24 h. After cooled to room temperature, the grafted magnetic particles were collected with a magnet and washed with organic solvents, and the washing details were provided in Table S1. Finally, the washed magnetic material was dried at 60 C overnight, and pulverized to obtain black powder (referred to as FSPV).
2.3.
Adsorbent characterization
The adsorbent constituents were identified by Fourier transform infra-red spectroscopy (FTIR, Tensor 27, Bruker, Germany), in which the solid samples were measured in transmission mode through KBr pellets and the liquid samples with attenuated total reflection accessory (ATR, MIRacle, Pike Technologies, Madison, WI, USA). The elemental contents and the chemical states of the prepared materials were characterized with X-ray photoelectron spectroscopy (XPS, Axis Ultra, Kratos Analytical Ltd., Manchester, UK) with a monochromatic Al Ka radiation source (225 W, 15 mA, 15 kV), and the organic contents were quantified by thermogravimetric analysis (TGA, Q50, TA Instruments, New Castle, DE, USA). The morphology and size of the nanoparticles were characterized by transmission electron microscope (TEM, Tecnai G2 T20, FEI, Hillsboro, OR, USA). Hysteresis loops were obtained with an alternating gradient magnetometer (AGM, 2900-04C, Princeton Measurements Corporation, Princeton, NJ, USA) to characterize the magnetic properties of the materials. Zetasizer Nano ZS90 (Malvern Instruments Ltd, Worcestershire, UK) was employed to measure the zeta potentials of the prepared adsorbent at various pH from 4 to 11.
2.4.
Hg(II) solution contained in a 100-mL Erlenmeyer flask under ultrasonic wave for 2 min. The flasks were sealed and shaken at 180 rpm under 25 C in a thermostatic orbit incubator for sufficient time to reach equilibrium. Unless stated otherwise, the initial pH of Hg(II) solution was adjusted to 7.0 ± 0.1 with NaOH, and 10 mM of NaCl served as the background electrolyte to represent the typical ionic strength in the prepared Hg(II) solutions. After the adsorption reached equilibrium, the suspension was separated with a magnet, and the supernatant was analyzed for Hg(II) concentration with atomic fluorescence spectroscopy (AFS9130, Titan Instruments, Beijing, China) and inductively coupled plasma optical emission spectroscopy (ICP-OES, Prodigy, Teledyne Leeman Labs, Hudson, NH, USA) at low concentrations (below 1 mg/L) and at high concentrations (above 1 mg/L), respectively. The amount of adsorbed Hg was calculated by conducting mass balance between initial and final Hg concentrations. All batch adsorption experiments were performed in triplicate. The adsorption isotherms of Fe3O4, Fe3O4@SiO2, and FSPV were first acquired by varying the initial Hg(II) concentration in appropriate ranges for comparison among the three materials, whereas, the modified adsorbent FSPV was solely investigated for all the rest adsorption experiments. In the adsorption kinetics study, 15 mg of FSPV were dispersed into 150 mL of 1 mg/ L Hg(II) solution, and a 1-mL aliquot was sampled and filtered through 0.22 mm polyethersulfone membrane at a series of time intervals. Effect of solution pH was investigated by adjusting the initial solution pH from 2 to 12 with HCl or NaOH, and meanwhile the release of Fe in the treated solution was determined. To investigate the effects of common cations 2 (Naþ, Kþ, Ca2þ, and Mg2þ) and anions (Cl, NO 3 , and SO4 ) on Hg(II) removal, adsorption experiments were performed in six salts solutions including NaCl, Na2SO4, NaNO3, KNO3, Ca(NO3)2, and Mg(NO3)2 with ionic strength from 0 to 100 mM. Humic acid stock solution was adjusted to pH 7 with 1 M HCl and then filtered through 0.22 mm polyethersulfone membrane prior to use. Total organic carbon (TOC) content of the filtered humic acid stock solution was measured by a combustion-type TOC meter (TOC-V CPN, Shimadzu, Kyoto, Japan).
2.5.
Adsorbent regeneration and reuse
Five consecutive adsorptionedesorption cycles were performed in triplicate to test the reusability of FSPV. Each adsorption process was carried out as in batch experiment for 12 h. Then, the adsorbent was magnetically separated from the solution and then washed gently with ultrapure water to remove the adhered solution. The collected adsorbent was then dispersed into 20 mL of 0.5 M HCl and shaken at 180 rpm shortly for 10 min to regenerate the adsorbent. Prior to the next cycle, the regenerated adsorbent was washed repeatedly with ultrapure water until the effluent was neutral (pH 7).
Adsorption experiments
All the containers and sample vials used in the adsorption experiments were glassware instead of plastic to minimize adsorption of Hg on the containers, and the glassware vessels were pretreated by soaking in 2 M HNO3 overnight prior to use. Generally, 5 mg of adsorbent were dispersed into 50 mL of
3.
Results and discussion
3.1.
Adsorbent characterization
The mass spectrum of pVIm (Fig. S1) showed that the molecular weight of pVIm distributed approximately from 600 to
w a t e r r e s e a r c h 6 9 ( 2 0 1 5 ) 2 5 2 e2 6 0
2300. Thus, the degree of polymerization of pVIm could be estimated to be between 5 and 16. This observation demonstrated that the 1-VIm has been successfully telomerized into a chain of the desired length. Comparison of the FTIR spectrum of FSPV with those of 1-VIm, Fe3O4, and Fe3O4@SiO2 (Fig. S2) clearly indicated that the pVIm oligomer has been successfully grafted onto the surface of Fe3O4@SiO2 core. To estimate the grafting yield of pVIm on FSPV, thermogravimetric analysis of Fe3O4, Fe3O4@SiO2, and FSPV was conducted and the results were presented in Fig. S3. Negligible weight loss was obtained for either Fe3O4 or Fe3O4@SiO2, whereas, two obvious weight loss stages were observed in the TGA curve of FSPV. The first stage in the temperature region from 30 to 200 C could be attributed to the loss of water in the sample, while the second stage in the temperature range from 230 to 700 C was due to the decomposition of the organic components (Lai et al., 2012). The organic content of FSPV was found to be 22.8%, which indicated the high grafting yield of pVIm. Surface element survey of FSPV by XPS (Table S2) showed that Si and N contents were 11.3 and 15.9% wt., respectively, which demonstrated the effective SiO2 coating and pVIm grafting on FSPV. In contrast, the mass concentration of Fe element on the surface of FSPV was found to be only 5.3%, which was significantly lower than the theoretical concentration in bulk Fe3O4 (72.4%). This observation implied that the Fe3O4 has been firmly coated by the pVIm-SiO2 shell. Thereby, the detection of Fe element was inhibited due to the limited probing depth of XPS. TEM images (Fig. 1) displayed that the particles of Fe3O4, Fe3O4@SiO2, and FSPV were all spherical in shape with size of 10e20 nm. The morphologies of Fe3O4 and Fe3O4@SiO2 were similar, whereas, it can be obviously observed that the FSPV particles were wrapped by a thin layer of translucent polymer coat. The hysteresis loops of Fe3O4, Fe3O4@SiO2, and FSPV (Fig. S4) indicated that all three materials were superparamagnetic, and the saturation magnetizations of Fe3O4, Fe3O4@SiO2, and FSPV were 67.5, 55.9, and 44.7 emu/g, respectively. Thus, the FSPV particles could be easily separated from water with a magnet in short time of 5 min (Fig. S5). The magnetic properties of FSPV could ensure its convenient and rapid separation from water with a simple magnetic
255
Fig. 2 e Hg(II) adsorption isotherms of Fe3O4, Fe3O4@SiO2, and FSPV at 25 C. Initial solution pH 7.0 ± 0.1, background electrolyte 10 mM NaCl, contact time 36 h.
process after application. The zeta potentials of FSPV gradually decreased from þ43.9 mV to 31.5 mV with increasing pH from 4 to 11 (Fig. S6), and the isoelectric point of FSPV was estimated at around pH 7.4.
3.2.
Adsorption isotherms
Hg(II) adsorption isotherms of Fe3O4, Fe3O4@SiO2, and FSPV were investigated and the results were shown in Fig. 2. The three isotherms were all fitted with both Langmuir and Freundlich models, respectively. The detailed information of the fitting models was provided in Supplementary Information (Text S1). The fitted parameters for all three isotherms with both models were summarized in Table S3. The experimental data of both Fe3O4 and FSPV were fitted well with Langmuir model (r2 ¼ 0.974 and 0.988, respectively), whereas Freundlich model could better describe the trend of isotherm of Fe3O4@SiO2 (r2 ¼ 0.993). The adsorption capacity of Fe3O4@SiO2 (0.34 mg/g) was smaller than that of Fe3O4 (1.14 mg/g). The decrease in adsorption capacity could be attributed to the relatively small affinity of the SiO2 shell for HgeCl complex (Bonnissel-Gissinger et al., 1999), which
Fig. 1 e TEM images of Fe3O4 (a), Fe3O4@SiO2 (b), and FSPV (c).
256
w a t e r r e s e a r c h 6 9 ( 2 0 1 5 ) 2 5 2 e2 6 0
Table 1 e Comparison of Hg(II) adsorption capacity of FSPV with those of other functionalized magnetic adsorbents. Adsorbent humic acid coated Fe3O4 dithiocarbamate functionalized magnetite particles thiol modified Fe3O4@SiO2 thiol-functionalized mesoporous silica coated Fe3O4 thiol-functionalized superparamagnetic carbon nanotubes thiol-functionalized magnetite/graphene oxide hybrid polymer-brush-grafted magnetic nanoparticles Rhodamine hydrazide modified Fe3O4 naphthalimide functionalized Fe3O4@SiO2 CoFe2O4-reduced graphene oxide poly(1-vinylimidazole)-grafted Fe3O4@SiO2 (FSPV) a b
Temperature ( C)
Adsorption capacity (mg/g)
Reference
20 22 r.t.a 22.5 25
6.0 7.0 4.0 6.0 6.5
pH
98 206 79 208 65.6
Liu et al. (2008) Figueira et al. (2011) Mandel et al. (2012) Hakami et al. (2012) Zhang et al. (2012)
n.a.b r.t.
n.a. neutral pH
290 59.4
Bao et al. (2013) Farrukh et al. (2013)
n.a. 25 25 25
n.a. n.a. 4.6 7.0
7.5 32 158 346
Wang et al. (2013) Zhu et al. (2013) Zhang et al. (2014) present study
Room temperature. Not available.
prevented the immobilization of Hg(II) on the surface of Fe3O4. In contrast, grafting of pVIm on Fe3O4@SiO2 significantly increased Hg(II) adsorption. Specifically, Hg(II) adsorption capacity of FSPV remarkably reached 346 mg/g, which was two orders of magnitude larger than that of Fe3O4, and three orders of magnitude larger than that of Fe3O4@SiO2. Clearly, the high grafting yield of pVIm onto FSPV could provide a large number of extra sites for Hg(II) adsorption, leading to the significant enhancement of adsorption capacity. Comparing to the large adsorption capacity of FSPV (346 mg/g), the trivial decrease in adsorption capacity (0.8 mg/g) induced by the SiO2 shell obviously could be negligible. Moreover, the adsorption capacity of FSPV was greater than those of other functionalized magnetic adsorbents reported previously (Table 1). The above results suggested that the FSPV fabricated in our study contained a large adsorption capacity towards Hg(II) and could serve as a promising adsorbent for Hg(II) removal from water.
3.3.
Adsorption kinetics
The adsorption kinetics of Hg(II) by FSPV was investigated and the results were depicted in Fig. 3. As can be seen, over 90% of Hg(II) was removed by FSPV within 2.5 h, and the adsorption reached equilibrium in 7 h. Similar trend of Hg(II) adsorption kinetics has also been observed by previous studies (Arkas and Tsiourvas, 2009; Farrukh et al., 2013). The kinetics experimental data were fitted with both Lagergren pseudo-first order kinetic model and the pseudo-second order kinetic model (Ho and McKay, 1999), respectively. The detailed information of the fitting models can be found in Supplementary Information (Text S2). As shown in Table S4, the calculated qe derived from the pseudo-second order kinetic model (9.35 mg/g) was closer to the experimental value (9.50 mg/g) relative to that from the pseudo-first order kinetic model (8.83 mg/g). Furthermore, the coefficient of determination (r2) of pseudo-second order kinetic model (0.964) was larger than that of the pseudo-first order model (0.943). Therefore, the adsorption kinetics of Hg(II) by FSPV could be better described with the pseudosecond order kinetic model. Moreover, the calculated rate constant (k2) of Hg(II) adsorption by FSPV was greater than those of other functionalized magnetic adsorbents (Table S5).
These observations indicated that the FSPV fabricated in our study had a good kinetics performance of Hg(II) adsorption.
3.4.
Effect of solution pH
The influence of initial solution pH (from 2 to 12) on the uptake of Hg(II) by FSPV was investigated and the results were displayed in Fig. 4. According to the thermodynamic calculation results with Visual MINTEQ software (Fig. S7), the Hg(II) species were all soluble without precipitation over the whole pH range under the experimental conditions. As can be seen from Fig. 4, the removal percentage of Hg(II) was above 94% over a broad pH range from 4 to 10. Only under extreme acidic (i.e. at pH 2 and 3) or extreme basic (i.e. at pH 11 and 12) circumstance, the removal percentage of Hg(II) dropped below 80%. The decrease in removal of Hg(II) at low pH has also been reported in previous studies, which employed other pVImcontaining adsorbents (Molina et al., 2001; Sun et al., 2013). C¸am et al. (2013) reported that the protonation constant (log KH) of the ligand 1-Vim was 3.61 at 25 C determined by potentiometric titration. The imidazole groups thus would be highly protonated at pH lower than 3. The protons would increasingly
Fig. 3 e Adsorption kinetics of Hg(II) by FSPV at 25 C. Initial solution pH 7.0, background electrolyte 10 mM NaCl, adsorbent dosage 0.1 g/L.
w a t e r r e s e a r c h 6 9 ( 2 0 1 5 ) 2 5 2 e2 6 0
Fig. 4 e Effect of solution pH on Hg(II) removal by FSPV at 25 C. Initial Hg(II) concentration 1.0 mg/L, background electrolyte 10 mM NaCl, adsorbent dosage 0.1 g/L, contact time 24 h.
compete with Hg(II) cations to be bonded by the imidazole groups with the decrease of pH (below 3), resulting in the decrease in the uptake of Hg(II) by FSPV under extreme acidic conditions (pH 2 and 3). The inhibition of Hg(II) removal in the high pH (>10) region could be attributed to the ligand competition of OH with imidazole for Hg. The affinity between Hg(II) and OH is very high, as can be seen from the large stability constant (log b2 ¼ 21.7). Consequently, under extreme basic conditions (pH 11 and 12), the mobility of Hg was significantly enhanced by forming soluble Hg(OH)2 species in the presence of large amounts of OH (Fig. S7), leading to the decrease in the adsorption of Hg(II) under these solution conditions. The concentrations of Fe released into solution after adsorption of Hg(II) at different solution pH were simultaneously determined and the results were shown in Table S6. The release of Fe from FSPV particles into the final solution was not detected at pH from 3 to 12. Only under very acidic circumstance (i.e. at pH 2), a small amount of released Fe (0.92 mg/L) was detected. The observations indicated that the SiO2 shell could effectively protect the Fe3O4 core from dissolution at low solution pH, and FSPV was a safe and stable adsorbent under the experimental conditions (except pH 2). Thus, considering both the material stability and the removal performance, the FSPV synthesized in our study is suitable for removal of Hg(II) from water over a wide range of pH from 4 to 10.
3.5.
Effect of coexisting ions
The effects of four types of cations (Naþ, Kþ, Ca2þ, and Mg2þ) 2 and three types of anions (Cl, NO 3 , and SO4 ) on the adsorption of Hg(II) by FSPV were investigated. The results (Fig. 5) demonstrated that all seven types of ions including four cations and three anions slightly increased the adsorption of Hg(II) onto FSPV. Meanwhile, enhancement of adsorption increased with increasing ionic strength up to 100 mM regardless of all seven examined ion types. It has been commonly reported that cations including Naþ, Kþ, Ca2þ, 2 Mg2þ, and anions such as NO 3 or SO4 have no obvious effect on Hg(II) adsorption (Wang et al., 2009a, 2009b; Carro et al.,
257
Fig. 5 e Effect of coexisting ions on Hg(II) removal by FSPV at 25 C. Initial Hg(II) concentration 1.0 mg/L, initial solution pH 7.0 ± 0.1, adsorbent dosage 0.1 g/L, contact time 24 h.
2010; Zhang et al., 2010; Cui et al., 2012; Zhou et al., 2014). Due to the strong complexation ability of Cl with Hg2þ to 2 form HgCl 3 and HgCl4 , many studies yet found that increasing the concentration of chloride ion (Cl) significantly inhibited the uptake of Hg(II) (Herrero et al., 2005; Wang et al., 2009a, 2009b; Carro et al., 2010; Zhang et al., 2010; Cui et al., 2012). In contrast, the removal efficiency of Hg(II) by FSPV fabricated in our study yet slightly increased with the copresence of four cations (Naþ, Kþ, Ca2þ, Mg2þ) and three anions 2 (Cl, NO 3 , and SO4 ). Since the enhanced removal of Hg(II) by FSPV with increasing ionic strength was obtained for all seven examined coexisting ions (cations and anions), we presume that the mechanism controlling the increased uptake of Hg(II) at higher ionic strength possibly could be attributed to the change in conformation of the grafted pVIm polymer. Wang et al. (2011) have thoroughly elucidated the conformational behavior of grafted weak polyelectrolyte chains under different ionic strength conditions. They found that the uncharged weak polyelectrolyte chains were in collapse state at low ionic strength, whereas, with increased ionic strength, the chains would swell to a relative extended conformation as the anions were adsorbed on the chains via nonelectrostatic forces. Similarly, we assume that the increase of solution ionic strength might also lead to the transformation of conformation of pVIm grafted on FSPV from tightly collapsed state to loosely extended form. Thus, more adsorption sites on the grafted pVIm are exposed to Hg(II) at elevated ionic strength, which would contribute to the increase of Hg(II) removal. Although the detailed mechanism driving to the enhanced Hg(II) uptake at high ionic strength still requires further exploration, the above observations yet clearly demonstrated that Hg(II) could be effectively removed by FSPV in the presence of commonly coexisting cations (Naþ, Kþ, Ca2þ, or Mg2þ) 2 and anions (Cl, NO 3 , or SO4 ). Furthermore, unlike the trend observed in most previous studies, the adsorption of Hg(II) on FSPV would be more favorable at higher ionic strength up to 100 mM, regardless of cation/anion types (including chloride ion).
258
3.6.
w a t e r r e s e a r c h 6 9 ( 2 0 1 5 ) 2 5 2 e2 6 0
Effect of humic acid
The effect of humic acid, which is a ubiquitous and representative natural organic matter (NOM) in the aquatic environment, on the adsorption of Hg(II) by FSPV was explored, and the results were listed in Table S7 in Supplementary Information. Briefly, the removal of Hg(II) remained almost constant (from 95.5% to 97.9%) in the presence of humic acid (from 0 to 8 mg/L as TOC). Obviously, the presence of humic acid had negligible influence on the removal of Hg(II) by FSPV. By serving as a competitive ligand to complex mercury, previous studies have shown that humic acid would increase the mobility of Hg(II) and thus could substantially inhibit the uptake of Hg(II) (Wang et al., 2009a, 2009b; do Nascimento and Masini, 2014). Since humic acid had high affinity for Hg(II) with large stability constant (log b) from around 11 to 19 (Gismera et al., 2007), humic acid-containing adsorbents therefore have also been extensively developed for Hg(II) removal from water (Liu et al., 2008; Zhang et al., 2010). However, our study showed that adsorption of Hg(II) by FSPV was not obviously affected by the presence of humic acid. It is reasonable to deduce that pVIm possibly has stronger affinity with Hg(II) than humic acid, or alternatively, it also has high affinity for Hg(II)-humic acid complex besides Hg(II). Under the latter circumstance, high removal of Hg(II) with the presence of humic acid by FSPV still could be achieved even if Hg(II) was complexed with humic acid. The above observation demonstrated that the synthesized FSPV could serve as an excellent adsorbent for removal of Hg(II) from water containing humic acid with concentration up to 8 mg/L as TOC.
3.7.
Adsorption mechanism
The mechanism of adsorption of Hg(II) by FSPV was investigated with XPS analysis of FSPV before and after adsorption of Hg(II), and the results were shown in Fig. S8, Table S2, and Table S8 in Supplementary Information. Comparison the survey spectra of FSPV versus that of FSPV-Hg in Fig. S8a showed that new peaks of Hg 4f, Cl 2p, Cl 2s, Hg 4d, and Hg 4p appeared in the spectrum of FPSV-Hg, which confirmed that Hg(II) was adsorbed onto the surface of FSPV. As can be seen from Fig. S8b, the N 1 s spectra could be divided into two peaks at 400.1 and 398.2 eV, which were assigned to the amine and imine nitrogen of the imidazole ring, respectively (Ang et al., 2000; Sun et al., 2013). After the adsorption of Hg(II), the maximum position of the peak at 400.1 eV slightly shifted by 0.2 eV (from 400.15 to 399.95 eV). Moreover, the peak at 400.1 eV was strengthened, while the peak at 398.2 eV was weakened. The above observations revealed that the nitrogen atom of the imidazole ring on the FSPV was responsible for the bonding with Hg(II), which was in agreement with previous study (Sun et al., 2013). The fitting results of Hg 4f (Fig. S8c) and Cl 2p (Fig. S8d) peaks indicated that the valence states of Hg and Cl remained þ2 and 1, respectively (Kadono et al., 1994; Devi et al., 2006). This observation confirmed that no redox reaction was involved during the adsorption of Hg(II) by FSPV. Element quantification results (Table S2) showed that the atomic ratio of Cl/Hg was 1.79, which implied that the adsorbed Hg(II) species were mainly HgCl2 and HgClOH. This observation displayed that the bonding of Hg with N did not
result in cleavage of HgeCl bond in HgCl2 and HgClOH. Clearly, the adsorption of Hg(II) by FSPV was not a ligand exchange process, and thus the speciation of Hg(II) would not change after the adsorption. Due to the absence of competition between the ligands, the presence of Cl could not inhibit the removal of Hg(II). Moreover, unlike other metals such as Naþ, Kþ, Ca2þ, and Mg2þ, 92% of the total Hg(II) was in the form of electroneutral species such as HgCl2 and HgClOH (Fig. S7), and thus could be adsorbed by FSPV via nonelectrostatic forces. Thereby, even in the copresence of other metals, FSPV fabricated in our study had high selectivity for Hg(II) adsorption and thus the uptake of Hg(II) was not inhibited.
3.8.
Adsorbent regeneration and reuse
The regeneration and reuse of FSPV were evaluated in five consecutive adsorption/desorption cycles, and the corresponding results were presented in Fig. 6. Preliminary experiment showed that the desorption of Hg(II) from FSPV with HCl was very fast, and equilibrium could be obtained in less than 10 min (data not shown). Thereby, the contact time of desorption process was set to be 10 min. High and stable adsorption efficiency was achieved during the regeneration and reuse processes. Specifically, the adsorption efficiency remained from 93.7% to 96.7% throughout the five consecutive cycles. Moreover, desorption efficiency was also around 90% throughout the five cycles except for the first cycle. The observation confirmed that the Hg(II)-loaded FSPV could be effectively regenerated with 0.5 M HCl in very short time (10 min), and the FSPV could be reused for efficient removal of Hg(II) from water in consecutive cycles. Meanwhile, Fe release was not detected in the treated water. During desorption with HCl, negligible release of Fe was observed. Moreover, the release concentration obviously decreased with the number of cycles (Table S9). A small proportion of Fe3O4 might not be firmly coated by SiO2. After the dissolution of this proportion of Fe3O4, FSPV tended to be stable and thus the decreased
Fig. 6 e Hg(II) adsorption and desorption efficiencies with FSPV in five consecutive cycles at 25 C. Adsorption conditions: initial Hg(II) concentration 1.0 mg/L, initial solution pH 7.0 ± 0.1, background electrolyte 10 mM NaCl, adsorbent dosage 0.1 g/L, contact time 12 h. Desorption conditions: 0.5 M HCl, dosage 4 mL/mg, contact time 10 min.
w a t e r r e s e a r c h 6 9 ( 2 0 1 5 ) 2 5 2 e2 6 0
concentration of Fe was observed with increasing desorption cycles. It is reasonable to deduce that FSPV would be increasingly stable if further cycles were conducted. It should be noted that the total Fe loss was 1.78% throughout the five cycles, which was insignificant for the adsorbent mass. Hence, SiO2 coating on FSPV successfully inhibited the dissolution of Fe during acid regeneration process. The above observations clearly showed that after Hg(II) adsorption, FSPV could be effectively regenerated with a rapid and simple process. Clearly, FSPV fabricated in our study had good potential for repeated utilization for removal of Hg(II) from water.
4.
Conclusions
For the first time, FSPV was synthesized by grafting poly(1vinylimidazole) oligomer onto Fe3O4@SiO2 particles as an efficient adsorbent to remove Hg(II) from water. Containing a magnetic core, the FSPV particles could be separated from water with a simple magnetic process. The Hg(II) adsorption capacity of FSPV was 346 mg/g, which was two orders of magnitude larger than that of Fe3O4, and three orders of magnitude larger than that of Fe3O4@SiO2. The removal of Hg(II) slightly increased with the increase of ionic strength up to 100 mM regardless of cation (Naþ, Kþ, 2 Ca2þ, or Mg2þ) or anion (Cl, NO 3 , or SO4 ) types. Unlike the trend observed in previous studies, the presence of Cl also promoted the uptake of Hg(II) by FSPV. The nitrogen atom of the imidazole ring was responsible for the bonding with Hg(II). The adsorption of Hg(II) by FSPV was not a ligand exchange process. Thus, the removal of Hg(II) could not be inhibited even in the presence of other strong ligands such as Cl. FSPV could be effectively regenerated after Hg(II) adsorption with a rapid and simple process, and exhibited good potential for repeated utilization for removal of Hg(II) from water.
Acknowledgments This work was supported by the National Natural Science Foundation of China under Grant No. 21377006 and the program for New Century Excellent Talents in University under Grant No. NCET-13-0010.
Appendix A. Supplementary data Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.watres.2014.11.030.
references
Ang, A.K.S., Kang, E.T., Neoh, K.G., Tan, K.L., Cui, C.Q., Lim, T.B., 2000. Low-temperature graft copolymerization of 1-vinyl imidazole on polyimide films with simultaneous lamination to
259
copper foils - effect of crosslinking agents. Polymer 41 (2), 489e498. Arkas, M., Tsiourvas, D., 2009. Organic/inorganic hybrid nanospheres based on hyperbranched poly(ethylene imine) encapsulated into silica for the sorption of toxic metal ions and polycyclic aromatic hydrocarbons from water. J. Hazard. Mater. 170 (1), 35e42. Bao, J., Fu, Y., Bao, Z., 2013. Thiol-functionalized magnetite/ graphene oxide hybrid as a reusable adsorbent for Hg2þ removal. Nanoscale Res. Lett. 8, 486. Boening, D.W., 2000. Ecological effects, transport, and fate of mercury: a general review. Chemosphere 40 (12), 1335e1351. Bonnissel-Gissinger, P., Alnot, M., Lickes, J.P., Ehrhardt, J.J., Behra, P., 1999. Modeling the adsorption of mercury(II) on (hydr)oxides II: a-FeOOH (goethite) and amorphous silica. J. Colloid Interface Sci. 215 (2), 313e322. Brooks, P., Davidson, N., 1960. Mercury(II) complexes of imidazole and histidine. J. Am. Chem. Soc. 82 (9), 2118e2123. _ C¸am, T., Osman, B., Kara, A., Demirbel, E., Bes‚irli, N., Irez, G., 2013. Complexation of Zn2þ with vinyl imidazole containing IMAC adsorbent: potentiometric, kinetic, and thermodynamic studies. Hacet. J. Biol. Chem. 41 (3), 207e219. Carro, L., Anagnostopoulos, V., Lodeiro, P., Barriada, J.L., Herrero, R., Sastre de Vicente, M.E., 2010. A dynamic proof of mercury elimination from solution through a combined sorption-reduction process. Bioresour. Technol. 101 (23), 8969e8974. Cui, H., Qian, Y., Li, Q., Zhang, Q., Zhai, J., 2012. Adsorption of aqueous Hg(II) by a polyaniline/attapulgite composite. Chem. Eng. J. 211, 216e223. Devi, P.S.R., Kumar, S., Verma, R., Sudersanan, M., 2006. Sorption of mercury on chemically synthesized polyaniline. J. Radioanal. Nucl. Chem. 269 (1), 217e222. do Nascimento, F.H., Masini, J.C., 2014. Influence of humic acid on adsorption of Hg(II) by vermiculite. J. Environ. Manag. 143, 1e7. Farrukh, A., Akram, A., Ghaffar, A., Hanif, S., Hamid, A., Duran, H., Yameen, B., 2013. Design of polymer-brush-grafted magnetic nanoparticles for highly efficient water remediation. ACS Appl. Mater. Interfaces 5 (9), 3784e3793. Figueira, P., Lopes, C.B., Daniel-da-Silva, A.L., Pereira, E., Duarte, A.C., Trindade, T., 2011. Removal of mercury (II) by dithiocarbamate surface functionalized magnetite particles: application to synthetic and natural spiked waters. Water Res. 45 (17), 5773e5784. Gismera, M.J., Procopio, J.R., Sevilla, M.T., 2007. Characterization of mercury e humic acids interaction by potentiometric titration with a modified carbon paste mercury sensor. Electroanalysis 19 (10), 1055e1061. Hakami, O., Zhang, Y., Banks, C.J., 2012. Thiol-functionalised mesoporous silica-coated magnetite nanoparticles for high efficiency removal and recovery of Hg from water. Water Res. 46 (12), 3913e3922. Herrero, R., Lodeiro, P., Rey-Castro, C., Vilarino, T., de Vicente, M.E.S., 2005. Removal of inorganic mercury from aqueous solutions by biomass of the marine macroalga Cystoseira baccata. Water Res. 39 (14), 3199e3210. Ho, Y.S., McKay, G., 1999. Pseudo-second order model for sorption processes. Process Biochem. 34 (5), 451e465. Hu, H., Wang, Z., Pan, L., 2010. Synthesis of monodisperse Fe3O4@silica core-shell microspheres and their application for removal of heavy metal ions from water. J. Alloys Compd. 492 (1e2), 656e661. Jiang, G.B., Shi, J.B., Feng, X.B., 2006. Mercury pollution in China. Environ. Sci. Technol. 40 (12), 3672e3678. Kadono, K., Kinugawa, K., Tanaka, H., 1994. X-ray photoelectronspectroscopy of ZnCl2 based glasses. Phys. Chem. Glasses 35 (2), 59e64.
260
w a t e r r e s e a r c h 6 9 ( 2 0 1 5 ) 2 5 2 e2 6 0
Lai, B., Yeh, C., Chen, D., 2012. Surface modification of iron oxide nanoparticles with polyarginine as a highly positively charged magnetic nano-adsorbent for fast and effective recovery of acid proteins. Process Biochem. 47 (5), 799e805. Langford, N.J., Ferner, R.E., 1999. Toxicity of mercury. J. Hum. Hypertens. 13 (10), 651e656. Li, P., Feng, X.B., Qiu, G.L., Shang, L.H., Li, Z.G., 2009. Mercury pollution in Asia: a review of the contaminated sites. J. Hazard. Mater. 168 (2e3), 591e601. Liu, J., Zhao, Z., Jiang, G., 2008. Coating Fe3O4 magnetic nanoparticles with humic acid for high efficient removal of heavy metals in water. Environ. Sci. Technol. 42 (18), 6949e6954. Loux, N.T., 1998. An assessment of mercury-species-dependent binding with natural organic carbon. Chem. Speciat. Bioavailab. 10 (4), 127e136. Mandel, K., Hutter, F., Gellermann, C., Sextl, G., 2012. Modified superparamagnetic nanocomposite microparticles for highly selective Hg-II or Cu-II separation and recovery from aqueous solutions. ACS Appl. Mater. Interfaces 4 (10), 5633e5642. Miretzky, P., Fernandez Cirelli, A., 2009. Hg(II) removal from water by chitosan and chitosan derivatives: a review. J. Hazard. Mater. 167 (1e3), 10e23. Molina, M.J., Gomez-Anton, M.R., Rivas, B.L., Maturana, H.A., Pierola, I.F., 2001. Removal of Hg(II) from acid aqueous solutions by poly(N-vinylimidazole) hydrogel. J. Appl. Polym. Sci. 79 (8), 1467e1475. Pan, S., Shen, H., Xu, Q., Luo, J., Hu, M., 2012. Surface mercapto engineered magnetic Fe3O4 nanoadsorbent for the removal of mercury from aqueous solutions. J. Colloid Interface Sci. 365 (1), 204e212. Song, J., Kong, H., Jang, J., 2011. Adsorption of heavy metal ions from aqueous solution by polyrhodanine-encapsulated magnetic nanoparticles. J. Colloid Interface Sci. 359 (2), 505e511. Sun, J., Chen, Z., Ge, M., Xu, L., Zhai, M., 2013. Selective adsorption of Hg(II) by g-radiation synthesized silica-graft-vinyl imidazole adsorbent. J. Hazard. Mater. 244, 94e101. Takafuji, M., Ide, S., Ihara, H., Xu, Z.H., 2004. Preparation of poly(1-vinylimidazole)-grafted magnetic nanoparticles and their application for removal of metal ions. Chem. Mater. 16 (10), 1977e1983.
Tang, S.C.N., Lo, I.M.C., 2013. Magnetic nanoparticles: essential factors for sustainable environmental applications. Water Res. 47 (8), 2613e2632. Wang, J., Deng, B., Chen, H., Wang, X., Zheng, J., 2009a. Removal of aqueous Hg(II) by polyaniline: sorption characteristics and mechanisms. Environ. Sci. Technol. 43 (14), 5223e5228. Wang, J., Deng, B., Wang, X., Zheng, J., 2009b. Adsorption of aqueous Hg(II) by sulfur-impregnated activated carbon. Environ. Eng. Sci. 26 (12), 1693e1699. Wang, J., Zheng, S., Shao, Y., Liu, J., Xu, Z., Zhu, D., 2010. Aminofunctionalized Fe3O4@SiO2 core-shell magnetic nanomaterial as a novel adsorbent for aqueous heavy metals removal. J. Colloid Interface Sci. 349 (1), 293e299. Wang, Q.R., Kim, D., Dionysiou, D.D., Sorial, G.A., Timberlake, D., 2004. Sources and remediation for mercury contamination in aquatic systems - a literature review. Environ. Pollut. 131 (2), 323e336. Wang, X., Liu, G., Zhang, G., 2011. Conformational behavior of grafted weak polyelectrolyte chains: effects of counterion condensation and nonelectrostatic anion adsorption. Langmuir 27 (16), 9895e9901. Wang, Z., Wu, D., Wu, G., Yang, N., Wu, A., 2013. Modifying Fe3O4 microspheres with rhodamine hydrazide for selective detection and removal of Hg2þ ion in water. J. Hazard. Mater. 244, 621e627. Zhang, C., Sui, J., Li, J., Tang, Y., Cai, W., 2012. Efficient removal of heavy metal ions by thiol-functionalized superparamagnetic carbon nanotubes. Chem. Eng. J. 210, 45e52. Zhang, Y., Li, Q., Sun, L., Tang, R., Zhai, J., 2010. High efficient removal of mercury from aqueous solution by polyaniline/humic acid nanocomposite. J. Hazard. Mater. 175 (1e3), 404e409. Zhang, Y.K., Yan, L.G., Xu, W.Y., Guo, X.Y., Cui, L.M., Gao, L., Wei, Q., Du, B., 2014. Adsorption of Pb(II) and Hg(II) from aqueous solution using magnetic CoFe2O4-reduced graphene oxide. J. Mol. Liq. 191, 177e182. Zhou, Y., Zhang, J., Luo, X., Lin, X., 2014. Adsorption of Hg(II) in aqueous solutions using mercapto-functionalized alkali lignin. J. Appl. Polym. Sci. 131 (18). Zhu, B., Zhao, J., Yu, H., Yan, L., Wei, Q., Du, B., 2013. Naphthalimide-functionalized Fe3O4@SiO2 core/shell nanoparticles for selective and sensitive adsorption and detection of Hg2þ. Chem. Eng. J. 219, 411e418.
