Simultaneous adsorption and degradation of Zn²⁺ and Cu²⁺ from wastewaters using nanoscale zero-valent iron impregnated with clays

Abstract

Clays such as kaolin, bentonite and zeolite were evaluated as support material for nanoscale zero-valent iron (nZVI) to simultaneously remove Cu²⁺ and Zn²⁺ from aqueous solution. Of the three supported nZVIs, bentonite-supported nZVI (B-nZVI) was most effective in the simultaneous removal of Cu²⁺ and Zn²⁺ from a aqueous solution containing a 100 mg/l of Cu²⁺ and Zn²⁺, where 92.9 % Cu²⁺ and 58.3 % Zn²⁺ were removed. Scanning electronic microscope (SEM) revealed that the aggregation of nZVI decreased as the proportion of bentonite increased due to the good dispersion of nZVI, while energy dispersive spectroscopy (EDS) demonstrated the deposition of copper and zinc on B-nZVI after B-nZVI reacted with Cu²⁺ and Zn²⁺. A kinetics study indicated that removing Cu²⁺ and Zn²⁺ with B-nZVI accorded with the pseudo first-order model. These suggest that simultaneous adsorption of Cu²⁺and Zn²⁺ on bentonite and the degradation of Cu²⁺and Zn²⁺ by nZVI on the bentonite. However, Cu²⁺ removal by B-nZVI was reduced rather than adsorption, while Zn²⁺ removal was main adsorption. Finally, Cu²⁺, Zn²⁺, Ni²⁺, Pb²⁺ and total Cr from various wastewaters were removed by B-nZVI, and reusability of B-nZVI with different treatment was tested, which demonstrates that B-nZVI is a potential material for the removal of heavy metals from wastewaters.

Introduction

Remediation of heavy metals has attracted attention owing to the fact that they entail severe risks for the environment and human health (Järup 2003; Gupta et al. 2009, 2012a; Jain et al. 2004). The United States Environmental Protection Agency (USEPA) proposed a report on priority pollutants, listing 129 kinds of severe hazard contaminants, including copper and zinc (Keith and Telliard 1979; Gupta and Sharma 2003). Various conventional techniques have been developed for the treatment of heavy metals in wastewaters, such as chemical precipitation (Lin et al. 2005), oxidation–reduction (Chen et al. 2008; Gupta et al. 2007a,b), ion exchange (Keane 1998), adsorption (Bhattacharyya and Gupta 2008; Gupta et al. 2010, 2011a; Gupta and Nayak 2011; Gupta et al. 2006, 2012a; Saleh and Gupta 2012), biosorption (Sarioglu et al. 2009; Gupta et al. 2009; Gupta and Rastogi 2009 Gupta et al. 2012b, c), and electrochemical methods (Feng et al. 2007; Gupta et al. 2007c). However, some of these techniques have limitations due to their low efficiency and high operational cost. An alternative method known as nanoremediation, which entails using various nanomaterials to transform and detoxify pollutants, has received attention during the last decade (Nowack and Bucheli 2007; Saleh and Gupta 2011; Gupta et al. 2011b). Of these, nanoscale zero-valent iron (nZVI) is the most widely used for in situ remediation due to its large surface area, rapid reactivity and better injectability into aquifers. nZVI has been proved to be effective in the remediation of various pollutants in aqueous solution, including chlorinated hydrocarbon, chlorinated benzenes, nitrocompounds and heavy metals (Zhang 2003; Li et al. 2006).

Challenges remain in the real application of nZVI, including the aggregation of nZVI particles, which has resulted from a large surface area being present and ultimately leads to a decrease in the reactivity of nZVI. To address this issue, strategies have been developed such as impregnation of the nanoparticles on the surface of support materials (Shahwan et al. 2010; Shi et al. 2011). The removal of Cr₂O₇ ⁻² and Pb²⁺ by polymeric resin-supported nZVI has been reported (Ponder et al. 2000). More recently, the removal of Pb²⁺ using kaolin supported nZVI has been reported (Zhang et al. 2011), where a concept of the simultaneous adsorption of Pb²⁺ on kaolin and the degradation of Pb²⁺ by nZVI on the kaolin supported nZVI has been proposed. Despite clay such as kaolin can be used to remove heavy metals from aqueous solution (Jiang et al. 2010). However, the adsorbed contaminants could be released in to the environment and pose potential risks to ecosystems. The simultaneous adsorption and degradation of Pb²⁺ occurring on the kaolin supported nZVI materials suggest that this is a promising method for remediation of metals ions. In addition, heavy metals in electroplating wastewater contain various heavy metals as co-contaminants. A few studies have been published on simultaneous removal of heavy metals by supported nZVI. Hence, removal behaviours of heavy metals and the relevant mechanism may help to understand simultaneous removal of heavy metals, where the mechanism is still unclear.

Mineral clays are low-cost, abundant and stable, and therefore kaolin and bentonite were used as support material for nZVI to remove metals Cr₂O₇ ⁻² and Pb²⁺ and (Shi et al. 2011; Zhang et al. 2011) in our previous reports, but there is no report on clay as the support nZVI to examine these support materials affecting the removal of metals. In this report, mineral clays such as kaolin, bentonite and zeolite were evaluated as support materials for nZVI and used to simultaneously remove heavy metals such as Cu²⁺ and Zn²⁺. Hence, the aims of this study were: (1) synthesis of the support materials of nZVI; (2) characterization of the materials with scanning electronic microscopy (SEM), EDX and BET-N₂ techniques; (3) kinetics of Cu²⁺ and Zn²⁺ removal by B-nZVI; and (4) demonstrating firstly the removal of heavy metals from wastewater using B-nZVI, and secondly the reusability of B-nZVI.

Materials and methods

Materials and chemicals

Bentonite and kaolin used in this study are natural low-grade minerals without any modification, and were supplied by Longyan Kaolin Co. Ltd. (Fujian, China). These materials were ground and sieved through a 200-mesh after being dried at 80 °C overnight. The zeolite was chemically pure with a granular size of 200 mesh. All other regents used in this study were analytically pure and distilled water was used in all preparations. A highly concentrated stock solution of Cu²⁺ and Zn²⁺ was prepared by dissolving Cu(NO₃)₂ or Zn(NO₃)₂ with distilled water simultaneously or separately and stored under 5 °C.

Synthesis of nZVI and supported nZVI

The nZVI and supported nZVIs were prepared using conventional liquid-phase methods by reducing ferric iron with borohydride, which necessitated mineral clay acting as a support material (Shahwan et al. 2010; Zhang et al. 2011). Support material (2.00 g) was initially placed into a three-necked open flask, and a ferric solution produced by dissolving ferric chloride hexa-hydrate (9.66 g) in an ethanol/water solution (50 ml, 4:1 v/v) was added and stirred for 10 min. Subsequently, a freshly prepared NaBH₄ solution (3.54 g of NaBH₄ in 100 ml) was added dropwise into the mixture and constant stirred for 20 min. The whole process described above was performed under a N₂ atmosphere with vigorous stirring so that the B-nZVI did not oxidize. The synthetic reaction occurred as below (Shi et al. 2011)

$$ 4\mathrm{F}{{\mathrm{e}}^{3+ }}+\mathrm{BH}_4^{-}+3{{\mathrm{H}}_2}\mathrm{O}\to 4\mathrm{F}{{\mathrm{e}}^{2+ }}+{{\mathrm{H}}_2}\mathrm{BO}_3^{-}+4{{\mathrm{H}}^{+}}+2{{\mathrm{H}}_2}\uparrow $$

(1)

$$ 2\mathrm{F}{{\mathrm{e}}^{2+ }}+\mathrm{BH}_4^{-}+3{{\mathrm{H}}_2}\mathrm{O}\to 2\mathrm{F}{{\mathrm{e}}^0}\downarrow +{{\mathrm{H}}_2}\mathrm{BO}_3^{-}+4{{\mathrm{H}}^{+}}+2{{\mathrm{H}}_2}\uparrow $$

(2)

$$ 4\mathrm{F}{{\mathrm{e}}^{3+ }}+3\mathrm{BH}_4^{-}+9{{\mathrm{H}}_2}\mathrm{O}\to 4\mathrm{F}{{\mathrm{e}}^0}\downarrow +3{{\mathrm{H}}_2}\mathrm{BO}_3^{-}+12{{\mathrm{H}}^{+}}+6{{\mathrm{H}}_2}\uparrow $$

(3)

The formed suspension was filtered and the black nanoscale precipitate was washed three times with pure ethanol and dried overnight at 75 °C under vacuum (Shahwan et al. 2010). The theoretical mass fraction of mineral clay in synthesized B-nZVI was 50 %, and nZVI was prepared under identical conditions but with mineral clay omitted. The supported nZVI was labeled as B-nZVI, K-nZVI or Z-nZVI, when the support material was bentonite, kaolin or zeolite, respectively. nZVI and supported nZVI samples were stored in brown, sealed bottles in a dry environment and were not acidified prior to use. In subsequent experiments, unsupported nZVI and B-nZVI with different mass ratios of bentonite versus nZVI, which were theoretically calculated to be 0:1, 1:1, 2:1, 3:1 and 4:1, were prepared by varying the initial bentonite loadings at the beginning of the synthesis.

Characterizations and measurements

SEM images and energy dispersive spectroscope (EDS) patterns of various materials were obtained with a Philips-FEI XL30 ESEM-TMP (Philips Electronics Co., Eindhoven, the Netherlands) instrument.

The specific surface areas (SSA) of nZVI, B-nZVI and bentonite were measured using the BET-N₂ method using Micromeritics’ ASAP 2020 Accelerated Surface Area and Porosimetry Analyzer (Micromeritics Instrument Corp., Georgia, USA).

Batch experiments

In order to select the support material, B-nZVI, K-nZVI and Z-nZVI (ratio of support material/nZVI was 1:1) were used to remove Cu²⁺ and Zn²⁺, respectively. Next, 0.075 g supported-nZVI was added into 25-ml mixed solutions with the initial concentrations of both Cu²⁺ and Zn²⁺ at 100 mg/l without pH adjustment.

To determine the optimal mass ratios of bentonite/nZVI in B-nZVI, which were set at 0:1, 1:1, 2:1, 3:1 and 4:1 in the synthesis procedure, B-nZVI with different mass ratios were tested for their efficiency in removing Cu²⁺ and Zn²⁺ with the initial concentration at 100 mg/l. The theoretic dosage of iron was fixed at 1 g/l. In addition, bentonite was applied to remove Cu²⁺ and Zn²⁺ independently to examine the role that bentonite played in the B-nZVI.

Batch experiments were conducted to investigate the kinetics of Cu²⁺ and Zn²⁺ removal via B-nZVI (2:1). The effects of different initial concentrations of Cu²⁺ and Zn²⁺ ions and temperatures were tested, which were set at 20, 50, 70, 100 mg/l and 20 °C, 25 °C, 30 °C, and 35 °C, respectively. Each batch experiment was mixed with 25 ml Cu²⁺ and Zn²⁺ solution. The initial concentrations of both ions were set at 100 mg/l in the rotary shaker at 35 °C and 250 rpm, and then filtered through 0.45-μm MCE membranes prior to establishing the concentrations of Cu²⁺ and Zn²⁺ at selected timed intervals.

To investigate the capacity of B-nZVI (2:1) for the removal of heavy metal ions from wastewater, two electroplating wastewaters and one dyeing wastewater were collected from the manufacturer (Fuzhou, China). The wastewater was centrifuged at 3,000 rpm for 10 min to remove all solid impurities prior to determining the initial pH and initial concentrations of various heavy metal ions. Next, 25 ml wastewater was added to the batch of centrifugal tubes containing 0.10 g of B-nZVI (2:1), which were kept on a rotary shaker at 35 °C and 250 rpm for 2 h. Then, the pH and concentration of each heavy metal ion were determined.

Reuse experiments were done to evaluate the reusability of B-nZVI for removing Cu²⁺ and Zn²⁺ from mixed solution. B-nZVI (2:1) (0.1 g) was mixed with 20 or 50 mg/l Cu²⁺ and Zn²⁺ solution (25 ml), respectively, for 1 h (35 °C and 250 rpm) before being centrifuged at 3,000 rpm. All supernates were transferred carefully into clean containers to determine the residual concentrations of Cu²⁺ and Zn²⁺ while the used B-nZVI left in the bottom was mixed with another 25 ml Cu²⁺ and Zn²⁺ solution directly or after being treated. The used B-nZVI was treated by mixing with 25 ml distilled water and then 2 % HNO₃ for 15 min, respectively. Each series of reuse was successively carried out three times.

The concentrations of Cu²⁺, Zn²⁺ and other heavy metal ions were found using a flame atomic absorbance spectrometer (Varian AA 240FS, USA). In order to ensure the accuracy and reliability of the data, batch experiments were carried out at least in triplicate.

Results and discussion

Synthesis of various clays-supported nZVI

Cu²⁺ and Zn²⁺ removal using clays-supported nZVI

Studies have been shown where different clays such as kaolin (Zhang et al. 2009) and bentonite (Shi et al. 2011) were used to support nZVI, but no report has described the effect on the impregnation of nZVI using various clays. Hence, bentonite, kaolin and zeolite were used to support nZVI in this study. The initial pH value of Cu²⁺ and Zn²⁺ mixed solution was around 3.9 without adjustment. Figure 1 shows that the removal capacity of Cu²⁺ and Zn²⁺ using such various supported nZVIs are in order of decreasing effectiveness: bentonite support nZVI (B-nZVI) > kaolin support nZVI (K-nZVI) > zeolite support nZVI (Z-nZVI). It can be seen that B-nZVI was the most efficient material for removing both Cu²⁺ and Zn²⁺. The removal of 92.9 % Cu²⁺ and 58.3 % Zn²⁺ was observed after 3 h using B-nZVI, while only 48.1 % Cu²⁺ and 7.6 % Zn²⁺ were removed by Z-nZVI. However, the removal of Cu²⁺ using various support materials of nZVI was much more efficient than that of Zn²⁺.

Fig. 1

The removal of Cu²⁺ and Zn²⁺ by supported nZVI with different support materials. a Cu²⁺ ion; b Zn²⁺ ion

The poor removal of Cu²⁺ and Zn²⁺ using Z-nZVI resulted from the poor dispersion of nZVI on the surface of zeolites due to the porous material of zeolite having three-dimensional structures (Mohan and Pittman 2006). In addition, the zeolite used in this study with a micrometer-scaled diameter made the dispersion of the nZVI particles on zeolite ineffective. In this case, most nZVI particles existed without being supported. Kaolin and bentonite used in this study were natural minerals, where bentonite had a major component of montmorillonite >90 %. Compared to the 1:1 layer structure of kaolin, montmorillonite had a three-layer structure composed of silica–alumina–silica units, leading to an easier cleavage and expansion between the layers in aqueous solution. The net layer charges of kaolin and montmorillonite are 0 and −0.8 charge/unit cell, respectively (Bhattacharyya and Gupta 2008). This resulted in an interaction between the core–shell structure of nZVI and the shell composed of iron oxide and hydroxide, since the surface of the nZVI particles had positive net charges (Li and Zhang 2006, 2007). As a result, bentonite was more favorable for dispersing nZVI particles through electrostatic interactions. In addition, bentonite emerged as being more effective in improving the activity of nZVI since it has a flexible lamellar structure. It was therefore selected as the best support material for nZVI for further study.

Furthermore, a more efficient removal of Cu²⁺ than that of Zn²⁺ was observed as mentioned above. This is because the standard reduction potential of Cu²⁺ (Φ ⁰ _Cu ²⁺ _/Cu ⁰ = 0.337 V) is more positive than that of Zn²⁺(Φ ^θ _Zn ²⁺ _/Zn ⁰ = −0.762 V), meaning that Cu²⁺ is much easier to be reduced compared with Zn²⁺ (Ladd 2004). On the other hand, the adsorption of both Cu(II) and Zn(II) onto the negatively charged bentonite surface or the nZVI surface, which may include electrostatic interactions and specific surface bonding (Bhattacharyya and Gupta 2008; Li and Zhang 2007) were possible. However, adsorption could be dominated in the removal of Zn²⁺ while reduction may play a main role in the removal of Cu²⁺. This is also supported by the XPS spectra of nZVI after reacting with various heavy metal ions in aqueous solution (Li and Zhang 2007).

Mass ratios of bentonite/nZVI and their influence on Cu²⁺ and Zn²⁺removal

Bentonite can be used to absorb Cu²⁺ and Zn²⁺ with a low removal capacity (Bhattacharyya and Gupta 2008). To examine the adsorption of Cu²⁺ and Zn²⁺ on the bentonite, batch experiments were conducted utilizing co-adsorption of Cu²⁺ and Zn²⁺, and results showed that only 5.5 % of Cu²⁺ and 9.8 % Zn²⁺ adsorbed on bentonite. This indicates that bentonite can be as adsorbent to remove a small part of Cu²⁺ and Zn²⁺ — this is due to the decrease of available adsorption sites after being used as the support material for nZVI —— and the removal of Cu²⁺ and Zn²⁺ is dominated by the absorption and reduction by nZVI and bentonite using B-nZVI (Zhang 2003; Bhattacharyya and Gupta 2008).

The removal efficiencies of Cu²⁺ and Zn²⁺ using B-nZVI are shown in Fig. 2, which was prepared with different mass ratios of bentonite/nZVI (0:1, 1:1, 2:1, 3:1 and 4:1). It can be seen that the removal efficiencies of Cu²⁺ and Zn²⁺ varied according to the mass ratios of bentonite/nZVI, which increased as the mass ratio rose from 0:1 to 2:1. The highest removal efficiency of Cu²⁺ and Zn²⁺ was approximately 76.5 % and 42.3 %, respectively, using B-nZVI with mass ratios of bentonite/nZVI (2:1). This can be interpreted by the SEM images of unsupported nZVI and B-nZVI (Figs. 3 and 4). They show that the unsupported nanoparticles were severely aggregated, and the aggregation decreased as the proportion of bentonite increased.

Fig. 2

Removal efficiencies of Cu²⁺ and Zn²⁺ ion using B-nZVI with different mass ratios of bentonite/nZVI

Fig. 3

SEM images of B-nZVI with different mass ratios of bentonite/nZVI. a Bentonite; b B-nZVI (1:1); c B-nZVI (2:1); d B-nZVI (3:1); e B-nZVI (4:1)

Fig. 4

EDX patterns of various solid samples. a Bentonite; b freshly synthesized B-nZVI (2:1); c B-nZVI (2:1) after reaction with Cu²⁺ solution; d B-nZVI (2:1) after reaction with Zn²⁺ solution

The decrease in aggregation led to an increase in the SSA of the nanoparticles, and hence an increase in the number of active sites (Zhang et al. 2011). However, the removal efficiencies were equally poor when the mass ratios rose to 3:1 and 4:1, which were approximately 50.2 % of the efficiencies by B-nZVI (2:1). This was because the high content of bentonite blocked the formation of nZVI during synthesis, and consequently resulted in a decline of the amount of iron nanoparticles dispersed on bentonites (Shahwan et al. 2010). Additionally, the higher dispersion of nZVI accelerated the oxidation and deterioration of zero-valent iron (Wang et al. 2010). Hence, B-nZVI (2:1) was used in the subsequent study.

Characterization

The SEM images of bentonite and B-nZVI are presented in Fig. 3, which showed that the nanoparticles that adhered to the surface of bentonite are generally spherical in shape and have diameters within 100 nm. However, nZVI accumulated into a bulky nanocluster, which can be attributed to their natural tendency to remain in a more thermodynamically stable state since they have a huge SSA as previously reported (Zhang et al. 2009, 2011). Figure 3 shows that the proportion of separate iron nanoparticles that scatter on the surface of bentonite seems to increase as the ratio of bentonite increases. The surface area of the bentonite support increased with increasing bentonite content, leading to decreased aggregation and improved nanoparticle dispersion (Shi et al. 2011). This occurred even when the nZVI continued to interconnect with one another and formed a chain-like morphology. Bentonite was demonstrated to be effective in reducing the aggregation and accelerating the dispersion of nanoparticles, which was consistent with the observation of kaolinite used as a support material (Shahwan et al. 2010).

The SSAs for bentonite, nZVI, and B-nZVI were 6.03, 54.04 and 26.91 m²/g, respectively. Despite a low SSA of B-nZVI being obtained compared to that of nZVI, a high rate of removal of Cu²⁺ and Zn²⁺ was achieved using B-nZVI. Contributing to this was the decrease in nZVI aggregation and a subsequent increase in the reactivity of nZVI after the dispersion of nZVI on the bentonite. In turn, this led to the stabilization of nZVI nanoparticles. This was also supported by previous studies on the reduction of nZVI aggregation and increase in reactivity using various support materials for nZVI nanoparticles (Ponder et al. 2000; Shahwan et al. 2010).

The EDS spectra of various materials are illustrated in Fig. 4. The EDS analysis of bentonite is shown in Fig. 4a, where the main composition of bentonite is 56.20 wt.% O, 27.59 wt.% Si and 9.15 wt.% Fe, while other all elements combined such as such as C, Mg, K and Fe are less than 5 %. Similar results obtained from EDS analysis of bentonite were observed (Vieira et al. 2010). However, Fig. 4b clearly demonstrates that Fe peaks are significantly increased in freshly synthesized B-nZVI (2:1) due to the dispersion of nZVI nanoparticles on bentonites, where the peaks of Fe are 48.31 and 20.82 wt.%, respectively. The increase in oxygen in freshly synthesized B-nZVI can be contributed to the oxidation of the nZVI surface in the atmosphere due to its core–shell structure (Li and Zhang 2007). As shown in Fig. 5c and d, the existence of Cu and Zn on the surface of B-nZVI after reaction was observed, confirming the adsorption of Cu²⁺ and Zn²⁺ onto the B-nZVI. The weight percentage of O in B-nZVI was 29.13 % before reacting and increased to 42.80 % and 52.28 %, respectively, after reacting with Cu²⁺ and Zn²⁺ solution. This could be ascribed to the oxidation process of the zero-valent iron in aqueous solution to form the FeOOH (Li and Zhang 2007).

Fig. 5

The effects of the initial concentrations of Cu²⁺ and Zn²⁺ on their removal by B-nZVI (2:1). a Cu²⁺; b Zn²⁺ (35 °C, C ₀ (B-nZVI (2:1)) = 3 g/l, 250 rpm)

Condition experiments and kinetics study

The pseudo-first-order model can be used to describe the reduction of Cu²⁺ and Zn²⁺ by nZVI as follows (Zhang et al. 2009):

$$ \ln \frac{c}{c_0 }=-{k_{\mathrm{obs}}}t $$

(4)

where c (mg/l) and c ₀ (mg/l) are the instantaneous concentration and initial concentration of contaminant, respectively, and k _obs (min⁻¹) is the observed rate constant.

Kinetics study of the simultaneous removal of Cu²⁺ and Zn²⁺ using B-nZVI under different conditions suggested that the removal of both ions was in accordance with the pseudo-first-order model with all correlation coefficients (R ²) higher than 0.99. The observed reaction rate constant can be obtained by calculating the slope of the linear regression slope as described in the following sections.

Effect of initial concentrations of Cu²⁺ and Zn²⁺

By plotting ln(c/c ₀) versus t, the kinetics of Cu²⁺ and Zn²⁺ removal using B-nZVI under initial concentrations are presented in Fig. 5, which was described using the pseudo-first-order model. As shown in Fig. 5, for both ions the observed rate constant decreased as the initial concentrations increased. On the other hand, a lower rate of Zn²⁺ removal was obtained compared to Cu²⁺, which was consistent with the previous section in the comparison study. For instance, k _obs decreased from 0.1620 to 0.0568 min⁻¹ as the initial Cu²⁺ concentration increased from 20 to 100 mg/l, and the equilibrium time rose from 15 to 20 min, while the k _obs of Zn²⁺ removal fell from 0.1427 to 0.0111 min⁻¹. The equilibrium time rose from 20 to 40 min. Since the removal of both Cu²⁺ and Zn²⁺ was a surface-mediated process (Weber 1996; Li and Zhang 2007), this was attributed to the fact that the more the heavy metal ions approached the nZVI surface, the faster Fe⁰ was oxidized into Fe(III). It formed a passivation layer on the surface of nZVI and subsequently resulted in the degeneration of reactivity (Yuan et al. 2009).

Effect of temperature

The k _obs of Cu²⁺ and Zn²⁺ removal under different temperatures was achieved by plotting ln c/c ₀ versus time, which revealed that k _obs increased as temperature rose. A regression line can be achieved by plotting each k _obs versus reciprocal of corresponding temperature logarithmically with correlation coefficients (R ²) higher than 0.95. This allowed computation of E _a (Zhang et al. 2009):

$$ \ln {k_{\mathrm{obs}}}=-\frac{{{E_{\mathrm{a}}}}}{RT }+\ln {A_0} $$

(5)

where E _a (kJ/mol) is apparent activation energy, R (J/(mol K)) is the thermodynamic constant, T (K) is thermodynamic temperature, and A ₀ represents the pre-exponential factor with the same dimension as k _obs.

The E _a of Cu²⁺ and Zn²⁺ using B-nZVI was 23.8 and 46.6 kJ/mol, respectively. The activation energy greater than 21.0 kJ/mol was proposed for chemically controlled adsorption processes (Geng et al. 2009). However, the low activation energy for Cu²⁺ suggests that the rate determining step of Cu²⁺ removal by B-nZVI was reduced rather than adsorption, while Zn²⁺ removal is the main adsorption. This was supported in the previous section when Cu²⁺ and Zn²⁺ were removed using various materials. Hence, it was concluded that surface reactions including adsorption and chemical reaction controlled the removal of Cu²⁺ and Zn²⁺ under experimental conditions.

On the basis of the results described above, the mechanism for removing Cu²⁺ and Zn²⁺ from aqueous solutions by using B-nZVI involves two strategies: one is that Cu²⁺ and Zn²⁺ in aqueous solution are adsorbed on the surface of B-nZVI since bentonite can absorb Cu²⁺ and Zn²⁺ (Mohan and Pittman 2006), and another is a scenario in which most of the adsorbed Cu²⁺ was reduced by Fe⁰ or its corrosion products (Zhang 2003). However, Cu²⁺ removal by B-nZVI was reduced rather than adsorption, while Zn²⁺ removal was the main adsorption as Zn²⁺ has a more negative standard reduction of Zn²⁺/Zn⁰ than Fe²⁺/Fe⁰; it cannot be reduced by Fe⁰ into Zn⁰ in this reaction system, and it deposited on the B-nZVI in the form of Zn²⁺. A possible mechanism is proposed below:

$$ \mathrm{F}{{\mathrm{e}}^0}+2{{\mathrm{H}}_2}\mathrm{O}\to \mathrm{F}{{\mathrm{e}}^{2+ }}+{{\mathrm{H}}_2}+2\mathrm{O}{{\mathrm{H}}^{-}}\left( {\mathrm{in}\;\mathrm{basic}\;\mathrm{solution}} \right) $$

(6)

$$ \mathrm{or}\;\mathrm{F}{{\mathrm{e}}^0}+2{{\mathrm{H}}^{+}}\to \mathrm{F}{{\mathrm{e}}^{2+ }}+{{\mathrm{H}}_2}\left( {\mathrm{in}\;\mathrm{acid}\;\mathrm{solution}} \right) $$

(7)

$$ \mathrm{C}{{\mathrm{u}}^{2+ }}\left( {\mathrm{Z}{{\mathrm{n}}^{2+ }}} \right) + \mathrm{Bentonite}\to \mathrm{C}{{\mathrm{u}}^{2+ }}\left( {\mathrm{Z}{{\mathrm{n}}^{2+ }}} \right)-\mathrm{Bentonite}\left( {\mathrm{adsorption}} \right) $$

(8)

$$ \mathrm{C}{{\mathrm{u}}^{2+ }}-\mathrm{Bentonite}\left( {\mathrm{adsorption}} \right)+2\mathrm{e}\to \mathrm{C}{{\mathrm{u}}^0}-\mathrm{Bentonite} $$

(9)

In addition, bentonite as a support material could stabilize and disperse nZVI as well as prevent nZVI from aggregation and thereby increase the reactivity of the nZVI despite the fact that B-nZVI has a lower surface area than nZVI; these also are confirmed by the evidence presented in this study.

Removal of Cu²⁺ and Zn²⁺ from wastewaters by B-nZVI and B-nZVI reuse

Whether B-nZVI can be used to remove heavy metal ions and B-nZVI can be re-used to treat wastewaters containing heavy metal ions, were tested here. As listed in Table 1, the concentrations of total Cr, Pb²⁺, Cu²⁺, Zn²⁺ and Ni²⁺ ions in the wastewater were measured before and after reacting with B-nZVI. It can be seen that all heavy metals in wastewaters were undetected after reacting with 4 g/l of B-nZVI, except for Cu²⁺ and Zn²⁺ in electroplating wastewater I, which have high initial concentrations. Highly concentrated total Cr in electroplating wastewater I was completely removed because the reduction of Cr(VI) was favorable due to the much more positive standard reduction potential of Cr(VI)/Cr(III) (Φ ⁰ _{Cr(VI)/Cr(III)} = 1.23 V) than Fe²⁺/Fe⁰. However, the initial pH values of three wastewaters were 2.0, 3.1 and 6.5, respectively, which increased to 4.7, 5.2 and 8.3 after reaction mainly because of the formation of OH⁻ when Fe⁰ reacted with H₂O (Chen et al. 2008).

Table 1 Remediation of three different types of actual electroplating wastewater using B-nZVI

The reusability of B-nZVI in removing Cu²⁺ and Zn²⁺ is illustrated in Fig. 6 in terms of the amount removed by each unit mass of B-nZVI. Clearly, the heavy metals were completely removed with different initial concentrations when B-nZVI was first utilized. However, in the reuse cases, compared to the untreated B-nZVI, the removal of Cu²⁺ and Zn²⁺ decreased when B-nZVI treated either with distilled water or 2 % HNO₃ was undertaken. However, less amounts of Cu²⁺ and Zn²⁺ were removed when using treated B-nZVI with 2 % HNO₃. This could be ascribed to the excessive erosion by HNO₃ forming the iron oxide and reducing the reactivity of nZVI. The most durable reactivity was achieved using untreated B-nZVI with more than 50 % being removed even after three-time reuse, except for 50 mg/l Zn²⁺ solution. It indicated that a higher rate of removal of Cu²⁺ was obtained in all cases compared to Zn²⁺. Furthermore, the reusability of B-nZVI was proven to be more effective when the initial heavy metal concentration was low, such as 20 mg/l. This can be explained by the limitation of its removal capacity, which was consistent with the reuse of kaolinite-supported nZVI to remove Cu²⁺ and Co²⁺ (Shahwan et al. 2010).

Fig. 6

Recycling of B-nZVI using different types of treatment. a Cu²⁺, 20 mg/l; b Zn²⁺, 20 mg/l; c Cu²⁺, 50 mg/l; d Zn²⁺, 50 mg/l

Conclusions

In this study, B-nZVI has been used to effectively remove Cu²⁺ and Zn²⁺ compared to nZVI supported by zeolite or kaolin. The reactivity of B-nZVI varied with the mass ratio of bentonite/nZVI, and the highest removal was achieved by B-nZVI (2:1). SEM images showed that the aggregation of nZVI decreased as the proportion of bentonite increased, which led to an increase in the SSA of nZVI. The removals of Cu²⁺ and Zn²⁺ by B-nZVI were consistent with the pseudo-first-order model kinetics, and k _obs rose as the initial concentrations of heavy metals decreased and temperature increased. B-nZVI can be used to remove various heavy metals from electroplating wastewaters. The reusability of B-nZVI was effective in removing Cu²⁺ and Zn²⁺ after being used four times. B-nZVI showed potential as a remediation agent that can be applied for in situ wastewater treatment.