Adsorption of Heavy Metals on Layered Double Hydroxides (LDHs) Intercalated with Chelating Agents

Layered double hydroxides (LDHs) are lamellar ionic compounds containing a positively charged layer and exchangeable anions in the interlayer. In this study, LDHs intercalated with chelating agents were synthesized by anion exchange reaction. The materials synthesized in this work were characterized by chemical analysis, Fourier transform infrared spectroscopy (FT-IR), scanning electron microscopy (SEM), and powder X-ray diffraction (XRD) to confirm their properties. Adsorption experiments from aqueous solutions containing known amounts of some heavy metallic ions onto the adsorbent were explored in a batch system. The amount of metallic ions adsorbed by LDHs intercalated with EDTA and precursor LDHs were determined by inductively coupled plasma-atomic emission spectrometry (ICP-AES) and atomic absorption spectrometry (AAS). In order to examine the adsorption capacity of LDHs intercalated with chelating agents, the adsorption experiment was investigated under the optimum condition. The data were applied to Langmuir and Freundlich isotherm models. The pseudo-second-order kinetic model was more adequate to describe the kinetic in this case. LDHs intercalated with chelating agents synthesized in this work can be promising adsorbents for heavy metals. It is very significant information from the viewpoint of environmental protection.


Introduction
Recently, environmental pollution is one of the most serious problems in the world due to its deep effect on the future of human beings. Then the investigation for resolving the problem of the environmental pollution in the world began to draw major public attention [1,2]. It is well known that the major forms of environmental pollution include air pollution, water pollution, soil pollution, and so on. Among them, water pollution is the most serious due to its liquidity which may bring other pollutions. Surface water pollution and groundwater contamination are some of the environmental problems today. One of the cases of environmental pollution is due to heavy metal contaminants such as copper, lead, cadmium, chromium, arsenic, zinc, etc. Also, heavy metals are concerned because of their strong toxicity even at low concentrations. Based on the type of mining, the kinds and the concentrations of metal ions are many and varied. Heavy metal ions have high toxicity and poor biodegradability for plants and animals at higher concentrations [3,4].
In recent years, clay minerals have been aroused increasing interest as adsorbents by virtue of their properties, which make them attractive materials for adsorbing heavy metal ions. Their abundance in nature, low cost, and good cation adsorptive properties, a result of their negatively charged layers and high specific surface areas, make them suitable for adsorption of metal ions [5,6]. LDH used in this paper are the antitypes of clay minerals. Layered double hydroxides (LDHs) are lamellar ionic compounds containing a positively charged layer and exchangeable anions in the interlayer. They consist of brucite-like layers and are represented by the general formula [ M 1−x where cationic M II and M III are divalent and trivalent metals and occupy the octahedral holes in the brucite-like layer. An − is the interlayer exchangeable anions, which is located in the hydrate layered galleries, and x is the layer charge density [7][8][9]. The layered structure of LDHs is shown in Figure 1.
The ethylenediaminetetraacetic acid (EDTA) is a chelating agent widely used in industry and agriculture. It forms strong complexes with the ratio 1:1 between heavy metal ions and ligand. The EDDS (N, N′-1, 2-Ethanediylbis-1-Aspartic Acid) is also a chelating agent, which may offer a biodegradable alternative to EDTA and is currently used on a large scale in numerous applications [10,11]. The structure of this two chelating agents were shown in Figure 2.
Considering the structure of LDHs, it is suggested that these compounds can be intercalated with different polydentate ligands. Recently, the study using LDHs modified with chelating agents as the potential adsorbents of heavy metals from aqueous solution has been reported [12,13]. The aim of this work is at first to synthesize and to characterize LDHs intercalated with EDTA or EDDS and to study the uptake of heavy metals (Cu 2+ , Pb 2+ , Cd 2+ ) by these hybrid compounds. The following five kinds of compounds synthesized in this work are ZnAl-NO 3 (L1), ZnAl-EDTA (L2), MgAl-NO 3 (L3), MgAl-EDTA (L4), and MgAl-EDDS (L5). To confirm the effect of intercalation with EDTA, the adsorption of metallic ions onto L1 and L2 is also compared. This study investigated the adsorption ability of LDHS as adsorbent for Pb, Cu, and Cd from aqueous solution. Finally, the further

Synthesis of the adsorbents
The synthesis of LDHs intercalated with EDTA or EDDS includes two steps: (1) the preparation of the precursor LDHs (L1 or L4) and (2) the anion exchange reaction of this compound with chelating agents [14]. All the synthesis was purged with N 2 to avoid CO 2 uptake from atmosphere.
• Synthesis of L2, L3, and L5 L2 was synthesized as follows. Under a N 2 atmosphere, 0.015 mol of EDTA or EDDS was added to the 150 mL of suspended solution of L1. Then, the mixing solutions were agitated at 70°C for 8 h under a certain pH degree, then separated by centrifugation, washed until neutral, and then dried at 60°C overnight [12,17]. L5 was synthesized by L4 as the similar method for L2.

Characterization of these adsorbents
Elemental chemical analyses of C, H, and N in LDHs were carried out using an elemental analyzer instrument (JMC10, J-SCIENCE LAB CO., Ltd.). After dissolving the sample by HNO 3, the amount of metallic ions in LDHs was obtained by ICP-MS (Agilent HP 4500, Thermo). Infrared spectra were obtained using the KBr disc method, with wavenumbers from 400 to 4000 cm −1 on a FT-IR (FTIR-4200, Jasco, Japan). XRD (X-ray powder diffraction) of LDHs samples were carried out on a RINT2500HR-PC (RIGAKU Corporation) using Cu Kα radiation in the scanning range of 2-80°. N 2 adsorption and desorption isotherms were employed to determine the specific surface area by the specific surface area analyzers (AUTOSORB-1, Quabtachrome Inc., USA). The surface morphology of LDHs was surveyed using scanning electron microscopy (SEM; JSM-5800, JEOL, Japan). The element distribution and the component analysis were also analyzed by electron probe micro analyzer (EPMA; 1600, Shimadzu Corporation).

Adsorption experiments
For obtaining the optimum conditions regarding the adsorption of heavy metal, the batch experiments were studied by varying pH, contact time, adsorbent dose, and initial concentration on the adsorption of heavy metal [18][19][20]. The adsorption experiments of Cu(II) and Pb(II) using L2 and L3 were carried out. A certain amount of L2 or L3 was contacted with 30 mL of an aqueous solution containing known initial each metal ion (nitrate salts) ranging from 0.1 to 2 g L −1 . Sorption experiments were conducted in the pH range of 2-6, contact time from 30 minutes to 6 h, temperature from 25 to 40°C, and adsorbent dosage 5-40 mg. The pH of each solution was adjusted using 0.1 mol L −1 NH 4 OH and 0.1 mol L −1 HNO 3 . The adsorption capacities of Cu(II) or Pb(II) on L1, L2, and L3 were compared with that of commercial LDHs: DHT-4A ([Mg 4.5 Al 2 (OH) 13 CO 3 ⋅3.5H 2 O], Kyowa Chemical Industry Co., Ltd), which is abbreviated as L0 below.
The adsorption experiments of Cd(II) were also carried out similar as the method below. The experiment using heavy metallic ions solution without the adsorbent was also performed to identify potential loss of heavy metallic ions during the process such as precipitation. To confirm the effect of intercalation with EDTA, the adsorption of Cu(II), Pb(II) Cd(II) onto L4 and L5 are also compared.
The suspension containing the adsorbent and each of the above metallic solution was filtered through a 0.10 μm membrane filter (Mixed Cellulose Ester 47 mm, Advantec MFS, Inc.) to remove each metallic ion that have been adsorbed into the adsorbent. Then, the concentration of Cu(II) or Pb(II) in the filtrate was determined with an atomic absorption spectrophotometer (AAS), and the concentration of Cd(II) in the filtrate was determined by inductively coupled plasma-atomic emission spectrophotometer (ICP-AES) (SPS 1500, Seiko Instrument Inc).

Data analysis
For data analysis, various equilibrium, kinetic, and thermodynamic models (equations) were employed to interpret the data and establish the extent of adsorption. The metallic ions uptake by each adsorbent was calculated using the Eq. (1): (1) where Q is the adsorption capacities at equilibrium (μg g −1 ), C o and C e are the initial and equilibrium concentrations of metallic ions in a batch system, respectively (mg L −1 ), V is the volume of the solution (L), and W is the weight of each adsorbent (g) [15,16].

Adsorption isotherms
In adsorption processes, it is necessary and critical for the equilibrium isotherm studies to predict the behavior of pollutant adsorption onto the sorbent surfaces. Two common adsorption models, Langmuir and Freundlich isotherm models, were applied to evaluate the adsorption data obtained in this study.
The Langmuir adsorption model is based on the assumption that maximum adsorption corresponds to saturated monolayer of solute molecules on the adsorbent surface [21,22]. Langmuir model is given by Eq. (2): where C e is the concentration of metallic ions in batch system at equilibrium (mg L −1 ), q e is the amount of adsorption of metallic ions at equilibrium (mg g −1 ), q max is the maximum adsorption capacity on the surface of adsorbent (mg g −1 ), and K L is the equilibrium adsorption constant (L mg −1 ). A plot of C e /q e versus C e gives a straight line (Y = A + BX) with slope of 1/q max , and the intercept is 1/(K L q max ); The linearized Freundlich model isotherm is represented by the following equation: where K F and 1/n indicate the adsorption capacity and the adsorption intensity of the system, respectively. The plots of q e versus C e in log scale can be plotted to determine values of 1/n and K F depicting the constants of Freundlich model. The greater the value of the n, the more favorable is the adsorption [15,16,23].

Kinetic model
The kinetic data can be used to determine the time required for adsorption equilibrium and provide useful data to improve the efficiency of the adsorption model and develop predictive models [24,25]. In this work, pseudo-first-order and pseudo-second-order models were applied for modeling the adsorption process. The pseudo-first-order model is expressed as the Eq. (4): where q t and q e (μg g −1 ) are the metal amount adsorbed at time at t (h) and equilibrium, respectively, and k 1 is the rate constant of the pseudo-first-order adsorption (h −1 ).
The linear form of the pseudo-second-order rate equation is given as follows: where k 2 (g μg −1 h −1 ) is the pseudo-second-order rate constant of the adsorption [15,16,25].

Chemical analysis
The chemical analysis of LDH samples is shown in Table 1. The molar ratio of M II ∕ M III in L1 or L4 is nearly 2 which is well fitted to the expected formula. However, L2, L3, or L5 has lower M II ∕ M III ratio than L1 or L4. The presence of polydentate ligand (Zn-EDTA or Al-EDTA complex) can result in this decrease of the ratio. Moreover, the decrease suggests that octahedron in hydroxyl layer has a partial dissolution (pK sp (Zn(OH) 2 ) = 13.7, pK sp (Al(OH) 3 ) = 32.7, pK sp (Mg(OH) 2 ) = 12.7) during the anion exchange reaction which is performed at pH 5-6 [12,17,26]. The C/N of L5 is lower than that of EDTA ligand (5), and the little gap between them may be mainly due to the registration of nitrate ions in the interlayer. The more nitrate ions are included in LDHs, the lower the C/N value is [27].

FT-IR spectra
The FT-IR spectra of L1 and L2 and L3 are shown in Figure 3, and that of L4 and L5 are shown in Figure 4. Typical M-OH (M-metallic ions) vibration modes due to the hydroxide layer between 400 and 1000 cm −1 are found in both Figures 3 and 4.
The very sharp peak at 1385 cm −1 in Figures 3(a) and 4(a) is attributed to the NO 3 − stretching vibration. The NO 3 − stretching vibration at 1385 cm −1 is not observed from Figures 3(b) to 4(b). It may be due to the group which is hidden by the band at 1394 cm −1 [26,28]. The absorption bands at 1600 and 1394 cm −1 are characteristics of the symmetrical and asymmetrical vibration of COO-groups. The position of these bonds is similar to the spectrum of LDHs which is reported by Parida et al. [29] and [30]. It is found that EDTA has been intercalated into the interlayer successfully, although a certain amount of -NO 3 may still retain in the compound judging from the results of chemical analysis. The wide band at around 3450 cm −1 may be attributed to the -H bonding stretching vibrations of -OH groups and water molecules. The band at 1623 cm −1 of L1 and L4 is assigned to water bending vibration [8,26].

XRD patterns
XRD patterns of L1 and L2 and L3 are shown in Figure 5, and those of L4 and L5 are shown in Figure 6. They are typical XRD patterns of LDHs. The strong diffraction peaks at low angle, assigned to basal planes (003), (006), (009), were sharp and symmetric compared to the peaks at high angle, which are characteristics of clay mineral shaving a layered structure [29][30][31]. From the XRD pattern, the basal spacing (d) values of sample were calculated by using Bragg equation and the angle of peak (003). Then the gallery height was obtained by subtraction from the basal spacing to the layer width (0.48 nm) [30]. The basal spacing and the gallery height of L1, L2, L4, and L5 are shown in Table 2. It indicates that the intercalation of EDTA into NO 3 -LDHs gives rise to an increase of basal spacing. This basal spacing could identify the existence of EDTA, because it is close to the dimensions of EDTA complexes (0.9 nm −1 nm) founded by single crystal XRD of M-EDTA (M-metallic ions) compound [12,17,30,32].

SEM micrographs
SEM images of all composite synthesized in this work are shown in Figure 7. These adsorbents have clear plate-like morphology, which is typical for LDHs [33]. The intercalated product particles are more homogeneous than the precursor product  which may be due to the hydrogen bonding on the layer. Hydrogen bonding makes soft agglomeration occur on the surface of the LDHs, and after the chelating agent, anion replaces the nitrate ions between the layers, the hydrogen bonding between the hydroxyl groups is reduced, and the aggregation is weakened to a certain extent.
The inhomogeneous surface of the adsorbent indicated that a large amount of metal salt attached to the surface of the hydrotalcite in an excessive state, its unique    layered structure, resulting in removal of heavy metal ions in the aqueous solution not only by interlayer anion and heavy metal cation interaction but also rely on the role of surface adsorption and sedimentation.

Electron probe microanalyzer (EPMA)
Element distribution analysis of L1, L2, L4, and L5 by EPMA is shown in Figure 8. After the ion exchange, the element distribution of N decreased obviously (by comparing red parts in these pictures), and this decrease is observed in both MgAl-LDHs (a, b) and ZnAl-LDHs (c, d). Furthermore, it is found that the moles of divalent metals are at least equal to or greater than that of the trivalent metals [34,35], which is consistent with the results of chemical analysis. Figure 9 has shown the specific surface area of the product. Specific surface area of L2 and L3 are bigger than that of L1, and that of L5 is bigger than L4, which may be attributed to intercalation of EDTA or EDDS. Specific surface area of L0 is bigger than that of L1; it is due to the difference of their particle size.

Adsorption of Cu(II) or Pb(II) onto L1, L2, L3, and L0
The adsorption capacities of Cu(II) or Pb(II) onto L1, L2, L3, and L0 are compared in Figure 10. The adsorption efficiency of Cu 2+ was larger than that of Pb 2+ for the same absorbent, which could be attributed to their stability constant (EDTA-Cu, 18.7; EDDS-Cu, 18.4; EDTA-Pb, 18.0; EDDS-Pb, 12.7) [12,26]. That is to say, it can be considered that the large adsorption capacity is obtained when the stability constant of chelate-metal is high. By comparing among adsorbents used in this work, the order of the adsorption capacity is L2 > L3 > L0 > L1. The higher adsorption efficiency of L0 than L1 may be attributable to its high specific surface area.

Adsorption of Cu(II), Pb(II), and Cd(II) onto L4 and L5
In order to confirm the effect of the intercalation with chelate agents on the adsorption capacity of metals, the adsorption experiments of some metallic elements    onto L1 and L2 are compared. The adsorption of Cd(II), Cu(II), and Pb(II) onto these LDHs under the optimum condition are shown in Figures 11-13, respectively.
Both LDHs were found to take up Cd(II), Cu(II), and Pb(II) from aqueous solutions, and the uptake was found to increase with time. The adsorption capacity of both LDHs for Cd(II), Cu(II), and Pb(II) increased rapidly during the initial stages, and thereafter it increased gradually. It is generally found that the time needed for L5 to reach equilibrium was shorter than that for L4. From the adsorption experiment, the improvement of adsorption capacity by intercalation was observed. On the other hand, the adsorption capacity of Cu(II) and Pb(II) at equilibrium was    higher than that of Cd(II). It is considered that heavy metal was removed by LDHs including two mechanisms: chemical precipitation and chelation [16]. In the first case, the hydroxyl anions compete with chelating agents for the precipitation of metal hydroxides at higher pH, and divalent ions are usually selectively dissolved. In the second case, the adsorption affinity is generally determined by the stability constant of the corresponding complex [36][37][38].

Adsorption isotherms
The adsorption isotherms for Cu(II) or Pb(II) were obtained under the optimum adsorption conditions (i.e., pH 6, contact time 120 minutes, temperature 25°C, and adsorbent dosage 10 mg). The adsorption isotherms of Cu(II) or Pb(II) onto L2 and L3 were analyzed using Langmuir and Freundlich equations and were shown in Figures 14  and 15, respectively. From Figure 14, the linear correlation coefficient (R 2 ) of L2 (Pb 2+ ), L3 (Pb 2+ ), L2 (Cu 2+ ), and L3 (Cu 2+ ) and other parameter for Langmuir and Freundlich isotherms model were shown in Table 3, respectively. That is, L2 and L3 synthesized in this work are well fitted by Freundlich adsorption isotherms models.

Kinetic model
The kinetic isotherms for Cu(II) or Pb(II) were obtained under the optimum adsorption conditions (i.e., pH 6, concentration 200 ppm, temperature 25°C and  Table 4 which showed that adsorption process followed pseudo-second-order rather than pseudo-first-order model. The second order kinetic models plot for the adsorption of Cu(II) or Pb(II) on L2 or L3 is shown in Figure 16. The experimentally calculated values of q e at various concentrations were in a good agreement with theoretical calculated values. Also, the values of correlation coefficients (R 2 ) for the pseudo-second-order kinetic model was nearly 1, indicated that pseudo-second-order kinetic model was better obeyed.

Comparison of the adsorption capacities of the materials with other sorbents
The comparison of maximum adsorption capacity of these LDHs for Cu(II) in a present study with that of another adsorbents in previous literatures [39] are  Table 4. The kinetic fit parameters for Cu(II) and Pb(II) adsorbed on L2 or L3.

Figure 16.
The correlation of experimental data to pseudo-second-order models.
presented in Table 5. Moreover, Table 6 shows the comparison of adsorption capacity of Pb(II) by other adsorbents reported in the literature. As seen in Tables 5 and 6, the adsorption capacity of these LDHs for Cu(II) and Pb(II) in this work is on a level with that of another adsorbents in previous works.

Comparison of the adsorption capacities of LDHs in other literature Cu 2+ .
Adsorbents Q (mg g −1 )