Effective Removal of Pb(II) from Aqueous Media by a New Design of Cu-Mg Binary Ferrite.

Metal oxides and their composites have been extensively studied as effective adsorbents for the removal of heavy metals from aqueous solutions in environmental remediation. In this work, Cu0.5Mg0.5Fe2O4 was synthesized by a co-precipitation method followed by calcination (900 °C) and investigated for Pb(II) adsorption. The resultant samples were characterized by various analytical techniques including X-ray diffraction, N2 adsorption-desorption, scanning electron microscopy, thermogravimetric analysis, and Fourier transform infrared spectroscopy. The results revealed that single-phase cubic spinel was obtained by the calcination of as-synthesized samples at a temperature of 900 °C. Cu0.5Mg0.5Fe2O4 ferrite is a mesoporous material with a surface area, a total pore volume, and an average pore size of 41.3 m2/g, 0.2 cm3/g, and 15.1 nm, respectively. Pb(II) adsorption on Cu0.5Mg0.5Fe2O4 fitted well to the Langmuir model, indicating monolayer adsorption with a maximum capacity of 57.7 mg/g. The pseudo-second-order kinetic model can exactly describe Pb(II) adsorption with the normalized standard deviation (Δq) of 1.24%. The obtained results confirmed that the Cu0.5Mg0.5Fe2O4 ternary oxides exhibit a high adsorption capacity toward Pb(II), thanks to the increase in active adsorptive sites of ferrite.

Introduction

Spinel ferrites have recently emerged as a new class of adsorbents that are preferred for water treatment.[1−3] These materials have a high surface area and great active sites for interaction with contaminants, and therefore offer a high adsorption capacity. Spinel ferrites possess special superparamagnetic properties (SPM), allowing them to be easily recovered from the reaction mixture by application of an external magnetic field.[4,5] Spinel ferrites have been investigated to remove various contaminants in water such as organic compounds,[6−8] nutrient salts,[9] and toxic metals.[10−13]

Spinel ferrites have a general chemical formula of AFe2O4, where A is metallic ions such as Mn2+, Fe2+, Co2+, Ni2+, Zn2+, and so forth usually located in tetrahedral sites and Fe3+ positioned at the octahedral sites of the ferrite structure. Depending on the position of metal ions, three possible spinel ferrite structures, that is, normal, inverse, and mixed spinel ferrite can be available. In the normal structure, M2+ are located at tetrahedral sites, while Fe3+ are located at octahedral sites. In the inverse structure, Fe3+ are distributed equally at both sites; however, M2+ are located at octahedral sites. In the mixed structure, both ions can occupy both sites. These structures contain a number of active sites of surface charge, dipoles, and −OH groups, which could create interactions with contaminants.[3,14] Interactions that are responsible for adsorption could be chemical bonding, ion-exchange, inner-sphere or outer-sphere complex formation, van der Waals forces, hydrogen bonding, and dipole–dipole and π–π interactions depending on the nature of contaminants, adsorbents, and adsorption conditions. For example, chemical ion-exchange and formation of an outer-sphere complex are identified as a major adsorption mechanism for Pb(II) adsorption on Co0.6Fe2.4O4 at pH < 7, but at pH > 7, inner-sphere surface complexes are the dominant adsorption mechanism.[15]

Lead is a highly toxic metal; drinking water contaminated with lead ions may cause a number of health consequences such as anemia, gastro-intestinal colic, brain damage, diseases of liver and kidney, and cancer. Wastewater from the lead-acid battery industry is one of the major sources that is responsible for lead contamination. There are many possible methods to remove lead from water.[16] Recent studies revealed that spinel ferrites and their composites possess excellent Pb(II) adsorption.[17] The Pb(II) adsorption capacity of Co0.6Fe2.4O4 was 80.3 mg/g at 45 °C[15] and that of carbon nanotubes–CoFe2O4–NH2 was 66.3 mg/g at 30 °C; however, it increased to 140.4 mg/g by functionalizing with chitosan.[18] Carbon nanotubes–iron oxide magnetic composites have a Pb(II) adsorption capacity of 105.6 mg/g at 20 °C,[19] while amino-functionalized Fe3O4 adsorbs only 40.1 mg/g at 25 °C. CoFe2O4–reduced graphene oxide composites exhibited a high Pb adsorption capacity of up to 299.4 mg/g at 25 °C and pH 5.3.[20] MgFe2O4 also possesses a high adsorption capacity of 125.0 mg/g, but its graphene oxide composite can adsorb up to 142.8 mg/g.[21]

Copper ferrite (CuFe2O4) has attracted much attention from researchers for water purification because of its high SPM, low-cost, and high stability.[1] A wide range of contaminants from organic compounds to heavy metal ions can be effectively removed by CuFe2O4; however, its Pb(II) adsorption is relatively poor. A study by Tu et al. revealed that a maximum Pb2+ adsorption capacity of CuFe2O4 is 17.83 mg/g at 25 °C and pH 4.5,[22] while another work done by Al Yaqoob et al. showed that it can reach 31.1 mg/g at pH 12 and 25 °C.[23] Therefore, it is very interesting to think of an efficient way to improve the Pb(II) adsorption of CuFe2O4 in order to extend its potential application. One of the most popular methods to improve its adsorption is to tailor functional groups such as the amino group or graphene oxide on ferrite particles.[18,19,21,24−27] Besides, the modification can also be conducted via doping or partial substitution of Cu2+ in CuFe2O4 with different metal ions.[28] Camacho-González et al. showed a possibility to improve Pb(II) adsorption on CuFe2O4 by doping with Zn2+;[29] however, the research in this direction is still very limited. A literature survey revealed that ZnFe2O4 has a high Pb2+ adsorption capacity of up to 96.9 mg/g; meanwhile, MgFe2O4 posed to be the best ferrite for Pb2+ adsorption with a maximum capacity of 125.0 mg/g.[21] Accordingly, it is expected that doping or partially replacing Cu2+ in CuFe2O4 with Mg2+ could enhance its Pb2+ adsorption; however, the Pb2+ adsorption on a Cu–Mg binary ferrite (Cu0.5Mg0.5Fe2O4) has not been investigated. Therefore, the objective of this study is to synthesize a Cu–Mg binary ferrite and examine its Pb(II) adsorption behavior. The Pb(II) adsorption performance on the ferrite is analyzed, and the adsorption improvement that resulted from magnesium substitution is evaluated.

Results and Discussion

Characterization

The X-ray diffraction (XRD) patterns of Cu0.5Mg0.5Fe2O4 samples calcined at different temperatures are shown in Figure a. All diffraction peaks are attributed to the Fd3̅m space group in the cubic structure, which are in good agreement with the standard patterns of both CuFe2O4 (JCPDS 01-077-0010) and MgFe2O4 (JCPDS 01-089-3084). XRD results revealed that the crystallinity of samples was greatly affected by the calcination temperature. The spinel structure was not clearly observed at temperatures lower than 400 °C. However, the intensities of characteristic peaks of the spinel structure at 35.5, 43.2, and 62.7 °C could be observed as the temperature increased to above 500 °C, and a single spinel phase was obtained at 900 °C. This confirmed that increasing calcination temperature is favorable for Cu0.5Mg0.5Fe2O4 synthesis and reaches an optimum at 900 °C. Scanning electron microscopy (SEM) images (Figures b and S1) show that Cu0.5Mg0.5Fe2O4 has a cube-like shape with an average particle size of approximately 30 nm. This is also exhibited in the particle size histogram shown in Figure S2. This is because the particles size observed in the SEM image is less than 100 nm. The Debye–Scherrer equation (eq ) can also be utilized to calculate Cu0.5Mg0.5Fe2O4 particles as followswhere K is a constant dependent on the crystallite shape, θ is the diffraction angle degree (one-half of 2θ, which is the position of the peak), λ is the X-ray wavelength, and B represents the full width at half maximum of the peak. The average crystal size calculated from the Debye–Scherrer equation is approximately 29.5 nm, which is consistent with the particle size obtained from SEM images. The composition of the as-synthesized sample was estimated based on energy-dispersive X-ray spectroscopy (EDS) analysis (Figure c), and the result suggested a Fe/Mg/Cu molar ratio of 2:0.65:0.5, which is relatively consistent with the nominal value of theoretical calculation. Fourier transform infrared spectroscopy (FTIR) analysis showed a typical vibration peak at 576 cm–1, which is attributed to the Fe–O stretching mode of spinel ferrite (Figure d). Peaks observed at 1630, 2368, and 3434 cm–1 correspond to the vibration of adsorbed water and CO2. The results further confirmed the successful formation of Cu0.5Mg0.5Fe2O4.

Figure 1

XRD patterns (a), representative SEM image of the sample calcined at 900 °C (b), EDS analysis (c), and FTIR spectrum (d) of the Cu0.5Mg0.5Fe2O4 sample.

The result of thermogravimetric analysis (TGA) of the sample is shown in Figure a. The TGA curve revealed a significant mass loss that occurred below 200 °C. This is attributed to the loss of moisture and physically adsorbed water. The mass reduction then extended to approximately 375 °C because of the loss of structured water, mostly the condensation of −OH groups to release water molecules. No mass change observed at above 375 °C indicated the high thermal stability of the resultant Cu0.5Mg0.5Fe2O4 spinel ferrite.

Figure 2

TGA profile (a), nitrogen adsorption isotherm (b), and pore size distribution (b, inset) of Cu0.5Mg0.5Fe2O4 samples.

In order to be employed as an adsorbent, it requires Cu0.5Mg0.5Fe2O4 spinel ferrite to have a large number of adsorption sites, which is accessible to adsorbates. In other words, the material should possess a reasonable surface area, pore volume, and pore size. The result of nitrogen adsorption/desorption study (Figure b) showed that Cu0.5Mg0.5Fe2O4 spinel ferrite is a mesoporous material with an adsorption isotherm of type IV according to IUPAC classification. The isotherm is associated with type H1 hysteresis, indicative of a narrow distribution of uniform mesopores and limited networking effects.[30] These results suggest that the mesopores are likely the porous channels created by the arrangement of uniform Cu0.5Mg0.5Fe2O4 cubic crystals as primary building units in large aggregates. These pores can be clearly observed in the SEM image (Figure b), showing a large aggregate formed by interconnecting Cu0.5Mg0.5Fe2O4 nanocubes. The Cu0.5Mg0.5Fe2O4 spinel ferrite has a surface area, a total porous volume, and an average pore size of 41.3 m2/g, 0.2 cm3/g, and 15.1 nm, respectively. With the high surface area and porosity, the adsorbent is expected to have the high adsorption sites exposed, and therefore has a high adsorption capacity.

A representative magnetic hysteresis loop of the Cu0.5Mg0.5Fe2O4 sample is shown in Figures a and Figure S3. The saturation magnetization (Ms) and coercivity (Hc) of the Cu0.5Mg0.5Fe2O4 sample are 23.1 emu/g and 83 Oe, respectively. This relatively high saturation magnetization of Cu0.5Mg0.5Fe2O4 ternary oxides can be considered as a superparamagnetic material, meaning that the material can be facilely separated by using an external magnet. The adsorbent separation by the application of an external magnetic field is observed in Figure b. The obtained results revealed that the as-synthesized material can be magnetically separated and re-dispersed in water and potentially employed as a magnetic adsorbent for water treatment.

Figure 3

Magnetic hysteresis loop of the Cu0.5Mg0.5Fe2O4 sample measured at room temperature (a) and magnetic separation using a magnet (b).

Adsorption Study

Effect of pH on Pb(II) Adsorption

The adsorption behavior of the Cu–Mg binary ferrite toward Pb(II) with various degrees of Mg substitution in CuFe2O4 was briefly investigated, and is shown in Figure S4. It is obvious that the relative adsorption capacities of Cu–Mg binary ferrite toward Pb(II) increase along with the increase of Mg substitution degree. However, after the Mg substitution degree reaches 0.5, the further increase of Mg substitution witnessed the insignificant increase of adsorption capacities. Thus, hereafter, the Cu0.5Mg0.5Fe2O4 ternary oxides are considered as optimized composition for the following investigation.

In most adsorption systems, the solution pH is a very important factor, which greatly affects the adsorption performance of the adsorbent. The pH, at the same time, regulates the metal species existing in the solution and modifies the surface charge of the adsorbent. As shown in Figure a, Pb(II) ions are predominant at pH ≤ 6; however, they convert to hydroxide species (PbOH+, Pb(OH)2, and Pb(OH)3–) at pH > 6. Based on the precipitation constant, K = 1.2 × 10–15, and the initial concentration of Pb(II) in this study, C = 10–200 mg/L, the precipitation will not occur at pH ≤ 8.[31] For example, at the initial Pb(II) concentration of C = 20 mg/L (≈9.66 × 10–5 mol/L), it is expected that Pb(OH)2 precipitation can begin at pH ≈ 8.5. Accordingly, the precipitation does not exist at low Pb(II) concentrations (10–200 mg/L) and low pH (pH ≤ 8.) To further validate the above calculation, the precipitation of Pb(OH)2 with certain Pb(II) concentration at various pH values were investigated. The results showed that no precipitation of Pb(OH)2 and no decrease in Pb(II) concentration were observed in pH ≤ 7 solution, and with pH > 7, negligible amount of Pb(II) was decreased. This was probably the low concentration of the studied Pb(II) solution. High pH not only generates hydroxide species in solution but also enhances the deprotonation reaction on the surface of the adsorbent, which results in the increase of the negative surface charge of the adsorbent. Different adsorbents may respond to the pH change differently depending on their chemical composition and microstructure; therefore, there is no common adsorption pattern for all types of adsorbents. Pb2+ adsorption on Fe3O4, for example, can be improved with pH increasing from 2 to 6, but after being functionalized with amino-groups, it reached a maximum at pH 5, then reduced at a higher pH.[32]

Figure 4

(a) Effect of pH on the Pb2+ species in solution (adapted with permission from ref (15), Copyright 2015 Elsevier) and its adsorption on Cu0.5Mg0.5Fe2O4 and (b) a function of pH on the adsorption performance (the experiments were conducted at room temperature; adsorbent dose: 0.1 g/L; and initial Pb2+ concentration: 20 mg/L).

Table presents the optimal pH of different ferrites and functionalized ferrites toward Pb(II) adsorption. In general, the improvement in Pb(II) adsorption along with the increase in pH can be observed in most of the investigated ferrite adsorbents, and a similar trend can be seen in this study for Cu0.5Mg0.5Fe2O4, as exhibited in Figure b. Pb(II) adsorption increased at solution pH from 3 to 7, and its high adsorption capacity was maintained at pH from 7 to 11. Obviously, Pb(II) adsorption on Cu0.5Mg0.5Fe2O4 is strongly pH-dependent. The increase in adsorption capacity with solution pH from 3 to 7 is related to both lead species existing in solution and surface functional groups of the adsorbent.[15,34] The adsorbent comprises different metal oxides, and in the aqueous media, those metal oxides could be hydrolyzed to form neutral (−OH) groups. Depending on pH, the adsorbent surface could acquire a charge by protonation or deprotonation of the −OH groups. At low pH, the protonation reaction occurs, as shown in eq . This reaction may compete with the adsorption of Pb(II) onto the adsorbent (eq ), resulting in low adsorption capacity.[15,35] At high pH, deprotonation occurs as seen in reaction 4 creating the ≡S–O– sites which have strong affinity with Pb(II), and as a result, its adsorption onto the adsorbent improved with the increase in pH. Even though the high Pb(II) adsorption on Cu0.5Mg0.5Fe2O4 was observed until a pH of 11, pH 7 will be used for all later experiments to eliminate any possible influence caused by Pb(OH)2 precipitation.

Table 1

Optimal pH of Different Adsorbents toward Pb(II) Adsorption

no	adsorbent	pH investigated	optimal pH	references
1	Co0.6Fe2.4O4	2–12	7	(15)
2	CoFe2O4–NH2	2–6	6	(18)
3	chitosan-functionalized CoFe2O4–NH2	2–6	6	(18)
4	CoFe2O4–reduced graphene oxide composite	2–7	5.3	(20)
5	MgFe2O4–NH2	2–10	4	(25)
6	Fe3O4	2–6	6	(32)
7	amino-functionalized Fe3O4	2–6	5	(32)
8	Fe3O4/chitosan/graphene oxide composite	1–7	5	(33)
9	NiFe2O4	2–8	5	(34)
10	Mg0.5Cu0.5Fe2O4	3–11	7	this study

Adsorption Isotherm

One of the most important characteristics for evaluating the performance of the adsorbent is the maximum quantity of the adsorbate that an adsorbent can uptake. To determine the adsorption capacity of Cu0.5Mg0.5Fe2O4, the adsorption isotherm has been investigated to collect the adsorption equilibrium. The adsorption isotherm of Pb(II) on the Cu0.5Mg0.5Fe2O4 sample was investigated by mixing the adsorbent with Pb(II) solution for 3 h at room temperature in which the Pb(II) concentration varied from 6.6 to 84 mg/L at pH 7. The obtained adsorption isotherm is shown in Figure a. The Pb(II) concentration at equilibrium reached 0.09 mg/L as its initial concentration is 6.6 mg/L. The removal efficiency increased along with the decrease in the initial Pb(II) concentration and reached over 98.5% at the initial Pb(II) concentration below 29.3 mg/L (Figure b). This result revealed that Cu0.5Mg0.5Fe2O4 could be an effective adsorbent for Pb(II) removal in practical applications, where the Pb(II) concentration in wastewater is usually lower than 30 mg/L.

Figure 5

Adsorption isotherm (a), adsorption efficiency (b), and a linear plot of isotherm data based on Langmuir (c) and Freundlich (d) models.

To further investigate the adsorption characteristics of the adsorbent, the obtained data were fitted with Langmuir and Freundlich isotherm equations, expressed as followswhere qm (mg/g) is the maximum adsorption capacity, b is the binding constant which relates to the heat of adsorption, and kf and n are Freundlich constants.

The Langmuir model provides the simplest and most useful isotherm for the characterization of chemical adsorption based on several assumptions that the structure of the adsorbent is homogeneous; the adsorbed molecules or atoms are held at definite, localized sites; each site can accommodate only one molecule or atom; and the energy of adsorption is constant over all sites and no interaction occurs between neighboring adsorbates. In contrast, the Freundlich model is used to describe the heterogeneous systems and reversible adsorption, which does not restrict the monolayer adsorption. The linear plots of experimental data based on the Langmuir and Freundlich models are shown in Figure c,d, respectively. The Langmuir model has a better fit with a regression coefficient R2 = 0.9847 in comparison with R2 = 0.9711 for the Freundlich model. The maximum adsorption capacity of the sample calculated from the Langmuir model is 57.44 mg/g. The high adsorption capability of Cu0.5Mg0.5Fe2O4 toward Pb(II) could be ascribed to increasing surface area, porosity, and particle size. This indicates that the Pb(II) ions are adsorbed on Cu0.5Mg0.5Fe2O4 as a monolayer, and chemisorption is the major adsorption mechanism.

Compared to the published data on Pb(II) adsorption on CuFe2O4 (Table ), it could be inferred that the Pb(II) adsorption was improved by partially replacing Cu2+ with Mg2+ in the ferrite structure. However, it is really hard to explain the contribution of Mg2+ in the adsorption improvement, because the crystalline structure of CuFe2O4 is not significantly changed due to Mg2+ doping.[22,23] CuFe2O4 can exist in a tetragonal or cubic structure depending on the annealing temperature. The tetragonal structure is stable at room temperature, and the transformation to cubic one can be observed at around 360–420 °C.[36,37] Previous studies demonstrated that the cubic structure of CuFe2O4 and Mg0.5Cu0.5Fe2O4 possesses a larger unit cell volume; this could enhance the adsorption and ion exchange capacity of the ferrite.[38,39] In this study, both reference CuFe2O4[22,23] and the synthesized Mg0.5Cu0.5Fe2O4 existed in the cubic structure, therefore the influence of the crystalline structure on the improvement in adsorption performance can be eliminated.

Table 2

Comparison of Pb(II) Adsorption Capacities of Synthesized Cu0.5Mg0.5Fe2O4 with Reported CuFe2O4

no	adsorbent	surface area (m2/g)	pore volume (cm3/g)	particle size (nm)	adsorption capacity (mg/g)	optimal pH	temperature (°C)	references
1	CuFe2O4	48.3	0.08	≈50	17.8	4.5	25	(22)
2	CuFe2O4			52.8	31.1a	12	25	(23)
3	Mg0.5Cu0.5Fe2O4	41.3	0.2	29.5	57.7	7	25	this study

Adsorption capacities were taken from experimental data.

In fact, the maximum adsorption capacity varies greatly with the characteristics of the adsorbent and testing conditions, particularly, temperature and pH. To compare the maximum adsorption capacity of Cu0.5Mg0.5Fe2O4 ferrite with others reported in the literature, the capacities generated at room temperature (25 °C) and optimal pH were selected to minimize the influence of testing conditions. It was found that the same metal ferrite synthesized and examined by different investigators may differ in Pb(II) adsorption capacities. For example, MnFe2O4 prepared by Ren et al. adsorbed 69.1 mg/g at room temperature and pH 6,[35] but that prepared by Perez had a maximum adsorption capacity of 33.6 mg/g under similar conditions, pH 5.5 and room temperature.[40] Obviously, the active adsorptive sites, which are accessible to Pb2+, are varied in the same type of ferrite reported in different studies. This suggested that those ferrites may possess different properties, particularly, particle size, surface area, and porosity, which are likely the influencing parameters that cause the difference in adsorption capacity. In this study, the surface area of Cu0.5Mg0.5Fe2O4 (41.3 m2/g) is slightly lower than that of CuFe2O4 (48.3 m2/g); however, its pore volume is significantly higher (0.2 cm3/g) in comparison with the reported CuFe2O4 (0.08 cm3/g), as shown in Table . Additionally, the particle size of CuFe2O4 is much smaller; 29.5 nm compared to ≈50 nm. Therefore, it can be confirmed that the higher pore volume and smaller particle size are major characteristics that contribute to the improvement of Pb(II) adsorption on Cu0.5Mg0.5Fe2O4.

Apparently, the substitution of Cu2+ by Mg2+ in copper ferrite resulted in higher pore volume and smaller particle size. This is in good agreement with previous studies in which Mg2+ is used to substitute Fe2+ in the ferrite structure.[41−43] Substitution of Fe2+ with Mg2+ in Fe3O4 to form Mg0.27Fe2.5O4 increased both the surface area and the pore volume, from 162 to 438.2 m2/g and from 0.388 to 0.648 cm3/g, respectively.[42,43] As a result, the adsorption capacity of Mg0.27Fe2.5O4 toward As(III) reached 127 mg/g compared to 95 mg/g for Fe3O4.[43] It has been well known that the metal adsorption on ferrite is based on the reaction between metal ions and accessible adsorptive sites like ≡S–OH groups, as shown in reaction . The higher the porosity and smaller the particle size, more active sites are accessible, and therefore it is understandable that the adsorption capacity on ferrite can be improved by increasing the surface area, porosity, and particle size. This elucidated the improvement in Pb(II) adsorption followed by Mg2+ substitution on copper ferrite.

Adsorption Kinetics

The adsorption kinetics are very important properties that help predict the adsorption performance of an adsorbent, and an engineer needs those parameters to design an adsorption process. Here, the pseudo-first and pseudo-second order kinetic models and the intra-particle diffusion model were used to investigate the adsorption of Pb(II) ions onto Cu0.5Mg0.5Fe2O4. The linear equations of these models are given as follows (eqs –9)where k1 (L/min) is the rate constant of pseudo-first order adsorption; qe and q (mg/g) are the amount of Pb2+adsorbed on the adsorbent at equilibrium and at time t, respectively; k2 (mg/g·min) is the rate constant of pseudo-second order adsorption; kp is the intraparticle diffusion rate constant (mg/g·min); and c is the intercept of the intraparticle diffusion model.

The fitness between the experimental and calculated data was evaluated by the normalized standard deviation (Δq), which can be calculated using eq .where the subscripts “exp” and “cal” are experimental and calculated data from the kinetic model, respectively, and N is the number of data points.

The linearized plots of the investigated kinetic models are shown in Figure and the kinetic parameters are exhibited in Table . The results revealed that only the pseudo-second-order kinetic model described well the Pb(II) adsorption onto Cu0.5Mg0.5Fe2O4 with R2 = 0.9999, indicating typical chemical adsorption. The intra-particle diffusion model did not fit the experimental data, suggesting that the adsorption was not significantly influenced by the intra-particle diffusion. Because the pseudo-second-order kinetic model is the most fitting model, it was utilized to simulate the adsorption of Pb(II) on Cu0.5Mg0.5Fe2O4 and compared with experimental data. The experimental and calculated data for Pb2+ adsorption as a function of time are shown in Figure d. The results confirmed that the pseudo-second-order kinetic model can exactly predict the trend of Pb(II) adsorption on Cu0.5Mg0.5Fe2O4 with the normalized standard deviation (Δq) of 1.24%.

Figure 6

Linearized plot of log(qe – q) vs time (t) in the pseudo-first-order kinetic model (a) and t/q vs time (t) in the pseudo-second-order kinetics model (b), q vs t1/2 in the intra-particle diffusion model (c), and the experimental and calculated data from the pseudo-second-order kinetic model for Pb2+ adsorption onto Cu0.5Mg0.5Fe2O4 as a function of time (d).

Table 3

Constants and Correlation Coefficients for the Kinetic Models

models	parameters	unit	values
pseudo-first-order kinetic model	qe	mg/g	3.66
	k1	L/min	0.01598
	R2		0.91
pseudo-second-order kinetic model	qe	mg/g	38.2
	k2	mg/g·min	0.01308
	R2		0.9999
	Δq	%	1.2
intra-particle diffusion model	C		34.58
	k	mg/g·min	0.215
	R2		0.77

The recyclability of Cu0.5Mg0.5Fe2O4 ternary oxides toward Pb(II) adsorption was investigated using a bath experiment. The Cu0.5Mg0.5Fe2O4 ternary oxides were removed from solution using an external magnet and regenerated before the next test cycle by washing the adsorbent with 20 mL of 2 M NaOH solution. The reusability result of Cu0.5Mg0.5Fe2O4 as an adsorbent toward Pb(II) with five cycles of adsorption-regeneration process is shown in Figure S5. It can be clearly seen that only less than 10% in the decrease of removal efficiency was observed after 5 cycles, which indicates that the Cu0.5Mg0.5Fe2O4 ternary oxide is durable for Pb(II) removal.

The leaching of copper from Cu0.5Mg0.5Fe2O4 ternary oxide in an aqueous solution at different pH values was also recorded, as shown in Figure S6. The release of copper is relatively negligible, which indicates that the Cu0.5Mg0.5Fe2O4 ternary oxides are a stable and effective adsorbent for the removal of lead ions. Cu0.5Mg0.5Fe2O4 showed a good selectivity against common contaminated ions in water including Ca2+, Mg2+, K2+, and Na2+ ions (Figure S7). The results revealed that over 97% of Pb(II) was removed, while the removal of others was negligible at the initial concentration of 5 ppm.

Conclusions

In short, a binary copper–magnesium ferrite (Cu0.5Mg0.5Fe2O4) was successfully prepared by a co-precipitation method. Single-phase spinel ferrite was obtained by the calcination of as-synthesized samples at a temperature of 900 °C. The obtained Cu0.5Mg0.5Fe2O4 binary ferrite has an average particle size of 30 nm with a surface area of 41.3 m2/g, a total pore volume of 0.2 cm3/g, and an average pore size of 15.1 nm. Mesoporous Cu0.5Mg0.5Fe2O4 material exhibits an adsorption capacity of 57.7 mg/g toward Pb(II), which is much higher than that of CuFe2O4 reported in the literature. It is apparent that partially replacing Cu2+ by Mg2+ in the ferrite structure helps improve the Pb(II) adsorption capacity. CuFe2O4 has been extensively applied as an adsorbent for the removal of various contaminants including organic compounds and heavy metals because of its simplicity and low-cost fabrication; however, its Pb(II) adsorption is relatively poor; thus, the success of this study could provide an additional benefit for copper ferrite application with minimum modification on the ferrite synthesis process.

Experimental Section

Mg–Cu Ferrite Preparation

Copper magnesium ferrite Cu0.5Mg0.5Fe2O4 was synthesized by a co-precipitation method. FeCl3·6H2O, CuCl2·2H2O, and MgCl2·6H2O were dissolved in a 100 mL glass beaker containing 50 mL distilled water with the Fe3+/Cu2+/Mg2+ molar ratio of 2:0.5:0.5. The solution was stirred and heated to 80 °C with a magnetic stirrer for 1 h before a NaOH 5 M solution was added until a pH of 9–10. The solution was cooled down to the room temperature, and then copper magnesium ferrite was filtered and washed several times with distilled water. The solid was dried at 150 °C for 8 h and calcined at 500–900 °C for 3 h to obtain copper magnesium ferrites. To investigate the effect of Mg substitution degree on Pb(II) adsorption, the Mg2+/Cu2+ molar ratios were varied from 0 to 100% to generate ferrites with various Mg2+ concentrations with a formula of Cu1–MgFe2O4, where x is the molar fraction of Mg2+.

Sorbent Characterization

XRD was analyzed using an X’Pert PRO PANalytical with a 0.15405 nm Cu Kα radiation source. The surface area and porosity were measured by the nitrogen adsorption–desorption method using TriStar II Plus. Samples were degassed at 110 °C for 4 h and then nitrogen adsorption–desorption was carried out at −196 °C. Prior to measurement, samples were degassed for 4 h at 250 °C to remove the adsorbed components. The morphology of the adsorbent was studied by SEM using a Hitachi S-4600 equipped with an energy dispersive spectrometer for elemental analysis. TGA was performed from room temperature to 800 °C in an air atmosphere at a heating rate of 5 °C/min using a Thermogravimetric Analyzer (Netzsch STA 449 F3). FTIR measurements were conducted on a TENSOR II, Bruker. Magnetic properties were determined using a vibrating sample magnetometer.

Pb Adsorption Experiments

Pb(II) adsorption of the Cu0.5Mg0.5Fe2O4 adsorbent was investigated by a batch method at room temperature and pH 7. In a typical experiment, Cu0.5Mg0.5Fe2O4 (0.1 g) was added to 100 mL conical flasks containing 50 mL of Pb(II) at different concentrations from 10 to 200 ppm following by shaking at a speed of 100 rpm. After designated periods, supernatant samples were extracted and analyzed for Pb(II) concentration using an atomic absorption spectrometer (ContrAA 700, Analytik Jena). The adsorption capacity (qe) was determined by the following equation (eq )where C0 is the initial concentration of Pb(II) (mg/L), Ce is the final concentration of Pb2+ (mg/L), V is the volume of Pb(II) solution, and m is the mass of Cu0.5Mg0.5Fe2O4.