Synthesis of manganese dioxide/iron oxide/graphene oxide magnetic nanocomposites for hexavalent chromium removal

Abstract

Effective removal of heavy metal ions from wastewater is one of the important environment issues. Here, manganese dioxide/iron oxide/graphene-based magnetic nanocomposites (MnO₂/Fe₃O₄/graphene oxide and MnO₂/Fe₃O₄/graphene) were prepared by an easy and effective two-step reaction and investigated for the removal of hexavalent chromium ion from water. The materials were characterized by transmission electron microscopy, Fourier-transform infrared spectroscopy, X-ray diffraction, vibrating sample magnetometry, and Brunauer–Emmett–Teller surface area measurement. MnO₂/Fe₃O₄/graphene oxide (MnO₂/Fe₃O₄/GO) synthesized at pH 5.0 was the most effective adsorption material among other samples synthesized under different conditions. The Langmuir isotherm model was consistent with the experimental data at different pH values. MnO₂/Fe₃O₄/GO showed excellent adsorption capacities both at pH 5.0 (q_max = 175.4 mg g⁻¹) and pH 2.0 (q_max = 193.1 mg g⁻¹). Hexavalent chromium adsorption by MnO₂/Fe₃O₄/GO was pH dependent. The adsorption kinetic data were best described by a pseudo-second-order model. MnO₂/Fe₃O₄/GO was stable and easily recovered. These experimental results suggested that MnO₂/Fe₃O₄/GO had great potential as an economic and efficient adsorbent of heavy metals from wastewater.

---

1. Introduction

The continuously increasing heavy-metal pollution is gaining considerable attention. Among such metals, chromium is highly toxic to the environment.^1,2 Chromium exists in the environment mainly in two states: trivalent chromium (Cr³⁺) and hexavalent chromium (Cr⁶⁺). Cr³⁺ is an essential element in humans and is much less toxic, whereas Cr⁶⁺ is one of the extremely toxic heavy metals found in various industrial waters that can cause health problems, such as liver damage, pulmonary congestion, vomiting, and severe diarrhea.³ Therefore, the amount of chromium in effluents must be reduced to an acceptable level before releasing them into streams and rivers to protect human health and the environment. There are numerous methods in removing Cr⁶⁺, such as ion exchange, cyanide treatment, reverse osmosis, electrochemical precipitation, and adsorption.⁴ Adsorption is one of the most promising technologies because it is cost-effective, has little by product, and is simple to operate.^5,6 However, conventional adsorbents, such as biosorbents, activated carbon, and carbon nanotubes, often cause potential secondary pollution under acidic conditions, are not easily separated and are not sufficiently effective. On the other hand, most of the materials used for enhanced hexavalent chromium removal only exhibit high removal efficiency whenever pH is very low.^3,7,8 In the application of practical chemical engineering, a strong acidic system is dangerous to equipment, human, and environment. Thus, developing new materials with simple synthesis, easy separation, good stability under acidic conditions, and high adsorption capacity whenever pH is low or high is highly recommended.⁹

As an emerging carbon material with a unique two-dimensional conjugated chemical structure, graphene has attracted a great deal of attention in recent years owing to its better conductivity, more superior chemical stability, and much higher specific surface area^10–12 compared with carbon nanotubes, as well as its potential applications in dye sensitized solar cells,¹³ supercapacitors,¹⁴ and catalysts.¹⁵ However, the study of graphene-based adsorbent for wastewater treatment is still an emerging field. Unfortunately, graphene is hydrophobic and usually suffers from irreversible agglomeration in water because of the strong van der Waals interactions between neighboring sheets, which significantly reduces the surface area and is consequently not beneficial to the adsorption of contaminants. To prevent the restacking of graphene nanosheets, numerous studies focused on the intercalation of inorganic nanoparticles.¹⁰ Graphene oxide (GO) which usually was derived from exfoliation of graphite oxide has abundant functional groups on the surface, such as epoxy, hydroxyl, and carboxyl groups, making it useful for environmental applications. Therefore, features like large surface area and presence of surface functional groups make GO and their composites an attractive adsorbent candidate for water purification.

Recently, our group prepared a manganese dioxide/iron oxide/acid oxidized multi-walled carbon nanotube (MnO₂/Fe₃O₄/o-MWCNTs) magnetic nanocomposite. It exhibited a high Cr⁶⁺ removal efficiency whenever pH was lower than 6,¹⁶ and the idea of MnO₂ and Fe₃O₄ coated with GO may be utilized in synthesizing new nanocomposite materials which would have better performances in metal ions adsorption applications. The reusability and performance of a material like adsorption capacity, under both acidic and near neutral conditions, can be further enhanced by using a more advanced synthesis method.

In this study, MnO₂/Fe₃O₄/graphene-based magnetic nanocomposites were prepared using a very simple two-step reaction with one main reactant every step. The combination of the advantages of MnO₂ and Fe₃O₄ for the adsorption and magnetic properties of magnetic nanocomposites was developed. Magnetic nanocomposites were characterized using transmission electron microscopy (TEM), Fourier-transform infrared spectroscopy (FTIR) analysis, X-ray diffraction (XRD), Brunauer, Emmett, and Teller surface area measurement (BET) analyses, and vibrating sample magnetometry (VSM). The adsorption capacities of GO, Fe₃O₄/graphene oxide (Fe₃O₄/GO), Fe₃O₄/reduced graphene oxide (Fe₃O₄/RGO), MnO₂/Fe₃O₄/graphene oxide (MnO₂/Fe₃O₄/GO), and MnO₂/Fe₃O₄/reduced graphene oxide (MnO₂/Fe₃O₄/RGO), which were synthesized under different conditions, were compared to determine the most effective adsorbent material. To evaluate the abovementioned adsorption behaviors further, the adsorption isotherms, solution pH, kinetics, and reusability of the material were examined. The effects of contact time at different temperatures and pH values were also studied to investigate the adsorption capacities of adsorbents at the pH ranging from 2.0 to 5.0.

2. Materials and methods

2.1. Synthesis of graphite oxide (GO)

GO was prepared by modified Hummers method from natural flake graphite.¹⁷ Using concentrated H₂SO₄, H₃PO₄, and KMnO₄ for oxidation, and at the aid of ultrasonication, the graphite layers were exfoliated from each other. Then 30% H₂O₂ was added to the suspension to eliminate the excess KMnO₄. The desired products were rinsed with deionized water.

2.2. Synthesis of graphene-based magnetic nanocomposites

Graphene-based magnetic nanocomposites (Fe₃O₄/reduced graphene oxide and Fe₃O₄/graphene oxide) were prepared with the use of a easy and an effective one-pot solvothermal method.¹⁸ Dispersed in 45 mL of ethylene glycol (EG) were 50 mg of graphite oxide and 500 mg of iron acetylacetonate (Fe(acac)₃) that underwent ultrasonication for 30 min, and 30 μL of hydrazine hydrate was slowly added to the mixture while being stirred. Subsequently, the mixture was placed in an autoclave and heated at 180 °C for 16 h. The as-synthesized product was isolated by centrifugation, then washed three times with water and ethanol, and finally dried in a vacuum oven at 60 °C. These materials were denoted as “Fe₃O₄/RGO”. The synthesis of Fe₃O₄/graphene oxide nanocomposites (Fe₃O₄/GO) was carried out in the absence of hydrazine hydrate under the same conditions as the Fe₃O₄/GO.

2.3. Synthesis of manganese dioxide/iron oxide/graphene-based magnetic nanocomposites

MnO₂ was coated with Fe₃O₄/RGO or Fe₃O₄/GO by simple immersion into a KMnO₄ aqueous solution, which is similar with MnO₂ coated on carbon nanotubes.^19,20 First, 150 mL of 0.1 M KMnO₄ solution was heated at 80 °C using a circulator. Then, 0.1 g of Fe₃O₄/RGO or Fe₃O₄/GO was added to the solution with different pH values. The pH of the solution was controlled by HCl. During the synthesis, the temperature of the solution was maintained at 80 °C for 3 h using a circulator. The suspension was filtered and washed several times using deionized water and ethanol absolute, and then dried at 60 °C for 12 h. These materials were denoted as “MnO₂/Fe₃O₄/RGO” or “MnO₂/Fe₃O₄/GO”.

2.4. Characterization of different materials

The morphology of Fe₃O₄/GO and MnO₂/Fe₃O₄/GO was examined under transmission electron microscopy (TEM, TecnaiG²F30). The inorganic phases of the graphite, GO, Fe₃O₄/GO, and MnO₂/Fe₃O₄/GO, were examined using an X-ray diffraction (XRD; XRD-6000, Shimadzu, Japan). The functional groups of GO, Fe₃O₄/GO, and MnO₂/Fe₃O₄/GO were examined using a NEXUS 670 FTIR spectrometer (Nicolet Instrument Corporation, USA). Magnetic properties were determined by a vibrating sample magnetometry (VSM, LAKESHORE-7304) at room temperature. The BET surface areas were determined using nitrogen adsorption at liquid nitrogen temperatures under the relative pressures P/P₀ ranging between 0 and 1 (Sorptomatic 1990, Thermo, USA). The materials were dried fully in a vacuum oven at 60 °C for 12 h to make sure that water contained in the materials was completely removed before measuring the nitrogen adsorption isotherms at 77 K.

2.5. Adsorption experiments

Analytical grade potassium dichromate was used to prepare a 1000 mg L⁻¹ stock solution, which was further diluted to the required concentration before use. To determine the adsorption capacities of the different materials, GO, Fe₃O₄/RGO, Fe₃O₄/GO, and MnO₂/Fe₃O₄/GO were synthesized at an initial solution pH of 5 (denoted as “MnO₂/Fe₃O₄/GO (pH 5)”), and “MnO₂/Fe₃O₄/RGO” was synthesized at different initial solution pH values (pH 7, 5, and 1) (denoted as “MnO₂/Fe₃O₄/RGO (pH 7)”, “MnO₂/Fe₃O₄/RGO (pH 5)”, and “MnO₂/Fe₃O₄/RGO (pH 1)”, respectively) as adsorbents for Cr⁶⁺ removal. The different adsorbents were added to a 100 mL solution with initial Cr⁶⁺ concentrations from 50 mg L⁻¹ to 300 mg L⁻¹ (50 mg L⁻¹ intervals) at a pH of about 5.0, and stirred for 120 min at a temperature of 298 K. The two phases were separated by filtration using a 0.45 μm microporous membrane filter. The concentration of Cr⁶⁺ was analyzed by spectrophotometric method at a wavelength of 540 nm, using 1,5-diphenylcarbohydrazide as chromogenic reagent and H₂SO₄ and H₃PO₄ as buffering agent.¹⁶ The most effective material with the maximum adsorption capacity was chosen as the adsorbent in the following experiments and was kept constant. The pH values of 2, 4, and 5 with initial concentrations from 50 mg L⁻¹ to 300 mg L⁻¹ were chosen to evaluate the effect of pH on the Cr⁶⁺ adsorption process. Adsorption kinetic data were generated by adding the adsorbent to a 2000 mL solution (pH 2 and 5) with a Cr⁶⁺ concentration of 200 mg L⁻¹ at 275 K, 295 K, and 315 K. At predetermined time intervals, samples were collected using a 0.45 μm membrane filter. For the regeneration, after adsorption, desorption was carried out by washing the adsorbents with distilled water several times, and then the Cr⁶⁺-adsorbed adsorbent were immersed in 100 mL of 0.1 M NaOH for 120 min. Before the second adsorption, the adsorbent was treated with 0.1 M HCl solution for 120 min. The solid and liquid phases were separated by a magnet. This adsorption–desorption cycle was repeated five times using the same affinity adsorbents.

The equilibrium metal adsorption capacity was calculated for each Cr⁶⁺ sample using the following expression:

where q_e is the adsorbent adsorption capacity (mg g⁻¹), V is the sample volume (L), C₀ is the Cr⁶⁺ concentration (mg L⁻¹), C_e is the equilibrium Cr⁶⁺ concentration (mg L⁻¹), and m is the weight of the adsorbents (g).

3. Results and discussion

3.1. Characterization of different materials

The TEM images of Fe₃O₄/GO and MnO₂/Fe₃O₄/GO nanocomposite samples are shown in Fig. 1. Fig. 1(a) and (b) show the images of the Fe₃O₄/GO nanoparticles under different magnifications. It can be seen that the 150 nm Fe₃O₄ nanoparticles were well distributed on graphene oxide sheets, which were nearly flat and had a large area of up to several square micrometers. Some nanoparticles were slightly aggregated due to the loading degree close to saturation. Fig. 1(c) and (d) show the morphology of MnO₂/Fe₃O₄/GO under different magnifications. Fe₃O₄/GO was fully coated with a lot of very small, uniform, and continuous MnO₂, which was like a layer of MnO₂, had almost no gaps. These characteristics may be beneficial to the adsorption process.

Fig. 1 Representative TEM images of Fe₃O₄/GO and MnO₂/Fe₃O₄/GO under different magnifications: (a and b) Fe₃O₄/GO, and (c and d) MnO₂/Fe₃O₄/GO.

The XRD of the graphite, GO, Fe₃O₄/GO, and MnO₂/Fe₃O₄/GO, are presented in Fig. 2. The diffraction peak of natural graphite occurred near the 2θ = 26.34° location. It can be seen from the pattern that the peak is sharp and high, which shows high crystallinity and degree of order. In the spectrogram of GO, the diffraction peak (002) of graphite crystal disappeared, indicating that the graphite had been completely oxidized. In its place, the diffraction peak (002) of GO occurred. The characteristic XRD peak of GO appeared at 2θ = 10.54°.²¹ The XRD of Fe₃O₄/GO exhibited a new peak at 2θ = 22.53°, which belonged to the Fe₃O₄ (111),¹⁸ and the diffraction peak at 2θ = 10.92° was assigned to the (002) plane of GO. For MnO₂/Fe₃O₄/GO, the diffraction intensity almost disappeared, implying the high oxidation extent of the carbon atoms in the framework by MnO₂. It should be noted that the XRD of MnO₂/Fe₃O₄/GO was difficult to identify because of the poor crystallinity of MnO₂. The same results were also found in ref. 22, which synthesized graphene nanoplate-MnO₂ composites by KMnO₄ solution.

Fig. 2 XRD patterns of graphite, GO, Fe₃O₄/GO, and MnO₂/Fe₃O₄/GO.

The FTIR spectra of GO, Fe₃O₄/GO, and MnO₂/Fe₃O₄/GO are shown in Fig. 3. Adsorption peaks appearing at 1622, 1400, 1250, and 1073 cm⁻¹ in the FTIR spectra of pure GO are attributed to aromatic C=C, carboxyl O=C–O, epoxy C–O, and alkoxy C–O stretching vibrations, respectively, whereas the peaks located at 1730 cm⁻¹ and 3429 cm⁻¹ correspond to the C=O stretching vibrations of the –COOH groups, and –OH stretching vibration, respectively.^9,23 All these bands related with the oxygen-containing functional groups almost remained in the FTIR spectra of Fe₃O₄/GO and MnO₂/Fe₃O₄/GO. The additional peak at 605 cm⁻¹ in Fig. 3 can be ascribed to the lattice adsorption characteristic of Fe₃O₄.¹⁸ For MnO₂/Fe₃O₄/GO, a broad band was observed in the composite at the low-frequency region, at around 530 cm⁻¹. The presence of this band is attributed to the Mn–O, Mn–O–Mn, and Fe–O–Fe vibrations.^16,24

Fig. 3 FTIR transmission spectra of GO, Fe₃O₄/GO, and MnO₂/Fe₃O₄/GO.

The magnetic properties of the Fe₃O₄/GO and MnO₂/Fe₃O₄/GO nanocomposites are depicted in Fig. 4. Their saturation magnetization is 26.04 emu g⁻¹ and 11.86 emu g⁻¹, respectively. The saturation magnetization value of MnO₂/Fe₃O₄/GO is 11.86 emu g⁻¹, which is lower than that of the Fe₃O₄/GO. The decrease in the saturation magnetization can be mainly attributed to the existence of MnO₂. However, its maximum saturation magnetization (11.86 emu g⁻¹) is enough to maintain their magnetic recovery performance. Fig. 5 shows that MnO₂/Fe₃O₄/GO is attracted quickly toward an external magnetic field.

Fig. 4 Magnetic hysteresis curves of Fe₃O₄/GO and MnO₂/Fe₃O₄/GO.

Fig. 5 The photograph demonstrates that MnO₂/Fe₃O₄/GO can be dispersed in ethanol (A) and separated from ethanol using a magnet (B).

The surface area, pore volume and pore size of MnO₂/Fe₃O₄/GO (60.1 m² g⁻¹, 0.18 cm³ g⁻¹ and 12.2 nm) are evidently much lower than those of Fe₃O₄/GO (119.5 m² g⁻¹, 0.41 cm³ g⁻¹ and 13.2 nm). Compared with the unsupported single Fe₃O₄ and MnO₂ particles, the larger surface areas of MnO₂/Fe₃O₄/GO nanocomposites are due to the large surface area of GO.^25,26 The formation of MnO₂ blocks the pore entrances, thereby decreasing the surface area. Nevertheless, MnO₂ introduced on the surface of Fe₃O₄/GO provides numerous sorption sites, thereby increasing the sorption capacities of MnO₂/Fe₃O₄/GO. It is well known that the adsorption properties not only depend on the high specific surface area of the adsorbent, but also depend on the pore properties of the adsorbent. Generally speaking, the higher surface area and pore size are, the higher the adsorption capacity is. However, the adsorption capacities of different materials mainly depend on the amount and kinds of functional groups on the adsorbent surface. In this work, MnO₂/Fe₃O₄/GO is the most effective removal materials compared with the others materials because MnO₂/Fe₃O₄/GO provides more binding sites. Therefore, the adsorption capacities of different materials not only depend on the surface area and pore properties, but also depend on the functional groups on the adsorbent surface.

3.2. Adsorption capacity of different materials

The preparation of MnO₂/Fe₃O₄/RGO or MnO₂/Fe₃O₄/GO nanocomposites was carried out by a very simple two-step reaction, with one main reactant every step (Fe(acac)₃, step one; KMnO₄, step two). This simple method provides a good technique in the preparation of MnO₂/Fe₃O₄/RGO or MnO₂/Fe₃O₄/GO. It should be pointed out that the initial solution pH is evidently important during the synthesis process. In this study, the weight of MnO₂/Fe₃O₄/RGO or MnO₂/Fe₃O₄/GO increased as the initial synthesis solution pH decreased, and the color of MnO₂/Fe₃O₄/RGO or MnO₂/Fe₃O₄/GO became deeper as the initial synthesis solution pH decreased (Fig. S2†). The color of the solution after the reaction became shallower as the initial synthesis solution pH decreased, as shown in Fig. S3.† It means that the loading level of MnO₂ on Fe₃O₄/RGO or Fe₃O₄/GO was affected by the initial synthesis solution pH and this result is identical with the previously presented result, wherein MnO₂ coated on carbon nanotubes was also affected by the initial synthesis solution pH.²⁷

MnO₂ was coated on Fe₃O₄/GO or Fe₃O₄/RGO through a direct redox reaction between the GO or RGO and MnO₄⁻:¹⁹

MnO₄⁻ + 4H⁺ + 3e⁻ → MnO₂ + 2H₂O (2)

Another self-limiting reaction of aqueous permanganate with graphene that enabled deposition of birnessite-type MnO₂ in between the graphene sheets must have occurred. The reaction between carbon and KMnO₄ can be described as follows:²⁸

4MnO₄⁻ + 3C + H₂O ⇔ 4MnO₂ + CO₃²⁻ + 2HCO₃⁻ (3)

The organic functional groups on GO can improve the adsorption capacity of the adsorbent. In this study, the organic functional groups on GO also influenced the synthesis process; the Mn⁷⁺ ions in the solution were adsorbed onto the surface of GO by carboxyl groups through electrostatic attraction, and these ions are in situ deoxidized to MnO₂. The reaction can be written as follows:²⁹

4MnO₄⁻ + 4H⁺ → 4MnO₂ + 2H₂O + 3O₂↑ (4)

The adsorption capacities of different materials were investigated, and the results are presented in Fig. 6. The adsorption capacity of Fe₃O₄/GO or Fe₃O₄/RGO and MnO₂/Fe₃O₄/RGO or MnO₂/Fe₃O₄/GO were much higher than that of GO because the MnO₂ and Fe₃O₄ were good adsorbents for Cr⁶⁺.^16,30 The adsorption capacity of Fe₃O₄/GO was higher than Fe₃O₄/RGO, and MnO₂/Fe₃O₄/GO (pH 5) was higher than MnO₂/Fe₃O₄/RGO (pH 5), suggesting that organic functional groups on GO, such as hydroxyl, carboxyl, and carbonyl groups, were expected to be the active centers for Cr⁶⁺.⁹ The low loading level of MnO₂ (at an initial synthesis solution pH 7.0) led to the worsening of the adsorption capacities of MnO₂/Fe₃O₄/RGO compared with those at pH 5.0. However, the excess MnO₂ load hindered the adsorption of organic functional groups, GO and Fe₃O₄, by effectively decreasing the volume; this is the reason why the adsorption capacity of MnO₂/Fe₃O₄/RGO (pH 5) was greater than MnO₂/Fe₃O₄/RGO (pH 1). Therefore, the adsorption capacities of different materials mainly depend on the functional groups and loading level of MnO₂ on the adsorbent surface. In this study, the most effective removal of Cr⁶⁺ occurred with MnO₂/Fe₃O₄/GO at an initial synthesis solution of pH 5.0.

Fig. 6 Adsorption capacities of different materials toward Cr⁶⁺.

In this work, a very important adsorption group is MnO₂ because the adsorption capacity of MnO₂/Fe₃O₄/GO is much better than that of Fe₃O₄/GO. The surface groups of MnO₂ are amphoteric and can function as an acid or a base. The oxide surface can undergo protonation and deprotonation in response to changes in solution pH. When the solution pH is low, MnO₂ would be protonated to form ≡MnOH. Given that Cr⁶⁺ mainly exists in the form of negatively charged HCrO₄⁻ in the acidic solution, it could be adsorbed by positively charged MnO₂/Fe₃O₄/GO through electrostatic attraction. Meanwhile, the positively charged MnO₂ possibly involves an exchange reaction of Cr³⁺ (Cr⁶⁺) with ≡MnOH, and the lone pair electrons of manganese oxide may induce surface complexation reactions with Cr³⁺ (Cr⁶⁺).

3.3. Adsorption isotherms

Batch equilibrium adsorption experiments were used for adsorption assessment through plots of adsorption isotherms. The maximum adsorption capacity can be obtained from adsorption isotherms. The Langmuir and Freundlich isotherms are the most common isotherms used to describe solid–liquid adsorption.³¹

Langmuir adsorption isotherms have been successfully applied to many other real adsorption processes. A basic assumption of Langmuir theory is that adsorption takes place at specific homogeneous sites within the adsorbent. It is assumed that once a molecule occupies a site, no further adsorption can occur at that site. Theoretically, a saturation value is reached and no further sorption can occur. A linear form of this expression is:³²

where q_e (mg g⁻¹) is the amount of solution adsorbed per unit mass of the adsorbent, C_e (mg L⁻¹) is the solute equilibrium concentration, q_m (mg g⁻¹) is the maximum adsorbate amount that forms a complete monolayer on the surface, and b (L mg⁻¹) is the Langmuir constant related to adsorption heat. When C_e/q_e is plotted against C_e and the data are regressed linearly, the q_m and b constants can be calculated from the slope and the intercept (Fig. S4†).

The efficiency of the adsorption can be expressed by the dimensionless equilibrium parameter R_L, which is defined as follows:^33,34

where b (L mg⁻¹) is the Langmuir constant and C₀ is the initial Cr⁶⁺ concentration (mg L⁻¹).

The values of R_L indicate the isotherm shapes, which can be unfavorable (R_L ≥ 1) or favorable (0 ≤ R_L ≤ 1).³⁵

The Freundlich isotherm assumes that adsorption occurs on heterogeneous surfaces by multilayer adsorption, and the amount of adsorbate adsorbed increases infinitely with an increasing concentration, which is expressed as follows:³⁶

where q_e and C_e have the same definitions as those in the Langmuir equation cited above, and K_F and n are the Freundlich constants related to the adsorption capacity and adsorption intensity, respectively. When ln q_e is plotted against ln C_e and the data are regressed linearly, the 1/n and K_F constants are determined from the slope and the intercept (Fig. S6†).

The results of adsorption isotherms are listed in Table 1. The R_L values have been found to be in the 0 to 1 range (Fig. S5†), which indicates the suitability of MnO₂/Fe₃O₄/GO as Cr⁶⁺ adsorbents. The adsorption data could be described well by the Langmuir model rather than Freundlich model, suggesting that the energies of the active sites on the surface of MnO₂/Fe₃O₄/GO are similar, where no interaction has been observed between the adsorbed molecules and the surface of MnO₂/Fe₃O₄/GO. It can be seen from the values of q_m (mg g⁻¹) in Table 1, that the maximum Cr⁶⁺ adsorption capacity increased from 175.4 mg g⁻¹ to 193.1 mg g⁻¹ as the pH decreased from 5.0 to 2.0. This phenomenon will be further discussed in Section 3.4. The q_m value of Cr⁶⁺ adsorption onto MnO₂/Fe₃O₄/GO in pH 5.0 is 175.4 mg g⁻¹, which is close to the q_m value of Cr⁶⁺ adsorption onto MnO₂/Fe₃O₄/GO at pH 2.0 (193.1 mg g⁻¹), and this maximum adsorption capacity is higher than many adsorbents. The most important result is that MnO₂/Fe₃O₄/GO has an excellent adsorption capacity at the pH ranging from 2.0 to 5.0, which is significant in the application of practical chemical engineering.

Table 1 Parameters of the Langmuir and Freundlich isotherms of Cr⁶⁺ adsorption onto MnO₂/Fe₃O₄/GO

pH	Langmuir isotherm				Freundlich isotherm
	q m (mg g^−1)	b	r 1 ^2	R L	1/n	K F	r 2 ^2
5.0	175.4	0.0102	0.993	0.247–0.663	0.499	8.398	0.930
4.0	177.1	0.0128	0.988	0.206–0.609	0.384	16.401	0.913
2.0	193.1	0.0127	0.989	0.208–0.612	0.394	16.817	0.908

---

3.4. Effect of solution pH on the adsorption

The solution pH has been identified as the most important variable governing metal adsorption onto the adsorbent. To investigate the adsorption capacity of MnO₂/Fe₃O₄/GO under both acidic and near neutral conditions, pH values of 2.0, 4.0, and 5.0 at initial concentrations from 50 mg L⁻¹ to 300 mg L⁻¹ were chosen to evaluate the effect of pH on the Cr⁶⁺ adsorption process.

Fig. 7 shows that the Cr⁶⁺ adsorption capacity increases as the pH increased from 2.0 to 5.0. At a lower pH, the adsorption effect is high because predominant Cr⁶⁺ species mainly exists in monovalent HCrO₄⁻ form, which is then gradually converted to divalent CrO₄²⁻ and Cr₂O₇²⁻ as pH increases. The adsorption free energy of HCrO₄⁻ is lower than that of CrO₄²⁻ and Cr₂O₇²⁻; and consequently, HCrO₄⁻ is more favorably adsorbed than CrO₄²⁻ and Cr₂O₇²⁻ at the same concentration.³⁷ As the pH increases, the MnO₂/Fe₃O₄/GO surface becomes increasingly deprotonated so that the amount of positive surface charges is significantly decreased, leading to a reduction in the adsorption capacity of Cr⁶⁺. Thus, the adsorption quantities of Cr⁶⁺ at a lower pH are larger than that of at higher pH.

Fig. 7 Effect of solution pH on chromium ions adsorbed onto MnO₂/Fe₃O₄/GO.

However, the Cr⁶⁺ adsorption capacity of MnO₂/Fe₃O₄/GO at pH 5.0 is close to the Cr⁶⁺ adsorption capacity of MnO₂/Fe₃O₄/GO at pH 2.0 and pH 4.0. Compared with the Cr⁶⁺ adsorption capacity of MnO₂/Fe₃O₄/GO at pH 2.0, the Cr⁶⁺ adsorption capacity of MnO₂/Fe₃O₄/GO at pH 5.0 still possessed more than 85% of the adsorption capacities. It should be pointed out that the pH values of the Cr⁶⁺ solution with the initial concentrations from 50 mg L⁻¹ to 300 mg L⁻¹ were about 5.0 ± 0.2 without any added acid or alkali. It means that MnO₂/Fe₃O₄/GO has a high adsorption capacity at the pH ranging from 2.0 to 5.0. It is known that Cr⁶⁺ is reduced to Cr³⁺ in the acidic medium, and the reaction is shown as follows:³⁷

HCrO₄⁻ + 7H⁺ + 3e⁻ ↔ Cr³⁺ + 4H₂O (8)

Cr₂O₇²⁻ + 6e⁻ + 3H⁺ ↔ 2Cr³⁺ + 7H₂O (9)

The reason why many adsorbents have high adsorption capacities for Cr⁶⁺ is that these reactions^3,8 often occur at a low pH, and rarely occur when pH is higher than 3.0, leading to a decreased adsorption capacity of adsorbents. We measured the total Cr concentration in a solution and found that Cr³⁺ exists at pH 2.0, and the amounts of total Cr and Cr⁶⁺ concentration at pH 5.0 are more or less the same, which may indicate that the reduction of Cr⁶⁺ to Cr³⁺ ions occur rarely at a higher pH. This finding explains why the MnO₂/Fe₃O₄/GO has a higher adsorption capacity in the acidic medium. These results are similar with the previous studies.^37,38

3.5. Adsorption kinetics

Adsorption kinetics, demonstrating the solute uptake rate, is one of the most important factors which represent the adsorption efficiency of the MnO₂/Fe₃O₄/GO and therefore, determines their potential applications. The effect of the contact time of MnO₂/Fe₃O₄/GO on the adsorption capacity of Cr⁶⁺ at different temperatures with pH 5.0 and 2.0 is shown in Fig. 8. Increasing the contact time obviously increases the Cr⁶⁺ adsorption. At pH 5.0, the equilibrium capacity is 108.4 mg g⁻¹ at 275 K, 111.5 mg g⁻¹ at 295 K, and 122.4 mg g⁻¹ at 315 K, and at pH 2.0, the equilibrium capacity is 131.2 mg g⁻¹ at 275 K, 135.7 mg g⁻¹ at 295 K, and 144.8 mg g⁻¹ at 315 K. As shown in Fig. 8, the results indicate that the equilibrium time exhibited differences as the temperature increased from 275 K to 315 K, and the pH decreased from 5.0 to 2.0. At pH 5.0, the adsorption equilibrium in 275 K was observed at 120 min, however, the adsorption in 295 K and 315 K almost reached the equilibrium at 90 min. At pH 2.0, the equilibrium of the adsorption process in 275 K, 295 K, and 315 K were all observed at 90 min. The contact was set to 120 min in each experiment to ensure that each adsorption process reached equilibrium. The adsorption process was faster at a lower pH, possibly because more HCrO₄⁻ exists in a solution at pH 2.0, and the adsorption free energy of HCrO₄⁻ is more favorably adsorbed than CrO₄²⁻ and Cr₂O₇²⁻ at the same concentration (Section 3.4), leading to the adsorption reaching equilibrium faster at pH 2.0. The adsorbent showed a good performance in adsorption during the 120 min. Quantifying the changes in adsorption with time requires an appropriate kinetic model. The pseudo-first-order equation, the second-order rate equation, and the pseudo-second-order rate equation, are expressed as follows:^39–41andwhere q_e and q_t (mg g⁻¹) are the adsorption capacities at equilibrium and at time t (min), respectively; k₁ (min⁻¹), k₂ (g mg⁻¹ min⁻¹), and k₃ (g mg⁻¹ min⁻¹) are the pseudo-first-order, second-order, and pseudo-second-order rate adsorption constants, respectively.

Fig. 8 Effect of contact time on chromium ions adsorbed onto MnO₂/Fe₃O₄/GO at different temperatures and pH values.

The straight line plots of log(q_e − q_t) versus t are 1/(q_e − q_t) versus t and t/q_tversus t for the three kinetic models. The pseudo-second-order rate constants were used to calculate the initial adsorption rate h (mg g⁻¹ min⁻¹) as follows:³⁹

h = k₃q_e² (13)

The kinetic parameters and correlation coefficients of Cr⁶⁺ at different temperatures (pH 5.0 and 2.0, respectively) are presented in Table 2. The q_e is the theoretical value, which is calculated using the kinetic models. From the value of r₃ (pseudo-second-order kinetic model), the experimental data fitted the pseudo-second-order kinetic model (≥0.981) better than the pseudo-first-order kinetic model and second-order rate kinetic model. The q_e calculated from the pseudo-second-order kinetic model increased correspondingly with the temperature and pH (pH 5.0: 143.4 mg g⁻¹ at 275 K, 144.1 mg g⁻¹ at 295 K, 157.7 mg g⁻¹ at 315 K; pH 2.0: 158.0 mg g⁻¹ at 275 K, 159.2 mg g⁻¹ at 295 K, 167.5 mg g⁻¹ at 315 K). The q_e calculated from the pseudo-first-order kinetic model and second-order rate kinetic model were much higher or much lower than the experimental q_e,exp values, and the adsorption capacities calculated from the pseudo-first-order kinetic model and second-order rate kinetic model did not always increase with increasing temperature and decreasing pH, which were different from the experimental data. These results indicate that the adsorption process could be optimized under the pseudo-second-order kinetic model. The initial adsorption rate increased correspondingly with the temperature and pH within a given adsorption system because of the higher temperature and lower pH, which were beneficial to the initial adsorption process. The adsorbent system can be well described by the pseudo-second-order kinetic model, suggesting that the adsorption may be the rate-limiting step involving valence forces through the sharing or exchange of electrons between the adsorbent and adsorbate.⁹

Table 2 Kinetic parameters of Cr⁶⁺ adsorption on MnO₂/Fe₃O₄/GO at various temperatures and pH values

pH	T (K)	q e,exp	Pseudo-first-order kinetic model			Second-order kinetic model
			k 1	q e	r 1	k 2 (×10^−3)	q e	r 2
5.0	275	108.4	0.0202	65.2	0.843	8.81	2.4	0.918
	295	111.5	0.0251	131.2	0.975	2.86	22.5	0.737
	315	122.4	0.0445	285.1	0.933	3.84	0.7	0.889
2.0	275	131.2	0.0353	182.3	0.989	6.45	4.6	0.860
	295	135.7	0.0244	97.1	0.949	3.67	30.5	0.783
	315	144.8	0.0194	98.7	0.901	9.32	2.4	0.847

---

pH	T (K)	Pseudo-second-order kinetic model
		k 3 (×10^−4)	q e	h	r 3
5.0	275	1.148	143.4	2.363	0.983
	295	1.357	144.1	2.777	0.982
	315	1.173	157.7	2.918	0.981
2.0	275	1.706	158.0	4.258	0.991
	295	2.119	159.2	5.373	0.986
	315	2.128	167.5	5.971	0.993

---

3.6. Regeneration of saturated adsorbents

For practical application, recycling and regeneration of the adsorbent is indispensable because the better repeated availability of advanced adsorbents may reduce the overall cost of the adsorbent. Given that the adsorption of Cr⁶⁺ ions on the MnO₂/Fe₃O₄/GO is pH-dependent, and that the lower pH is beneficial for the Cr⁶⁺ adsorption, the desorption of Cr⁶⁺ ions from the adsorbent can be achieved by increasing the system pH values. Therefore, for the reusability study, experiments were conducted at an alkaline condition.⁴² It can be seen that the MnO₂/Fe₃O₄/GO still possessed more than 84% of the adsorption capacities for Cr⁶⁺ after four cycles of reuse, which slightly decreased to 76% at the fifth cycle, reflecting the high adsorption and stability of MnO₂/Fe₃O₄/GO (Fig. 9).

Fig. 9 Reusability of MnO₂/Fe₃O₄/GO for chromium-ion adsorption.

3.7. Comparison of adsorbent performance with literature data

The q_max values of the MnO₂/Fe₃O₄/GO were compared with the Cr⁶⁺ adsorption capacities reported in the literature for other adsorption (Table 3). MnO₂/Fe₃O₄/GO has an excellent maximum adsorption capacity among all the absorbents, not only at pH 2.0 but also at pH 5.0. The chromium removal capacities of other adsorbents at pH 1.0 or 2.0, such as cetyltrimethylammonium bromide modified graphene, sawdust, activated carbon modified with nitric acid, activated carbon derived from rubber wood saw dust, Fe₃O₄/polypyrrole composite, graphene/MgAl-layered double hydroxides nanocomposite, and MnO₂/Fe₃O₄/carbon nanotubes were lower than that of MnO₂/Fe₃O₄/GO at pH 2.0,^{10,16,38,42–44} suggesting that MnO₂/Fe₃O₄/GO has an excellent maximum adsorption capacity in a strong acidic system. It should be pointed that the maximum adsorption capacity of MnO₂/Fe₃O₄/GO at pH 5.0 is still higher than the maximum adsorption capacity of most Cr⁶⁺ absorbents at pH 1.0, 2.0, and 5.0. It means that the MnO₂/Fe₃O₄/GO has an excellent adsorption capacity at the pH ranging from 2.0 to 5.0, and it is significant in the application of practical chemical engineering because a strong acidic system is dangerous for equipment, human and environment. Thus, the comparison suggests that MnO₂/Fe₃O₄/GO has a great potential as heavy metal ion adsorbent in wastewater treatment.

Table 3 Maximum adsorption capacities of Cr⁶⁺ on MnO₂/Fe₃O₄/GO and other adsorbents

Adsorbates	Adsorbents	q max (mg g^−1)	pH	Reference
Cr^6+	Cetyltrimethylammonium bromide modified graphene	21.6	2.0	40
	Sawdust	41.5	1.0	35
	Activated carbon modified with nitric acid	16.1	2.0	41
	Activated carbon derived from rubber wood saw dust	65.0	2.0
	Activated carbon derived from acrylonitrile–divinylbenzene	101.2	2.0
	Fe3O4/polypyrrole composite	91.6	2.0	39
	Graphene/MgAl-layered double hydroxides nanocomposite	172.6	2.0	8
	MnO2/Fe3O4/carbon nanotubes	186.9	2.0	14
	Polyacrylonitrile/FeCl2 composite	11.7	5.0	39
	Wood sawdust coated by polypyrrole	1.3	5.0
	MnO2/Fe3O4/GO	175.4	5.0	Present work
	MnO2/Fe3O4/GO	193.1	2.0	Present work

---

4. Conclusions

Manganese dioxide/iron oxide/graphene oxide magnetic nanocomposites (MnO₂/Fe₃O₄/GO) was successfully prepared by a straightforward method. The MnO₂/Fe₃O₄/GO, which was synthesized at pH 2.0, was the most effective material for removing Cr⁶⁺ ions compared with other materials. This adsorbent has a unique property that combined the high adsorption capacity of MnO₂, the high surface area of GO, and magnetic features of Fe₃O₄. As far as we know, this study is the first to use MnO₂/Fe₃O₄/GO as an adsorbent for chromium removal. The initial solution pH has an apparent effect on the adsorption capacity, which increased as the pH decreased. The kinetic data of the adsorption fitted well with the pseudo-second-order kinetic model. The maximum adsorption capacity calculated from the Langmuir isotherm model of the MnO₂/Fe₃O₄/GO was 193.1 mg g⁻¹ at pH 2.0 and 175.4 mg g⁻¹ at pH 5.0. This finding indicated that MnO₂/Fe₃O₄/GO had excellent maximum adsorption capacity under acidic and near neutral conditions. Desorption results showed that the adsorption capacity can remain up to 76% after five times of usage. Therefore, MnO₂/Fe₃O₄/GO nanocomposite was a promising adsorbent for hexavalent chromium and had great potential as an economical and efficient adsorbent of heavy metal from wastewater.

References

1. Y. Sun, Q. Yue, Y. Mao, B. Gao, Y. Gao and L. Huang, J. Hazard. Mater., 2014, 265, 191–200.
2. K. K. Krishnani, S. Srinives, B. C. Mohapatraa, V. M. Boddu, J. Haof, X. Mengf and A. Mulchandani, J. Hazard. Mater., 2013, 252–253, 99–106.
3. J. Hu, C. Chen, X. Zhu and X. Wang, J. Hazard. Mater., 2009, 162, 1542–1550.
4. A. B. Albadarin, C. Mangwandia, A. a. H. Al-Muhtaseb, G. M. Walkera, S. J. Allen and M. N. M. Ahmad, Chem. Eng. J., 2012, 179, 193–202.
5. Y. Liu, C. Luo, J. Sun, H. Li, Z. Sun and S. Yan, J. Mater. Chem. A, 2015, 3, 5674–5682.
6. Y. Liu, G. Cui, C. Luo, L. Zhang, Y. Guo and S. Yan, RSC Adv., 2014, 4, 55162–55172.
7. E. Malkoc, Y. Nuhoglu and M. Dundar, J. Hazard. Mater., 2006, 138, 142–151.
8. X.-j. Hu, J.-s. Wang, Y.-g. Liu, X. Li, G.-m. Zeng, Z.-l. Bao, X.-x. Zeng, A.-w. Chen and F. Long, J. Hazard. Mater., 2011, 185, 306–314.
9. L. Fan, C. Luo, M. Sun and H. Qiu, J. Mater. Chem., 2012, 22, 24577–24583.
10. X. Yuan, Y. Wang, J. Wang, C. Zhou, Q. Tang and X. Rao, Chem. Eng. J., 2013, 221, 204–213.
11. C. Houa, Q. Zhang, Y. Li and H. Wang, J. Hazard. Mater., 2012, 205–206, 229–235.
12. A. Dimiev, D. V. Kosynkin, A. Sinitskii, A. Slesarev, Z. Sun and J. M. Tour, Science, 2011, 331, 1168–1172.
13. H.-K. Seo, M. Song, S. Ameen, M. S. Akhtar and H. S. Shin, Chem. Eng. J., 2013, 222, 464–471.
14. J. Li, H. Xie and Y. Li, J. Power Sources, 2013, 241, 388–395.
15. S. Li, S. Guo, H. Yang, G. Gou, R. Ren, J. Li, Z. Dong, J. Jin and J. Ma, J. Hazard. Mater., 2014, 270, 11–17.
16. C. Luo, Z. Tian, B. Yang, L. Zhang and S. Yan, Chem. Eng. J., 2013, 234, 256–265.
17. D. C. Marcano, D. V. Kosynkin, J. M. Berlin, A. Sinitskii, Z. Sun, A. Slesarev, L. B. Alemany, W. Lu and J. M. Tour, ACS Nano, 2010, 4, 4806–4814.
18. X. Shen, J. Wu, S. Bai and H. Zhou, J. Alloys Compd., 2010, 506, 136–140.
19. S. B. Ma, K. Y. Ahn, E. S. Lee, K. H. Oh and K. B. Kim, Carbon, 2007, 45, 375–382.
20. J. Wei, Y. Liu, Y. Ding, C. Luo, X. Du and J. Lin, Chem. Commun., 2014, 50, 11938–11941.
21. N. A. Travlou, G. Z. Kyzas, N. K. Lazaridis and E. A. Deliyanni, Chem. Eng. J., 2013, 217, 256–265.
22. H. Huang and X. Wang, Nanoscale, 2011, 3, 3185–3191.
23. R. D. Martínez-Orozcoa, H. C. Rosu, S.-W. Lee and V. Rodríguez-González, J. Hazard. Mater., 2013, 263, 52–60.
24. B. Yang, Q. Gong, L. Zhao, H. Sun, N. Ren, J. Qin, J. Xu and H. Yang, Desalination, 2011, 278, 65–69.
25. X. Luo, C. Wang, S. Luo, R. Dong, X. Tu and G. Zeng, Chem. Eng. J., 2012, 187, 45–52.
26. P. I. Girginova, A. L. Daniel-da-Silva, C. B. Lopes, P. Figueira, M. Otero, V. S. Amaral, E. Pereira and T. Trindade, J. Colloid Interface Sci., 2010, 345, 234–240.
27. C. Luo, R. Wei, D. Guo, S. Zhang and S. Yan, Chem. Eng. J., 2013, 225, 406–415.
28. Z. Fan, J. Yan, T. Wei, L. Zhi, G. Ning, T. Li and F. Wei, Adv. Funct. Mater., 2011, 21, 2366–2375.
29. X. Huang, C. Pan and X. Huang, Mater. Lett., 2007, 61, 934–936.
30. S. R. Chowdhury, E. K. Yanful and A. R. Pratt, J. Hazard. Mater., 2012, 235–236, 246–256.
31. A. Mittal, L. Kurup and J. Mittal, J. Hazard. Mater., 2007, 146, 243–248.
32. Y. Yu, J. G. Shapter, R. Popelka-Filcoff, J. W. Bennett and A. V. Ellis, J. Hazard. Mater., 2014, 273, 174–182.
33. D. Suna, X. Zhang, Y. Wu and X. Liu, J. Hazard. Mater., 2010, 181, 335–342.
34. N. Azouaoua, Z. Sadaoui, A. Djaafri and H. Mokaddem, J. Hazard. Mater., 2010, 184, 126–134.
35. L. Zhang, T. Xu, X. Liu, Y. Zhang and H. Jin, J. Hazard. Mater., 2011, 197, 389–396.
36. S. Zhang, C. Liu, Z. Luan, X. Peng, H. Ren and J. Wang, J. Hazard. Mater., 2008, 152, 486–492.
37. S. Chen, Q. Yue, B. Gao and X. Xu, J. Colloid Interface Sci., 2010, 349, 256–264.
38. S. Gupta and B. V. Babu, Chem. Eng. J., 2009, 150, 352–365.
39. Y.-S. Ho, J. Hazard. Mater., 2006, 136, 681–689.
40. Y. S. Ho and G. McKay, Chem. Eng. J., 1998, 70, 115–124.
41. Q. Wen, Z. Chen, J. Lian, Y. Feng and N. Ren, J. Hazard. Mater., 2012, 209–210, 226–232.
42. X. Feng, Z. Yan, N. Chen, Y. Zhang, Y. Ma, X. Liu, Q. Fan, L. Wang and W. Huang, J. Mater. Chem. A, 2013, 1, 12818–12825.
43. Y. Wu, H. Luo, H. Wang, C. Wang, J. Zhang and Z. Zhang, J. Colloid Interface Sci., 2013, 394, 183–191.
44. G. Huang, J. X. Shi and T. A. G. Langrish, Chem. Eng. J., 2009, 152, 434–439.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra06455d

---