Introduction of α-MnO₂ nanosheets to NH₂ graphene to remove Cr⁶⁺ from aqueous solutions

Abstract

The planar structure of the designed α-MnO₂–NH₂–RGO hybrid was prepared and characterized and used to remove hexavalent chromium ions (Cr⁶⁺) from aqueous solutions. The characterization of the novel adsorbent was carried out using the Fourier transform infrared spectrum (FTIR), X-ray diffraction (XRD), transmission electron microscopy (TEM), and the Brunauer–Emmett–Teller (BET) theory. Cr⁶⁺ adsorption efficiency was investigated as a function of the pH of Cr⁶⁺ solution, contact time and temperature. The results showed that the adsorption capacity strongly depended on the pH and that the adsorption equilibrium data were best described by the Freundlich isothermal model. The maximum sorption capacity towards Cr⁶⁺ was 371 mg g⁻¹. The kinetic adsorption was fitted to the pseudo-second-order kinetics model and it indicated that the adsorption mechanism is physical or chemical sorption on heterogeneous materials. The thermodynamic parameters were obtained, and the results showed that the adsorption process is spontaneous and exothermic. The adsorption capacity of α-MnO₂–NH₂–RGO can remain as high as 81% after five usage cycles. As a result, these findings propose that α-MnO₂–NH₂–RGO could be used as an outstanding adsorbent for Cr⁶⁺.

---

1. Introduction

Chromium has been considered to be one of the most common pollutants because of its widespread use in industrial processes.¹ The use of chromium in pigment manufacture, leather tanning, metal finishing, ink manufacturing, and automobile spare parts are the main sources of chromium-based pollution. The acute toxicity, mutagenicity and carcinogenicity of chromium are strongly dependent on the oxidation state.² Chromium exists in the environment in both trivalent [Cr(III)] (Cr³⁺) and hexavalent [Cr(VI)] (Cr⁶⁺) forms. However, hexavalent chromium Cr⁶⁺ is more toxic and more mobile in nature than trivalent [Cr(III)]. Because of non-biodegradation and easy accumulation in living beings,³ it can cause various diseases and disorder generating fatal effects, such as corrosion of skin, respiratory tract infection, lung carcinoma, severe diarrhea, and hemorrhage.⁴ The World Health Organization (WHO) recommends a maximum allowable level of 50 ppb total chromium content in drinking water. The U.S. Environmental Protection Agency established a guideline of 100 ppb maximum contaminant level for total chromium content in drinking water.^3,5

To meet environmental regulations, effluents or water polluted with chromium must be treated before discharge. The technologies available for controlling Cr⁶⁺ in the environment include ion exchange,⁶ chemical precipitation,⁷ reverse osmosis,⁸ and adsorption.⁴ Adsorption is considered to be an attractive and cost-effective method for removal, recovery and recycling of metals from wastewater. The biosorbents can adequately remove heavy-metal ions, they do not show mechanical strength because of low selectivity.⁹ As for the chemical reduction precipitation process, hexavalent chromium is reduced to trivalent chromium.¹⁰ However, this process is expensive, the treatment is incomplete, and the water still has high Cr⁶⁺ concentration. These defects have improved the applications for Cr⁶⁺ removal from effluents or wastewater. Thus, there is a pressing need to develop new materials with high specific surface area, high adsorption capacity, easy sorption and stability in extreme conditions.¹¹

Being a one-atom-thick sheet of carbon atoms packed in two-dimensional (2D) honeycomb lattices, graphene possesses a large theoretical specific surface area (2630 m² g⁻¹),¹² superior electronic and outstanding chemical stability.¹³ Currently, graphene has enjoyed significant attention in the preparation of graphene-based composites because it provides outstanding support.^14,15 Among these nanocomposites, considerable attention has been focused on integrating graphene with other inorganic particles for the fabrication of composites or hybrids. In brief, it is believed that the adhesion of inorganic materials to graphene can prevent the graphene sheets from aggregating and keep the specific surface area and pore volume at high levels, which is essential for applications such as adsorption processes and photocatalysis. On the other hand, graphene can efficiently stabilize inorganic particles to prevent aggregation, and the performance of these inorganic materials could be enhanced through graphene loading.^16–19

Recent studies have showed that oxide minerals, ubiquitous in soils, rocks and sediments, serve as natural sinks for contaminants in the environment.^20,21 Nanosized manganese oxide (MnO₂) is among the most reactive minerals that have a large specific surface area, outstanding oxidizing/adsorption abilities, and excellent stability under acidic conditions.²² It provides an effective scavenging pathway for heavy metals in poisonous systems.

In general, the surface charge of MnO₂ is negative, and it can be used as a dominant adsorbent for the removal of heavy metals from water effluents. However, as the medium of filtration, pure manganese oxide is uneconomical and disadvantageous and possesses a physical characteristic that MnO₂ attenuates numerous heavy-metal ions via adsorption, ion exchange, and co-precipitation. Thus, it is highly desirable to explore the support of MnO₂ that can increase the removal efficiency for heavy ions with a low cost and environmentally benign nature.²³

In the present work, we demonstrate a novel hybrid nanostructure consisting of distinct ultrathin nanosheets for greatly enhanced adsorption performance. Partially, the structures based on the α-MnO₂ nanosheets integrated on NH₂ graphene sheets can introduce more active adsorption sites and bring additional interfaces at the hybridized interlayer areas to promote the adsorption process. Moreover, the α-MnO₂ integrated on NH₂ graphene can potentially tailor the distance between each layer of densely stacked graphene and open up the interlayer space to allow for more Cr⁶⁺ ions to adsorb effectively into the hybridized film. Consequently, the hybrid nanostructural design may enhance the adsorption of Cr⁶⁺, because it enlarges the specific surface area and improves the distribution channel of graphene. Hierarchically structured α-MnO₂–NH₂–RGO is obtained by the chemical integration of α-MnO₂ nanosheets on NH₂ graphene because of the strong electrostatic interaction in the process solution treatment. It is evidenced that the water-exfoliated δ-MnO₂ nanosheet is electronegative, and the DMF-exfoliated NH₂ graphene is electropositive. The produced composite was characterized by Fourier transform infrared spectrum (FTIR), X-ray diffraction (XRD), transmission electron microscopy (TEM), and the Brunauer–Emmett–Teller (BET) theory. To further evaluate adsorption behaviors, the pH of the Cr⁶⁺ solution, contact time and temperature were studied. In addition, the adsorption kinetics model, rate-limiting mechanism and the thermodynamics of the hybrids were also examined.

2. Experimental

2.1 Materials

Natural flake graphite was obtained from Nanyang Boxing Mining Co., Ltd. (China). Other chemicals (analytical grade) were purchased from AiHua Fine Chemicals Co., Ltd. (China) and used as received without any further purification.

2.2 Preparation of graphene oxide (GO)

Graphene oxide was synthesized by chemical oxidation in an improved Hummers' method.²⁴ In brief, 500 mg of graphite flakes was added into the mixture of concentrated H₂SO₄–H₃PO₄ (60 : 6.7 ml). Then, 3.0 g KMnO₄ was slowly added with vigorous stirring and produced a slight exothermic reaction at 35–40 °C. The mixture was stirred vigorously for about 12 h while maintaining a temperature of 323 K. The reaction was then cooled to room temperature and poured into 400 ml of ice with 1.5 ml of 35% hydrogen peroxide aqueous solution. Afterwards, the dispersion was filtered, washed with 200 ml water, HCl and absolute ethanol in succession and finally dried under vacuum conditions. The NH₂ graphene was obtained via a hydrothermal method with ethylenediamine.²⁵ Typically, a certain amount of GO was added into the ethanol solution (50 ml) with ultrasonication for 1 h, forming a homogeneous suspension. The ethylenediamine solution was subsequently dropped into the suspension (pH = 12.5) under vigorous stirring. The solution was then sealed in a Teflon-lined, stainless-steel autoclave and placed in an oven at 160 °C for 24 h, and then was cooled to room temperature naturally. The precipitate was collected and washed with deionized water. The yielded product was dried at 80 °C in air for further use.

The bulk MnO₂ was synthesized using a reported method.²⁶ In a typical reaction, 20 ml of a pre-blended solution of 0.6 M tetramethylammonium hydroxide (TMA OH) and 30 wt% hydrogen peroxide were added into 10 ml of 0.3 M MnCl₂·4H₂O aqueous solution within 15 s. The solution turned dark brown immediately, showing the evidence that Mn²⁺ was vigorously stirred for 12 h at room temperature. Afterwards, the precipitate was washed with water and methanol and dried in a vacuum oven at 60 °C. In our case, the layered H-type birnessite MnO₂ and NH₂ graphene can be exfoliated by water and DMF solution. Then, the two solution were mixed and sealed in a Teflon-lined, stainless-steel autoclave and heated at 160 °C for 8 h. The resultant product was washed thoroughly with deionized water and ethanol and dried at 60 °C in an oven.

2.3 Adsorption experiment

Cr⁶⁺ was used as an objective pollutant to evaluate the adsorptive performance of the as-prepared materials. Analytical grade potassium dichromate was used to prepare the 1000 mg l⁻¹ stock solution, which was further diluted to the required concentration before use. To investigate the adsorption capacity of the various adsorbents, 0.02 g of different materials were added into 100 ml of a 300 mg l⁻¹ Cr⁶⁺ solution at a fixed pH of 2.0. For the equilibrium, the initial Cr⁶⁺ concentrations were varied in the range of 50–300 mg l⁻¹. After a specified time, the fixed amounts of the suspension were withdrawn and filtered by 0.45 μm membrane filter for further analysis.

The pH effect experiments were performed at the initial Cr⁶⁺ concentration of 300 mg l⁻¹ by adjusting the initial pH in the range of 2.0–11.0 using 0.1 mol l⁻¹ HCl and NaOH. Kinetic studies were conducted using 0.1 g α-MnO₂–NH₂–RGO in 500 ml of the adsorbent solution (100 mg l⁻¹, 200 mg l⁻¹, and 300 mg l⁻¹). Thermodynamic tests were performed at different temperatures (275 K, 298 K, and 328 K) with various solutions ranging from 50 mg l⁻¹ to 300 mg l⁻¹. The adsorbance amount was calculated as follows:

where q_e is the amount of Cr⁶⁺ adsorbed by the as-prepared materials (mg g⁻¹) at any time (t), c₀ is the initial Cr⁶⁺ concentration (mg l⁻¹), c_t is the Cr⁶⁺ concentration after a fixed time (mg l⁻¹), v is the sample volume (l), and m is the weight of the adsorbents (g). For the regeneration, 0.02 g of α-MnO₂–NH₂–RGO was first added into 100 ml of 300 mg l⁻¹ Cr⁶⁺ solution within 150 min. After adsorption, we separated the adsorbents with high-speed centrifugation and washed them several times with distilled water, after which the adsorbents were immersed in 100 ml of 0.1 M NaOH solution for 150 min. Before the second adsorption, the adsorbent was treated with 0.1 M HCl solution for 150 min. Regenerated α-MnO₂–NH₂–RGO was used for adsorption in the five cycles.

3. Results and discussion

3.1 Characterization of different materials

Fig. 1 showed the typical XRD patterns of GO, NH₂–RGO, bulk MnO₂ and α-MnO₂–NH₂–RGO, conforming the ultrathin sheet structure of α-type MnO₂ on the NH₂ graphene. It is apparent that the most intensive peak of GO at around 2θ = 9.8° was in correspondence with the (001) reflection, and the interlayer spacing (0.91 nm) was considerably larger than that of graphite (0.34 nm) because of the introduction of oxygen-containing functional groups on the pristine graphite sheets.^27,28 The diffraction peaks of NH₂–RGO (25°) explained that the GO was reduced by ethylenediamine. The weak broad peaks of bulk MnO₂ at around 12.23°, 37.8°, and 63.64°, corresponding to the (001), (101), and (110) crystal planes of δ-MnO₂, showed a low degree of crystallinity. The α-MnO₂–NH₂–RGO nanocomposite corresponded to the (110), (200), (310), (211), (411) and (521) crystal planes of α-MnO₂.²⁹ Hydrothermal treatment facilitated the phase transition from δ-MnO₂ to α-MnO₂. These findings are in good agreement with the MnO₂ crystal growth theory.^30,31

Fig. 1 XRD patterns of GO, NH₂–RGO, bulk MnO₂, and α-MnO₂–NH₂–RGO.

The FTIR spectra were investigated to affirm the process for the preparation of α-MnO₂–NH₂–RGO hybrids. From Fig. 2, in the case of GO, the characteristic peaks appearing at 3416 cm⁻¹ and 1408 cm⁻¹ were assigned to the O–H stretching and bending vibrations of intercalated water. The absorption bands of oxygen groups at 1228 cm⁻¹, 1624 cm⁻¹ and 1738 cm⁻¹ were ascribed to the C–OH/C–O–C groups, carbonyl moieties, and C=O stretching of the –COOH, respectively.³² In the FT-IR spectrum of NH₂–RGO and α-MnO₂–NH₂–RGO, the amplitudes of the oxygen-containing functional characteristic bands decrease to a great degree, or the peaks even disappeared. This revealed that the GO was reduced after the hydrothermal treatment. In addition, the new peaks located at 1583 cm⁻¹ of NH₂–RGO and α-MnO₂–NH₂–RGO were ascribed to the stretching vibration of N–H (in the C–NH group).³³ Sharp peaks were observed at around 600–500 cm⁻¹, which can be attributed to Mn–O and Mn–O–Mn vibrations in the MnO₆ octahedral.³⁴ A new peak at around 725 cm⁻¹ was observed in α-MnO₂–NH₂–RGO, and it was attributed to the bending vibration of the Mn–O–C bond because of the interactional reaction between carbon and MnO₄⁻ ions. These results confirm that an intimate contact exists between graphene and MnO₂.³⁵

Fig. 2 FTIR spectra of GO, NH₂–RGO, bulk MnO₂, and α-MnO₂–NH₂–RGO.

The morphology of the as-prepared materials was investigated using TEM. Fig. 3a and b show flake-like shapes of GO with wrinkles. Fig. 3c and d show that MnO₂ with a sheet-like shape is homogeneously and tightly integrated on the NH₂ graphene surfaces, and the ultrathin MnO₂ sheets have widths ranging from 10 to 30 nm. It can be observed that the graphene sheets with large areas and flat morphologies served as ideal microscopic substrates to host the α-MnO₂ nanosheets (Fig. 3d) for maximum areal utilization. Furthermore, the existence of NH₂–RGO and MnO₂ in the nanocomposite has been verified by the peaks of C, N, Mn, and O in the EDX analysis, as exhibited in Fig. 3e.

Fig. 3 TEM images of (a) GO, (b) NH₂–RGO, (c and d) α-MnO₂–NH₂–RGO, (e) δ-MnO₂ sheets, and (f) EDX patterns of α-MnO₂–NH₂–RGO.

According to the N₂ adsorption–desorption isotherms (Fig. 4) of the GO, NH₂–RGO, bulk MnO₂ and α-MnO₂–NH₂–RGO, it conformed to the type-IV curves with an H₃ hysteresis loop, and the shapes indicated meso- and macro-porous characteristics.³⁶ From Table 1, it can be observed that the α-MnO₂–NH₂–RGO composite had a higher specific surface area (117 m² g⁻¹) than that of the GO (41 m² g⁻¹) and NH₂–RGO (57 m² g⁻¹), but lower than the bulk MnO₂ (212 m² g⁻¹), because the bulk MnO₂ was analyzed without exfoliation. The results indicated that ultrathin MnO₂ sheets were dispersed on the surface of the NH₂–RGO, which can present the aggregation of graphene sheets.³⁵ The α-MnO₂–NH₂–RGO showed a mesoporous structure with an average pore size of 14.256 nm. The high specific surface area and mesoporous structure of α-MnO₂–NH₂–RGO make them very promising candidates for the sorption of Cr(VI) in wastewater.

Fig. 4 Nitrogen sorption isotherms for (1) bulk MnO₂, (2) α-MnO₂–NH₂–RGO, and (3) NH₂–RGO (4) GO.

Table 1 Pore characteristics of GO, NH₂–RGO, bulk MnO₂, and α-MnO₂–NH₂–RGO

Materials	Surface area (m^2 g^−1)	Pore volume (cm^3 g^−1)	Pore size (nm)
GO	40.76	0.07435	5.2689
NH2–RGO	57.16	0.3165	20.965
Bulk MnO2	211.97	1.03948	14.537
α-MnO2–NH2–RGO	117.59	0.3904	14.256

---

3.2 Adsorption capacities of different materials

The adsorption capacities of different nanocomposites were discussed, and the results are presented in Fig. 5. The adsorption capacity of NH₂–RGO was greater than that of GO, indicating that the amino groups on the NH₂–RGO were expected to be the active centers for Cr⁶⁺. The adsorption amount of bulk MnO₂ is higher than that of GO, because manganese oxides have larger surface areas and higher affinities than GO. The α-MnO₂–NH₂–RGO nanocomposites were the most effective removal materials, compared with the others, because a greater number of binding sites were provided. The MnO₂ loading level on the surface of NH₂–RGO was suitable for Cr⁶⁺ adsorption, and the morphology of α-MnO₂ was uniform and thin, and more binding sites were provided. Thus, the adsorption amounts of various nanocomposites primarily relied on the functional groups, as well as on the morphology of the adsorbent surface.

Fig. 5 Adsorption capacity of various materials for Cr⁶⁺.

3.3 Effect of pH on Cr⁶⁺ adsorption

Adjustment of the initial Cr⁶⁺ pH value is a crucial parameter, which dominated the adsorption process, particularly the adsorption capacity. This parameter caused a change in the surface charge of the adsorbent, transformation of the chromium species and other ions present in the solution, and the degree of dissociation of functional groups on the active sites of the sorbent.³⁷ From Fig. S8,† we can clearly observe the XRD without any change. Thus, we think that the composite has stability under strong acidic conditions. Fig. 6 shows that the Cr⁶⁺ sorption capacity decreases with an increase in pH from 2.0 to 11.0. The Cr⁶⁺ species may be represented in an aqueous solution primarily because HCrO⁴⁻ and Cr₂O₄²⁻ were the predominant species with a pH lower than 6.8, and CrO₄²⁻ was prime form of Cr⁶⁺ when the pH is higher than 6.8.^38,39 The uptake of hexavalent chromium at a lower pH indicated that the predominant component (HCrO⁴⁻) of Cr⁶⁺ necessitated one exchange site from the α-MnO₂–NH₂–RGO for the adsorption to occur. On the contrary, at high pH, the divalent forms of Cr⁶⁺ (Cr₂O₇²⁻, CrO₄²⁻) were almost excited and needed two adsorption sites from the α-MnO₂–NH₂–RGO to occur for the adsorption.³⁷ Thus, the adsorption amount of α-MnO₂–NH₂–RGO was considerably larger at lower pH than that at higher pH. Furthermore, the amino groups in α-MnO₂–NH₂–RGO triggered the protonation and deprotonation due to the interaction mechanism. The reaction in eqn (1) illustrated that the protonation of amino groups enhanced at a low-solution pH, causing the α-MnO₂–NH₂–RGO surface to be more positively charged, leading to the binding of anionic Cr⁶⁺ species strongly. With increasing solution pH, the ability of –NH₂ to be protonated weakened, and OH⁻ ions may be adsorbed to the surface of the adsorbent through hydrogen bonds leading to more coulombic repulsion of metals,⁴⁰

Sur–NH₂ + H⁺ = Sur–NH₃⁺ (1)

Sur–NH₂ + OH⁻ = Sur–NH₂⋯OH⁻, (2)

where Sur denoted the surface of α-MnO₂–NH₂–RGO. In addition, chromium was an active metal, and the binding of Cr⁶⁺ with the α-MnO₂–NH₂–RGO may be attributed to the partial conversion of hexavalent chromium to the reduced speciation of trivalent chromium. The reduction of Cr⁶⁺ into Cr³⁺ by reducing the substrate (CxOH) on the occurrence of redox reactions between the surface groups and the Cr⁶⁺ at low pH values^41,42 is expressed as

HCrO₄⁻ + 7H⁺ + 3e⁻ ↔ Cr³⁺ + 4H₂O, (3)

Cr₂O₇²⁻ + 6e⁻ + 3H⁺ ↔ 2Cr³⁺ + 7H₂O. (4)

Fig. 6 Effect of the pH on Cr(VI) adsorption on the α-MnO₂–NH₂–RGO.

In order to affirm that Cr³⁺ coexists with Cr⁶⁺, we measured the concentration of total chromium and found that Cr³⁺ exists in the solution. Hydroxyl and carboxyl groups in adsorbents are suspected to bind Cr³⁺ ions by surface complexation and cation exchange mechanisms.⁴¹

In brief, the adsorption of Cr⁶⁺ is a complicated process, and the mechanism includes electrostatic attraction and complexation ion exchange.⁴³

3.4 Adsorption isotherm

Adsorption isotherms are the mathematical models that describe the solid–liquid adsorption system. Based on a set of assumptions, isotherms are primarily related to the homogeneity/heterogeneity of adsorbents. In addition, the type of coverage and how adsorbates could interact with the adsorbent are involved. The experimental adsorption equilibrium data were fitted on the Langmuir, Freundlich and Dubinin–Radushkevich (D–R) isotherms to represent the adsorption equilibrium. Fig. 7 shows the adsorption isotherms for Cr⁶⁺ on the surface of α-MnO₂–NH₂–RGO at various temperatures. The Langmuir model is derived from several assumptions, such as the uniform energies of the adsorption site and no transmigration of adsorbates on the surface of materials. The model takes the following linear form:⁴⁴

where q_e (mg g⁻¹) was the adsorption capacity at an equilibrium state on adsorbent and q_m and b (l mg⁻¹) signified the maximum adsorption amount and the apparent energy of the adsorption, respectively. C_e and C₀ (mg l⁻¹) represent the equilibrium concentration and initial concentration of Cr⁶⁺, respectively (Fig. S1†). The R_L value can demonstrate whether the isothermal type is favorable (0 < R_L < 1) or unfavorable (R_L > 1). The Freundlich isotherm, which is appreciable to heterogeneous surfaces and multilayer adsorption, is defined as follows:⁴⁵
where K_F (mg g⁻¹ (l mg⁻¹)^1/n) and n signified the adsorption capacity and the adsorption intensity, respectively. The Freundlich constants K_F and 1/n were obtained from the slope and intercept of the equation process (Fig. S2†). The D–R isotherm was applied to estimate the mean free energy of adsorption (E) in order to confirm whether the process is physical adsorption or chemical adsorption. Dubinin and Radushkevich⁴⁶ have proposed that the adsorption occurs on a single uniform pore. In this connection, the D–R isotherm is used extensively, because the assumption does not contain the homogeneous surface or similar adsorption energy.⁴⁷

ln q_e = ln q_m − βε²

where β (mol² kJ⁻²) is a constant related to sorption mean free energy, and ε is the polony potential (Fig. S3†). The mean free energy of adsorption can be calculated out using the following relationship:

Fig. 7 Adsorption isotherms for Cr⁶⁺ on the surface of α-MnO₂–NH₂–RGO at various temperatures.

From Table 2, we can see that the hexavalent Cr⁶⁺ ions were favorably adsorbed on the α-MnO₂–NH₂–RGO, and the adsorption amounts of Cr⁶⁺ obtained were 333 mg g⁻¹, 350 mg g⁻¹, and 371 mg g⁻¹ at 275 K, 298 K and 328 K, respectively. The sorption amount attained with the α-MnO₂–NH₂–RGO was considerably higher than that gained with numerous other reported adsorptions. In consideration of the correlation coefficient as a criterion for the goodness of fit in the adsorption process, the Freundlich isothermal adsorption exhibited an improved correlation (R² = 0.996, 0.993, 0.997) than the Langmuir (R² = 0.924, 0.928, 0.933) and D–R (R² = 0.81, 0.85, 0.84) isothermal models, which demonstrated that the Freundlich isothermal adsorption system was more ideal. Table 2 indicates that the adsorption of Cr⁶⁺ is a multi-adsorption process, because the Freundlich isothermal adsorption model assumes a multilayer adsorption and the stronger binding sites are occupied first by the adsorbates.⁴⁸ Furthermore, the K_F and n were calculated from Fig. S2.† The values of K_F are 34.33, 45.84, and 55.26, whereas n are 2.5, 2.8, and 2.98. The n can be used to predict the adsorption characteristics. For n < 1, the adsorption is unfavorable, n between a value of 1 and 2 adsorption is defined as moderately difficult, and the value of n between 2 and 10 considered as having good adsorption.⁴⁹ The n determined in our studies represents good adsorption that was in accordance with the larger value of K_F. The parameter E (kJ mol⁻¹) indicated the adsorption mechanism, which is physical or chemical. If the magnitude of E is between 8 and 16 kJ mol⁻¹, the adsorption process occurs chemically, whereas the value of E < 8 kJ mol⁻¹ corresponds to a physical adsorption process.^50,51 From the calculated value of the mean free adsorption energy, it indicated that the adsorption mechanism of Cr⁶⁺ on the α-MnO₂–NH₂–RGO may be a physical adsorption. However, the D–R isotherm was poorly fitted to the adsorption process with a regression coefficient. Combining with the reference of pseudo-second-order kinetics, it could suggest that the adsorption process may occur by a physical–chemical mechanism.⁵²

Table 2 Parameters of the Langmuir, Freundlich and D–R isotherms for Cr⁶⁺ adsorption onto α-MnO₂–NH₂–RGO

T (K)	Langmuir isotherm				Freundlich isotherm				D–R isotherm
	qm	b	r1^2	RL	n	KF	r2^2	qm	β (mol^2 J^−2)	E (J mol^−1)	r3^2
275	333	0.0306	0.924	0.417–0.106	2.5	34.33	0.996	257	7.61 × 10^−5	81	0.81
298	350	0.0298	0.928	0.402–0.1	2.8	45.84	0.993	284	6.27 × 10^−5	89	0.85
328	371	0.0405	0.933	0.331–0.076	2.98	55.26	0.997	302	3.97 × 10^−5	112	0.84

---

3.5 Adsorption kinetics

The kinetics of adsorption on the α-MnO₂–NH₂–RGO for various initial Cr⁶⁺ concentrations are shown in Fig. 8. Possible mechanisms can be investigated in accordance with several reaction processes such as chemical, diffusion control and mass-transport kinetics. Pseudo-first-order and pseudo-second-order models and the intraparticle diffusion model were employed to analyze the kinetics of Cr⁶⁺ adsorption. The pseudo-first-order model is expressed by the equation as⁵³
where q_e and q_t represented the amount of Cr⁶⁺ adsorbed (mg g⁻¹) at the equilibrium time and time t (min), respectively. k₁ (min⁻¹) is the rate constant (Fig. S4†). The pseudo-second-order rate equation is given as⁵⁴
where k₂ was the pseudo-second-order rate constant, and q_e, q_t and t had the same definitions as those in the pseudo-first-order kinetics equation (Fig. S5†). The initial adsorption rate h (mg g⁻¹ min)⁵⁵ is expressed in the following manner:

h = kq_e².

Fig. 8 The adsorption kinetics on the α-MnO₂–NH₂–RGO about various initial Cr⁶⁺ concentrations.

The intra-particle diffusion model is given as⁵⁶

q_t = k_pt^0.5 + c,

where k_p was the intraparticle diffusion rate constant (mg g⁻¹ min^−1/2). The kinetic parameters under various conditions are listed in Table 3. The correlation coefficient (R²) of the pseudo-second order model (0.993, 0.995, 0.994) was considerably higher than that of the pseudo-first order model (0.85, 0.94, 0.946), and the calculated value of q_e from the pseudo-second-order model (209 mg g⁻¹, 261 mg g⁻¹, 346 mg g⁻¹) fitting were in accordance with those of the experimental ones (194 mg g⁻¹, 256 mg g⁻¹, 334 mg g⁻¹). These results indicated that the adsorption conforms to the pseudo-second-order reaction mechanism, and the reaction rate of the entire adsorption process is controlled by chemical adsorption rather than mass transfer.⁵⁷ The initial adsorption rate increased with an increase in initial Cr⁶⁺ concentration, possibly because a more efficient utilization of the adsorption capacity of the adsorbent was expected at higher initial concentrations due to the greater driving force.⁵⁸ When the linear plots of q versus t^0.5 passed through the origin, the intraparticle diffusion played a significant role in the sole rate-limiting process. Intra-particle diffusion from the solid–liquid interface to the interior of the solid particle played a significant role in Cr⁶⁺ adsorption onto the surface of α-MnO₂–NH₂–RGO. When the linear plots of q vs. t^0.5 passed through the origin, the intra-diffusion was the sole rate-limiting process. Fig. 9 shows that intraparticle diffusion is not the only rate-limiting step, although it is involved in the adsorption, and boundary layer diffusion also has an effect of adsorption to some extent.

Table 3 Kinetic parameters for Cr⁶⁺ adsorption onto α-MnO₂–NH₂–RGO

C0 (mg l^−1)	qe.exp	Pseudo-first-order kinetic model			Pseudo-second-order kinetic model
		k1	qe	r1^2	k2	qe ( × 10^−4)	h	r2^2
100	194	0.0303	275	0.85	1.20	209	5.24	0.993
200	256	0.026	325	0.94	1.23	261	8.38	0.995
300	334	0.0352	398	0.95	1.32	346	15.8	0.994

---

Fig. 9 Weber–Morris plots for the kinetic modeling of Cr⁶⁺ adsorbed onto α-MnO₂–NH₂–RGO at various temperatures.

3.6 Thermodynamic studies

Thermodynamic parameters provide in-depth information about internal energy changes that are associated with adsorption. The adsorption standard Gibbs free energy changes (ΔG°), the average standard enthalpy change (ΔH°), and the standard entropy charge (ΔS°) can be obtained from the thermodynamic equilibrium constant K₀ with a change in the temperature of the Cr⁶⁺ adsorption onto α-MnO₂–NH₂–RGO to evaluate the adsorption process. K₀ is calculated as follows:⁵⁹where a_s is the activity of the adsorbed Cr⁶⁺, a_e is the activity of the Cr⁶⁺ at equilibrium, C_s (mmol g⁻¹) is the capacity of Cr⁶⁺ adsorbed per α-MnO₂–NH₂–RGO mass, C_e (mmol ml⁻¹) is the equilibrium concentration, v_s is the activity coefficient of the adsorbed Cr⁶⁺, and v_e is the activity coefficient of Cr⁶⁺ in solution. As the concentration of the Cr⁶⁺ approaches unity, eqn (1) is reduced to the following form,K₀ was obtained by plotting ln(C_s/C_e) versus C_s and extrapolating C_s to zero (Fig. S6†). The adsorption standard Gibbs free energy changes (ΔG°) can be obtained as follows:

ΔG° = −RT ln K₀,

where R was the gas constant (8.314 kJ mol⁻¹ k⁻¹), and T was the absolute temperature in Kelvin. The values of ΔH°, ΔS° for Cr⁶⁺ adsorption were obtained from slope and intercept of linear plot of ln K₀ versus 1/T (Fig. S7†) using the following equation⁶⁰

Thermodynamic parameters are listed in Table 4. The negative values of ΔG° demonstrate that the adsorption process was spontaneous. In addition, with increased temperature, the ΔG° value becomes more negative, revealing that higher temperature facilitated the spontaneity due to a greater driving force of adsorption or activation of the adsorbent surface.⁶¹ The calculated ΔG° values suggest that the adsorption process was primarily a physical process. However, chemical adsorptions may also occur because the calculated value of ΔG° range from 20 kJ mol⁻¹ to 80 kJ mol⁻¹, which is the process in which physical adsorption combines with the chemical adsorption. The obtained result was in good agreement with the result found with the D–R isotherm. The positive value of ΔH° revealed that the interaction of Cr⁶⁺ adsorbed by α-MnO₂–NH₂–RGO was an endothermic process. The positive cause was that the metal ions are well hydrated and will need breaking of the hydration bond to facilitate adsorption, which consequently requires higher energy. Thus, higher temperatures accelerate the dehydration process. The positive standard change (ΔS°) suggested an increased disorder at the solid–solution interface.⁶²

Table 4 Thermodynamic parameters for Cr⁶⁺ adsorption onto α-MnO₂–NH₂–RGO

T (K)	ln K0	ΔG° (kJ mol^−1)	R^2	ΔH° (kJ mol^−1)	ΔS° (J mol^−1 K^−1)	R^2
275	9.525	21.78	0.969	6.73	103.7	0.996
298	9.746	24.15	0.991
328	10.01	27.29	0.992

---

3.7 Regeneration of α-MnO₂–NH₂–RGO

From a practical/industrial point view, the recycling and the reuse of the adsorbent is an economic necessity. Considering that the nanocomposites showed poor adsorption ability at high pH, alkaline condition treatment is likely to be a suitable method for the regeneration of α-MnO₂–NH₂–RGO. Thus, a 0.1 M NaOH solution was used, and the adsorption–desorption cycles were replicated five times using the same batch of α-MnO₂–NH₂–RGO (Fig. 10). The as-prepared composite still exhibited more than 81% adsorption capacity after five cycles of reuse, indicating that α-MnO₂–NH₂–RGO possess good regeneration and reusability for Cr⁶⁺ adsorption. This was consistent with the abovementioned result that higher pH depressed the adsorption of Cr⁶⁺, but was beneficial for its recovery.

Fig. 10 The regeneration of α-MnO₂–NH₂–RGO for Cr⁶⁺ adsorption.

3.8 Adsorption mechanism

The Cr⁶⁺ sorption of the α-MnO₂–NH₂–RGO might primarily occur via the electrostatic attraction route in the early stage, whereas it became chemical adsorption via the covalent bonding route in the late stage.²⁶ In an aqueous solution, Cr⁶⁺ ions exist in the forms of Cr₂O₇²⁻ and HCrO₄⁻, and the amine groups on the sorbent were protonated in order to form surface complexes with the chromium species via electrostatic attraction, which explains the pH effect on the adsorption (Fig. 6). Moreover, the foremost sorption group was MnO₂, because the sorption amount of α-MnO₂–NH₂–RGO was better than that of NH₂–RGO (Fig. 5). The MnO₂ was amphoteric and can feature with weak acidic or weak basal characteristics. The protonation and deprotonation of the oxide surface was controlled by the pH of the Cr⁶⁺ solution. Therefore, when the pH was low, the MnO₂ could be efficiently protonated from =MnOH with a positive charge,⁶³ and an anion HCrO₄⁻ could thus be bound through electrostatic attraction. In addition, the redox reactions may proceed either sequentially and/or in parallel, particularly under acidic conditions. Furthermore, an ion exchange would occur among Cr₂O₇²⁻, HCrO₄⁻, and OH⁻ on the α-MnO₂–NH₂–RGO generating the ion exchange, thereby constituting the adsorption mechanism.⁶⁴

3.9 Comparison of adsorbent performance with literature data

To evaluate the effectiveness of α-MnO₂–NH₂–RGO as a potential adsorption for Cr(VI), the maximum adsorption capacity q_m was compared with those of other adsorbents reported previously. From Table 5, the maximum adsorption amount of α-MnO₂–NH₂–RGO was higher than the values of graphene/MgAl, monodisperse Fe₃O₄, iron nanoparticle-decorated graphene, Fe₃O₄–graphene, and several surface-modified MnO₂ nanoparticles such as MnO₂, manganese oxide nanofibers, MnO₂-modified expanded graphite, MnO₂ nanowires-diatomite, and MnO₂/Fe₃O₄/o-MuCNTs, but lower than polyethyleneimine–graphene oxide (GO) and UV-active functionalized graphene oxide. Compared with the MnO₂, manganese oxide nanofibers, MnO₂-modified expanded graphite and MnO₂/Fe₃O₄/o-MWCNTs, the α-MnO₂–NH₂–RGO displayed a higher surface area, which could provide more active adsorption sites. Moreover, the crystallization of MnO₂ thin films on the surface of RGO also generated more active oxygen, and the resultant interface contributed to the excellent adsorption. Moreover, the introduction of functional groups, particularly metal oxide and amino groups, can enhance the adsorption capacity and dispersion of graphene. It is noteworthy that α-MnO₂–NH₂–RGO is very stable under acidic conditions. For the aforementioned reasons, α-MnO₂–NH₂–RGO can be considered a promising absorbent for the effective removal of Cr(VI).

Table 5 Maximum adsorption capacities of Cr⁶⁺ on α-MnO₂–NH₂–RGO and other adsorbents

Adsorbates	Adsorbents	qmax (mg g^−1)	Ref.
Cr^6+	Calcined graphene/MgAl-layered double hydroxides	177.6	65
	Amine-grafted alumino–siloxane hybrid	48.96	66
	40% aliquat 336/60% PVC PIM	50.85	67
	MnO2	51.996	68
	Manganese oxide nanofibers	11.8	69
	High saturation magnetization monodisperse Fe3O4 hollow microsphere	180	70
	MnO2-modified expanded graphite	0.3011	71
	Iron nanoparticle-decorated graphene	162	72
	MnO2 nanowires in situ grown on diatomite	197.6	30
	MnO2/Fe3O4/o-MWCNTs	186.9	73
	Graphene oxide	65.2	10
	Iron nanoparticle decorated graphene	162	74
	Polyethyleneimine–graphene oxide (GO)	539.5	75
	UV-active functionalized-graphene oxide	393.8	76
	α-MnO2–NH2–RGO	371	Present work

---

4. Conclusion

In summary, we have fabricated for the first time a novel, high-performance adsorbent by chemically integrating α-MnO₂ nanosheets on NH₂ graphene due to strong electrostatic interaction. The results exhibited that the adsorption capacity was strongly dependent on the pH and initial Cr⁶⁺ concentrations, the adsorption data were found to fit well with the Freundlich isotherm model. The adsorption equilibrium was achieved within 100 min, and the maximum adsorption capacity was 371 mg g⁻¹. The adsorption kinetics was found to follow the pseudo-second-order equation. By calculating the thermodynamic parameters, it indicated that the adsorption process was spontaneous and exothermic. The D–R model showed that sorption had a physical–chemical nature. The adsorption capacity of α-MnO₂–NH₂–RGO can remain as high as 81% after five cycles of usage. The results obtained in this study illustrated that the α-MnO₂–NH₂–RGO is expected to be an effective and economical adsorbent for Cr(VI) removal from an aqueous system.

Acknowledgements

We thank classmates for giving many suggestions on this work.

References

1. J. Johnson, L. C. Wel and A. E. Graedel, Environ. Sci. Technol., 2006, 40, 7060–7069.
2. Y. P. Huang, H. Ma, S. G. Wang, M. W. Shen, R. Guo, X. Y. Cao, M. F. Zhu and X. Y. Shi, ACS Appl. Mater. Interfaces, 2012, 4, 3054–3061.
3. W. J. Jiang, M. Pelaez, D. D. Dionysiou, M. H. Entezari, D. Tsoutsou and K. O'Shea, Chem. Eng. J., 2013, 222, 527–533.
4. K. G. Bhattacharyya and S. S. Gupta, Ind. Eng. Chem. Res., 2006, 45, 7232–7240.
5. D. H. Thomas, J. S. Rohrer, P. E. Jackson, T. Pak and J. N. Scott, J. Chromatogr. A, 2002, 956, 255–259.
6. S. Rengaraj, C. K. Joo, Y. H. Kim and J. Yi, J. Hazard. Mater., 2003, 102, 257–275.
7. N. M. Lee, H. Carlsson, H. Aspegren, T. Welander and B. Andersson, Water Sci. Technol., 1996, 34, 101–109.
8. A. Hafez and S. El-Manharawy, Desalination, 2004, 165, 141–151.
9. D. Zhang, S. Y. Wei, C. D. Kaila, X. Su, J. Wu, A. B. Karki, D. P. Young and Z. H. Guo, Nanoscale, 2010, 2, 917–919.
10. L. Fan, C. N. Luo, M. Sun and H. M. Qiu, J. Mater. Chem., 2010, 22, 24577–24583.
11. L. L. Fan, C. N. Luo, M. Sun and H. M. Qiu, J. Mater. Chem., 2012, 22, 24577–24583.
12. P. Y. Dong, Y. H. Wang, L. N. Guo, B. Liu, S. Y. Xin, J. Zhang, Y. R. Shi, W. Zeng and S. Yin, Nanoscale, 2012, 4, 4641–4649.
13. K. P. Loh, Q. L. Bao, P. K. Ang and J. X. Yang, J. Mater. Chem., 2010, 20, 2277–2289.
14. J. W. Qin, M. H. Cao, N. Li and C. W. Hu, J. Mater. Chem., 2011, 21, 17167–17174.
15. X. Y. Yan, X. L. Tong, Y. F. Zhang, X. D. Han, Y. Y. Wang, G. Q. Jin, Y. Qin and X. Y. Guo, Chem. Commun., 2012, 48, 1892–1894.
16. B. J. Li, H. Q. Cao and G. Yin, J. Mater. Chem., 2011, 21, 13765–13768.
17. M. Willander, K. U. Hasan, O. Nur, A. Zainelabdin, S. Zaman and G. Amin, J. Mater. Chem., 2012, 22, 2337–2350.
18. B. J. Li, H. Q. Cao, G. Yin, Y. X. Lu and J. F. Yin, J. Mater. Chem., 2011, 21, 10645–10648.
19. Y. H. Xue, H. Chen, D. S. Yu, S. Y. Wang, M. Yardeni, Q. B. Dai, M. M. Guo, Y. Liu, F. Lu, J. Qu and L. M. Dai, Chem. Commun., 2011, 47, 11689–11691.
20. E. J. Kim, C. S. Lee, Y. Y. Chang and Y. S. Chang, ACS Appl. Mater. Interfaces, 2013, 5, 9628–9634.
21. R. P. Han, W. H. Zou, H. K. Li, Y. H. Li and J. Shi, J. Hazard. Mater., 2006, 102, 934–942.
22. R. P. Han, W. H. Zou, Z. P. Zhang, J. Shi and J. J. Yang, J. Hazard. Mater., 2006, 137, 384–395.
23. Z. W. Zhao, J. Liu, F. Y. Cui, H. Feng and L. L. Zhang, J. Mater. Chem., 2012, 22, 9052–9057.
24. D. C. Marcano, D. V. Kosynkin, J. M. Berlin, A. Sinitskii, Z. Z. Sun, A. Slesarev, L. B. Alemany, W. Lu and J. M. Tour, ACS Nano, 2010, 4, 4806–4814.
25. L. Zhang, H. Z. Li, Y. Liu, Z. Tian, B. Yang, Z. B. Sun and S. Q. Yan, RSC Adv., 2014, 4, 48703–48711.
26. L. L. Peng, X. Peng, B. R. Liu, C. Z. Wu, Y. Xie and G. H. Yu, Nano Lett., 2013, 13, 2151–2157.
27. S. Chen, J. W. Zhu, X. D. Wu, Q. F. Han and X. Wang, ACS Nano, 2010, 4, 2822–2830.
28. C. Xu, X. D. Wu, J. W. Zhu and X. Wang, Carbon, 2007, 46, 386–389.
29. JCPDS 00-044-0141.
30. Y. C. Du, G. W. Zheng, J. S. Wang, L. P. Wang, J. S. Wu and H. X. Dai, Microporous Mesoporous Mater., 2014, 200, 27–34.
31. X. L. Huang, N. T. Hu, R. G. Gao, Y. Yu, Y. Y. Wang, Z. Yang, E. Siu, W. Kong, H. Wei and Y. F. Zhang, J. Mater. Chem., 2012, 22, 22488–22495.
32. Y. Liu, R. J. Deng, Z. Wang and H. T. Liu, J. Mater. Chem., 2012, 22, 13619–13624.
33. A. B. Yuan and Q. L. Zhang, Electrochem. Commun., 2006, 8, 1173–1178.
34. M. C. Hu, K. S. Hui and K. N. Hui, Chem. Eng. J., 2014, 254, 237–244.
35. G. D. Jiang, Z. F. Lin, C. Chen, L. H. Zhu, Q. Chang, N. Wang, W. Wei and H. Q. Tang, Carbon, 2011, 49, 2693–2701.
36. M. S. Hosseini, A. H. Bandegharaei, H. Raissi and F. Belador, J. Hazard. Mater., 2009, 169, 52–57.
37. L. Tang, Y. Fang, Y. Pang, G. M. Zeng, J. J. Wang, Y. Y. Zhou, Y. Y. Deng, G. D. Yang, Y. Cai and J. Chen, Chem. Eng. J., 2014, 254, 302–312.
38. Y. Pang, G. M. Zeng, L. Tang, Y. Zhang, Y. Y. Liu, X. X. Lei, Z. Li, J. C. Zhang, Z. F. Liu and Y. Q. Xiong, Chem. Eng. J., 2011, 175, 222–227.
39. A. M. Yusof and N. A. N. N. Malek, J. Hazard. Mater., 2009, 162, 1019–1024.
40. Q. Q. Zhong, Q. Y. Yue, B. Y. Gao, Q. Li and X. Xu, Chem. Eng. J., 2013, 229, 90–98.
41. J. Hu, C. L. Chen, X. X. Zhu and X. K. Wang, J. Hazard. Mater., 2009, 162, 1542–1550.
42. G. R. Xu, J. N. Wang and C. J. Li, Chem. Eng. J., 2012, 198–199, 310–317.
43. J. Anandkumar and B. Mandal, J. Hazard. Mater., 2009, 168, 633–640.
44. A. Kurniawan, H. Sutiono, N. Indraswati and S. Ismadji, Chem. Eng. J., 2012, 189–190, 264–274.
45. M. Brdar, M. Šćiban, A. Takači and T. Došenović, Chem. Eng. J., 2012, 183, 108–111.
46. N. D. Hutso and R. T. Yang, Adsorption, 1997, 3, 189–195.
47. N. Ünlü and M. Ersoz, J. Hazard. Mater., 2006, 136, 272–280.
48. S. Haider and S. Y. Park, J. Membr. Sci., 2009, 328, 90–96.
49. M. Atienza-Martínez, I. Fonts, J. Ábrego, J. Ceamanos and G. Gea, Chem. Eng. J., 2013, 222, 534–545.
50. I. Kiran, T. Akar, A. S. Ozcan, A. Ozcan and S. Tunali, Biochem. Eng. J., 2006, 31, 197–203.
51. M. S. Onyango, Y. Kojima, O. Aoyi, E. C. Bernardo and H. Matsuda, J. Colloid Interface Sci., 2004, 279, 341–350.
52. F. Raji and M. Pakizeh, Appl. Surf. Sci., 2013, 282, 415.
53. M. Chairat, S. W. Rattanaphani, J. B. Bremnerc and V. Rattanaphani, Dyes Pigm., 2008, 76, 435–439.
54. Y. S. Ho and G. McKay, Process Biochem., 1999, 34, 451–465.
55. E. Demirbas, N. Dizge, M. T. Sulak and M. Kobya, Chem. Eng. J., 2009, 148, 480–487.
56. N. Li, Z. Mei and X. Wei, Chem. Eng. J., 2012, 192, 138–145.
57. R. Sitko, E. Turek, B. Zawisza, E. Malicka, E. Talik, J. Heimann, A. Gagor, B. Feist and R. Wrzalik, Dalton Trans., 2013, 5682–5689.
58. Q. Q. Zhong, Q. Q. Yue, Q. Li, B. Y. Gao and X. Xu, Carbohydr. Polym., 2014, 111, 788–796.
59. R. Niwas, U. Gupta, A. A. Khan and K. G. Varshney, Colloids Surf., A, 2000, 164, 115–119.
60. L. Ai, C. Y. Zhang and Z. L. Chen, J. Hazard. Mater., 2011, 192, 1515–1524.
61. R. P. Han, Z. Lu, W. H. Zou, D. T. Wang, J. Shi and J. J. Yang, J. Hazard. Mater., 2006, 137, 480–488.
62. M. Bhaumik, A. Maity, V. V. Srinivasu and M. S. Onyango, J. Hazard. Mater., 2011, 190, 381–390.
63. R. P. Han, W. H. Zou, Z. P. Zhang, J. Shi and J. J. Yang, J. Hazard. Mater., 2006, 137, 384–395.
64. X. F. Sun, Y. Ma, X. W. Liu, S. G. Wang, B. Y. Gao and X. M. Li, Water Res., 2010, 44, 2517–2524.
65. X. Y. Yuan, Y. F. Wang, J. Wang, C. Zhou, Q. Tang and X. B. Rao, Chem. Eng. J., 2013, 221, 204–213.
66. V. Linsha, P. S. Suchithra, A. P. Mohamed and S. Ananthakumar, Chem. Eng. J., 2013, 220, 244–253.
67. C. V. Gherasim and G. Bourceanu, Chem. Eng. J., 2013, 220, 24–34.
68. C. Noubactep, K. B. D. Btatkeu and J. B. Tchatchueng, Chem. Eng. J., 2011, 178, 78–84.
69. A. H. Qusti, J. Ind. Eng. Chem., 2014, 20, 3394–3399.
70. Y. B. Liu, Y. Q. Wang, S. M. Zhou, S. Y. Lou, L. Yuan, T. Gao, X. P. Wu, X. J. Shi and K. Wang, ACS Appl. Mater. Interfaces, 2012, 4, 4913–4920.
71. H. Y. Jin, J. Yuan, H. Y. Hao, Z. J. Ji, M. Liu and S. Hou, Mater. Lett., 2013, 110, 69–72.
72. H. Jabeen, V. Chandra, S. Jung, J. W. Lee, K. S. Kim and S. B. Kim, Nanoscale, 2011, 3, 3583–3585.
73. C. Luo, Z. Tian, B. Yang, L. Zhang and S. Q. Yan, Chem. Eng. J., 2013, 224, 256–265.
74. H. Jabeen, V. Chandra, S. Jung, J. W. Lee, K. S. Kim and S. B. Kim, Nanoscale, 2011, 3, 3583–3585.
75. J. H. Chen, H. T. Xing, H. X. Guo, W. Weng, S. R. Hu, S. X. Li, Y. H. Huang, X. Sun and Z. B. Su, J. Mater. Chem. A, 2014, 2, 12561–12570.
76. D. Dinda, A. Gupta and S. K. Saha, J. Mater. Chem. A, 2013, 1, 11221–11228.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra04545b

---