Mn-Doped Spinel for Removing Cr(VI) from Aqueous Solutions: Adsorption Characteristics and Mechanisms

In this study, the manganese (Mn) was doped in the MnFe2O4 crystal by the solid-phase synthesis method. Under the optimum conditions (pH = 3), the max removal rate and adsorption quantity of Cr(VI) on MnFe2O4 adsorbent obtain under pH = 3 were 92.54% and 5.813 mg/g, respectively. The DFT calculation results indicated that the adsorption energy (Eads) between HCrO4− and MnFe2O4 is −215.2 KJ/mol. The Cr(VI) is mainly adsorbed on the Mn atoms via chemical bonds in the form of HCrO4−. The adsorption of Mn on the MnFe2O4 surface belonged to chemisorption and conformed to the Pseudo-second-order equation. The mechanism investigation indicated that the Mn in MnFe2O4 has an excellent enhancement effect on the Cr(VI) removal process. The roles of Mn in the Cr(VI) removal process included two parts, providing adsorbing sites and being reductant. Firstly, the Cr(VI) is adsorbed onto the MnFe2O4 via chemisorption. The Mn in MnFe2O4 can form ionic bonds with the O atoms of HCrO4−/CrO42−, thus providing the firm adsorbing sites for the Cr(VI). Subsequently, the dissolved Mn(II) can reduce Cr(VI) to Cr(III). The disproportionation of oxidized Mn(III) produced Mn(II), causing Mn(II) to continue to participate in the Cr(VI) reduction. Finally, the reduced Cr(III) is deposited on the MnFe2O4 surface in the form of Cr(OH)3 colloids, which can be separated by magnetic separation.


Introduction
Chromium (Cr) widely exists in the surface and underground water as a common poisonous heavy metal. The Cr in water mainly comes from industrial discharges, such as leather tanning, electroplating, metal plating, and chemical manufacturing [1][2][3]. There are two primary species of Cr, trivalent (Cr (III)) and hexavalent (Cr (VI)) ions in the aqueous system [4], and the toxicity and mobility of Cr are highly dependent on its existing species [5]. It is widely believed that the toxicity of Cr (VI) is much higher than that of Cr (III), and Cr(VI) is highly carcinogenic and teratogenic [6]. The hypertoxic Cr(VI) in the natural environment could cause a range of diseases and pose a major threat to humans and ecology [7,8]. Therefore, many present researchers have focused on the field of Cr(VI) reduction [9,10]. Among this research, the adsorption-reduction method for Cr(VI) removal has received the most extensive research [11]. Furthermore, some biosorbents have been developed for heavy-metal removal [12].
The current Cr(VI) reduction methods primarily include chemical, electrochemical, photocatalytic, and biological reduction processes [13][14][15][16]. Most of the literature has shown that adsorption is a prerequisite for Cr(VI) reduction [17]. Many adsorbents have been prepared to achieve the adsorption-reduction reaction of Cr(VI) on their surfaces [18][19][20]. Among these adsorbents, metallic oxide, especially spinel-type materials, has been most extensively investigated due to its good magnetic properties, easy synthesis, low cost, and chemical stability. Many published literature studies have verified that Cr(VI) can be reduced by certain low-valence metal ions such as Fe(II) and Mn(II) in spinel materials [21,22].
In the previous work, our group invented a novel high-temperature solid-phase synthesis method to prepare the MnFe 2 O 4 adsorbent [26]. To begin with, the ferromanganese ore and hematite ore powders were evenly mixed, and the stoichiometric ratio of Mn and Fe in mixed raw materials was 1:2. The roasting temperature and time were controlled at 1200 • C and 1 h. The mixture briquettes were broken and ground to −0.074 mm, accounting for 70 wt% after roasting. Then the MnFe 2 O 4 particles and the gangue minerals were separated by magnetic separation. The obtained MnFe 2 O 4 particles continued to be nano-ground to the size of 200 nm (d0.5 = 200 nm). The nano-MnFe 2 O 4 particles were used as a magnetic adsorbent for the Cr(VI) removal.
All chemicals used in this research were of analytical grade. The simulated Cr(VI)bearing solution was prepared using K 2 Cr 2 O 7 (Sinopharm Group Co., Ltd., Beijing, China). The pH was adjusted with 0.1 M HCl and 0.1 M NaOH solutions.

Adsorption Characteristics
Adsorption experiments were conducted under isothermal conditions using the batch method. The isothermal adsorption experiments were conducted to investigate the Cr(VI) adsorption characteristics of the MnFe 2 O 4 adsorbent. A certain weight of the MnFe 2 O 4 adsorbent was added to an Erlenmeyer flask with a 50 mL Cr(VI) solution to conduct the adsorption experiments. The mixed solutions were oscillated for 24 h to ensure the Cr(VI) achieved equilibrium. Then the filtrates were obtained by filtering through a 0.45 µm membrane filter. The equilibrium concentration of Cr(VI) was measured by ICP- where q e is the adsorbing amount of Cr(VI) on the MnFe 2 O 4 adsorbent (mg·g −1 ); C o and C e are the initial and equilibrium concentrations of the Cr(VI) solution (mg·L −1 ), respectively; V is the volume of the Cr(VI)-bearing solution; and m is the mass of the MnFe 2 O 4 adsorbent. The pH effect tests were conducted within the pH range of 1-10. The pH of the solution was adjusted by 0.1 M HCl and 0.1 M NaOH. The initial Cr(VI) concentration and MnFe 2 O 4 dosage were controlled at 10 mg/L and 3 g/L, respectively. The adsorption kinetics of Cr(VI) on the MnFe 2 O 4 were investigated by changing the oscillation time of the Cr(VI) adsorption. The oscillation times were changed from 5 to 720 min in this research, and the initial Cr(VI) concentration was controlled at 10 mg/L. The Cr(VI) adsorption data at the various oscillation time were analyzed according to the following two models, the pseudo-first-order equation and the pseudo-second-order equation. The linear forms of these two models are listed below [30]: where q e and q t are the adsorption amounts of Cr(VI) at equilibrium and a certain time t (mg·g −1 ), respectively; k 1 and k 2 are the rate constants of the pseudo-first-order equation and the pseudo-second-order equation, respectively. The isothermal adsorption experiments were conducted with the initial Cr(VI) concentration range of 1-100 mg/L. The obtained equilibrium concentrations were fitted via the Langmuir and Freundlich models. The linearized Langmuir and Freundlich isotherm models are expressed as the following equations [31]: where q e (mg·g −1 ) is the amount of Cr(VI) adsorbed per weight unit of the MnFe 2 O 4 adsorbent after equilibrium; C e (mg·L −1 ) is the equilibrium concentration of Cr(VI) in the solution; q m (mg·g −1 ) is the maximum adsorption capacity of the adsorption; K L is a constant related to the heat of adsorption (L·mg −1 ); K F (mg 1−n ·L n ·g −1 ) is a constant related to the extent of adsorption; and 1/n is related to the intensity of adsorption.

Characterization
The X-ray powder diffraction (XRD) patterns of the synthetic MnFe 2 O 4 before and after nano-scale grinding were detected by a diffractometer with Cu Kα radiation (RIGAKU D/Max 2500, Tokyo, Japan). The tube current and voltage were 250 mA and 40 kV, the scanning range was 10-70 • , and the step size was 0.02 • . A vibrating sample magnetometer was adopted to measure the magnetism properties of the MnFe 2 O 4 adsorbent (PPMS Dynacool, Quantum Design, San Diego, CA, USA). The element composition and binding energies of atomic orbital on the MnFe 2 O 4 surface before and after adsorbing Cr(VI) were investigated by the XPS test (XPS, Mono 650 µm, 200 W, Al Kα, pass energy 20 eV). The scanning step size was 1.0 eV for the full scan and 0.1 eV for the narrow scan. Zeta potential measurements were conducted using a zetasizer under a wide pH range from 3 to 10 at 25 • C, and the pH value was adjusted using 0.1 M HCl and NaOH. The particle size distribution was characterized by a laser particle analyzer (Malvern Mastersizer 2000, Malvern, UK). The specific surface area and pore volume of the MnFe 2 O 4 adsorbent were measured by nitrogen adsorption-desorption on a specific surface area with a pore analyzer.
The slab supercell that consists of a (2 × 2) unit cell of MnFe 2 O 4 (100) was constructed as the substrate. A vacuum of 20 Å was added to eliminate the interaction between two periodic slabs and H atoms were attached to O on the surface in acidic conditions (MnFe 2 O 4 -H (100)). The calculations were conducted under density functional theory, and the Vienna ab initio simulation (VASP) code 1, 2 was used to implement. The initial adsorption site was predicted by the simulated annealing method for locating a good approximation to the global minimum. Projector Augmented Wave (PAW) pseudopotentials are used with a cutoff energy of 400 eV for plane wave expansions. The GGA-PBE5 was used as an exchange-correlation function. The atomic relaxation was performed with a k-mesh of 1 × 1 × 1 until the force on each atom is less than 0.02 eV/Å. Moreover, van der Waals forces between the adsorbate and substrate were also corrected by the DFT-D26 dispersion corrections. After relaxation, a gamma-center k-mesh 2 × 2 × 1 was adopted for the ground state total energy calculation. The adsorption energy (Eads) was evaluated by: where E(total), E(sub), and E(ads) are the total energy of the where A i and A i 0 are the self-consistent charge and the atomic charge of atom i, respectively. n is the total number of studied adsorbates. The negative value represents electron dissipation to the substrate and positive ∆chg represents electron accumulation on the adsorbate.

Characterization of MnFe 2 O 4 Adsorbent
The effects of nano-scale grinding on the crystallinity and crystal integrity of MnFe 2 O 4 were investigated by X-ray diffraction patterns. Figure 1a shows that the diffraction peaks at 29.95 • , 35.33 • , 56.42 • , and 61.94 • were considered to belong to the 220, 311, 511, and 440 crystal faces of MnFe 2 O 4 [35], respectively. After grinding, the crystallinity of MnFe 2 O 4 worsened and certain diffraction peaks disappeared. This result indicated that the strong mechanical grinding would damage the MnFe 2 O 4 crystal to some extent. The magnetization shown in Figure 1d exhibited that the magnetization intensity of MnFe 2 O 4 decreased sharply from 63.38 to 23.80 emu/g after grinding. The main reason is that the magnetism of MnFe 2 O 4 is the macroscopic manifestation of the magnetic properties of many tiny magnetic domains inside it, and as the particle size decreases, the number of internal magnetic domains decreases, resulting in a weakening in the magnetism of nano-size MnFe 2 O 4 particles. Fortunately, the magnetic saturation curves displayed in Figure 1d manifested that the saturation magnetic of nano-MnFe 2 O 4 still retained 20 emu/g, illustrating that the MnFe 2 O 4 adsorbent still had good magnetism and could be recovered by magnetic separation. The BET test manifested that the specific surface area of the MnFe 2 O 4 adsorbent could achieve 31.997 m 2 /g after nano-scale grinding. The zeta potential results exhibited in Figure 1c

Effect of the Initial pH
In this study, the Cr(VI) adsorption tests were conducted under pH values of 1 to 10. As shown in Figure 2a, the Cr(VI) removal first increased within the pH range of 1-3 and then continued to decline with an increasing pH. The Cr(VI) reduction in the acidic system conformed to the following equations [36,37]:

Effect of the Initial pH
In this study, the Cr(VI) adsorption tests were conducted under pH values of 1 to 10. As shown in Figure 2a, the Cr(VI) removal first increased within the pH range of 1-3 and then continued to decline with an increasing pH. The Cr(VI) reduction in the acidic system conformed to the following equations [36,37]: The hydrogen ions participated in the Cr(VI) reduction, and this indicated that the acidic system is favorable for Cr(VI) reduction. Furthermore, according to the zeta potential tests of MnFe 2 O 4 , the electrostatic attraction between the MnFe 2 O 4 and chromate is more reliable in a low-pH environment. However, when in a very strong acidic system (pH = 1 and 2), the reduction product Cr(III) could not be deposited on the MnFe 2 O 4 surface by forming the Cr(OH) x compound due to the neutralization. This is the main reason that the Cr(VI) removal under pH = 1 and 2 is lower than that under pH = 3. Materials 2023, 16, x FOR PEER REVIEW 6 of 15 The hydrogen ions participated in the Cr(VI) reduction, and this indicated that the acidic system is favorable for Cr(VI) reduction. Furthermore, according to the zeta potential tests of MnFe2O4, the electrostatic attraction between the MnFe2O4 and chromate is more reliable in a low-pH environment. However, when in a very strong acidic system (pH = 1 and 2), the reduction product Cr(III) could not be deposited on the MnFe2O4 surface by forming the Cr(OH)x compound due to the neutralization. This is the main reason that the Cr(VI) removal under pH = 1 and 2 is lower than that under pH = 3.

Adsorption Kinetics
The time needed to reach the adsorption equilibrium was a significant value that decided the Cr(VI) removal efficiency. In this experiment, the Cr(VI) adsorption kinetics was determined by measuring the adsorption amount of Cr(VI) at various times. The adsorption kinetic curve of Cr(VI) is displayed in Figure 3a, and the fitting curves of the pseudofirst-order and pseudo-second-order equations are displayed in Figure 3b,c, respectively. The fitting parameters of the pseudo-first-order equation and pseudo-second-order models are given in Table 1.

Adsorption Kinetics
The time needed to reach the adsorption equilibrium was a significant value that decided the Cr(VI) removal efficiency. In this experiment, the Cr(VI) adsorption kinetics was determined by measuring the adsorption amount of Cr(VI) at various times. The adsorption kinetic curve of Cr(VI) is displayed in Figure 3a, and the fitting curves of the pseudo-first-order and pseudo-second-order equations are displayed in Figure 3b,c, respectively. The fitting parameters of the pseudo-first-order equation and pseudo-secondorder models are given in Table 1.    The adsorption kinetic curve of Cr(VI) is displayed in Figure 3a, and the fitting curves of the pseudo-first-order and pseudo-second-order equations are displayed in Figure 3b,c, respectively. The fitting parameters of the pseudo-first-order equation and pseudo-secondorder models are given in Table 1. The correlation coefficient (R 2 ) of these two models indicated that the pseudo-second-order model was more suitable to describe the Cr(VI) adsorption kinetics on MnFe 2 O 4 adsorbent. The result showed that the Cr(VI) adsorption achieved equilibrium in 480 min, and further extending the adsorption time could hardly affect the adsorption amount of Cr(VI). The above results manifested that Cr(VI) adsorption on the MnFe 2 O 4 adsorbent belonged to chemisorption [38].

Adsorption Isotherm
As many papers reported, the adsorption isotherm model fitting was a common tool to understand the adsorption type between the adsorbent and adsorbate [39]. The Langmuir and Freundlich isotherm models fitting lines and parameters of Cr(VI) adsorption on the MnFe 2 O 4 surface are shown in Figure 4 and Table 2, respectively.  The adsorption kinetic curve of Cr(VI) is displayed in Figure 3a, and the fitting curves of the pseudo-first-order and pseudo-second-order equations are displayed in Figure 3b,c, respectively. The fitting parameters of the pseudo-first-order equation and pseudo-second-order models are given in Table 1. The correlation coefficient (R 2 ) of these two models indicated that the pseudo-second-order model was more suitable to describe the Cr(VI) adsorption kinetics on MnFe2O4 adsorbent. The result showed that the Cr(VI) adsorption achieved equilibrium in 480 min, and further extending the adsorption time could hardly affect the adsorption amount of Cr(VI). The above results manifested that Cr(VI) adsorption on the MnFe2O4 adsorbent belonged to chemisorption [38].

Adsorption Isotherm
As many papers reported, the adsorption isotherm model fitting was a common tool to understand the adsorption type between the adsorbent and adsorbate [39]. The Langmuir and Freundlich isotherm models fitting lines and parameters of Cr(VI) adsorption on the MnFe2O4 surface are shown in Figure 4 and Table 2, respectively.   The results illustrated that the max adsorption quantity of Cr(VI) is 5.813 mg/g. The correlation coefficient (R 2 ) of the Langmuir model is higher than that of the Freundlich model, showing that the Cr(VI) adsorption on the MnFe 2 O 4 surface conformed to the Langmuir model. It was likely that the Cr(VI) adsorption tended to be monolayer adsorption [40]. Compared with other adsorbents, the adsorption capacity of the adsorbent in this study is lower. The main reason is that there are some impurities when using natural minerals for synthesis. What is more, the solid-phase synthetic MnFe 2 O 4 particles are coarse, while after grinding, their specific surface area is still much smaller than that of the liquid-phase synthetic MnFe 2 O 4 particles.

Cr(VI) Adsorption Analyses
The Cr(VI) removal process on the MnFe 2 O 4 surface primarily included the adsorption and reduction process. The Cr(VI) could only be reduced after being adsorbed on the MnFe 2 O 4 surface. As much of the literature has reported, the DFT calculation is a powerful tool to investigate the interaction reaction between two kinds of molecules [41,42]. In this study, the MnFe 2 O 4 (100) surface, a typical low-index surface, was used to investigate the role of Mn in the Cr(VI) adsorption process. Figure 5a shows model, showing that the Cr(VI) adsorption on the MnFe2O4 surface conformed to the Langmuir model. It was likely that the Cr(VI) adsorption tended to be monolayer adsorption [40]. Compared with other adsorbents, the adsorption capacity of the adsorbent in this study is lower. The main reason is that there are some impurities when using natural minerals for synthesis. What is more, the solid-phase synthetic MnFe2O4 particles are coarse, while after grinding, their specific surface area is still much smaller than that of the liquid-phase synthetic MnFe2O4 particles.

Cr(VI) Adsorption Analyses
The Cr(VI) removal process on the MnFe2O4 surface primarily included the adsorption and reduction process. The Cr(VI) could only be reduced after being adsorbed on the MnFe2O4 surface. As much of the literature has reported, the DFT calculation is a powerful tool to investigate the interaction reaction between two kinds of molecules [41,42]. In this study, the MnFe2O4 (100) surface, a typical low-index surface, was used to investigate the role of Mn in the Cr(VI) adsorption process. Figure 5a shows the optimum structure of spinel-type MnFe2O4, the bond distances of Fe-Fe, Fe-Mn, Fe-O, and Mn-O are 3.011 Å, 3.530 Å, 4.760 Å, and 3.687 Å, respectively. However, according to the stern double-layer theory [43], when MnFe2O4 is placed in an aqueous solution, a layer of hydrogen or hydroxyl groups adsorbed on the surface of MnFe2O4 molecules exists, which is called the "positioning ion" [44]. In this study, the Cr(VI) adsorption process primarily occurred in the acidic solution, and the MnFe2O4-H molecule was dominant compared with MnFe2O4-OH. We built up a new structure of MnFe2O4-H by causing the hydrogen ion to adsorb on the surface of the MnFe2O4 molecule. The optimum structure of MnFe2O4-H is displayed in Figure 5b. The bond distances of Fe-H, Mn-H, and O-H were 1.500 Å, 1.610 Å, and 1.100 Å, respectively.   Table 3.
Next, the groups of HcrO4 − , CrO4 2− , and Cr2O7 2− were selected to conduct DFT calculations. The optimum structures of MnFe2O4-H-HcrO4 − , MnFe2O4-H-CrO4 2− , and MnFe2O4-H-Cr2O7 2− are shown in Figure 6. Moreover, the adsorption energies and the bond distances of HcrO4 − , CrO4 2− , and Cr2O7 2− on MnFe2O4-H (100) are listed in Table 3.   [30] As shown in Table 4, the adsorption energies of HcrO4 − , CrO4 2− , and Cr2O7 2− on the MnFe2O4-H surface were −215.2, −200.7, and −116.7 kJ/mol, respectively. The adsorption energy is an indicator of system stability calculated by subtracting the system's energy before and after adsorption [45]. In general, the higher the adsorption energy is, the more stable the system after adsorption is [46]. Therefore, it can be easily observed that HcrO4 − was the most easily adsorbed on the MnFe2O4 adsorbent. Moreover, it can also be observed that from the bond distances displayed in Table 3, in the MnFe2O4-H-HcrO4 − system, the O atom of HcrO4 − formed a chemical bond with the Mn atoms of MnFe2O4, providing high adsorption energy. As we know, the adsorption process is an essential link to the subsequent reduction [47,48]. The DFT calculation indicated that the HcrO4 − could be efficiently adsorbed by the Mn atoms of the MnFe2O4 via the chemical bonds. This efficient adsorption of Cr(VI) created a favorable environment for the subsequent Cr(VI) reduction by the Mn(II). Subsequently, the charge density distribution was calculated by VASP software to clarify the chemical bond type between the chromate and the MnFe2O4-H.   [30] As shown in Table 4, the adsorption energies of HcrO 4 − , CrO 4 2− , and Cr 2 O 7 2− on the MnFe 2 O 4 -H surface were −215.2, −200.7, and −116.7 kJ/mol, respectively. The adsorption energy is an indicator of system stability calculated by subtracting the system's energy before and after adsorption [45]. In general, the higher the adsorption energy is, the more stable the system after adsorption is [46]. Therefore, it can be easily observed that HcrO 4 − was the most easily adsorbed on the MnFe 2 O 4 adsorbent. Moreover, it can also be observed that from the bond distances displayed in Table 3, in the MnFe 2 O 4 −H−HcrO 4 − system, the O atom of HcrO 4 − formed a chemical bond with the Mn atoms of MnFe 2 O 4 , providing high adsorption energy. As we know, the adsorption process is an essential link to the subsequent reduction [47,48]. The DFT calculation indicated that the HcrO 4 − could be efficiently adsorbed by the Mn atoms of the MnFe 2 O 4 via the chemical bonds. This efficient adsorption of Cr(VI) created a favorable environment for the subsequent Cr(VI) reduction by the Mn(II). Subsequently, the charge density distribution was calculated by VASP software to clarify the chemical bond type between the chromate and the MnFe 2 O 4 -H.
In general, the charges would be distributed at both ends of the bond in the electrostatic interaction or ionic bond [49]. In contrast, for the covalent bond, the charge would be distributed in the middle of the bond [50]. The charge density distribution shown in Figure 7 indicates In general, the charges would be distributed at both ends of the bond in the electrostatic interaction or ionic bond [49]. In contrast, for the covalent bond, the charge would be distributed in the middle of the bond [50]. The charge density distribution shown in Figure 7 indicates that the charges of the Mn/Fe-O bonds in the MnFe2O4-H-HcrO4 − system were distributed at the ends of the bonds. This implied that the Mn/Fe-O bond belonged to the ionic bond. Similarly, CrO4 2− was also adsorbed on the Mn atom of the MnFe2O4 adsorbent via the ionic bond, while during the Cr2O7 2− adsorption process, the bonding had little effect on the charge distribution of the two molecules. Therefore, it could be suggested that the adsorption of Cr2O7 2− on the MnFe2O4 surface is a type of physical adsorption via electrostatic attraction.

Cr(VI) Reduction Analyses
The chemical states on the MnFe2O4 surface before and after the Cr(VI) adsorption were investigated by the XPS test to explain the reduction mechanisms of Cr(VI). The XPS full-scanning spectra of the original MnFe2O4 adsorbent and MnFe2O4 adsorbed Cr(VI) under pH = 3 and 7 are shown in Figure 8. The main atom contents on the MnFe2O4 adsorbent surface before and after adsorbing Cr(VI) are displayed in Table 5.

Cr(VI) Reduction Analyses
The chemical states on the MnFe 2 O 4 surface before and after the Cr(VI) adsorption were investigated by the XPS test to explain the reduction mechanisms of Cr(VI). The XPS full-scanning spectra of the original MnFe 2 O 4 adsorbent and MnFe 2 O 4 adsorbed Cr(VI) under pH = 3 and 7 are shown in Figure 8. The main atom contents on the MnFe 2 O 4 adsorbent surface before and after adsorbing Cr(VI) are displayed in Table 5. In general, the charges would be distributed at both ends of the bond in the electrostatic interaction or ionic bond [49]. In contrast, for the covalent bond, the charge would be distributed in the middle of the bond [50]. The charge density distribution shown in Figure 7 indicates that the charges of the Mn/Fe-O bonds in the MnFe2O4-H-HcrO4 − system were distributed at the ends of the bonds. This implied that the Mn/Fe-O bond belonged to the ionic bond. Similarly, CrO4 2− was also adsorbed on the Mn atom of the MnFe2O4 adsorbent via the ionic bond, while during the Cr2O7 2− adsorption process, the bonding had little effect on the charge distribution of the two molecules. Therefore, it could be suggested that the adsorption of Cr2O7 2− on the MnFe2O4 surface is a type of physical adsorption via electrostatic attraction.

Cr(VI) Reduction Analyses
The chemical states on the MnFe2O4 surface before and after the Cr(VI) adsorption were investigated by the XPS test to explain the reduction mechanisms of Cr(VI). The XPS full-scanning spectra of the original MnFe2O4 adsorbent and MnFe2O4 adsorbed Cr(VI) under pH = 3 and 7 are shown in Figure 8. The main atom contents on the MnFe2O4 adsorbent surface before and after adsorbing Cr(VI) are displayed in Table 5.   As shown in Figure 8, there were four primary XPS peaks, 284.8, 532.8, 638.7, and 710.9 eV, belonging to the C1s, O1s, Mn2p, and Fe2p atomic orbitals, respectively. After the Cr(VI) adsorption, the Cr2p peak at 573.6 eV emerged. The results of atom contents showed that the Cr atom content of MnFe 2 O 4 -Cr at pH = 3 was significantly higher than that of MnFe 2 O 4 -Cr at pH = 7 due to the high adsorption quantity of the Cr(VI) under an acidic environment. It is worth noting that the Mn2p atom content declined after Cr(VI) adsorption, implying that the Mn atom in MnFe 2 O 4 particles was involved in the Cr(VI) reduction reaction. It was inferred that the divalent Mn was oxidized to trivalent Mn by Cr(VI) and entered the solution, resulting in a decrease in the atom content of Mn on the surface of the MnFe 2 O 4 adsorbent.
To further investigate the adsorption mechanisms of Cr(VI), the high-resolution spectra of Mn2p and Cr2p were conducted. From the Mn2p spectrum of the original MnFe 2 O 4 adsorbent shown in Figure 9a, there is only one main peak at 641.5 eV of Mn-O considered to belong to the MnFe 2 O 4 . However, after the Cr(VI) adsorption, the peak of MnO 2 emerged, indicating that the Mn(II) on the MnFe 2 O 4 surface might be oxidized to tetravalence Mn by Cr(VI). In fact, Mn(II) was oxidized to Mn(III) first by Cr(VI). However, the Mn(III) was extremely unstable in an acidic solution and was prone to the disproportionation reaction to form Mn(II) and MnO 2 . The resulting Mn(II) could continue to participate in the Cr(VI) reduction, and MnO 2 was deposited on the MnFe 2 O 4 surface. The XPS results proved that Mn(II) acted as a reductant in the Cr(VI) reduction process.  As shown in Figure 8, there were four primary XPS peaks, 284.8, 532.8, 638.7, and 710.9 eV, belonging to the C1s, O1s, Mn2p, and Fe2p atomic orbitals, respectively. After the Cr(VI) adsorption, the Cr2p peak at 573.6 eV emerged. The results of atom contents showed that the Cr atom content of MnFe2O4-Cr at pH = 3 was significantly higher than that of MnFe2O4-Cr at pH = 7 due to the high adsorption quantity of the Cr(VI) under an acidic environment. It is worth noting that the Mn2p atom content declined after Cr(VI) adsorption, implying that the Mn atom in MnFe2O4 particles was involved in the Cr(VI) reduction reaction. It was inferred that the divalent Mn was oxidized to trivalent Mn by Cr(VI) and entered the solution, resulting in a decrease in the atom content of Mn on the surface of the MnFe2O4 adsorbent.
To further investigate the adsorption mechanisms of Cr(VI), the high-resolution spectra of Mn2p and Cr2p were conducted. From the Mn2p spectrum of the original MnFe2O4 adsorbent shown in Figure 9a, there is only one main peak at 641.5 eV of Mn-O considered to belong to the MnFe2O4. However, after the Cr(VI) adsorption, the peak of MnO2 emerged, indicating that the Mn(II) on the MnFe2O4 surface might be oxidized to tetravalence Mn by Cr(VI). In fact, Mn(II) was oxidized to Mn(III) first by Cr(VI). However, the Mn(III) was extremely unstable in an acidic solution and was prone to the disproportionation reaction to form Mn(II) and MnO2. The resulting Mn(II) could continue to participate in the Cr(VI) reduction, and MnO2 was deposited on the MnFe2O4 surface. The XPS results proved that Mn(II) acted as a reductant in the Cr(VI) reduction process.  The high-resolution Cr2p spectra of MnFe 2 O 4 -Cr depicted in Figure 10 could be resolved into three individual peaks: The groups of HCrO 4 − /CrO 4 2− (577.5 and 587.2 eV), Cr 3+ (576.2 and 585.9 eV), and Cr 2 O 7 2− (578.1 and 587.8 eV) [51]. Under the pH = 3 condition, the XPS spectrum showed that the main peak of Cr2p on the MnFe 2 O 4 -Cr surface is Cr(III). When the pH value increased to 7, the dominant species of Cr is Cr(VI). It could be seen that from the XPS results, the Cr adsorbed on the MnFe 2 O 4 surface, mainly in the form of Cr(III) under pH = 3. This result confirmed that most of the Cr(VI) was reduced to Cr(III) by Mn(II) and formed the Cr(OH) 3  The high-resolution Cr2p spectra of MnFe2O4-Cr depicted in Figure 10 could be resolved into three individual peaks: The groups of HCrO4 − /CrO4 2− (577.5 and 587.2 eV), Cr 3+ (576.2 and 585.9 eV), and Cr2O7 2− (578.1 and 587.8 eV) [51]. Under the pH = 3 condition, the XPS spectrum showed that the main peak of Cr2p on the MnFe2O4-Cr surface is Cr(III). When the pH value increased to 7, the dominant species of Cr is Cr(VI). It could be seen that from the XPS results, the Cr adsorbed on the MnFe2O4 surface, mainly in the form of Cr(III) under pH = 3. This result confirmed that most of the Cr(VI) was reduced to Cr(III) by Mn(II) and formed the Cr(OH)3 compound on the MnFe2O4 surface. The rest of the Cr adsorbed onto the MnFe2O4 surface in the form of Cr(VI) via electrostatic adsorption. According to the XPS results, we could speculate that the Cr(VI) was adsorbed on the surface of the MnFe2O4 adsorbent in the form of chromate. Subsequently, most of the chromate was reduced to Cr(III) by Mn(II).

The Roles of Mn in Cr(VI) Adsorption and Reduction Process
According to the above results, it could be inferred that the roles of Mn in the Cr(VI) removal process could be divided into two parts, providing the adsorbing sites and being a reductant. The Cr(VI) removal process is explained in Figure 11. In the Cr(VI) adsorption process, the Mn atoms of the MnFe2O4 could form a firm ionic bond with the O atoms of HCrO4 − and CrO4 2− , thus providing the adsorbing sites for

The Roles of Mn in Cr(VI) Adsorption and Reduction Process
According to the above results, it could be inferred that the roles of Mn in the Cr(VI) removal process could be divided into two parts, providing the adsorbing sites and being a reductant. The Cr(VI) removal process is explained in Figure 11. The high-resolution Cr2p spectra of MnFe2O4-Cr depicted in Figure 10 could be resolved into three individual peaks: The groups of HCrO4 − /CrO4 2− (577.5 and 587.2 eV), Cr 3+ (576.2 and 585.9 eV), and Cr2O7 2− (578.1 and 587.8 eV) [51]. Under the pH = 3 condition, the XPS spectrum showed that the main peak of Cr2p on the MnFe2O4-Cr surface is Cr(III). When the pH value increased to 7, the dominant species of Cr is Cr(VI). It could be seen that from the XPS results, the Cr adsorbed on the MnFe2O4 surface, mainly in the form of Cr(III) under pH = 3. This result confirmed that most of the Cr(VI) was reduced to Cr(III) by Mn(II) and formed the Cr(OH)3 compound on the MnFe2O4 surface. The rest of the Cr adsorbed onto the MnFe2O4 surface in the form of Cr(VI) via electrostatic adsorption. According to the XPS results, we could speculate that the Cr(VI) was adsorbed on the surface of the MnFe2O4 adsorbent in the form of chromate. Subsequently, most of the chromate was reduced to Cr(III) by Mn(II).

The Roles of Mn in Cr(VI) Adsorption and Reduction Process
According to the above results, it could be inferred that the roles of Mn in the Cr(VI) removal process could be divided into two parts, providing the adsorbing sites and being a reductant. The Cr(VI) removal process is explained in Figure 11. In the Cr(VI) adsorption process, the Mn atoms of the MnFe2O4 could form a firm ionic bond with the O atoms of HCrO4 − and CrO4 2− , thus providing the adsorbing sites for In the Cr(VI) adsorption process, the Mn atoms of the MnFe 2 O 4 could form a firm ionic bond with the O atoms of HCrO 4 − and CrO 4 2− , thus providing the adsorbing sites for the Cr(VI) adsorption. Then, the adsorbed Cr(VI) was reduced by the Mn(II) dissolved on the MnFe 2 O 4 surface. In the reduction process, the dissolved Mn(II) acted as a reductant. The Mn(II) first oxidized to Mn(III) by Cr(VI). The unstable Mn(III) would cause the disproportionated reaction and formed Mn(II) and MnO 2 in the aqueous system [52]. The reaction link of the Cr(VI) reduction by Mn(II) can be described by the following equations: Cr 3+ + 3OH − → Cr(OH) 3 (11) 2Mn 3+ + 2H 2 O → Mn 2+ + MnO 2 + 4H + After Cr(VI) reduction, the obtained Cr(III) was complexed with the hydroxy groups to form the inner-sphere complex [53]. The Cr(OH) 3 precipitate was deposited on the MnFe 2 O 4 surface and separated by magnetic separation.

Conclusions
The Cr(VI) could be removed efficiently by the MnFe 2 O 4 adsorbent. The adsorption process belonged to chemisorption. The mechanism research showed that the roles of Mn in the Cr(VI) removal process could be divided into two parts, providing adsorbing sites and being a reductant. The adsorption mechanisms of the Cr(VI) on the MnFe 2