Highly Efficient and Reusable Montmorillonite/Fe3O4/Humic Acid Nanocomposites for Simultaneous Removal of Cr(VI) and Aniline

Recyclable nanomaterials are in great need to develop clean technology for applications in the removal of water contaminants. In this work, easily separable montmorillonite/Fe3O4/humic acid (MFH) nanocomposites were fabricated through a facile hydrothermal route. It was found the adsorption ability and stability of MFH was significantly enhanced due to the synergistic effects between montmorillonite, Fe3O4 nanoparticles and humic acid. The MFH nanocomposites are highly efficient and recyclable as they can remove at least 82.3% of Cr(VI) and 95.1% of aniline in six consecutive runs. The adsorption mechanism was investigated by analyzing the kinetic parameters of pseudo first-order, pseudo second-order, and intraparticle diffusion models and describing the equilibrium isotherms of Langmuir and Freundlich models. Results indicated different adsorption mechanisms of Cr(VI) and aniline by MFH. The readily synthesized MFH nanocomposites can act as effective and practical materials for environmental applications.


Introduction
As one of the most toxic heavy metals and common contaminants in wastewater, Cr can be found in wastewater from painting, tanning, and metallurgical industries [1,2]. It generally exists in the form of Cr(III) or Cr(VI) with the latter hundreds of times more toxic than the former. Cr(VI) is highly hazardous to organs, such as skin, liver, and lung, and even mutagenic to organisms [3]. In addition, it causes liver damage, pulmonary congestion, and skin irritation. Converting Cr(VI) to Cr(III) is a common and efficient strategy to solve the problem of Cr(VI). Meanwhile, aniline is a group of emerging contaminants in wastewater from many industrial processes, such as painting, pharmacy, and rubber [4]. Aniline can greatly harm human and aquatic life as it is highly toxic and it can accumulate in the environment [5]. It can enter the body through inhalation, digestion and skin absorption, convert hemoglobin to methemoglobin and result in cyanosis [4]. In recent years, growing attention has been focused on aniline removal strategies. In practical situations, heavy metals like Cr(VI) and aniline can coexist in industrial wastewater [6]. Thus, it is significant to achieve simultaneous removal of Cr(VI) and aniline from industrial wastewater to prevent their damage to the ecosystem and human health.

Batch Adsorption Experiments
One-hundred mL water with Cr(VI) (0-40.0 mg·L −1 ) and aniline (0-100.0 mg·L −1 ) at pH = 3.0-11.0 was added into in 250 mL flasks. The value of pH was controlled by adding 0.05 M HCl or NaOH. The solution was agitated at 400 rpm and kept at 25 °C. 0.05-0.40 g·L −1 adsorbents were added in the flasks. Every 5 min during the experiment process (100 min), samples (1 mL) were taken out for UV analysis. After the nanocomposites were collected using a magnet, they were washed with 50% ethanol in water (v/v) several times and then dried. To evaluate the reusability, they were reused in the next five runs. Removal efficiency and equilibrium adsorption capacity qe (mg·g −1 ) were obtained using Equations (1) and (2). All experiments were performed in triplicate.

Batch Adsorption Experiments
One-hundred mL water with Cr(VI) (0-40.0 mg·L −1 ) and aniline (0-100.0 mg·L −1 ) at pH = 3.0-11.0 was added into in 250 mL flasks. The value of pH was controlled by adding 0.05 M HCl or NaOH. The solution was agitated at 400 rpm and kept at 25 • C. 0.05-0.40 g·L −1 adsorbents were added in the flasks. Every 5 min during the experiment process (100 min), samples (1 mL) were taken out for UV analysis. After the nanocomposites were collected using a magnet, they were washed with 50% ethanol in water (v/v) several times and then dried. To evaluate the reusability, they were reused in the next five runs. Removal efficiency and equilibrium adsorption capacity q e (mg·g −1 ) were obtained using Equations (1) and (2). All experiments were performed in triplicate.

Characterization
As shown in Figure 1a, bare Fe 3 O 4 NPs tended to aggregate, which would reduce specific surface area, dispersity, mobility, and adsorption ability. Following the synthesis procedure in Section 2.2, HA-coated Fe 3 O 4 NPs were dispersed onto Mt layers to overcome this problem. After sonication, Mt of fewer layers (See Figure 1b) can be obtained, which possessed higher specific surface area than normal Mt powder. As both Mt and HA were negatively charged in the form of Na-Mt and humic acid sodium salt respectively, HA was less likely to compete with  Figure 1e. Therefore, the MFH nanocomposites investigated in this paper were prepared according to the optimum mass ratio of 1:4 (Fe 3 O 4 :Mt). The S BET and V tp of as-prepared MFH were increased significantly compared with those of Mt and bare Fe 3 O 4 NPs. MFH enriched the pollutants by adsorption and significantly improved the removal efficiency [34,35].  Figure 1b) can be obtained, which possessed higher specific surface area than normal Mt powder. As both Mt and HA were negatively charged in the form of Na-Mt and humic acid sodium salt respectively, HA was less likely to compete with Fe3O4 NPs to bind with Mt. The electrostatic attraction between Fe3O4 NPs and Mt, and between Fe3O4 NPs and HA, along with the electrostatic repulsion between Mt and HA, functioned as the main interaction forces to disperse HA-coated Fe3O4 NPs onto Mt and maintain the structure stability. The ratio of HA-coated Fe3O4 NPs loaded on Mt was controlled by varying the ratio between Mt and Fe sources.
In Figure 1c-f, different mass ratios of HA-coated Fe3O4 NPs were loaded onto Mt. According to Table 1, MFH (Fe3O4:Mt = 1:4) had the highest specific surface area (SBET) among MFH of different mass ratios. With the increase of Fe3O4:Mt ratio in MFH, the total pore volume (Vtp) of MFH increased, and the average pore diameter (Dp) narrowed (Fe3O4:Mt < 1:4) and then widened (Fe3O4:Mt > 1:4). Compared with MFH in Figure 1c,d, the larger SBET and Vtp of MFH in Figure 1e was due to more Fe3O4 NPs loaded onto Mt. However, when Mt was overloaded with Fe3O4 NPs, the aggregation of Fe3O4 NPs would result in loss of SBET and increase of Dp compared with MFH in Figure 1e. Therefore, the MFH nanocomposites investigated in this paper were prepared according to the optimum mass ratio of 1:4 (Fe3O4:Mt). The SBET and Vtp of as-prepared MFH were increased significantly compared with those of Mt and bare Fe3O4 NPs. MFH enriched the pollutants by adsorption and significantly improved the removal efficiency [34,35].    in MFH. It could be explained by that as the insertion of larger hydrolyzed iron species replaced Na + in initial Mt, the planar stress increased and thus resulted in the shift to lower angle in XRD pattern [36,37]. Note that as an organic component HA does not diffract and thus cannot be distinguished by XRD. The XRD patterns of Mt, Fe3O4 NPs and MFH are shown in Figure 2. For the XRD pattern of Mt, there were four diffraction peaks at 2θ = 20.4°, 35.9°, 54.1° and 62.2°. As for bare Fe3O4 NPs, the characteristic peaks at 2θ = 30.2°, 35.8°, 43.3°, 52.6°, 57.2°, and 62.6° can be found. In the XRD pattern of MFH, the typical diffraction peaks of both Mt and Fe3O4 NPs can be seen, demonstrating their high crystallinity and the successful dispersion of Fe3O4 NPs onto Mt. Particularly, the peak of Mt at 2θ = 20.4° showed a slight shift to 2θ = 19.6° in MFH. It could be explained by that as the insertion of larger hydrolyzed iron species replaced Na + in initial Mt, the planar stress increased and thus resulted in the shift to lower angle in XRD pattern [36,37]. Note that as an organic component HA does not diffract and thus cannot be distinguished by XRD. To clarify the coating of HA on Fe3O4 NPs, FTIR was carried out. In Figure 3a, the band at ~1030 cm −1 was attributed to the stretching vibration absorption of Si-O of Mt [38]. In the Fe3O4 curve (Figure 3b), the main absorption was observed at around 582 cm −1 , corresponding to Fe-O bending vibration [39]. In Figure 3c, the typical C=O stretching of carboxylate in HA was located at ~1680 cm −1 , and the band (~1350 cm −1 ) can be attributed to -CH2-scissoring [40]. In Figure 3d, all the characteristic bands of Mt, Fe3O4 NPs and HA mentioned above can be seen, verifying the formation of MFH. The C=O stretching in MFH (featured at ~1550 cm −1 ) showed blue shift compared with that of free HA (~1680 cm −1 ). This change can also provide evidence that the carboxylate anions of HA interacted with the surface of Fe3O4 NPs. To clarify the coating of HA on Fe 3 O 4 NPs, FTIR was carried out. In Figure 3a, the band at~1030 cm −1 was attributed to the stretching vibration absorption of Si-O of Mt [38]. In the Fe 3 O 4 curve (Figure 3b), the main absorption was observed at around 582 cm −1 , corresponding to Fe-O bending vibration [39]. In Figure 3c, the typical C=O stretching of carboxylate in HA was located at~1680 cm −1 , and the band (~1350 cm −1 ) can be attributed to -CH 2 -scissoring [40]. In Figure 3d, all the characteristic bands of Mt, Fe 3 O 4 NPs and HA mentioned above can be seen, verifying the formation of MFH. The C=O stretching in MFH (featured at~1550 cm −1 ) showed blue shift compared with that of free HA (~1680 cm −1 ). This change can also provide evidence that the carboxylate anions of HA interacted with the surface of Fe 3 O 4 NPs.
~1680 cm −1 , and the band (~1350 cm −1 ) can be attributed to -CH2-scissoring [40]. In Figure 3d, all the characteristic bands of Mt, Fe3O4 NPs and HA mentioned above can be seen, verifying the formation of MFH. The C=O stretching in MFH (featured at ~1550 cm −1 ) showed blue shift compared with that of free HA (~1680 cm −1 ). This change can also provide evidence that the carboxylate anions of HA interacted with the surface of Fe3O4 NPs.   Figure 4g, the Cr 2p 3/2 peak of MFH after removal experiments can be split into two peaks at 574.3 and 573.2 eV (the characteristic peaks of Cr(VI) and Cr(III)), respectively, suggesting that both Cr(VI) and Cr(III) existed on MFH surface [41,42]. Therefore, it can be concluded that Cr(VI) can be reduced to Cr(III) species (e.g., Cr(OH) 3 ) and precipitated on MFH surface.  (Figure 4f) was distinguished, the existence of organic HA can also be verified. In summary, XPS indicated that the oxidation of Fe3O4 NPs was prevented by coating HA on Fe3O4 NPs surface and dispersing HA-coated Fe3O4 NPs onto Mt layers. As shown in Figure 4g, the Cr 2p3/2 peak of MFH after removal experiments can be split into two peaks at 574.3 and 573.2 eV (the characteristic peaks of Cr(VI) and Cr(III)), respectively, suggesting that both Cr(VI) and Cr(III) existed on MFH surface [41,42]. Therefore, it can be concluded that Cr(VI) can be reduced to Cr(III) species (e.g., Cr(OH)3) and precipitated on MFH surface. The total organic carbon (TOC) content of raw HA was measured to be 43.81% (mass fraction). According to the dosage of HA used in MFH synthesis, the TOC content of as-prepared MFH was calculated to be 5.68%, which was very close to the measured value (5.34%). Besides, according to ICP-OES, the contents of Si, Al and Fe in MFH were 4.24%, 1.60% and 50.13%, respectively. As the contents of Si and Al in raw Mt were 24.18% and 8.86% respectively, theoretical contents of Si, Al and Fe in MFH were calculated to be 4.13%, 1.51%, and 49.4%, respectively. Therefore, the chemical The total organic carbon (TOC) content of raw HA was measured to be 43.81% (mass fraction). According to the dosage of HA used in MFH synthesis, the TOC content of as-prepared MFH was calculated to be 5.68%, which was very close to the measured value (5.34%). Besides, according to ICP-OES, the contents of Si, Al and Fe in MFH were 4.24%, 1.60% and 50.13%, respectively. As the contents of Si and Al in raw Mt were 24.18% and 8.86% respectively, theoretical contents of Si, Al and Fe in MFH were calculated to be 4.13%, 1.51%, and 49.4%, respectively. Therefore, the chemical compositions of as-prepared MFH were close to the initial dosage, which suggested the firm combination between Mt, Fe 3 O 4 NPs and HA.
To study the recyclability, magnetic hysteresis loops of MFH were plotted in Figure 5. The saturation magnetization (Ms) of Fe 3 O 4 NPs and MFH were measured to be 74.3 and 51.1 emu·g −1 , respectively. The decrease of Ms was resulted from the total contents of Mt and HA in MFH, which turned out to be 31.2% after calculation [28]. The composition value agreed well with that the initial dosage and calculation value of ICP-OES and TOC (30.6%). With a relatively high Ms, MFH can be easily separated from water by a magnet.

. Simultaneous Removal of Cr(VI) and Aniline
In Figure 6, the performance of MFH for removing Cr(VI) and aniline was demonstrated an mpared with that of Mt, Fe3O4 NPs and HA. The removal efficiency of both Cr(VI) and aniline wa FH > Fe3O4 > Mt > HA. 84.8% of Cr(VI) and 89.2% of aniline could be removed by MFH in about 4 d 50 min, respectively. By comparing with the longer time and lower removal efficiency by Fe3O s, Mt and HA, the enhanced adsorption ability of MFH can be verified. As stated above, th ecific surface area of MFH (98.86 m 2 ·g −1 ) was much larger than that of both Mt (53.72 m 2 ·g −1 ) an 3O4 NPs (77.24 m 2 ·g −1 ), which enriched the pollutants by adsorption and significantly improve removal efficiency. Mt can adsorb contaminants in wastewater and further enhance the adsorptio ility of Fe3O4 NPs. Coating HA onto Fe3O4 NPs greatly improved the stability and dispersity an rther adsorbed more heavy metals and organic pollutants. Besides, HA also worked as reductant t uce Cr(VI) to Cr(III) [29][30][31]. Therefore, the synergistic effects between Fe3O4 NPs, Mt and H ntributed greatly to the improvement of adsorption ability and removal efficiency in MFH.

Simultaneous Removal of Cr(VI) and Aniline
In Figure 6, the performance of MFH for removing Cr(VI) and aniline was demonstrated and compared with that of Mt, Fe 3 O 4 NPs and HA. The removal efficiency of both Cr(VI) and aniline was MFH > Fe 3 O 4 > Mt > HA. 84.8% of Cr(VI) and 89.2% of aniline could be removed by MFH in about 40 and 50 min, respectively. By comparing with the longer time and lower removal efficiency by Fe 3 O 4 NPs, Mt and HA, the enhanced adsorption ability of MFH can be verified. As stated above, the specific surface area of MFH (98.86 m 2 ·g −1 ) was much larger than that of both Mt (53.72 m 2 ·g −1 ) and Fe 3 O 4 NPs (77.24 m 2 ·g −1 ), which enriched the pollutants by adsorption and significantly improved the removal efficiency. Mt can adsorb contaminants in wastewater and further enhance the adsorption ability of Fe 3 O 4 NPs. Coating HA onto Fe 3 O 4 NPs greatly improved the stability and dispersity and further adsorbed more heavy metals and organic pollutants. Besides, HA also worked as reductant to reduce Cr(VI) to Cr(III) [29][30][31]. Therefore, the synergistic effects between Fe 3 O 4 NPs, Mt and HA contributed greatly to the improvement of adsorption ability and removal efficiency in MFH. the removal efficiency. Mt can adsorb contaminants in wastewater and further enhance the adsorption ability of Fe3O4 NPs. Coating HA onto Fe3O4 NPs greatly improved the stability and dispersity and further adsorbed more heavy metals and organic pollutants. Besides, HA also worked as reductant to reduce Cr(VI) to Cr(III) [29][30][31]. Therefore, the synergistic effects between Fe3O4 NPs, Mt and HA contributed greatly to the improvement of adsorption ability and removal efficiency in MFH.

Adsorption Kinetics
Adsorption kinetics helps to understand the mechanism, and the obtained data can be used to build mathematical models for interpreting interactions. To figure out the mechanisms in adsorption, pseudo first-order, pseudo second-order and intraparticle diffusion models (Equations (3)-(5)) were used for fitting the data from experiments.

Adsorption Kinetics
Adsorption kinetics helps to understand the mechanism, and the obtained data can be used to build mathematical models for interpreting interactions. To figure out the mechanisms in adsorption, pseudo first-order, pseudo second-order and intraparticle diffusion models (Equations (3)-(5)) were used for fitting the data from experiments.
q e and q t (mg·g −1 ): the mass of contaminants adsorbed on adsorbent at equilibrium and t, respectively; k 1 (min −1 ), k 2 (g·mg −1 ·min −1 ) and k i (mg·g −1 ·min −1/2 ): the pseudo first-, second-order and the intraparticle diffusion rate constants, respectively; C: a constant. These three models were tested for the adsorption of Cr(VI) and aniline on MFH in this study. The results are shown in Table 2. Seen from Table 2, for Cr(VI), the pseudo first-order model showed the best fit and a close value of q e,cal to q e,exp . As for aniline, R 2 of the pseudo second-order model (R 2 = 0.982) was the highest, indicating that the adsorption of aniline by MFH followed the pseudo second-order kinetic model. It suggested that chemisorption may limit the rate of aniline adsorption. Meanwhile, the intraparticle diffusion model (R 2 = 0.975) also exhibited a good fit. Therefore, the adsorption of aniline on MFH may mainly occur in two steps: first, diffusion from water to MFH surface; second, the intraparticle diffusion between the pores of MFH. Cr-bentonite [43], activated carbon [44] and multi-walled carbon nanotubes [45] with similar mechanisms to MFH can be referenced for better understanding.

Adsorption Isotherm
For the evaluation of adsorption capacity and the interpretation of mechanism, Langmuir and Freundlich models were utilized in this study. The former (Equation (6)) is based on the assumptions of monolayer adsorption and identical adsorption sites. The latter (Equation (7)) describes heterogeneous surface and adsorption sites [43,46].
ln q e = ln K + 1 n ln c e (7) q m (mg·g −1 ): the maximum capacity for monolayer adsorption; b (L·mg −1 ): a constant; K (mg 1−(1/n) L 1/n ·g −1 ): a constant; n: an empirical parameter [47,48]. According to Table 3, the adsorption of Cr(VI) and aniline on MFH followed Langmuir and Freundlich models, indicating monolayer and multilayer adsorption, respectively. As displayed in Table 4, MFH possessed higher adsorption capacity towards both Cr(VI) and aniline than some reported magnetic materials [49][50][51][52].  Figure 7 shows that removing of both Cr(VI) and aniline was affected by pH. The adsorption rate was slowed dramatically if pH increased. Zero point charge pH (pH ZPC ) of MFH in this study was measured to be 4.3 (±0.4) by acid-base titration method [53]. Solid surfaces of MFH had positive and negative charges when pH was higher and lower than pH ZPC , respectively. As the surface charge can be affected by pH, the surface of MFH will become more positively charged when pH decreases and thus will have stronger electrostatic attraction with Cr(VI) anions. Adsorption is often affected by electrostatic repulsion and attraction between adsorbent and adsorbate [30]. Stronger electrostatic repulsion due to higher pH will lead to less adsorption of Cr(VI) anions. As Equation (8) shows, studies have revealed that lower pH could enhance the redox potential of Cr(VI)/Cr(III), making Cr(VI) more easily to be reduced to Cr(III) [30]. Although unmodified Fe 3 O 4 has reduction capacity and can reduce Cr(VI), Fe 2+ in HA-Fe 3 O 4 turned out not participate in Cr(VI) reduction [30,54,55]. On the contrary, higher pH leads to stronger electrostatic repulsion between them and thus the decrease of adsorption [2]. Moreover, at higher pH, the adsorption of Cr(III) species (e.g., Cr(OH) 3 ) onto MFH surface reduces the adsorption ability towards contaminants [56]. On the other hand, aniline will be more protonated at lower pH (Equation (9)), which strengthens the electrostatic attraction with the negative carboxylic ions of HA in MFH. Increasing pH will significantly reduce the interaction between aniline and MFH. It can be concluded that lower pH is preferred to remove both Cr(VI) and aniline by MFH in wastewater. This tendency agrees well with previous studies on their individual removal [49][50][51][52]56].

Effect of MFH Dosage
The effect of MFH dosage on the simultaneous removal of Cr(VI) and aniline was investigated, as shown in Figure 8. The results indicate that no great improvement was achieved at a higher dosage, which should provide more adsorption sites and thus contribute to contaminants removal. However, excessive MFH cannot effectively improve the adsorption efficiency due to the concentration limit of contaminants. Therefore, 0.10 g·L −1 turns out to be the optimized dosage of MFH for both Cr(VI) and aniline.

Effect of MFH Dosage
The effect of MFH dosage on the simultaneous removal of Cr(VI) and aniline was investigated, as shown in Figure 8. The results indicate that no great improvement was achieved at a higher dosage, which should provide more adsorption sites and thus contribute to contaminants removal. However, excessive MFH cannot effectively improve the adsorption efficiency due to the concentration limit of contaminants. Therefore, 0.10 g·L −1 turns out to be the optimized dosage of MFH for both Cr(VI) and aniline. as shown in Figure 8. The results indicate that no great improvement was achieved at a higher dosage, which should provide more adsorption sites and thus contribute to contaminants removal. However, excessive MFH cannot effectively improve the adsorption efficiency due to the concentration limit of contaminants. Therefore, 0.10 g·L −1 turns out to be the optimized dosage of MFH for both Cr(VI) and aniline.

Mutual Effect between Cr(VI) and Aniline
Experiments were carried out to find whether Cr(VI) and aniline affected the removal of each other. As shown in Figure 9, when c0(aniline) and c0(Cr(VI)) were increased from 0 to 200 mg·L −1 and 0 to 100 mg·L −1 respectively, their corresponding removal efficiency barely improved. Since the upper limits of their initial concentrations were relatively high compared with actual wastewater, it can be assumed that they have no remarkable influence on the adsorption of each other by MFH. It is most likely that they are adsorbed by MFH in different mechanisms (Cr(VI), Langmuir, pseudo first-order kinetic; aniline, Freundlich, pseudo second-order kinetic), and the adsorption force is much stronger than the interaction between them.

Mutual Effect between Cr(VI) and Aniline
Experiments were carried out to find whether Cr(VI) and aniline affected the removal of each other. As shown in Figure 9, when c 0 (aniline) and c 0 (Cr(VI)) were increased from 0 to 200 mg·L −1 and 0 to 100 mg·L −1 respectively, their corresponding removal efficiency barely improved. Since the upper limits of their initial concentrations were relatively high compared with actual wastewater, it can be assumed that they have no remarkable influence on the adsorption of each other by MFH. It is most likely that they are adsorbed by MFH in different mechanisms (Cr(VI), Langmuir, pseudo first-order kinetic; aniline, Freundlich, pseudo second-order kinetic), and the adsorption force is much stronger than the interaction between them.

Effect of NaCl Content
Inorganic salts of high contents, especially NaCl, can also be found in actual wastewater. In general, NaCl mainly comes from the salinity in initial water or that produced by adding chemical agents in the process. Since the high salinity may have influence on the removal of contaminants in actual wastewater, the effect of NaCl was also studied. Seen from Figure 10, the efficiency of Cr(VI) removal was slightly reduced with NaCl content increasing from 0 to 30% in wastewater. Therefore, it can be concluded that NaCl in wastewater has low effect on Cr(VI) removal, which follows Langmuir monolayer adsorption. On the other hand, the removal efficiency of aniline dramatically decreased with NaCl content increasing from 0% to 30% in wastewater. Negatively charged Cl − may compete with carboxylic ions of HA in MFH to bind with the protonated aniline at low pH, and positively charged Na + may compete against the protonated aniline for adsorption sites. As a result, the diffusion of aniline to MFH exterior surface and even the intraparticle diffusion in pore structures is impeded. Thus, the aniline removal efficiency drops and the influence will become

Effect of NaCl Content
Inorganic salts of high contents, especially NaCl, can also be found in actual wastewater. In general, NaCl mainly comes from the salinity in initial water or that produced by adding chemical agents in the process. Since the high salinity may have influence on the removal of contaminants in actual wastewater, the effect of NaCl was also studied. Seen from Figure 10, the efficiency of Cr(VI) removal was slightly reduced with NaCl content increasing from 0 to 30% in wastewater. Therefore, it can be concluded that NaCl in wastewater has low effect on Cr(VI) removal, which follows Langmuir monolayer adsorption. On the other hand, the removal efficiency of aniline dramatically decreased with NaCl content increasing from 0% to 30% in wastewater. Negatively charged Cl − may compete with carboxylic ions of HA in MFH to bind with the protonated aniline at low pH, and positively charged Na + may compete against the protonated aniline for adsorption sites. As a result, the diffusion of aniline to MFH exterior surface and even the intraparticle diffusion in pore structures is impeded. Thus, the aniline removal efficiency drops and the influence will become more significant when the content of NaCl in wastewater increases.
Inorganic salts of high contents, especially NaCl, can also be found in actual wastewater. In general, NaCl mainly comes from the salinity in initial water or that produced by adding chemical agents in the process. Since the high salinity may have influence on the removal of contaminants in actual wastewater, the effect of NaCl was also studied. Seen from Figure 10, the efficiency of Cr(VI) removal was slightly reduced with NaCl content increasing from 0 to 30% in wastewater. Therefore, it can be concluded that NaCl in wastewater has low effect on Cr(VI) removal, which follows Langmuir monolayer adsorption. On the other hand, the removal efficiency of aniline dramatically decreased with NaCl content increasing from 0% to 30% in wastewater. Negatively charged Cl − may compete with carboxylic ions of HA in MFH to bind with the protonated aniline at low pH, and positively charged Na + may compete against the protonated aniline for adsorption sites. As a result, the diffusion of aniline to MFH exterior surface and even the intraparticle diffusion in pore structures is impeded. Thus, the aniline removal efficiency drops and the influence will become more significant when the content of NaCl in wastewater increases.

Reusability
In a single run, the batch experiment was carried out as described in Section 2.4 and then the MFH were recycled from solution using a magnet, washed with 50% ethanol in water (v/v) several times and dried. The reusability was evaluated by reusing the reclaimed MFH in the next five runs. As Figure 11 shows, ≥82.3% of Cr(VI) and ≥95.1% of aniline can be removed in six consecutive cycles, demonstrating excellent reusability and structural stability.

Reusability
In a single run, the batch experiment was carried out as described in Section 2.4 and then the MFH were recycled from solution using a magnet, washed with 50% ethanol in water (v/v) several times and dried. The reusability was evaluated by reusing the reclaimed MFH in the next five runs. As Figure 11 shows, ≥82.3% of Cr(VI) and ≥95.1% of aniline can be removed in six consecutive cycles, demonstrating excellent reusability and structural stability.

Conclusions
In this study, montmorillonite/Fe3O4/humic acid (MFH) nanocomposites were easily fabricated through a hydrothermal route. The coating of HA onto Fe3O4 NPs and dispersion of HA-coated Fe3O4 NPs onto Mt can protect Fe3O4 NPs from oxidation and inhibit their aggregation. The synergistic effects between Mt, Fe3O4 NPs and HA turned out to effectively enhance the reactivity and efficiency of Fe3O4 NPs for the simultaneous removal of Cr(VI) and aniline from wastewater. The adsorption of Cr(VI) and aniline by MFH obey different mechanisms, Cr(VI)-Langmuir isotherm, pseudo first-order kinetic; aniline-Freundlich isotherm, pseudo second-order kinetic. Furthermore, the MFH nanocomposites possess good separation ability, stability, and reusability, making them promising materials for applications in water contaminants removal.

Conclusions
In this study, montmorillonite/Fe 3