Chitosan-modified magnetic carbon nanomaterials with high efficiency, controlled motility, and reusability—for removal of chromium ions from real wastewater

Hexavalent chromium Cr(VI) is one of the most hazardous oxygen-containing anions to human health and the environment. Adsorption is considered to be an effective method for the removal of Cr(VI) from aqueous solutions. Based on an environmental perspective, we used renewable biomass cellulose as carbon source and chitosan as functional material to synthesize chitosan-coated magnetic carbon (MC@CS) material. The synthesized chitosan magnetic carbons were uniform in diameter (~ 20 nm) and contain a large number of abundant hydroxyl and amino functional groups on the surface, meanwhile owning excellent magnetic separation properties. The MC@CS exhibited high adsorption capacity (83.40 mg/g) at pH 3 and excellent cycling regeneration ability when applied to Cr(VI) removal in water, removal rate of Cr(VI) (10 mg/L) was still over 70% after 10 cycles. FT-IR and XPS spectra showed that electrostatic interaction and reduction with Cr(VI) are the main mechanisms of Cr(VI) removal by MC@CS nanomaterial. This work provides an environment-friendly adsorption material that could be reused for the removal of Cr(VI) in multiple cycles.


Introduction
Chromium Cr(VI), classified as one of the most toxic pollutants in wastewater, is commonly found in surface water and groundwater due to its widespread use in industries such as electroplating, printing, and dyestuffs (Wachiraphorn et al. 2019). In general, chromium ions are mainly present in Responsible Editor: Tito Roberto Cadaval Jr Highlights • Magnetic carbon modified by chitosan (MC@CS) was applied as a novel remediation material to remove hexavalent chromium (Cr(VI)). • A high capacity (83.4 mg/g) for the removal of Cr(VI) was observed.
• Cr(VI) removal was an adsorption-reduction process by MC@ CS.
• MC@CS showed good magnetic separation performance and excellent reproducibility. aqueous solutions in the form of Cr(VI) and Cr(III), with Cr(VI) being over 100 times more toxic than Cr(III) (Owlad et al. 2009). The World Health Organization (WHO) stipulates that the maximum permissible concentration of Cr(VI) in drinking water is 50 μg/L, and the United States Environmental Protection Agency (EPA) has set a limit of 1.00 μg/L for total chromium in drinking water (Zhu et al. 2012).
Hence, it is essential to develop economical, eco-friendly, and effective water treatment technologies to remove Cr(VI) from aqueous solutions. It is worth noting that the reduction of Cr(VI) to Cr(III) or immobilization of Cr(VI) in the effluent is a very important pollution management strategy (Yang et al. 2014). Currently, electrocatalytic chemical precipitation (Ruotolo et al. 2006), ion exchange (Yunqing et al. 2007), redox treatment (Verma and Sarkar 2020), physical chemical adsorption (Arslan et al. 2010), and reverse osmosis (Hafez and El-Mariharawy 2004) are all considered effective as the methods for the removal of heavy metal ions. Among these techniques, adsorption is considered to be the most suitable method for removal due to its high efficiency (Lee et al. 2020) because its simplicity of operation and ease of regeneration by eluent. A range of adsorbents including activated carbon (Qiu et al. 2014;Tz et al. 2020;Wang et al. 2020), zerovalent iron (ZVI) , metal oxides (Pinakidou et al. 2016), and zeolites (Silva et al. 2008) have been successfully applied. Activated carbon is widely researched and used in large quantities due to its high specific surface area and excellent adsorption properties (Mohan and Pittman 2006). However, activated carbon materials remain a lot problems, such as the difficult to recycle, degradation of material properties after multiple cycles, unclear mechanism of Cr adsorption in the case of complex ions, and the high cost of carbon materials such as carbon nanotubes and graphene, causing limitation in broad application in water treatment.
Magnetic carbon nanomaterials alleviate the above issue properly owing to its excellent adsorption and separation properties simultaneously (Li et al. 2016a, b). Recently, many attempts were made to prepare the raw and functionalized magnetic carbon nanomaterials to remove the pollutants in the contaminated water environment. For example, Cui and Atkinson successfully produced the magnetic mesoporous Fe/C composite materials from waste glycerol with excellent removal efficiency for Cr(VI) and separation performance (Cui and Atkinson 2019). Chitosan is the product of partial acetylation of the natural polysaccharide chitin, which has various functions such as biodegradability, biocompatibility, non-toxicity, and anti-bacterial (Suh and Matthew 2000). Similar to this paper, Aslam et al. (2020) used chitosan to modify carbon nanotubes and graphene to obtain adsorbent materials with better performance, The material synthesized by Aslam and other researchers increased the adsorption capacity to 119 mg/g, but the use of carbon nanotubes and graphene resulted in a high cost of the material, and there is room for improving the cycling performance of the material. Zhu et al. (2012) successfully prepared chitosan-modified graphitized polycarbonate nanotubes for efficient removal of organic dyes from aqueous solutions, but the removal of inorganic metal ions by this material has not been investigated. Considering the recoverability and reusability of magnetic carbon, it is necessary to prepare the magnetic carbon or functionalized magnetic carbon materials with stable adsorption and regeneration properties. Wu et al. (2017) successfully synthesized Fe(III) crosslinked chitosan material and used it to explore its adsorption performance in low concentration Cr(VI) (less than 20.0 mg/L). The results show that Fe(III)-CBs has a good prospect in purifying low concentration Cr(VI) water with a pH range from 2.0 to 6.0.
In order to solve the problem of high cost of carbon nanotubes and graphene materials, it is especially important to find a low-cost and reproducible material to replace graphene and carbon nanotubes for the commercialization of carbon adsorption materials. It is more environment-friendly that selecting natural raw materials for the preparation of absorbents as a low consumption, low toxicity, and sustainable solution. Cellulose is one of the most widely distributed and abundant polysaccharides in nature, accounting for more than 50% of the carbon content of the plant world (Klemm et al. 2005). Cellulose is also the oldest and most abundant natural polymer on the planet, making it an inexhaustible and the most valuable natural renewable resource for mankind. Using cellulose as a source of carbon can help to make efficient use of biomass. Transforming this widely available substance into useful products makes good sense from an environmental and economic perspective (Moon et al. 2011).
Herein, we have prepared a chitosan-coated magnetic carbon (MC@CS) material using cellulose as the carbon source. Chitosan magnetic carbon nanospheres have a uniform diameter (about 20 nm) and a large number of hydroxyl and amino functional groups on the surface, with an excellent magnetic separation performance. When applied to chromium removal, the chitosan magnetic carbon nanospheres have a high adsorption capacity (83.4 mg/g) and an outstanding cycling regeneration capacity. After ten cycles, there was still more than 70% removal of low concentrations of Cr (10 mg/L). The adsorption mechanism study showed that both electrostatic interaction and reduction of chitosan-modified magnetic nanoparticles were responsible for the removal of Cr(VI).

Characterization
Morphologies of samples were characterized by field-emission scanning electron microscope (SEM) (TESCAN MIRA4, Czekh) and transmission electron microscopy (TEM) (FEI TF20, USA). Fourier transform infrared spectrometer (FT-IR) was recorded using Bruker Tensor27 infrared spectrometer. The specific Brunauer-Emmett-Teller (BET) surface area and pore size distribution were measured on a MICROMERIT-ICS INSTRUMENT CORP ASAP2460 by nitrogen adsorption at 77.4 K. The pore size distribution of the samples was calculated by the Barrett-Joyner-Halenda (BJH) method using nitrogen desorption isotherms. The XRD patterns were recorded on a Rigaku D/Max 2400 diffractometer employing Cu Ka radiation run at 40 kV and 100 mA. UV-Vis absorption spectra were obtained on an UV-T2600 spectrometer (Shanghai Youke, China). Atomic absorption spectrometry data were acquired on an EWAI AA7020 spectrometer (East-west analysis, China). X-ray photoelectron spectroscopy (XPS) (Thermo fisher Scientific K-Alpha, USA) is employed to analyze the presence of valence and composition of elements in the adsorbent. Magnetic properties were analyzed by a vibrating sample magnetometer (VSM) (LakeShore 7307, USA) at room temperature.

Preparation of magnetic carbon (MC)
Magnetic carbon nano-absorbents were prepared by reference to the method of Qiu et al. (2014). Typically, 4.0 g Fe(NO 3 ) 3 ·9H 2 O was firstly dissolved in 20 mL ethanol, in which 4.0 g cellulose was then added. The mixed suspension was mechanically stirred at 400 rpm for 3 h at room temperature to homogenize the Fe(NO 3 ) 3 in the cellulose. The well-mixed suspension was then heated in a water bath at 50 °C to completely evaporate ethanol. The remaining solid (Fe(NO 3 ) 3 /cellulose) was dried in a vacuum oven overnight. The Fe(NO 3 ) 3 /cellulose was loaded in a tube furnace and heated to 600 °C at a heating rate of 5 °C/min under nitrogen and held for 1 h, then cooled naturally to room temperature. Finally, the product was rinsed with deionized water and finally dried overnight in a vacuum oven at 60 °C.

Preparation of chitosan-coated magnetic porous carbon (MC@CS)
Chitosan (0.1 ~ 0.9 g) was dissolved in 20 mL of 5% acetic acid solution and stirred for 1 h to dissolve it well. Meanwhile, MC (0.5 g) was sonicated and dispersed in 80 mL of methanol solution and then 0.1 g of trimethyl glycine was added. After mixing the above two solutions, 8 mL of glutaraldehyde (20 wt.%) was added into the mixed solution and keep the whole stirred in a water bath at 60 °C for 6 h. When the reaction was completed, the solid was separated magnetically, washed thoroughly with ethanol and deionized water, and dried under vacuum at 60 °C for 12 h to obtain chitosan-modified magnetic carbon (MC@CS) adsorbent.
The synthesis process is shown in the following Fig. 1.

Batch adsorption experiments
Batch adsorption experiments were conducted to investigate the adsorption performance of chitosan-modified magnetic carbon nanospheres on Cr(VI). Typically, MC@CS with different chitosan ratios was added to the Cr(VI) solution and sonicated for a period of time at room temperature (25 °C). The Cr(VI) removal efficiency was investigated in detail for different Cr(VI) concentrations (5 ~ 60 mg/L), nano-sorbent dosage (0.125 ~ 4 g/L), treatment times (0 ~ 480 min), and pH values (2.0 ~ 10.0, measured using a pH meter). HCl (1 mol/L) and NaOH (1 mol/L) solutions were used to adjust the pH of the solutions. The adsorption kinetic tests were performed by adding synthetic MC@CS adsorbent (150 mg) into different concentrations of Cr(VI) solutions (150 mL) and the pH value was adjusted at 3.0 with treatment times ranging from 0 to 480 min. For the adsorption isotherm study, 20 mg of MC@CS nanospheres was added to Cr(VI) solution (40 mL) at initial concentrations of 5-60 mg/L and initial temperatures were set to 298 K, 308 K, and 318 K. The concentration of Cr(VI) was determined by a Shimadzu UV-2550 UV-vis spectrophotometer using the diphenyl carbazide method.
where C 0 (mg/L) is the initial Cr(VI) concentration, and C t (mg/L) is the Cr(VI) concentration in solution at the contact time t.
The amount of Cr(VI) adsorbed on the magnetic mesoporous carbon nanospheres was calculated according to the following Eq. (2).
where V (mL) represents the volume of chromium solution, and m (mg) stands for the mass of the used adsorbent.
For regeneration, the Cr(VI)-adsorbed MC@CS was collected by a magnet and regenerated with 40 mL of 0.05 M HCl for 12 h and then washed with high-purity water until neutral pH. The regenerated MC@CS were reused in the next cycle of the adsorption experiment. The adsorption of Cr(VI) was carried out in the same experimental condition with an initial concentration of Cr(VI) as 10 mg/L at pH 3. The adsorption-desorption processes were conducted for ten cycles. The experimental data obtained are the average values after 3 repeated measurements.

Actual wastewater experiments
In order to investigate the treatment prospects of chitosanmodified magnetic carbon nanomotors in actual wastewater and to compare with the adsorption effect in the laboratory, the actual electroplating wastewater was retrieved from a factory in Yunnan for experimental study, and the elements contained in the actual electroplating wastewater were measured as shown in Table 1. The chromium content in this electroplating wastewater is high, reaching 55.67 mg/L, which needs to be treated urgently in order to reach the standard discharge of "Comprehensive Sewage Discharge Standard (GB8978-1996)." The concentrations of other elements are also shown in the table. The pH value of this wastewater was measured between 6 and 7 with a pH meter and was light yellow.
The concentration of Cr(VI) in this actual wastewater was high. In the experiment, 40 mL of the actual wastewater solution was taken, the pH was adjusted at about 3, and the amount of adsorbent dosing was increased appropriately to observe the removal rate of Cr(VI) from the actual wastewater.

Characterization of MC@CS
The morphology of the MC and MC@CS composite was examined by SEM. The SEM image of MC ( Fig. 2A) represents a homogeneous distribution of spherical particles with an average diameter of 10-20 nm, indicating that the formation of Fe 3 O 4 nanoparticles is uniformly and densely arranged in the cellulose carbonization. The MC@CS SEM   2B) shows that the part of functionalized magnetic carbon has a smooth surface, indicating that chitosan is successfully wrapped around the surface of the magnetic carbon. The TEM images ( Fig. 2C and D) further verify the fact that a fairly homogeneous particle size of magnetic carbon was uniformly encapsulated in a hyaline substance. Figure 2D shows that the spherical particles after encapsulation of chitosan are mostly about 15 nm in diameter. It was also found that some of the magnetic carbons were agglomerated, which might be caused by the stirring speed not being vigorous enough during the functionalization reaction and the addition of trimethyl glycine which exacerbated the cross-linking reaction between chitosan and magnetic carbon (Hui et al. 2004). The specific surface area and pore size distribution of the MC and MC@CS were obtained by the nitrogen adsorption and desorption isotherms (Fig. 3A). The type-IV adsorption-desorption isotherm curves with the hysteresis loop indicate the property of narrow pores in the material. The BJH pore size distribution of MC@CS ( Fig. 3B) shows that most of the mesopore diameters are in the range of 0-10 nm, and the presence of mesopores will facilitate ion translocation to the adsorption sites. As shown in Table 2, the BET specific surface area of MC and MC@CS are 319.62 and 36.48 m 2 /g, respectively. The decrease in specific surface area mainly caused by the formation of chitosan gel. Figure 4A shows the WA-XRD patterns of MC and MC@ CS. The diffraction peaks of both are similar, indicating that the functionalization of chitosan has no effect on the crystalline form of MC. Diffraction peaks at 24.3° for both which is attributed to the (002) plane reflection of amorphous carbon . The diffraction peaks of two at 30.21°, 35.60°, 43.24°, 57.12°, and 62.70° are corresponding to their indices (220), (311), (400), (511), and (440), indicating that the Fe 3 O 4 nanocrystals are successfully formed (Taberna et al. 2006). These two points clearly explain the successful fabrication of the magnetic carbon by carbonization and the non-significant effect on the crystalline shape of the magnetic carbon after chitosan coating.
Elemental analysis is used to determine the elemental composition of the synthesized MC@CS (Fig. 4B). The weight percentages of C, O, and Fe in MC@CS were 78.41%, 9.66%, and 11.93%, respectively, where the Fe/C atomic ratio was 0.03 and the Fe/O ratio was 0.35, which was less than the atomic ratio in Fe 3 O 4 , indicating that the chitosan was successfully wrapped around the magnetic carbon surface, increasing the atomic weight of O.  Figure 5A shows the influence of chitosan addition amount on material removal effect. A gradient experiment of chitosan dosage was designed, with dosage of 0.1-0.9 g, respectively. Fe-CS represents the adsorption of chitosan   Fe3O4 material on Cr. In the experiment, the concentration of Cr solution is 20 mg/L, the dosage of solution is 20 mL, the adsorption time is 10 min, and the dosage of adsorbent is 10 mg. It can be seen from the experimental results that the proportion of chitosan plays a key role in the performance of the adsorption material. When the proportion is less than 1:1 (0.5 g chitosan + 0.5 g magnetite carbon), the continuous increase of chitosan is conducive to the improvement of the adsorption performance. However, when the proportion of chitosan exceeds 1:1, that is, the amount of 0.7 g and 0.9 g, the adsorption performance has not been further improved. According to the BET analysis of the material above, this is because chitosan will be coated on the surface of magnetite carbon, resulting in further reduction of the specific surface area of the material. When the proportion is 1:1, the diameter of the mesopore left is within the range of 0 ~ 10 nm, and the existence of the mesopore will promote the ion transport to the adsorption site and promote the synthesis MC@ CS adsorption of chromium containing metal anions by the material. Excess chitosan will further block the pores, resulting in no obvious adsorption effect or a certain reduction. The removal efficiency of chitosan Fe 3 O 4 material without carbon addition which named Fe-CS in Fig. 5A is obviously lower than that of Cr MC@CS. The material also shows MC@CS composite properties of materials. Figure 5B shows the performance of MC and MC@ CS for the removal of Cr(VI). The pre-modified material (MC) showed a gradual increase in removal with increasing dosing. However, the modified material also showed complete removal of Cr(VI) at low dosing rates. The modified material (MC@CS) had better adsorption performance than the pre-modified material. This can be explained by the large number of active sites on the surface of the chitosan-modified material. Chitosan itself contains a large number of functional groups with hydroxyl and amino groups, which can effectively cooperate with Cr(VI) through surface complexation reactions, resulting in a better removal rate of Cr(VI).

Optimization of adsorption influencing factors
The Cr(VI) removal efficiency of MC@CS was investigated in the range of pH 2 ~ 10. The results displayed in Fig. 6A show that the pH value of the solution has a significant effect on the Cr(VI) removal efficiency. The maximum removal efficiency of Cr(VI) is reached at pH 3, but as the pH value gradually rises to 9, the Cr(VI) removal efficiency decreases significantly. This phenomenon was not only due to the fact that the form of Cr(VI) present in solution varies with the pH value, but was also related to the surface properties of MC@CS (Kumari et al. 2014). When the solution is acidic, chromium ions are present in solution in two forms of chromic acid (H 2 CrO 4 ) when the pH value is less than 1 and chromic acid hydrogen ion (HCrO 4 − ) when the pH 1 ~ 6 (Zhao et al. 2016). At pH 2 ~ 6, the hydroxyl radicals and amino groups on the surface of the adsorbent were 2− in the solution. As the pH value in the solution continues to increase, the adsorbent solid surface becomes negatively charged, creating an electrostatic repulsion with the CrO 4 2− in solution resulting in difficult adsorption of chromium (Ansari et al. 2017). Therefore, the chromium removal rate keeps decreasing gradually while the pH 6 grew to 10. From this experiment, it can be obtained that the adsorption of chromium is mainly attributed to the electrostatic interaction between the chromate anion and the protonated amine under acidic conditions. The amount of sorbent is an important factor affecting the adsorption effect. The effect of the dosage on the removal efficiency of Cr(VI) in aqueous solutions is shown in Fig. 6B. The amount of adsorbent was increased from 2.5 to 80 mg and accordingly the removal rate increased from 54 to 99.8%. The removal rate raised sharply when the absorbent amount was lower than 10 mg, while the removal rate was stabilized at about 99.97% as the amount continuously added. In general, the higher the amount of adsorbent in solution, the more active sites that can bind to Cr(VI).
The effect of adsorption time on the removal rate of Cr(VI) was investigated experimentally. The removal rate of Cr(VI) increased with time at T = 298 K, an initial concentration of 20 mg/L and a stirring frequency of 150 rpm. As can be seen in Fig. 6C, the chitosan-modified magnetic carbon reached 95.74% within 20 min. After 40 min of adsorption, the removal rate started to stabilize. Thereafter, when the adsorption time reached 60 min, the removal rate reached almost 100%. When the time exceeded 60 min, the removal rate remained unchanged, at which time it was the adsorption equilibrium. The reason for this phenomenon was that all the Cr(VI) adsorption sites on the surface of the material were occupied by Cr(VI), resulting in the removal rate no longer increasing.
The experiments on the effect of the initial concentration of adsorption and temperature on the amount of adsorption were performed by prepared Cr(VI) ion solutions with concentration gradients, and the variation of adsorption amount versus the corresponding equilibrium concentration was obtained at T = 298 K, 308 K, and 318 K, respectively. As shown in Fig. 6D, the unit removal efficiency of Cr(VI) showed a gradual increase from 298 to 318 K and showed the best adsorption capacity at 318 K. In the removal rate experiments at different initial concentrations of 298 K, 308 K, and 318 K, it was found that the removal rate at 318 K was always higher than that at 298 K and 308 K. At each temperature, the removal rate decreased slowly with increasing initial concentration, but the adsorption capacity increased steadily to the saturation value.

Effect of co-ionization on Cr adsorption effect
There are usually a variety of other metal ions or anions in actual wastewaters containing chromium ions, so the effect of co-existing ions on the chromium removal efficiency of chitosan-modified magnetic carbon nanoparticles was Fig. 6 Influence factor on the removal of Cr(VI): effect of pH on Cr(VI) removal (A); effect of adsorbent amount on the removal of Cr(VI) (B); effect of adsorption time on removal rate (C); effect of initial concentration of adsorption and temperature on the amount of adsorption (D) further investigated. In this study, the main metal ions Fe 3+ , Mn 2+ , and Zn 2+ and the common interfering oxygencontaining anions PO 4 3− , SO 4 2− , CO 3 2− , F − , and Cl −1 in industrial wastewater were selected. The effect of coexisting ions at different concentrations (5 to 50 mg/L) on the removal rate of Cr(VI) was investigated. It can be seen from Fig. 7A that Fe 3+ , Mn 2+ , and Zn 2+ have a relatively obvious effect on the removal of chromium, while PO 4 3− , SO 4 2− , CO 3 2− , F − , and Cl −1 has almost no effect as shown in Fig. 7B. The chromium removal efficiency gradually diminished as the concentrations of Fe 3+ , Mn 2+ , and Zn 2+ increased, which had a combination with the hydroxyl and amino groups on the surface part of the adsorbent (Ngah et al. 2005;Jianying et al. 2015;Xiao-Qiang, Xiao-Fang, and Sheng n.d.). However, after coexisting ion concentration of 40 mg/L, the degree of influence gradually settled down, indicating that the adsorbent has a certain selectivity for Cr(VI). And PO 4 3− in solution preferentially combined with H + under acidic conditions, so it has little affect the removal of Cr(VI) at lower pH condition (Robertson 2003).
In order to further reveal the selectivity of adsorbent to Cr(VI), MS software was used to construct the adsorption model, and DMol 3 module was used to geometrically optimize each adsorption model to obtain the lowest energy system. The main process of adsorption energy calculation is to build the substrate → optimize the structure of the substrate and calculate the energy → optimize the adsorption part and calculate the energy → build and optimize the adsorption structure → calculate the energy properties. The commonly used calculation formula of adsorption energy is as follows: where E CS+M refers to the total energy of the system after adsorption, E CS refers to the energy of CS substrate, E M refers to the energy of metal ions, and E ads refers to the adsorption energy. The adsorption model and energy of each molecule are obtained by establishing the corresponding configuration in MS and calculating the DMOL 3 energy, as Fig. 8. The obtained energy values and the adsorption energy and parameters calculated are shown in Table 3.
According to the calculation of adsorption energy, the absolute value of adsorption energy between chitosan and HCrO 4 − under acidic conditions is greater than that of the other three heavy metal cations. Therefore, in theory, chitosan has a preferential effect on the adsorption of hydrogen chromate ions under acidic conditions. According to the experimental phenomenon of interference ion adsorption, the adsorption of Cr(VI) still exists with the increase of interference ion concentration, and the adsorption of Cr(VI) is not gradually lost with the increase of interference ion concentration. This is supported by the theoretical calculation results at the same time MC@CS. The preferential adsorption of Cr(VI) proves that it has a certain selective effect on the removal of Cr(VI).

Regeneration cycle study
The economic feasibility of sorbent materials for practical applications depends significantly on their regeneration capacity during the sorption-desorption process. The high saturation magnetism of MC@CS makes it susceptible to separate it in solution by applying a magnetic field. MC@CS material magnetically controlled motion and separation experiments we provide in the supplemental material, it is shown experimentally that MC@CS material can be separated easily by magnetic field and can be controlled by magnetic field is the present motion in the water body. In this experiment, the adsorbent in solution is therefore magnetically separated and then subjected to a desorption experiment. Firstly, 0.05 M HCl, 0.05 M NaOH, and deionized water were chosen as the eluent. Fifty milligrams of chromium adsorbed MC@ CS was immersed in 50 mL of the resolving solution and then placed in a shaker at 200 rpm for 12 h. After the completion of the desorption, the material was magnetically separated and the surface was rinsed with deionized water to neutral and then used to adsorb 100 mL of Fig. 7 Effect of different co-ion on Cr(VI) removal rate and adsorption capacity. Effect of common co-cations on the adsorption effect (A); effect of common co-occurring anions on the adsorption effect (B) 10 mg/L Cr(VI) solution. The Cr(VI) removal efficiencies for the three eluent phases are shown in Fig. 9A. It was found that 0.05 M HCl was the most effective solution for the regeneration (Bhaumik et al. 2011). Therefore, the stability of the Cr(VI) removal efficiency was investigated by using 0.05 M HCl as the solvent in the next cycle regeneration experiments. The degree of chromium removal decreased slightly after ten cycles, and the removal rate was still higher than 70% (Fig. 9B), indicating that MC@CS has an optimistic regeneration capacity. The reduced removal rate may be due to the degradability of chitosan.

Actual wastewater adsorption experiment
In order to investigate the treatment prospects of chitosanmodified magnetic carbon nanomotors in actual wastewater and to compare with the adsorption effect in the laboratory, the actual electroplating wastewater was retrieved from a factory in Yunnan for experimental study, and the elements contained in the actual electroplating wastewater were measured as shown in Table 3. The chromium content in the electroplating wastewater is high, reaching 55.67 mg/L, which needs to be treated urgently in order to reach the standard discharge of "Comprehensive Sewage Discharge Standard   -1996)." The concentrations of other elements are also shown in the table. The pH value of this wastewater was measured between 6 and 7 with a pH meter and the color was light yellow. Due to the high concentration of Cr(VI) in this actual wastewater, 40 mL of the actual wastewater solution was taken, the pH was adjusted to 3, and the amount of MC@ CS material dosing was increased appropriately to observe the removal rate of Cr(VI) in the actual wastewater. The treated data are shown in Fig. 10A. It can be observed that the concentration of Cr(VI) in the solution gradually decreases with the increase of the MC@CS material dosage, which indicates that the adsorption effect is good for this wastewater. When the dosage was up to 60 mg, the concentration of Cr(VI) in the solution was lower than 0.01 mg/L. When the dosage was increased to 70 mg, the removal rate of Cr(VI) in the solution reached 100%, and the treatment effect reached the WHO standard at this time. After the adsorption experiment with the dosage of 60 mg MC@CS material, filter the adsorbent and conduct regeneration test according to the method in the "Regeneration cycle study" section. Flush the adsorbent with 0.05 M HCl, and add it again to 40 mL of the same actual electroplating wastewater. Calculate the adsorption performance of the regenerated material. The results are shown in Fig. 10B. Compared with the laboratory results, the adsorption performance of the material in the actual wastewater has decreased, and the adsorption performance after each cycle has decreased to varying degrees compared with the laboratory results. In the laboratory experiment, the material still has a removal efficiency of more than 75% after 10 cycles. In the actual wastewater experiment, the adsorption performance has decreased to 68% after 10 cycles, because the ionic environment in the actual wastewater is more complex. It is impossible to clean all adsorption sites with 0.05 M HCl alone, which leads to the continuous reduction of material adsorption performance in the actual wastewater experiment.

Equilibrium modeling of the Cr adsorption
The calculation method of equilibrium modeling of the Cr absorption is given in the supply information. The rate-limiting step for the removal of Cr(VI) with chitosan-modified magnetic carbon was illustrated by adsorption kinetic studies. As shown in Fig. 11A, once the MC@CS was added to the solution, Cr(VI) was rapidly captured within 40 min. Subsequently, the rate of Cr(VI) adsorption slows down and equilibrium is reached after 60 min due to less target anions Cr(VI) in solution and fewer adsorption sites available on the chitosan-modified magnetic carbon. The kinetic data were fitted to a pseudo-first-order model and a pseudo-secondorder model, respectively (Fig. 11B). The kinetic parameters are listed in Table 3. The higher correlation coefficient (R 2 = 0.9991) obtained for the pseudo-second-order model, thus indicating that the rate control step is associated with the chemisorption of Cr(VI) . Due to the relatively small size of Cr(VI) and the short distance into the adsorption site on the chitosan-modified magnetic carbon , the rate constant of Cr(VI) is 0.0382 g/ (mg-min).

Adsorption isotherms
The calculation method of adsorption isotherms is given in the supply information. The experimental data and analytical results are shown in Fig. 12 and Table 4.
In combination with Fig. 11, it can be seen that experimental data fit Langmuir's linear isotherm model the most. The correlation coefficients for the fits of Langmuir isotherm equation in Table 5 are all above 0.991, which indicates that the adsorption behavior of the adsorbent on Cr(VI) basically obeys the Langmuir unimolecular layer adsorption (Hameed et al. 2008;Zhao et al. 2017a). It can also be derived from Table 5 that the maximum saturation adsorption capacity of Cr(VI) reaches 83.40 mg/g. It is worth noting that although the adsorption isotherm equation fits of Freundlich, Dubinin-Radushkevich, and Temkin are not as well as the Langmuir equation, their correlation parameters can still be used as reflect the adsorption pattern of the adsorbent for Cr(VI). Where 0.1 < 1/n < 0.5, indicating that the adsorption is a preferential adsorption. Table 5 shows the maximum  Figure 13A shows the FT-IR spectra of the synthetic material MC (Fig. 13A-a), MC@CS before ( Fig. 13A-b), after adsorption ( Fig. 13A-c), and after cycle regeneration ( Fig. 13A-d). All of which show Fe-O vibrational peaks of magnetite in the 559 cm −1 band (Shabnam and Ahmad 2015). After CS modification, several stretching vibrational absorption peaks of the C-OH of glucose appear in the 1031-1063 cm −1 band in both Fig. 13A-b, A-c, and A-d, indicating that CS was successfully functionalized on magnetic carbon surface (Cui and Atkinson 2019). The absorption peaks appearing at 2865 cm −1 and 2931 cm −1 in Fig. 13A-b, A-c, and A-d spectrum may correspond to the -CH 2 of the glutaraldehyde cross-linker added in the reaction. The shift of the -OH absorption peak of Fig. 13A-a ~ d at 3435 cm −1 becomes higher, indicating the presence of a large amount of hydroxyl groups on the surface of the material (Li et al. 2016a, b). The absorption peak at 1569 cm −1 without the adsorption ( Fig. 13A-a, A-b) of Cr(VI) shifts towards 1629 cm −1 after the adsorption reaction ( Fig. 13A-c, A-d), which is a variable angle absorption peak for -NH 2 , indicating that -NH 2 interacted with the chromate.

Removal mechanism analysis
In order to investigate the main mechanism of Cr(VI) removal by adsorbents, X-ray photoelectron spectroscopy (XPS) was applied to study the surface chemical compositions of the Cr-MC@CS. Figure 13B shows a scan of the entire area of the MC@CS surface of the adsorbent after full exposure to Cr (50 mg/L) solution, at pH = 3 and 20 °C. For MC@CS, the main elements of the surface are carbon (63.98%), oxygen (30.26%), nitrogen (2.71%), iron (1.16%), and chromium (1.88%). The detailed XPS surveys of the Fe-2p and Cr-2p regions are shown in Fig. 13C and D. For the chromium spectra shown in Fig. 13C, the photoelectron peaks for Cr 2p3/2 and 2p1/2 centered at 577.5 eV and 586.9 eV, respectively, which are similar to the previously reported (Tz et al. 2020;Wachiraphorn et al. 2019;Wang et al. 2018). This data suggests that the adsorbed Cr(VI) anion was reduced to Cr(III) after exposure to MC@CS. Cr(III) may be present on its solid surface in the form of Cr(OH) 3 , in combination with the previous article (He et al. 2019; Lu   Zhao et al. 2017b). The results clearly show that the Cr(VI) removal process involves the reduction of Cr(VI) to Cr(III) by the adsorption process. However, the batch adsorption experiments showed that a large amount of Cr(VI) was removed after contact with MC@CS; there is no significant Cr(VI) band appeared in the XPS spectrum (Fig. 13C, similar to the previous paper (Yang et al. 2014). The reason was probably that the reduction of Cr(VI) occurs on the surface of the material MC@CS, and the reduction products were loaded on the surface or combined with iron oxide, hiding the traces of the presence of Cr(VI). For the Fe-2p spectrum (shown in Fig. 13D), where the 711.1 eV and 725.1 eV binding energies of Fe-2p can be assigned to Fe 3 O 4 and Fe 2 O 3 (Huang et al. 2013), as well as the satellite between the two main peaks may be caused by the generation of iron salts. Moreover, the peak corresponding to Fe 0 hardly appears in Fig. 13D, which could be due to the oxidation of the Fe 0 on the carbon surface by oxygen and its participation in the redox reaction (Cr(VI)-Cr(III)) ).

The evaluation adsorption performance of MC@CS in comparison with similar adsorbents
For the removal of Cr ions, the maximum adsorbent capacities for the proposed adsorbent and various other known adsorbents were reported in Table 6. Because the application conditions of each material are different, all data are compared with the best value. It was obviously determined that the adsorption capacity of MC@CS was much higher than the previously reported ones. The magnetic adsorbent produced has a high surface area and still has a removal rate of over 70% after 10 cycles, allows it a high-performance adsorbent that can be used to remove Cr ions from aqueous solutions. In addition, it is possible to separate MC@CS from the aqueous environment by means of a magnet.

Conclusions
In this work, we used cellulose as the carbon source and chitosan as the modified material to synthesize a magnetic driven nano motor with selective adsorption and rapid separation of the oxygenated metal anion CrO 4 2− in wastewater-chitosan-modified magnetic carbon MC@CS.
Chitosan magnetic carbon nanospheres have a uniform diameter (about 20 nm) and a large number of abundant hydroxyl and amino functional groups on the surface, with an excellent magnetic separation performance. When applied to chromium removal, the chitosan magnetic carbon nanospheres have a high adsorption capacity (83.40 mg/g) and an outstanding cycling regeneration capacity. After ten cycles, the adsorption performance can still remain more than 75%. In the actual treatment of electroplating wastewater, Cr(VI) in the treated wastewater can be reduced to less than 0.05 mg/L by increasing the dosage, meeting the comprehensive wastewater discharge standard (GB8789-1996). The adsorption isotherm and kinetic analysis showed that the Langmuir isotherm and the pseudo-second-order kinetic model could describe the adsorption process better than other models, indicating that the adsorption of Cr(VI) by chitosan-modified magnetic carbon was mainly monolayer adsorption caused by chemical reactions. SEM-EDS, FTIR, and XPS characterization showed that both electrostatic interactions and reduction of chitosan-modified magnetic nanoparticles were responsible for the removal of Cr(VI). Chitosan-modified magnetic carbon is likely to be a promising adsorbent for the removal of Cr(VI).