Cadmium adsorption performance and mechanism from aqueous solution using red mud modified with amorphous MnO2

In this study, red mud modified by manganese dioxide(MRM) was utilized as an adsorbent to effectively remove Cd2+ from aqueous solution. The characteristics were analysed by SEM–EDS, XRD, BET, FTIR and XPS. Different factors that affected the Cd2+ removal on MRM, such as dosage, initial pH, initial Cd2+ concentration, were investigated using batch adsorption experiments. Simultaneously, the adsorption kinetics, adsorption isotherms and adsorption thermodynamics of Cd2+ were also investigated using adsorption experiments data. The characterization results showed that MRM had a rougher, larger specific surface area and pore volume (38.91 m2 g−1, 0.02 cm3 g−1) than RM (10.22 m2 g−1, 0.73 cm3 g−1). The adsorption experiments found that the equilibrium adsorption capacity of MRM for Cd2+ was significantly increased to 46.36 mg g−1, which was almost three times that of RM. According to the fitting results, the pseudo-second-order kinetic model described the adsorption process better than the pseudo-first-order kinetic model. The Langmuir model fitted the adsorption isotherms well, indicating that the adsorption process was unimolecular layer adsorption and the maximum capacity was 103.59 mg g−1. The thermodynamic parameters indicated that the adsorption process was heat-trapping and spontaneous. Finally, combined XPS and FTIR studies, it was speculated that the adsorption mechanisms should be electrostatic attachment, specific adsorption (i.e., Cd–O or hydroxyl binding) and ion exchange. Therefore, manganese dioxide modified red mud can be an effective and economical alternative to the removal of Cd2+ in the wastewater treatment process.

. The mineral composition of RM before and after loading was analysed by XRD, and the results are shown in Fig. 2a. The main phases of RM are minerals dominated by Si, Ca, and Al, such as Ca 3 Al 2 (SiO 4 )(OH) 8 (Katoite, Si-rich), Ca 2 SiO 4 (Calcium Silicate), Ca 5 (SiO 4 ) 2 CO 3 (Spurrite) and Ca 3 Al 2 O 6 (Calcium Aluminum Oxide). After the loading of MnO 2 , the peak intensities of the characteristic peaks contained in RM all weakened and some even disappeared, but no new characteristic peaks were generated. This may be because the MnO 2 loading of the amorphous phase weakened the intensity of the diffraction peaks of the other crystals 28 . It can be inferred that www.nature.com/scientificreports/ the MnO 2 loaded on the surface of the red mud in this study is a non-chemo-metric amorphous phase, which is possibly due to the change in the crystalline structure of the manganese dioxide during the preparation of the material 29 .The FTIR spectra of RM and MRM at wavelengths of 400-4000 cm −1 were analysed and the results are shown in Fig. 2b. It can be seen from the figure that the diffraction peaks of each functional group changed significantly after the loading of MnO 2 and new characteristic peaks appeared. A solid and broad absorption band appeared at 3432.58 and 3405.43 cm −1 for RM and MRM, respectively, which corresponds to the O-H stretching vibration of the hydroxyl group on the surface of the samples 30,31 . The band of MRM at 1635.13 cm −1 is attributed to the O-H stretching vibrations in Mn-OH 20,32 . The band at 1418.62 cm −1 appearing in RM is attributed to CO 3 2+ ,proving that RM contains many carbonate compounds 33,34 . However, the CO 3 2+ peak in the MRM shifts and decreases in intensity after loading MnO 2 . The peaks appearing at 996.62 and 992.96 cm −1 are caused by O-Si-O stretching vibrations 35,36 . The presence of Al-O stretching vibrations (684.62 cm −1 ), Si-O-Al stretching vibrations (620.98 cm −1 ) and Fe-O stretching vibration (458.98 cm −1 ) existing in RM disappeared in the MRM 37 , while a clear absorption peak at 532.56 cm −1 appeared in the MRM. This band at 532.56 cm −1 indicates the presence of Mn-O stretching vibratio 26 , further verifying the successful loading of MnO 2 . In conclusion, MRM has rich surface functional groups, which makes MRM have vigorous chemisorption activity.
The SEM micrographs of RM and MRM are shown in Fig. 3. It can be seen that the RM (Fig. 3a) surface shows the irregular distribution and the presence of large particle agglomerates and lamellar particles with many irregular small particles in their gaps. After MnO 2 loading (Fig. 3b), MRM was obviously covered by MnO 2 . Its surface was covered by many spherical, and flocculent particles appeared, showing a three-dimensional porous structure, which in turn provided more adsorption sites on the surface, pore channels and interlayer domains, thus improving the adsorption performance of MRM. Comparing the EDS energy spectra before and after modification (Fig. 4a,b), the content of manganese in the original red mud was not detected, but the content of  www.nature.com/scientificreports/ main metal elements in the red mud changed after MnO 2 loading. The increase of Mn content also indicated that the red mud had been successfully loaded with MnO 2 . Porosity and pore size distribution of RM and MRM were determined by nitrogen adsorption-desorption experiments, and the results are shown in Fig. 5. According to the standard of International Union of Pure and Applied Chemistry (IUPAC), the adsorption-desorption isothermal curves of both RM and MRM are of type VI (Fig. 5a). Both showed obvious hysteresis back loops due to capillary condensation, indicating the existence of mesoporous structures inside. Meanwhile, the closed hysteresis loops are similar to the H3-type hysteresis loops, indicating that the pore structure is very irregular. The specific surface area and pore size were calculated based on the BET model and the BJH model, respectively, and the results are shown in Table 1. The apparent increase in the specific surface area of MRM is due to the formation of manganese oxides on its surface, which is consistent with the successful loading of MnO 2 as shown in the SEM and EDS energy spectra. However, the average pore size of MRM is smaller than that of RM, which is since the loading of MnO 2 changes RM from dispersed particles to more dense particles and the number of pores increases. The pore size distribution curves (Fig. 5b) also show that the pore size distribution of MRM is broader and the number of pores increases. In addition, the increase in pore capacity and the number of pores facilitates the diffusion of metal ions into the interior of the adsorbent, which is also one of the advantages of MRM.    Fig. 6, it was found that the different ratios of MnO 2 and red mud had essential effects on the adsorption performance of MRM. The equilibrium adsorption capacity of the original red mud for Cd 2+ was 18.73 mg g −1 ; when the ratio of manganese dioxide to red mud was 1:4, the equilibrium adsorption capacity of the generated MRM for Cd 2+ in solution reached the highest, 44.93 mg g −1 , which was 26.20 mg g −1 higher than that of the original red mud for Cd 2+ ; the equilibrium adsorption capacities of the other five ratios were also higher than that of the original red mud. The improved adsorption capacity of the modified red mud was attributed to the surface loading of MnO 2 . The loading of MnO 2 on the red mud resulted in a higher surface activity, giving it a high adsorption affinity for the metal cation Cd 2+ . In addition, because the red mud material is a good carrier, MnO 2 can be better loaded and dispersed on its surface, enhancing the composite's adsorption performance 38 . When the ratio of MnO 2 to red mud was 1:1, the adsorption effect was not as good as the modified red mud with a ratio of 1:4. This may be due to the deposition of excess MnO 2 on the surface of the red mud, which leads to pore blockage and thus affects the adsorption effect 39 . Therefore, the modified red mud with a 1:4 ratio of MnO 2 to red mud was chosen for the following experiments.
Effect of the adsorbent dosage. The adsorbent dosage is one of the critical parameters to determine the adsorption capacity of the material at a fixed initial concentration. The adsorption capacity and removal efficiency curves of Cd 2+ on MRM are shown in Fig. 7. It can be seen from the figure that the removal efficiency of Cd 2+   www.nature.com/scientificreports/ showed an apparent increasing trend in the range of 0.2-4 g L −1 with the increase of the MRM dosage. This is because the adsorption sites available for MRM increased with the increase of the dosage, which led to the increase of the removal efficiency. When the dosage exceeded 2 g L −1 , the removal efficiency of Cd 2+ increased slowly and finally reached equilibrium. At the same time, the adsorption capacity of MRM decreased with the increase of adsorbent dosage. In general, when the action of adsorbent and heavy metal ions reached equilibrium, the removal rate no longer increased. Continuing to increase the dosage of adsorbent will reduce the adsorption efficiency of adsorbent per unit mass. Therefore, to save materials and consider the adsorption capacity and removal efficiency, the optimal dosage of adsorbent was determined to be 1 g L −1 for further experiments.
Effect of initial pH. The pH value is one of the most critical parameters to control the heavy metal adsorption process because it can influence the surface properties of the adsorbent and the ionic species in the solution. The effect of pH on the adsorption performance of MRM was analysed by varying the pH (2)(3)(4)(5)(6)(7)(8) and the results are shown in Fig. 8. The initial pH has a significant effect on the adsorption effect of Cd 2+ , and the general rule is that the adsorption rate increases with the increase of pH value. The removal rate of Cd 2+ increased from 21.18 to 86.70% when the pH was varied in the range of 2 to 6, and even the removal rate was close to 100% at pH 7 and 8. At lower pH (2)(3)(4), the presence of a higher concentration of H + in the solution caused the accumulation of a large amount of H + around MRM, which in turn led to a positive charge of manganese hydroxyl functional group due to protonation, as shown by the reaction equation in Eq. (1) 40 . The electrostatic repulsion was produced between free Cd 2+ in solution and positively charged MRM 41 . At the same time, high concentrations of H + competed with Cd 2+ for adsorption, resulting in a relatively low amount of adsorbed Cd 2+ . While at higher pH (4-6), there was less H + in the solution and the competition for adsorption was weaker. And the manganese hydroxyl functional group underwent deprotonation and became negatively charged (Eq. (2)), generating electrostatic repulsion with Cd 2+ . The resulting result is an increase in MRM adsorption efficiency. When pH > 7.4, Cd 2+ starts to produce precipitation, so the effect of pH on the adsorption performance of MRM under alkaline conditions is not discussed in this study.
The analysis of the Zeta potential of MRM at different pH conditions is shown in Fig. 9. It can be seen from the figure that before adsorption, the charge of MRM gradually increases with the increase of solution pH. The pH corresponding to when the zeta potential is zero is the isoelectric point of MRM (pHpzc = 2.2). After adsorption, the zeta potential of MRM decreases with the increase of pH and gradually stabilizes after pH > 4. From this, we can infer that Cd 2+ can be adsorbed and neutralize the negative charge on the MRM surface or form metal complexes 42 .
Effect of coexisting cations. The simulated test wastewater in this study does not contain other ions, while metal cations such as Na + , K + , Ca 2+ and Mg 2+ are very common in natural waters or actual wastewater and easily exhibit competitive adsorption with the target heavy metal ions 41 . The effects of different competing ions such as Na + , K + , Ca 2+ and Mg 2+ on Cd 2+ removal were investigated at an initial Cd 2+ concentration of 50 mg L −1 and a coexisting ion concentration of 0.2 mmol L −1 . The effect of coexisting ions on the adsorption of Cd 2+ by MRM is shown in Fig. 10. The results showed that Na + , K + , Ca 2+ and Mg 2+ in the solution all decreased the adsorption of Cd 2+ Compared with the solution without the addition of any coexisting cations, the inhibition magnitude of Cd 2+ adsorption by the same concentration of cations was in the order of Ca 2+ > Mg 2+ > K + > Na + , with Ca 2+ having the most substantial inhibition effect. This is due to the similar structure of Ca 2+ and Cd 2+ . The relatively small hydrated ionic radius of Ca 2+ makes it more easily to compete with Cd 2+ for adsorption sites, and Ca 2+ ions can rapidly bind to the active functional groups on the MRM 43,44 . More competing ions and impurities may exist in actual industrial wastewater and natural water bodies, such as many fibres, inorganic salts and pigments in paper industry wastewater. Therefore, the practical application of MRM requires further research on its removal effect on different types of wastewater.
Effect of the contact time and kinetics. Samples were taken for Cd 2+ concentration at contact times of 10, 20, 30, 60, 90, 120, 150, 180, 240, 300, 360, 420, 480 min. The effects of contact time on the adsorption performance of MRM and RM are shown in Fig. 11a. The adsorption capacity of both RM and MRM for Cd 2+ increased with time, but the adsorption capacity of MRM was always more significant than that of RM. RM reached saturation  www.nature.com/scientificreports/ at the contact time of 60 min, and the adsorption capacity was maintained at 18 mg g −1 or less. The adsorption process of Cd 2+ on MRM can be divided into two stages: the time at 0-180 min is the surface rapid adsorption stage, and the adsorption capacity of MRM increases to 44.72 mg g −1 ; after contact time more than 180 min, the adsorption of MRM on Cd 2+ entered the slow adsorption stage, the adsorption capacity maintained at about 45 mg g −1 , indicating that the adsorption equilibrium has been reached. This is because, at the early stage of adsorption, there are more free adsorption sites on the surface of MRM, the concentration gradient is large, and the adsorption rate is fast. However, with the increase of contact time, the number of free adsorption sites on the surface of MRM gradually becomes less, and the Cd 2+ already adsorbed on MRM diffuses into the interior of the  www.nature.com/scientificreports/ material along with the pore size, leading to the increasing diffusion resistance and the decrease of adsorption efficiency, and finally reaches the adsorption equilibrium. The adsorption kinetic model is one of the essential tools to study the adsorption mechanism. The adsorption mechanism of the adsorbent is explored by studying the rate of solute adsorption by the adsorbent and fitting the data using a kinetic model. This study used pseudo-first-order, pseudo-second-order, external diffusion kinetic, and intraparticle diffusion models to fit the obtained adsorption data. The specific equations are as follows. The correlation coefficient was used to determine the fitting effect, and the fitting results are shown in Table 2.
The pseudo-first-order kinetic model: The pseudo-second-order kinetic model: The external diffusion kinetic model: The intraparticle diffusion model: where q t (mg g −1 ) is the amount of adsorption on the adsorbent at time t, q e (mg g −1 ) is the equilibrium Cd 2+ adsorption capacity, k 1 (g mg −1 min −1 ) and k 2 (g mg −1 min −1 ) are the rate constants of pseudo-first-order adsorption and pseudo-second-order adsorption, respectively, c 0 is the initial Cd 2+ concentration, c t is the Cd 2+ concentration at time t, k ext (1/min) is the constant of external diffusion, k p (mg g −1 min 1/2 ) is the intraparticle diffusion rate constant. The fitting curves obtained are shown in Fig. 11a-f. From Table 2, the fit coefficients obtained from .the pseudo-first-order kinetic model and pseudo-secondorder kinetic model were 0.9733 and 0.9904, respectively. The pseudo-second-order kinetic model was more suitable to describe the adsorption process of MRM, indicating that the adsorption process was dominated by chemisorption. This chemisorption involved electron sharing or electron transfer between the adsorbate and the adsorbent. The model fit of RM showed that the pseudo-first-order kinetic model was more consistent with the adsorption process of Cd 2+ by RM, indicating that the adsorption process was physical adsorption. Applying the external diffusion model, we obtained the curves as Fig. 11c,d, and lnc t showed a linear relationship with t in the first 180 min with R 2 of 0.93, indicating that external diffusion was the dominant process in the initial adsorption process.
Thus, the intraparticle diffusion model analysed the rate-limiting phase on the Cd 2+ adsorption process. According to Fig. 11e,f, q t is linear for t 0.5 , but the straight line does not pass through the origin, indicating that intraparticle diffusion is not the only rate-limiting step, and there may be other steps involved in controlling the adsorption rate. The fitting curves were divided into three linear parts possessing their respective rate constants, which followed the order k p,1 > k p,2 > k p,3 . The first linearity represents the film diffusion process with the most considerable adsorption rate. This is due to the initial adsorption process in which the vacant adsorption sites on the MRM surface can rapidly bind to Cd 2+ . The second linearity is due to the intraparticle diffusion process. In this process, the surface adsorption sites gradually decrease, and the adsorption capacity starts to approach the maximum. Moreover, the reduction of adsorption sites causes the adsorbed Cd 2+ to enter the interior of the  Fig. 11 and Table 2, it can be seen that MRM has higher metal adsorption capacity than RM. MRM has more active sites, larger specific surface area and pore volume, which contributes to the sufficient contact between metal ions and MRM and effectively reduces the diffusion resistance of metal ions in solution. All these results indicate that the modification of biochar by MnO 2 is necessary and practical.
Effect of the initial Cd 2+ concentration and adsorption isotherm. The effect of the initial concentration of Cd 2+ on the adsorption effect of MRM is shown in Fig. 12. As can be seen from the figure, when the initial concentration of Cd 2+ was low (10-100 mg L −1 ), the adsorption capacity of Cd 2+ increased rapidly with the increase of concentration, and the removal rate decreased continuously. This is because there are enough adsorption sites on the MRM surface for Cd 2+ adsorption when the initial concentration of heavy metals is low. At the initial concentration of Cd 2+ higher than 150 mg L −1 , the adsorption capacity of Cd 2+ was unchanged due to the limited effective adsorption sites, and the maximum adsorption capacity of Cd 2+ is 97.85 mg g −1 .
The adsorption patterns of MRM on Cd 2+ were investigated at three temperatures of 25 °C, 45 °C and 65 °C, as shown in Fig. 13. The common isothermal adsorption models are mainly Langmuir model and Freundich model, which were fitted to the data, and the fitted parameters are shown in Table 3.
The Langmuir model:  www.nature.com/scientificreports/ where q e (mg g −1 ) is the equilibrium Cd 2+ adsorption capacity, q m (mg g −1 ) is the maximum Cd 2+ adsorption capacity, c e is the equilibrium Cd 2+ concentration of the solution, and K L (L mg −1 ) is the Langmuir adsorption constant. The essential characteristics of the Langmuir isotherm can be expressed in terms of the dimensionless constant separation factor or equilibrium parameter R L , which is expressed as follows 47 .
where c 0 is the initial Cd 2+ concentration and K L (L mg −1 ) is the Langmuir adsorption constant. For unfavorable adsorption, R L > 1; when R L = 1, linear adsorption happens; R L = 0 describe irreversible adsorption; favorable adsorption occurs while 0 < R L < 1 14 .
The Freundlich model: where q e (mg g −1 ) is the equilibrium Cd 2+ adsorption capacity, c e is the equilibrium Cd 2+ concentration of the solution, n is the Freundlich exponent, and K F (mg 1−1/n L 1/n g −1 ) is Freundlich adsorption constant. 1/n value indicates the strength of the effect of concentration on adsorption, with 1 < n < 10 indicating favourable adsorption 35,48 . From the fitted parameters in the Table 3, it can be seen that the Langmuir model better reflected the isotherms of the three tesmperatures than the Freundlich model, It indicated that the adsorption of MRM on Cd 2+ was a monomolecular layer adsorption process, and the adsorption sites on the surface of MRM were uniformly distributed. The theoretical maximum adsorption amounts obtained by the fitting were 103.5857, 116.0454 and 118.4734 mg g −1 , similar to the experimental results. The calculations showed that the separation coefficients R L in the Langmuir model were all in the range of 0 to 1, indicating that MRM was favourable for Cd 2+ adsorption.

Adsorption thermodynamics.
To investigate the thermodynamic behavior of Cd 2+ adsorption on MRM, the adsorption process of MRM on Cd 2+ was studied at three temperatures of 25 °C, 45 °C and 65 °C. The thermodynamic calculations of adsorption were performed using the temperature-dependent adsorption equation and Van't Hoff equation 40 .
The Gibbs free energy (ΔG) of the adsorption reaction was calculated with the following expression.
where q e (mg g −1 ) is the equilibrium Cd 2+ adsorption capacity, c e is the equilibrium Cd 2+ concentration of the solution, R (8.314 J mol −1 K −1 ) is the universal gas constant, T (K) is the absolute temperature and K f (L g −1 ) is the distribution coefficient. The enthalpy change of adsorption (ΔH) and the entropy change of adsorption (ΔS) are calculated as follows: The relevant thermodynamic parameters ΔG, ΔH and ΔS were calculated from the above expressions(Eqs. (10)-(12)) and Fig. 14. As shown in Table 4. ΔH values are negative, which indicates the adsorption of Cd 2+ by MRM is a heat-taking process. The values of ΔG are always negative at different temperatures and concentrations, representing that Cd 2+ removal is favourable and spontaneous. ΔS values are negative, indicating that the adsorption process has an increased degree of freedom between the solid-liquid interface, which may be related to the ion exchange process 49 . The positive values of ΔH and ΔS reveal that the adsorption is a heat absorption and entropy increase process, corresponding to the isotherm study.   www.nature.com/scientificreports/ Adsorption mechanism. To further determine the adsorption mechanism of Cd 2+ on MRM, FTIR and XPS were used to analyse the elemental composition and chemical morphology of MRM before and after the adsorption of Cd 2+ . As shown in Fig. 16, no new vibrational peaks appeared in the FTIR after Cd 2+ adsorption, but each functional group's peak intensity and wavenumber changed. The vibrational peak of -OH was weakened and shifted from 3405.43 to 3406.37 cm −1 , which indicated that the hydroxyl group in MRM was involved in the adsorption process. In addition, the adsorption of Cd 2+ led to a significant weakening and displacement of the Mn-OH absorption peak intensity, which was due to the entry of Cd 2+ into the MRM and the ion exchange with the protons on the O atom of the Mn-OH group 20 . Thus, the Mn-OH group played an essential role in Cd 2+ adsorption mainly through ion exchange 50 , demonstrating that MRM has some ion exchange capacity. The Mn-O functional group is similarly displaced, which may be due to the formation of an inner complex between the metal ion and MRM, in which a Cd-O bond was formed. As shown in Fig. 17a, after the adsorption of Cd 2+ , the absorption peak of Cd 3d appeared in the XPS spectrum, indicating that MRM successfully adsorbed Cd 2+ . By fitting, two peaks of Cd 3d 5/2 and Cd 3d 3/2 with characteristic binding energies of 405.23 eV and 411.97 eV, respectively, appeared in the XPS spectrum of Cd 3d (Fig. 17b). The appearance of the characteristic peak of Cd 3d 5/2 indicated the formation of endogenous compounds of Cd (-OCdOH), CdCO 3 , or Cd(OH) 2 51,52 . As shown in Fig. 17c,d, it was found that the Mn 2p spectra before and after the adsorption of MRM were very similar, with the characteristic peak binding energies of 653.88 and 642.11 eV for Mn 2p 3/2 and Mn 2p 5/2 , respectively, indicating that Mn (IV) was always present before and after the adsorption.
According to previous reports, the O 1s profile consists of three peaks corresponding to hydroxide oxygen (OH − ), lattice oxygen (O 2− ), and oxygen and hydroxide in molecular water (i.e., physisorbed, chemisorbed, and structural H 2 O and water in poor electrical contact with the mineral surface) 26,53 . As can be seen from Fig. 17e, before adsorption of Cd 2+ , the O 1s spectrum had three peaks at binding energies 530.112, 532.03, and 533.06 eV, corresponding to metal oxides (M-O), hydroxyl-containing metals (M-OH), and oxygen in adsorbed H 2 O, respectively. The intensity of the hydroxyl peak was significantly higher than the remaining two oxygen forms, confirming the presence of many hydroxyl groups on the surface of the MRM sample. After adsorption of Cd 2+ (Fig. 17f), the M-OH percentage content decreased from 55.47 to 36.85%, indicating that Cd 2+ interacts with M-OH to form hydroxy complexes (hydroxy composites) or ion exchange. In contrast, the M-O percentage content increased from 37.74 to 49.89%, indicating the formation of Mn-O-R (R: functional group) or Cd-O may be formed on the surface of MRM after adsorption. In addition, because the reaction of MRM with Cd 2+ consumed M-OH and increased Mn-O, this process formed H 2 O, thus increasing the percentage of H 2 O 54 . In summary, it can be concluded that the mechanism of cadmium adsorption by MRM may include ion exchange, internal sphere complexation and electrostatic attraction.
Comparison of MRM with other adsorbents for Cd 2+ removal. To further evaluate the adsorption capacity of MRM, the Cd 2+ removal effect of MRM and other commonly used heavy metal adsorbent materials were compared under the same experimental conditions(pH = 6, T = 25 °C, dosage = 1 g L −1 , t = 240 min and initial Cd 2+ concentration = 50 mg L −1 .), and the obtained results are shown in Table 5. From Table 5, it can be seen that the removal of Cd 2+ by MRM is better than many adsorbent materials reported in the literature, so MRM can be used for medium heavy metal removal in aqueous solutions, and since RM is an industrial waste and MnO 2 is a cheap metal oxide, MRM can be prepared at a small cost.  Table 6, from which we can see that the leaching toxicity of the red mud itself was weaker, and the leaching toxicity of the red mud modified by MnO 2 was lower than the former. The MRM leaching metal concentrations were all lower than the limits of the standards for drinkingwater quality of China (GB5749-2006). Even the content of such major elements as Al and Ca was less than the standard mass value in GB 5749-2006. Therefore, the data of leached metals indicated the safety of MRM for use in aqueous environment.

Conclusions
In this study, the behavior and mechanism were conducted using amorphous MnO 2 modified red mud to enhance the removal of cadmium, At the same time, the influencing factors of the adsorption capacity of Cd 2+ in different process conditions were studied through adsorption experiments, including the adsorption kinetics, adsorption isotherm and adsorption thermodynamics of Cd 2+ . Based on the findings of this study, the following conclusions and recommendations are drawn: 1. MnO 2 -modified red mud effectively improved the specific surface area and pore capacity and good adsorption effect on Cd 2+ . The adsorption performance of MRM on Cd 2+ was affected by the ratio of MnO 2 and red mud, pH, dosage, initial concentration of Cd 2+ , coexisting ions and contact time. 2. The pseudo-first-order kinetic model is more suitable than the pseudo-second-order kinetic model to describe the adsorption process of MRM, indicating that the chemisorption of Cd 2+ on the surface of MRM occurs through electron sharing or electron transfer. The Langmuir model had a higher fitting coefficient than the Freundlich model, indicating that the adsorption of MRM on Cd 2+ was a monomolecular layer adsorption process, and the surface adsorption sites are uniformly distributed. The fitted theoretical maximum adsorption amount is 103.5857 mg g −1 , and the separation coefficient R L is in the range of 0 to 1, indicating that MRM was favourable for Cd 2+ adsorption. The thermodynamic adsorption study indicated that the adsorption reaction was a spontaneous heat absorption process. 3. Combined XPS and FTIR studies, the adsorption mechanism of MRM on Cd 2+ was speculated to be that MRM relied on the ion exchange, internal sphere complexation and electrostatic attraction of its surface hydroxyl groups to achieve the adsorption of Cd 2+ . 4. Using MnO 2 to modify red mud not only provides a good method to recover red mud as a waste water treatment material, but also greatly improves the adsorption performance of the red mud, which can be used as an adsorbent to treat cadmium-containing wastewater.

Materials and methods
Sources of the RM and reagents.   Characterisation analysis. In this study, the crystal phase and composition of the sample was analysed by the X-ray diffractometer (XRD, RIGAKU, Japan), with the conditions of Cu Kα as the radiation, λ = 1.5406 Å, and scanning speed of 5° min −1 . The scanning electron microscope (SEM) and energy-dispersive spectroscopic (EDS, JSM-7500F, Japan) were obtained to investigate surface morphology and elemental composition changes of the samples. The specific surface area of the samples was measured by N 2 adsorption-desorption experiments at 200 °C with a degassing time of 6 h using a Mike 2460 physisorption instrument and calculated by the Brunauer-Emmett-Teller (BET). The pore size of the sample was measured and calculated by Barret-Joyner-Halender (BJH) method in the N 2 adsorption-desorption experiments. Fourier transform infrared spectrometer (FTIR, Nicolet 670, USA) was used to measure the infrared spectra of the samples and analyse the functional groups contained in the samples. In addition, the samples' X-ray photoelectron spectroscopy (XPS, Thermo Scientific K-Alpha, USA) was made with Al Kα (hv = 1486.6 eV) rays as the excitation source corrected for C1s = 284.80 eV binding energy to investigate the adsorption mechanism further.
Batch adsorption and detection of Cd 2+ . Batch adsorption experiments investigated the adsorption performance of MRM on Cd 2+ . Cd(NO 3 ) 2 ·4H 2 O was used as the cadmium source, and 1.386 g of Cd(NO 3 ) 2 ·4H 2 O was weighed and dissolved in 500 ml of deionised water to produce 1000 mg L −1 of cadmium stock solution.
In this study, the effects of different ratios of MnO 2 and red mud, dosage, initial pH, initial Cd 2+ concentration, coexisting ions, contact time and temperature on the adsorption performance of the materials were investigated sequentially. A certain amount of MRM was added to a 250 mL conical flask containing 100 mL Cd 2+ solution and then shaken in a shaker bath (7HZ-82, Changzh·ou) for 4 h at 200 rpm. The temperature range was 25-65 °C, the dosage range was 0.02-1.0 g, the pH range was 2-8, the initial Cd 2+ concentration range was 10-300 mg L −1 , the contact time range was 10-480 min, and the coexisting Na + , K + , Ca 2+ and Mg 2+ were selected at a concentration of 0.2 mmol L −1 . After oscillation, all samples were filtered through a 0.45 μm membrane, and the Cd 2+ concentration in the solution was measured by atomic absorption spectrophotometer (AA 6100, Techcomp) at 228.8 nm. The removal rate and equilibrium adsorption capacity of Cd 2+ were calculated according to Eqs. (14) and (15).
where E (%) is Cd 2+ removal efficiency, c 0 and c e (mg L −1 ) are initial and equilibrium Cd 2+ concentration in the solution, m (g) is the mass of MRM, q e (mg g −1 ) is equilibrium Cd 2+ adsorption capacity, and V(L) is solution volume.

Leaching metal test.
To examine the risk of toxic leaching of the material, the Chinese standard determination procedure (HJ/T299-2007) was adopted for MRM. A mixture of concentrated sulfuric acid and concen- www.nature.com/scientificreports/ trated nitric acid with a mass ratio of 2:1 was added to water (about 2 drops of the mixture in 1 L of water) to prepare a leaching agent with pH 3.20 ± 0.05. The components to be measured (red mud and MRM) and leaching agent were added at a liquid-solid ratio of 10:1 (kg L −1 ), shaken for 20 h at 120 rpm and 25 °C. The number of hazardous components leached from the supernatant was determined after centrifugation.
Desorption and regeneration studies. The MRM adsorbent was regenerated by five cycles (the adsorption test conditions were at pH = 6, T = 25 °C, dosage = 1 g L −1 , t = 240 min and initial Cd 2+ concentration = 50 mg L −1 ). The adsorbed MRM was adsorbed with 0.1 mol L −1 HCl to remove Cd 2+ , then washed three times with deionized water and dried to constant weight in a vacuum drying oven at 60 °C.

Data availability
All data included in this study are available from the corresponding author on reasonable request.