Magnetized inulin by Fe3O4 as a bio-nano adsorbent for treating water contaminated with methyl orange and crystal violet dyes

Current work focuses on fabricating a new bio-nano adsorbent of Fe3O4@inulin nanocomposite via an in-situ co-precipitation procedure to adsorb methyl orange (MO) and crystal violet (CV) dyes from aqueous solutions. Different physical characterization analyses verified the successful fabrication of the magnetic nanocomposite. The adsorbent performance in dye removal was evaluated by varying initial dye concentration, adsorbent dosage, pH and temperature in 5110 mg/L, 0.10.8 g/L, 111 and 283–338 K, respectively. Due to the pH of zero point of charge and intrinsic properties of dyes, the optimum pHs were 5 and 7 for MO and CV adsorption, respectively. The correlation of coefficient (R2) and reduced chi-squared value were the criteria in order to select the best isotherm and kinetics models. The Langmuir model illustrated a better fit for the adsorption data for both dyes, demonstrating the maximum adsorption capacity of 276.26 and 223.57 mg/g at 338 K for MO and CV, respectively. As well, the pseudo-second-order model showed a better fitness for kinetics data compared to the pseudo-first-order and Elovich models. The thermodynamic parameters exhibited that the dye adsorption process is endothermic and spontaneous, which supported the enhanced adsorption rate by increasing temperature. Moreover, the nanocomposite presented outstanding capacity and stability after 6 successive cycles by retaining more than 87% of its initial dye removal efficiency. Overall, the magnetized inulin with Fe3O4 could be a competent adsorbent for eliminating anionic and cationic dyes from water.


Synthesis of Fe 3 O 4 @inulin.
First, 60 mL of deionized water was added to a 250 mL two-mouth glass flask. The amount of 0.7 g of FeCl 2 ⋅4H 2 O was poured into the flask under stirring, and then the amount of 1.9 g of FeCl 3 ⋅6H 2 O was added to the solution. The mixture was stirred for 5 min until completely dissolved to obtain a uniform solution. Next, 0.75 g of inulin was slowly added to the solution in several steps under stirring for 25 min. The ambient temperature was raised to 85 °C under a nitrogen atmosphere, and 10 mL of NH 3 OH was injected dropwise by syringe for 70 min. The color of the solution turned black, indicating the formation of Fe 3 O 4 nanoparticles during the injection. The mixture was further stirred for 30 min to allow the remaining NH 3 OH to react. Next, the magnetized inulin nanoparticles were separated from the solution with a magnet and washed www.nature.com/scientificreports/ with deionized water four times to remove the unreacted materials from the nanocomposite. The magnetized inulin was transferred to an oven and placed at 55 °C for 15 h to dry. Finally, the powdered nanoparticles were obtained for the experiments and characterization. Figure 1c illustrates the molecular scheme of the synthesis procedure of Fe 3 O 4 @inlulin.
Characterization. The crystal network of the synthesized adsorbent nanoparticles was examined by an XRD analysis (STOE STADI-MP model, Germany) with Cu Ka radiation (= 1.54 Å) and 2θ of 10°-80°. The adsorbent functional groups were evaluated using the FTIR analysis (Spectrum Rx1 model, Perkin Elmer Co.) in 400-4000 cm −1 . The magnetization property of the prepared nanocomposite was assessed by a VSM device (LBKFB model, Iran) under the ambient condition and in the range of − 4000 to 4000 Oe. The structure, morphology and elemental distribution of the adsorbent were identified by SEM-EDX analysis (MIRA3 TESCAN). By heating the sample at 120 °C to eliminate any impurities in its structure, it was analyzed by BET analysis (Micromeritics, Model ASAP 2020, USA) at 77 K. Several NaNO 3 solutions (0.1 M) with 50 mL volume were added to different flasks at the ambient condition to measure the point of zero charges (PZC Adsorption experiments. The dye removal experiments have been conducted to determine the adsorption performance of synthesized magnetic nanocomposites as follows. The batch adsorption was carried out in an Erlenmeyer flask with a volume of 250 mL by adding 0.1-0.8 g of the magnetic bio-nano adsorbent in the controlled operational conditions for the isotherm and kinetics investigations. The initial concentration (C dye ) of MO and CV varied in the range of 25-200 mg/L. The volume of every solution was 100 mL, and the adsorption temperature was changed from 10 to 65C. The solution pH was considered in the range of 1-11 to find the best conditions for each dye removal. To do so, the calibration curves were first plotted for both dyes at different pHs. Next, the adsorption experiments were conducted at different pHs and various adsorption times (15-180 min, every 15 min sampling). Finally, the calibration curve with the pH proportional to the corresponding solution pH was used to measure the concentration of the remaining dye in the aqueous solution. The dye concentration was determined by UV-Vis spectrophotometer (Rayleigh/UV 2601) at a wavelength of 464 nm and 590 nm for MO and CV, respectively. The formulas of the removal efficiency and adsorption capacity (Q) are as follows 28 : in which, C 0 and C e (mg/L) are the initial and equilibrium concentrations of dyes in the aqueous solution, respectively. The terms m (g) and V (L) are the mass of the nanocomposite and the volume of the aqueous solution, respectively. Each experimental run was repeated three times, and all reported results are the averaged data with a deviation error of ± 5%.

Results and discussion
Adsorbent characterization. FTIR analysis is a practical tool for detecting the functional groups in the structure of materials. Figure 2a presents 30 , which is abundant in the inulin backbone. The adsorption band at about 3095 cm −1 is associated with the stretching vibration of the CH bond 31 . The peak at 1101 cm −1 is related to the asymmetric stretching vibration of the COC bond, and an absorption band at 1052 cm −1 is attributed to the stretching vibration of the CO bond 32 . Furthermore, the appearance of FeO peak in the Fe 3 O 4 @inulin spectrum confirms the successful fabrication of the nanocomposite.
The crystalline structure of Fe 3 O 4 @inulin nanocomposite studied by XRD analysis is shown in Fig. 2b 33 , the average crystalline size for Fe 3 O 4 @inulin bio-nano adsorbent is 10.9 nm.
SEM-EDX analysis of the Fe 3 O 4 @inulin adsorbent has been conducted to evaluate its morphology and elemental distribution, as presented in Fig. 2c,d. As shown in Fig. 2c at low magnification, the surface morphology of Fe 3 O 4 @inulin nanocomposite is rough, heterogeneous and irregular with some cavities. However, according to the SEM images at higher magnifications, the adsorbent surface depicts a relatively uniform, regular www.nature.com/scientificreports/ distribution of spherical nanoparticles on the Fe 3 O 4 @inulin nanocomposite surface, with a particle diameter range of 10-20 nm. Moreover, the elemental mapping in Fig. 2d illustrates a uniform distribution of all elements, including carbon, oxygen and iron, within the adsorbent structure, exhibiting the maximum content of oxygen with a contribution of 48.61%. N 2 adsorption/desorption isotherm of the synthesized Fe 3 O 4 @inulin nanoparticles is shown in Fig. 3a, resulting from the BET analysis. This curve exhibits the mesoporous structure of the adsorbent supported by the type IV isotherm based on the IUPAC classification 34 . Such mesopores are favorable for enhancing the specific surface area of the material. According to the calculations by the BJH method 35 , the specific surface area of 66.15 m 2 /g, pore volume of 0.181 cm 3 /g, and average pore diameter of 11.027 nm are obtained for Fe 3 O 4 @inulin adsorbent. The factor pH PZC shows the charge of the adsorbent surface, the pH value where the net surface charge is zero. The pH f against pH i values are plotted in Fig. 3b to determine this factor for the prepared Fe 3 O 4 @inulin composite, indicating a pH PZC of approximately 6. Therefore, the adsorbent surface will be negatively charged at pH values higher than 6, while it will be positively charged otherwise.
The TGA method has been used to assess the thermal resistance of magnetic Fe 3 O 4 @inulin in the temperature range of 100-600 °C under an air atmosphere. Figure 3c illustrates the weight loss curve of pure inulin and Fe 3 O 4 @inulin materials. Both substances have initially experienced a weight loss of about 9% due to the elimination of free water and moisture evaporation. The most significant decomposition rate related to inulin exists in the range of 240-310 °C with a reduction in its weight by about 50%, which might be associated with dehydration. However, the most considerable weight loss of the Fe 3 O 4 @inulin nanocomposite occurs at 240-285 °C with a reduction of approximately 20%, representing the primary degradation of magnetic composite possibly caused by selective dehydration. The residual weights until 600 °C for inulin and Fe 3 O 4 @inulin are about 25% and 65%, respectively. As a result, adding magnetic Fe 3 O 4 nanoparticles into the inulin matrix can significantly increase its thermal stability due to strong intermolecular interactions. The magnetic characteristic of magnetized inulin is evaluated by VSM analysis. Figure   Effect of operating conditions. The adsorption experiments were performed at different pHs (1-11) to find the best solution pH for each dye. The other operating factors were constant, including the temperature of 25 °C, the adsorbent dosage of 0.5 g/L, the dye concentration of 25 mg/L, and the adsorption time of 180 min. As known, molecular aggregation occurs in aqueous solutions when the pH value is very low, resulting in considerable aggregates; thus, it is reasonable to detect the optimum pH for every solution. Figure 4a shows the removal efficiency of both dyes at different solution pHs. The maximum adsorption rates of about 95% and 91% were obtained at a pH of 5 and 7 for MO and CV dyes, respectively. Such adsorption trends by pH variation can be explained by the pH ZPC curve, reported in Fig. 3b. The solution pH affects the net surface for amphoteric molecules containing both negative and positive charges. They become more positively or negatively charged by either gaining or losing protons. Because MO acts as a cationic dye in highly acidic solutions 36 , its removal efficiency is lower due to the electrostatic repulsion between cationic dye and negatively charged adsorbent surface. By further increasing the solution pH, the number of positive charges is reduced, achieving the maximum MO removal efficiency at pH 5. However, at a pH higher than 5, the excessive quantity of hydroxyl ions competing with MO molecules for adsorption sites and electrostatic repulsion will reduce its adsorption rate. Moreover, the color of the solution containing MO dye becomes shallow, not only because of the change in pH but also due to the association with the adsorption by the Fe 3 O 4 @inulin adsorbent. For CV adsorption, lower removal efficiency at an acidic pH might be due to excessive H + ions competing with CV dye cations for the adsorption sites 37 . Under a pH higher than the neutral solution, the CV removal efficiency does not change significantly. Therefore, the optimal pH value for CV removal was found to be 7. Figure 4b exhibits the effect of Fe 3 O 4 @inulin adsorbent dosage on the removal efficiency of both dyes under different concentrations in the range of 0.10.8 g/L. The optimal solution pH related to each dye was considered in this case study. The removal efficiencies by 0.5 g/L concentration of adsorbent are about 6.3 and 9.1 times when the adsorbent dosage is 0.1 g/L for MO and CV adsorption, respectively. This significant increment is due to the enhanced active functional groups and increased number of active binding sites in the nanocomposite. At dosages higher than 0.5 g/L, the removal efficiency increases insignificantly because of the saturation of adsorption sites. Hence, the optimum amount of adsorbent for adsorbing both dyes is selected as 0.5 g/L for the following examinations. www.nature.com/scientificreports/ where Q e is the amount of dye adsorbed at the equilibrium state, Q m is the Langmuir monolayer adsorption capacity, K L is the Langmuir isotherm constant, K F is the Freundlich isotherm constant, and n is the Freundlich exponent 39 . The term K D is the isotherm constant in the thermodynamic equilibrium constant in the adsorption process, R is the gas constant (8.314 J/mol K), and T is the temperature. Also, the terms K T and B T are the Temkin isotherm constant and a constant related to the heat of adsorption, respectively. The Langmuir isotherm is typically utilized when there is ideal monolayer adsorption on a homogeneous surface. The Freundlich isotherm is generally appropriate for nonideal adsorption on heterogeneous surfaces. The Freundlich isotherm is a purely empirical model, presuming that a great number and a variety of existing sites act simultaneously, each with a different free energy of sorption. The D-R model assumes that the adsorption equilibrium relation for a specific adsorbent-adsorbate combination can be expressed by the adsorption potential independent of temperature. The Temkin isotherm model presumes that the adsorption heat of all molecules is linearly reduced with the increment in coverage of the adsorbent surface and that adsorption is described by a uniform distribution of binding energies 40 . The isotherm modeling of MO and CV adsorption onto the Fe 3 O 4 @inulin composite is illustrated in Fig. 5 for two temperatures of 283 and 338 K. The results and fitting parameters related to all isotherm models at other temperatures within this range are reported in Table 1. In this study, the adsorbent dosage and adsorption time are 0.5 g/L and 180 min, respectively. At all temperatures, the Langmuir isotherm model exhibited a better prediction for the adsorption data related to both dyes due to a coefficient of determination (R 2 ) higher than 0.98 and the minimum reduced chi-squared value among all considered models. The temperature rise leads to an increase in the adsorption capacity of the nanocomposite, indicating that higher temperatures favor the MO and CV adsorption process because of its endothermic nature. Also, in the Freundlich isotherm model, the constant (K F ) value is higher at elevated temperatures, further demonstrating that the adsorption process is endothermic. This increment in adsorption rate can be explained by the temperature-induced breakage of some internal bonds at the composite active surface, leading to an enhanced number of adsorption sites, as well as the dye molecule diffusion into the composite surface cavities raised by a temperature trigger. Additionally, the values of 1/n in the Freundlich models at all temperatures are smaller than 1, demonstrating a favorable adsorption process.
Adsorption kinetics. Kinetics investigation for the pollutant adsorption from contaminated water can be worthwhile because it enables finding the adsorption equilibrium time, adsorption kinetics rate, and the adsorbate concentration in each phase after achieving the equilibrium state. Here, pseudo-first-order (PFO), pseudosecond-order (PSO), and Elovich kinetics models are used to better examine the adsorption mechanism of MO and CV dyes onto the Fe 3 O 4 @inulin nanocomposite. www.nature.com/scientificreports/ The PFO model assumes that the rate of variation in solute adsorption is directly proportional to the difference in saturated concentration and the amount of adsorptive solid adsorbed over time. The PSO kinetics describes that the adsorption rate is measured by the interaction between the adsorbate and adsorbent species. The Elovich model can estimate the surface and mass diffusion, activation and deactivation energy of the adsorption system. In this model, the adsorption rate of solute is exponentially reduced by increasing the adsorbed solute amount 41 . The non-linear equations of these models are as follows 42 : where Q t is the amount of dye adsorbed at time t, and the terms k 1 and k 2 are the rate constants for PFO and PSO models, respectively. Also, the terms and are the initial adsorption rate and the desorption constant in the Elovich model, respectively. The fitting results of kinetic data are shown in Fig. 6. Among these three models, the PSO model provides excellent predictions of the experimental kinetics data of the adsorption capacity of the prepared nanocomposite for both dyes. Table 2 presents the parameters resulting from fitting the kinetics models to the experimental data. The highest R 2 and lowest reduced chi-squared values confirm that the adsorption mechanism can be characterized by a PSO kinetics model. The applicability of this model implies that the adsorption of both dyes onto the Fe 3 O 4 @inulin adsorbent surface is a chemisorption process 43 , where the molecules of dyes have been bounded to the Fe 3 O 4 @inulin through surface exchange reactions. The dye adsorption mechanism by the prepared Fe 3 O 4 @inulin nanocomposite is also shown in Fig. 6b. The most contributive forces are electrostatic attraction and hydrogen bonding interactions (dipole-dipole and Yoshida). The contribution of each force in dye adsorption differs at different pHs, as its details are reported in the literature 44 . Thermodynamic study. The thermodynamic parameters of the enthalpy (ΔH°), entropy (ΔS°), and Gibbs free energy (ΔG°) were measured by the following correlations: www.nature.com/scientificreports/ where K c is the adsorption equilibrium constant. The thermodynamic data calculated from the above equations are reported in Table 3. The positive value of ΔH° proposed that the adsorption process was endothermic, which agreed with the increasing dye adsorption with temperature. The ΔG° with negative values indicated a spontaneous adsorption process. The ΔG° is reduced by increasing temperature, proposing that the adsorption process is more favorable at higher temperatures. Additionally, the ΔS° with positive quantities demonstrated enhanced randomness at the solution-solid interface because of several structural changes throughout the process. The surface of the composites and dye molecules in the aqueous solution was surrounded by hydration layers of (11)  www.nature.com/scientificreports/ water molecules. Throughout the adsorption process, the water molecules ordered in these hydration layers were disturbed and compelled, improving the degree of freedom in the adsorbent-dye interaction.
Adsorbent reusability. In addition to acceptable adsorption performance, easy reusability and multi-cycle utilization are critical factors in providing a feasible, scalable and efficient treatment system because it influences the adsorbent performance and reduces the operational cost. The ethanol and acetone with high dipole moments were harnessed as the desorption agents to regenerate the adsorbent. First, dye particles were adsorbed onto the Fe 3 O 4 @inulin. After the adsorption process, the dye-loaded particles of the adsorbent were regenerated via ethanol or acetone, exhaustively washed with DI water and then used for dye adsorption again. The Fe 3 O 4 @ inulin was reused six times to check its applicability in removing MO and CV dyes. The reusability results for the adsorption of both dyes are reported in Fig. 7 under their optimum solution pH and adsorbent dosage. Notably, the removal efficiency of Fe 3 O 4 @inulin magnetic composite decreases from 94.92 to 83.34% and 91.13-78.86% after six successive adsorption cycles for MO and CV dyes, respectively, maintaining more than 87% of their initial adsorption rates. The decrease in the adsorption performance with increasing cycle number might be due to the saturation of sorption sites by strongly adsorbed dye molecules. The Fe 3 O 4 @inulin exhibits high MO and CV removal rates with excellent reusability. Even though each sorbent has unique nature, intrinsic characteristics, advantages, and disadvantages, the adsorption capacity of the Fe 3 O 4 @inulin for MO and CV removal is compared with other bio-nano adsorbents. Table 4 compares different bio-adsorbents at their optimal solution pH. The best solution pH, isotherm and kinetics models are listed as well. As shown, the capacity of magnetized inulin is higher than that of similar magnetized biomaterials in removing both dyes. Such a performance approves that the Fe 3 O 4 @inulin is an excellent magnetized sorbent with good stability for adsorption of both anionic and cationic dyes under an appropriate solution pH.

Conclusion
Nowadays, dye-contaminated water has become a severe environmental concern due to growing industrialization worldwide. Thus, environmentalists and researchers should urgently consider developing efficient methods and materials to mitigate such problems. In this work, we synthesized a new bio-nano adsorbent, i.e., magnetized inulin with Fe 3 O 4 , to remove toxic anionic and cationic dyes from wastewater. Various physical and structural analyses were conducted to evaluate the fabrication of the magnetic nanocomposite. According to the FTIR spectra, the presence of the Fe-O bond approved the successful synthesis of Fe 3 O 4 @inulin with an average crystalline size of 10.9 nm obtained from its XRD pattern. The SEM-EDX results demonstrated a relatively uniform,  www.nature.com/scientificreports/ regular dispersion of spherical nanoparticles ranging from 10 to 20 nm on the composite surface. Based on the BET analysis, the adsorbent's specific surface area, pore volume, and average pore diameter were 66.15 m 2 /g, 0.181 cm 3 /g, and 11.027 nm, respectively. The VSM analysis showed that the nanocomposite's magnetic saturation, i.e., 40.21 emu/g, was about 34% lower than that for pure Fe 3 O 4 nanoparticles. Considering the pH PZC value of 6 and the anionic and cationic characteristics of both dyes, the optimal solution pH was 5 for MO and 7 for CV. According to the highest R 2 value (> 0.99) and the lowest reduced chi-squared, the Langmuir isotherm fitted better the experimental data with a maximum adsorption capacity of 276.26 mg/g for MO and 223.57 mg/g for CV at 338 K. Considering these statistical criteria, the pseudo-second-order model (R 2 > 0.99 and reduced chi-squared < 1) was a better kinetics model than pseudo-second-order and Elovich models in predicting the kinetics data of both dyes. In addition, the thermodynamic parameter demonstrated that the dye adsorption by Fe 3 O 4 @inulin was an endothermic and spontaneous process. Furthermore, after using the adsorbent six times, its removal efficiency maintained more than 87% of its initial adsorption rate for both pollutants, exhibiting excellent stability and reusability of the prepared nanocomposite. Overall, the comparison of the ability of different magnetized adsorbents in adsorbing cationic and anionic dyes verified that the Fe 3 O 4 @inulin could be a promising candidate for treating water contaminated with toxic dyes.

Data availability
Data are available with the permission of [Prof. Ali Maleki]. The data that support the findings of this study are available from the corresponding authors, [Prof. Ali Maleki