Recyclable Magnetic Iron Immobilized onto Chitosan with Bridging Cu Ion for the Enhanced Adsorption of Methyl Orange

Abstract

Chitosan (CS) is a natural and low-cost adsorbent for capturing metal ions and organic compounds. However, the high solubility of CS in acidic solution would make it difficult to recycle the adsorbent from the liquid phase. In this study, the CS/Fe₃O₄ was prepared via Fe₃O₄ nanoparticles immobilized onto a CS surface, and the DCS/Fe₃O₄-Cu was further fabricated after surface modification and the adsorption of Cu ions. The meticulously tailored material displayed the sub-micron size of an agglomerated structure with numerous magnetic Fe₃O₄ nanoparticles. During the adsorption of methyl orange (MO), the DCS/Fe₃O₄-Cu delivered a superior removal efficiency of 96.4% at 40 min, which is more than twice the removal efficiency of 38.7% for pristine CS/Fe₃O₄. At an initial MO concentration of 100 mg L⁻¹, the DCS/Fe₃O₄-Cu exhibited the maximum adsorption capacity of 144.60 mg g⁻¹. The experimental data were well explained by the pseudo-second-order model and Langmuir isotherm, suggesting the dominant monolayer adsorption. The composite adsorbent still maintained a large removal rate of 93.5% after five regeneration cycles. This work develops an effective strategy to simultaneously achieve high adsorption performance and convenient recyclability for wastewater treatment.

1. Introduction

Increasingly discharged industrial wastewater from textile mills and metal electroplating plants greatly endanger the whole ecosystem with the rapid development of industrialization [1]. Contaminants, especially heavy metal ions (e.g., Cu²⁺) and organic dyes (e.g., methylene orange, MO), seriously threaten living organisms even at trace levels through binding with enzymes and proteins [2,3]. Therefore, efficient wastewater treatment concerning Cu²⁺ and MO removal is urgently needed. At present, various methods were used to remove metal ions and organic compounds from solution, including chemical precipitation, ion exchange, membrane separation, electrochemical method, redox method and adsorption [4,5]. Among these approaches, due to simple operation, affordable price and high efficiency, adsorption was widely applied to uptake contaminants in wastewater [6].

Recently, natural organic or inorganic adsorbents have attracted increasingly attention owing to the plentiful sources, low cost and environmental friendliness compared with other artificial adsorbents, such as activated carbon, bentonite, and biochar [7]. Chitosan (CS) is a renewable polysaccharide from the deacetylation of natural chitin (in the skeleton of arthropods and cell wall of fungi), which the resource reserve ranks second in nature (only less than cellulose) [8]. The CS shows good biodegradation and functional activity, and it can be regarded as a promising adsorbent material [9]. However, the high solubility of CS in acidic solution would make it is difficult to recycle the CS-based adsorbent from the liquid phase. Using surface modification to introduce magnetic material (e.g., Fe₃O₄) into CS would address the recovery issue [10]. Some researchers reported the removal of heavy metal ions (such as Cu²⁺, Co²⁺, Ni²⁺, and Pb²⁺) by the magnetic CS material with considerable adsorption capacity [11,12]. Moreover, the modified magnetic CS was also applied as adsorbent to remove organic dyes (such as blue 21, food yellow 3, and acid yellow 23) from wastewater [13]. The above reports usually used the fresh CS-based adsorbent, while there was almost no report using the spent adsorbent after the adsorption of contaminants. It is still a huge challenge to adsorb organic dye using the spent CS-based adsorbent after the removal of heavy metal ion.

In this work, the Fe₃O₄ nanoparticles immobilized onto CS (CS/Fe₃O₄) were prepared via the facile coprecipitation method (Figure 1a). Using diethylenetriamine for the surface modification of CS to provide abundant amino groups (–NH₂), the CS transformed to DCS. During the adsorption of Cu ions, these amino groups could bridge Cu²⁺, and the DCS/Fe₃O₄ further formed the spent adsorbent (defined as DCS/Fe₃O₄-Cu). Such a meticulously tailored adsorbent exhibits several advantages [14,15]: (a) the adsorbent can maintain the stable particle structure in solution; (b) the magnetic adsorbent can be easily separated and recycled from the liquid phase environment; and (c) the bridging cationic Cu²⁺ can adsorb anionic organic matter (e.g., MO anions). Accordingly, the DCS/Fe₃O₄-Cu material delivered enhanced adsorption capability of MO compared with the pristine CS/Fe₃O₄ adsorbent. Furthermore, the kinetics, isotherm and thermodynamics were also investigated to reveal the adsorption mechanism. The purpose of the present study was to devise an available method for preparing a recyclable CS-based adsorbent with low cost and high adsorption performance.

2. Results and Discussion

2.1. Physical Properties of DCS/Fe₃O₄-Cu

Figure 2 presents the effects of adsorption conditions on Cu²⁺ removal efficiency for the DCS/Fe₃O₄ material [16]. From Figure 2a, the Cu²⁺ removal efficiency increased with the reduction in pH value. At pH = 5, the removal efficiency reached the saturated value (>80%). Figure 2b showed that the Cu²⁺ removal rate decreased with the increasement of initial Cu²⁺ concentrations (50–300 mg L⁻¹), while the corresponding adsorption capacity increased. As seen in Figure 2c, with the absorbent dosage increased from 1 to 8 g L⁻¹, the Cu²⁺ removal rate increased from 44.1% to 93.4%. With the prolonging of contact time, the Cu²⁺ removal efficiency gradually increased (Figure 2d). Under the optimal conditions of pH = 5, initial Cu²⁺ concentration = 100 mg L⁻¹, absorbent dosage = 4 g L⁻¹, and contact time = 300 min, the removal efficiency of Cu²⁺ could reach >80%. The spent adsorbent DCS/Fe₃O₄-Cu was collected and further used for the subsequent adsorption of MO. Similarly, the effects of adsorption conditions on Cu²⁺ removal efficiency for the CS/Fe₃O₄ were also conducted. As seen from Figure S1 (Supplementary Materials), the optimal conditions for CS/Fe₃O₄ to remove Cu²⁺ were pH = 5, initial Cu²⁺ concentration = 100 mg L⁻¹, absorbent dosage = 4 g L⁻¹, and contact time = 300 min, while the removal efficiency was only recorded as ~60%.

The X-ray diffraction (XRD) patterns of CS, CS/Fe₃O₄, DCS/Fe₃O₄, and DCS/Fe₃O₄-Cu are illustrated in Figure 3a. The CS sample showed a wide peak located at 2θ = ~21.5°, which was a typical characteristic of semi-crystalline CS [9]. For the CS/Fe₃O₄ sample, several new peaks appeared at around 30.3°, 36.2°, 43.2°, 57.4°, and 63.1°, corresponding to the (220), (311), (400), (511), and (440) planes of cubic Fe₃O₄ lattice (JCPDS. 75-0449) [17]. The characteristic peak of CS disappeared in the CS/Fe₃O₄ pattern due to the disordered crystal structures that resulted from interactions of Fe₃O₄ nanoparticles with the CS matrix [17]. For the DCS/Fe₃O₄ and DCS/Fe₃O₄-Cu, both samples maintained the diffraction peaks of Fe₃O₄, suggesting the structural integrity of composite materials. On the DCS/Fe₃O₄-Cu pattern, a new sharp peak at ~24.0° could be attributed to the bridging Cu²⁺ [18]. Using the Scherrer equation and Segal method, the grain size and crystallinity index could be calculated in Table S1. The wide diffraction peak of CS suggested its low crystallinity (grain size of 3.4 nm and crystallinity index of 63%). The DCS/Fe₃O₄-Cu sample exhibited high crystallinity (grain size of 27.9 nm and crystallinity index of 86%).

The Fourier-transform infrared (FTIR) spectra of CS, CS/Fe₃O₄, DCS/Fe₃O₄, CS/Fe₃O₄-Cu and DCS/Fe₃O₄-Cu samples are shown in Figure 3b. Two bands located at around 3480 and 2900 cm⁻¹ were assigned to the stretching vibration of hydroxyl groups (O–H) and C–H bond, respectively [9]. The bands near 1640 and 1080 cm⁻¹ could be related to the bending vibration of amino groups (–NH₂) and stretching vibration of the C–O bond in CS molecules [17]. Apart from the pure CS sample, the new band that appeared at approximately 600 cm⁻¹ in other four samples corresponded to the stretching vibration of the Fe–O bond [19]. For the DCS/Fe₃O₄ sample, a sharp peak at ~3420 cm⁻¹ was assigned to the stretching vibration of –NH₂ from diethylenetriamine molecule. After Cu²⁺ adsorption, the peak at 3420 cm⁻¹ disappeared, and the DCS/Fe₃O₄-Cu sample expressed a new band at about 890 cm⁻¹, attributing to the interaction between bridging Cu ions and –NH₂ groups (Cu–N) [20,21]. While for the CS/Fe₃O₄ sample, there was no similar Cu–N bond on the spectrum.

Liquid nitrogen adsorption–desorption isotherms of CS/Fe₃O₄, DCS/Fe₃O₄ and DCS/Fe₃O₄-Cu are shown in Figure 3c. All three samples presented the type IV isotherms with an obvious hysteresis loop. The hysteresis loop indicated the existence of abundant mesoporous structures (pore diameter from 2 to 50 nm) in materials [22]. According to the classical Brunauer–Emmett–Teller (BET) equation, the specific surface areas of CS/Fe₃O₄, DCS/Fe₃O₄ and DCS/Fe₃O₄-Cu samples were determined as 38.4, 28.7, and 25.8 m² g⁻¹, respectively. The pore texture parameters of above three samples can be found in Table S2. Moreover, Figure S2 shows the Barrett–Joyner–Halenda (BJH) pore size distributions of these samples [9]. The CS/Fe₃O₄ material presented the largest pore volume of 0.198 cm³ g⁻¹ with an average pore diameter of 20.6 nm, while the DCS/Fe₃O₄-Cu sample presented the smallest pore volume of 0.142 cm³ g⁻¹ (average pore diameter of 22.0 nm). The small specific surface area and pore volume of DCS/Fe₃O₄-Cu indicated that modified molecules blocked or occupied part of the pores of the CS/Fe₃O₄ material [22].

The thermal behaviors of the CS/Fe₃O₄, DCS/Fe₃O₄, and DCS/Fe₃O₄-Cu samples were studied by thermogravimetric (TG) analysis (Figure 3d). The results showed a small decline below 180 °C in the TG curves, corresponding to the evaporation of adsorbed water molecules. A significant decline ranging from 180 to 700 °C could be ascribed to the pyrolysis of chitosan [23]. Subsequently, the TG curves were basically stable above 700 °C. Among three samples, the CS/Fe₃O₄ and DCS/Fe₃O₄ expressed the large weight loss, suggesting their high carbon contents. For the DCS/Fe₃O₄-Cu sample, its small weight loss could be attributed to bridging Cu ions.

The magnetic hysteresis loops of CS/Fe₃O₄, DCS/Fe₃O₄ and DCS/Fe₃O₄-Cu are displayed in Figure 3e. The magnetic properties of adsorbents confirmed that Fe³⁺ and Fe²⁺ can form magnetic Fe₃O₄ particles in an alkaline environment. The saturation magnetizations of as-prepared samples were all beyond 27 emu g⁻¹, which were convenient for the separation and recycling of spent adsorbents from the liquid phase environment [17].

The zeta potential curves (Figure 3f) can give the pH_PZC values (pH value at zeta potential of 0) of DCS/Fe₃O₄ and DCS/Fe₃O₄-Cu. The sample surface exhibited negative charges when the pH > pH_PZC; and it exhibited positive charges when the pH < pH_PZC [24]. The pH_PZC value of the DCS/Fe₃O₄ material was fitted as 4.88. Due to the bridging Cu²⁺, the pH_PZC value of the DCS/Fe₃O₄-Cu sample shifted to 8.66. During the MO adsorption process, the MO solution exhibited weak basicity. In this case, the DCS/Fe₃O₄ adsorbent would have negative charges to repel MO anions. The positively charged DCS/Fe₃O₄-Cu adsorbent would incline to adsorb MO anions, and the small zeta potential (nearly zero) made DCS/Fe₃O₄-Cu particles agglomerate in solution (digital photo in Figure 1) [25].

The X-ray photoelectron spectroscopy (XPS) survey spectra of DCS/Fe₃O₄, CS/Fe₃O₄-Cu and DCS/Fe₃O₄-Cu samples are present in Figure 4a. Characteristic peaks at around 285, 399, 531 and 710 eV were attributed to the C 1s, N 1s, O 1s and Fe 2p, respectively [26]. A new peak that appeared at ~934 eV in the DCS/Fe₃O₄-Cu spectrum was assigned to Cu 2p.

The O 1s high-resolution XPS spectra in Figure 4b demonstrated three components: peaks at around 531.9, 530.3 and 528.9 eV could be assigned to the Fe–O, C–O and O–H, respectively [27]. The N 1s high-resolution XPS spectra of the DCS/Fe₃O₄ and CS/Fe₃O₄-Cu samples in Figure 4c could be deconvoluted into two peaks at 398.3 and 400.3 eV, which were possibly related to the pyrrole N of C–N and N–H, respectively [20,27]. A new peak appeared at 399.2 eV in the DCS/Fe₃O₄-Cu curve, which resulted from the coordination between Cu ions and amino groups (Cu–N) [20,28]. From Figure 4d, the Cu 2p high-resolution XPS spectrum of DCS/Fe₃O₄-Cu sample displayed two obvious spin-orbit doublet peaks of Cu 2p_3/2 (932.4 eV) and Cu 2p_1/2 (953.1 eV), corresponding to the bridging Cu ions [20,29]. While for the CS/Fe₃O₄-Cu sample, the weak Cu 2p peaks could be assigned to the trace Cu ions adsorbed on the adsorbent surface.

Figure 5 shows the scanning electron microscopy (SEM) images of CS, Fe₃O₄, CS/Fe₃O₄, DCS/Fe₃O₄ and DCS/Fe₃O₄-Cu samples. As shown in Figure 5a, the CS sample displayed the irregular particle morphology. The nano-Fe₃O₄ material presented an irregular nanosphere shape with the particle size of 50–100 nm (Figure 5b). For the CS/Fe₃O₄ sample, Figure 5c exhibits an agglomerated structure, in which numerous Fe₃O₄ particles are immobilized onto the surface of the chitosan matrix. The corresponding mapping images and energy-dispersive spectrometer (EDS) spectrum (Figure S3) of the CS/Fe₃O₄ sample confirmed the existence of the Fe element [12]. After diethylenetriamine modification and Cu²⁺ adsorption, the DCS/Fe₃O₄ (Figure 5d) and DCS/Fe₃O₄-Cu (Figure 5e) also maintained the original agglomerated structure. The mapping images (Figure 5f) and EDS spectrum (Figure S3) of the DCS/Fe₃O₄-Cu sample further verified the presence of the Cu element [30]. The sub-micron size of DCS/Fe₃O₄-Cu can make the adsorbent powder easily separated and recycled from solution [9].

2.2. Adsorption of MO on DCS/Fe₃O₄-Cu

The effect of light scattering from adsorbent particles was investigated by the UV-Vis absorbance spectra of MO. From Figure S4, with the prolonging of time, the intensity of the MO peak gradually decreased, and the peak shifted a little. All concentration data were calculated according to the peak position.

The comparison of MO removal efficiency (initial MO concentration of 40 mg L⁻¹, 25 °C) for CS/Fe₃O₄, DCS/Fe₃O₄ and DCS/Fe₃O₄-Cu adsorbents was exhibited in Figure 6a. For all curves, the removal rate increased rapidly in the first 10 min and then gradually reached the adsorption–desorption equilibrium status at an adsorption time of 40 min. Among three adsorbents, the pristine CS/Fe₃O₄ sample had the minimum removal efficiency (38.7%). Due to the bridging Cu cations, the DCS/Fe₃O₄-Cu material delivered the maximum removal rate of 96.4% for MO anions (more than twice that of CS/Fe₃O₄) [31].

Adsorption kinetics was used to investigate the effect of MO concentration. The classic pseudo first-order (Equation (1)) and pseudo second-order (Equation (2)) equations were given as [32]:

$$\ln \left(Q_{\mathrm{e}1}-Q_{\mathrm{t}}\right)=\ln Q_{\mathrm{e}1}-K_1·t$$

(1)

$$\frac{t}{Q_{\mathrm{t}}}=\frac{1}{K_2·Q_{\mathrm{e}2}^2}+\frac{t}{Q_{\mathrm{e}2}}$$

(2)

where Q_t (mg g⁻¹ min⁻¹) is the amount of MO adsorbed at time t (min), K₁ (min⁻¹) and K₂ (g mg⁻¹·min⁻¹) are the rate constants of pseudo first-order and pseudo second-order equations.

Using the linear fitting method, the kinetic parameters of MO adsorption onto the DCS/Fe₃O₄-Cu sample at 25 °C are listed in Table S3. With MO initial concentration increased from 10 to 100 mg L⁻¹, the pseudo-second-order plots all expressed the good fitting degree (Figure 6b), with a high correlation coefficient (R²) of ~0.998 [33]. Meanwhile, as seen in Figure S5, the fitting degree of the pseudo-first-order model was relatively low. The good fitting of the pseudo-second-order model indicated that the adsorption rate of MO molecules onto the DCS/Fe₃O₄-Cu adsorbent was dependent on the amount of adsorption sites rather than the MO concentration in solution [34].

Langmuir and Freundlich isotherms were usually applied to evaluate the adsorption process. The Langmuir model was based on an assumption of homogenous monolayer adsorption, while the Freundlich model was attributed to the heterogeneous multilayer adsorption [35]. The equations of Langmuir (Equation (3)) and Freundlich (Equation (4)) can be expressed as the following [36]:

$$\frac{C_{\mathrm{e}}}{Q_{\mathrm{e}}}=\frac{1}{Q_{\mathrm{m}}·K_{\mathrm{L}}}+\frac{C_{\mathrm{e}}}{Q_{\mathrm{m}}}$$

(3)

$$\ln Q_{\mathrm{e}}=\ln K_{\mathrm{F}}+\frac{\ln C_{\mathrm{e}}}{n}$$

(4)

where Q_m (mg g⁻¹) is the maximum adsorption capacity at the monolayer, and K_L (L mg⁻¹) is a constant related to the free energy of Langmuir adsorption; K_F (mg g⁻¹) (mg L⁻¹)ⁿ is the Freundlich constant, and n (dimensionless) is the heterogeneity factor.

Figure S6 showed the linear fitting data of MO adsorption onto the DCS/Fe₃O₄-Cu adsorbent at 25, 35 and 45 °C. From Figure S6a, the Langmuir model gave a high fitting degree (R² > 0.97 in Table S4), while the R² values of the Freundlich model were only recorded as 0.90–0.92 (Figure S6b). Accordingly, the Langmuir isotherm was more suitable to describe the current experimental data, revealing the mainly monolayer adsorption behavior for the MO capture [37]. As seen in Figure 6c, with the environment temperature increased from 25 to 45 °C, the corresponding adsorption capacities decreased (e.g., at an initial MO concentration of 100 mg L⁻¹, the Q_e value decreased from 144.60 mg g⁻¹ to 121.83 mg g⁻¹) [10].

The effect of temperature on the adsorption activity was further investigated via the thermodynamics analysis. Several representative thermodynamic parameters such as the Gibbs free energy change (ΔG, kJ mol⁻¹), enthalpy change (ΔH, kJ mol⁻¹), and entropy change (ΔS, J mol⁻¹ K⁻¹) were determined by Equations 5, 6 and 7 [23,28].

$$\Delta G=-R·T·\ln K$$

(5)

$$K=\frac{{\alpha }_{\mathrm{ad}}}{{\alpha }_{\mathrm{e}}}=\frac{{\gamma }_{\mathrm{ad}}·C_{\mathrm{ad}}}{{\gamma }_{\mathrm{e}}·C_{\mathrm{e}}}$$

(6)

$$\ln K=\frac{\Delta S}{R}-\frac{\Delta H}{R·T}$$

(7)

where R is the gas constant (8.314 J mol⁻¹ K⁻¹), T (K) is absolute temperature; K is the thermodynamic equilibrium constant, γ_ad and γ_e are activity coefficients, and C_ad is the MO concentration adsorbed on the sample (mg L⁻¹). In the dilute solution, the activity coefficient approached unity, and the K value could be determined graphically by plotting ln(C_ad/C_e) against C_ad and extrapolating to zero [19].

Figure S7 displays the linear fitting result of relationship C_ad-ln(C_ad/C_e). The K constants were calculated from the intercept with the Y-axis (4.275 at 25 °C, 2.449 at 35 °C, 1.642 at 45 °C). The ΔG values of 25, 35 and 45 °C could be obtained as −3.60, −2.29 and −1.31 kJ mol⁻¹, respectively (Table S5). Negative ΔG values revealed the spontaneous nature of the adsorption reaction [38]. Furthermore, the ΔG values were usually in the range of 0 to −20 kJ mol⁻¹ and −80 to −400 kJ mol⁻¹ for physical and chemical adsorptions, respectively [39]. In this case, the MO adsorption onto the DCS/Fe₃O₄-Cu adsorbent was the physical adsorption (electrostatic force between Cu cations and MO anions) [39].

From Figure 6d, the ΔH and ΔS values were obtained from the slope and intercept of 1/T-lnK plots. Using Equation 7, the ΔH and ΔS values were calculated as −37.78 kJ mol⁻¹ and −114.82 J mol⁻¹·K⁻¹. The negative ΔH value suggested the exothermic nature for this adsorption behavior, and low temperature was favorable for the adsorption reaction [38,39], which was consistent with the adsorption isotherms in Figure 5c.

2.3. Regeneration Study of DCS/Fe₃O₄-Cu

The good regeneration ability of the adsorbent is very important for the cost factor in industrial application [40]. Figure 7a expressed the removal rate of MO onto the DCS/Fe₃O₄-Cu sample during repeated absorption–desorption cycles (MO initial concentration of 40 mg L⁻¹, environment temperature of 25 °C). For the first absorption process, the DCS/Fe₃O₄-Cu adsorbent delivered a high removal efficiency of 96.4%. Owing to the convenient recyclability from the magnetic Fe₃O₄, the removal rate was still recorded as 93.5% after five regeneration cycles.

The morphology of the DCS/Fe₃O₄-Cu material after five absorption cycles (labeled as DCS/Fe₃O₄-Cu-MO) is displayed in Figure 7b. The DCS/Fe₃O₄-Cu-MO sample still maintained the original agglomerated structure after multiple times absorption–desorption processes [9,25]. From Figure 7c, the mapping images exhibited uniform distribution of C, Fe and Cu elements of the DCS/Fe₃O₄-Cu-MO particle. The S element mapping image further confirmed the capture of MO molecules.

Figure 7d presents the FTIR spectra of DCS/Fe₃O₄-Cu and DCS/Fe₃O₄-Cu-MO samples. It could be found that both samples kept original characteristic peaks of the O–H bond (3480 cm⁻¹), C–H bond (2900 cm⁻¹), –NH₂ bond (1640 cm⁻¹), C–O bond (1080 cm⁻¹), Fe–O bond (600 cm⁻¹), and Cu–N bond (~890 cm⁻¹). These characteristic peaks suggested the structural integrity of material during the absorption process [9]. After MO absorption, a new sharp peak located at 1600 cm⁻¹ could be corresponded to the N=N bond in an adsorbed MO molecule [23,28]. Two new sharp peaks at 1030 and 1116 cm⁻¹ were assigned to the S–O and S=O bonds in the MO molecule, respectively. These results also verified the successful absorption of MO onto the DCS/Fe₃O₄-Cu adsorbent.

The XPS survey spectra of DCS/Fe₃O₄-Cu and DCS/Fe₃O₄-Cu-MO samples are displayed in Figure 7e. The characteristic peaks of C 1s (285 eV), N 1s (399 eV), O 1s (531 eV), Fe 2p (710 eV), and Cu 2p (934 eV) were all retained after MO absorption. A new peak that appeared at ~167 eV on the DCS/Fe₃O₄-Cu-MO curve was the characteristic peak of S 2p [41]. The S 2p high-resolution XPS spectrum (Figure 7f) further displayed two peaks centered at 168.2 and 166.9 eV. The former peak could be attributed to the S–O bond, and the latter peak was indexed to be the S=O bond, respectively [42].

3. Materials and Methods

3.1. Materials and Reagents

Chitosan (CS, deacetylation degree of 96%), diethylenetriamine, epichlorohydrin (C₃H₅ClO), ethylenediamine tetraacetic acid (EDTA), FeCl₃·6H₂O, FeSO₄·7H₂O, CuSO₄·5H₂O, and methyl orange (MO) were purchased from Energy Chemical Industrial Inc. (Shanghai, China). Ammonia (NH₃·H₂O), acetic acid (CH₃COOH) and ethanol (C₂H₅OH) were purchased from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). All chemical reagents were analytical grade and used directly without further purification.

3.2. Synthesis of CS/Fe₃O₄

The CS/Fe₃O₄ was prepared by a one-step coprecipitation method, as shown in Figure 1a. Typically, 1 g of CS was dissolved in 200 mL of acetic acid solution with ultrasonic oscillation. Then, 4.7 g of FeCl₃·6H₂O and 2.4 g of FeSO₄·7H₂O were dissolved in 22 mL deionized water and added into the CS solution. After, 40 mL of NH₃·H₂O (28%) was dropwise added into solution under continuous stirring at 40 °C for 20 min. Subsequently, 6 mL of C₃H₅ClO was added into the solution at 60 °C for 3 h to form the CS/Fe₃O₄.

3.3. Synthesis of DCS/Fe₃O₄-Cu

As shown in Figure 1b, 1 g of CS/Fe₃O₄ and 3.1 g of diethylenetriamine were dissolved in 70 mL acetic acid solution plus 100 mL of deionized water under vigorous stirring at 80 °C for 12 h. The as-prepared DCS/Fe₃O₄ was added into CuSO₄·5H₂O solution to adsorb Cu ions (II). After washing with CH₃COOH/C₂H₅OH/deionized water and drying, the magnetic DCS/Fe₃O₄-Cu product could be obtained.

3.4. Characterization

The crystalline structures of materials were measured by the AXS D8 Advance X-ray diffractometer (XRD, Cu Kα source; Bruker, Karlsruhe, Germany). The element compositions of samples were identified via the X-ray photoelectron spectrometry (XPS) using an ESCA LAB MK-II X-ray photoelectron spectrometer (VG Scientific, St Leonards, UK). The surface functional groups of materials were confirmed by Fourier transform infrared (FTIR) spectroscopy via the KBr pellet using a Nicolet iS10 spectrometer (Thermal Fisher Scientific, Waltham, MA, USA). The surface morphologies of samples were investigated by the field-emission scanning electron microscopy (FESEM, SU8010, Hitachi, Tokyo, Japan). The thermal stabilities of materials were tested on the Mettler–Toledo (Zurich, Switzerland) thermogravimetric (TG) Stare ESI-0910 instrument from room temperature (25 °C) to 900 °C (heating rate of 3 °C min⁻¹) under Ar atmosphere. The specific surface areas and pore size distributions of samples were determined via N₂ adsorption–desorption experiments using the Brunauer–Emmett–Teller (BET) method by an Autosotrb-IQ2-MP-XR (Micromeritics, Norcross, GA, USA) gas sorption analyzer (samples were degassed at 200 °C in vacuum for 18 h). The magnetic properties of materials were measured by a vibrating-sample magnetometer (VSM, PPMS DynaCool 9, Quantum Design, San Diego, CA, USA) at room temperature, and the hysteresis loops were recorded in the field range of 30,000 Oe. The zeta potential of samples was measured by a Zetasizer Nano Z with a surface zeta potential accessory (Malvern, Worcs, UK).

3.5. Adsorption Experiments

To prepare the DCS/Fe₃O₄-Cu, a certain amount of the DCS/Fe₃O₄ was added into 30 mL of CuSO₄·5H₂O solution. The solution pH (2–6, adjusted using 0.1 M H₂SO₄ or 0.1 M NaOH), adsorbent dosage (1–8 g L⁻¹), initial Cu ions concentration (50–300 mg L⁻¹), and contact time (0–700 min) were investigated. Using a thermostatic shaker (SHZ-C, Suzhou, China) at 25 °C/150 rpm during the whole adsorption process, the concentration of Cu ions was determined by the inductively coupled plasma optical emission spectrometer (ICP-OES, Avio 500, PerkinElmer, Waltham, MA, USA).

Similarly, for the batch adsorption experiments of MO, 0.015 g of CS/Fe₃O₄, DCS/Fe₃O₄ or DCS/Fe₃O₄-Cu adsorbent was added into 30 mL of MO aqueous solution (MO initial concentration of 10–100 mg L⁻¹) at 25 °C. The MO concentration in solution was determined by an UV-3600 PLUS spectrophotometer (Shimadzu, Kyoto, Japan) at a wavelength of 460 nm. The removal efficiency (R, %) and adsorption capacity (Q_e, mg g⁻¹) can be calculated by the following Equation 8 and Equation 9, respectively.

$$R=\frac{C_0-C_{\mathrm{e}}}{C_0}\times 100\%$$

(8)

$$Q_e=\frac{\left(C_0-C_e\right)·V}{m}$$

(9)

where C₀ (mg L⁻¹) is the initial concentration of Cu²⁺ or MO, C_e (mg L⁻¹) is the equilibrium concentration, V (L) is the volume of solution, and m (g) is the mass of adsorbent.

3.6. Desorption and Regeneration

The EDTA solution, C₂H₅OH, and ultrapure H₂O were used as the desorption agents. For each adsorption/desorption cycle, the adsorbent (DCS/Fe₃O₄-Cu) after MO adsorption was separated and collected by a magnet. Adding a small amount of NaOH under ultrasonic oscillation for 20 min, the spent absorbent could be regenerated by thoroughly washing using the EDTA, C₂H₅OH plus ultrapure H₂O several times and drying. Such a regenerated DCS/Fe₃O₄-Cu adsorbent was further used for the removal of MO in solution.

4. Conclusions

In this study, the CS/Fe₃O₄ material was prepared via Fe₃O₄ nanoparticles immobilized onto the CS matrix. After diethylenetriamine modification and the adsorption of Cu²⁺, the CS/Fe₃O₄ further formed the DCS/Fe₃O₄-Cu. The composite material displayed the sub-micron size of an agglomerated structure with numerous Fe₃O₄ nanoparticles (particle size of 50–100 nm). Due to the electrostatic force between bridging Cu cations and MO anions, the DCS/Fe₃O₄-Cu adsorbent exhibited an enhanced adsorption performance. For example, the DCS/Fe₃O₄-Cu delivered an excellent removal efficiency of 96.4% for MO at 40 min, which was much higher than 38.7% of the pristine CS/Fe₃O₄. At 100 mg L⁻¹ MO initial concentration and 25 °C, the DCS/Fe₃O₄-Cu exhibited the maximum adsorption capacity of 144.60 mg g⁻¹. The experimental data were well explained by the pseudo-second-order model and Langmuir isotherm, indicating the dominant monolayer adsorption for MO. A thermodynamic study confirmed the exothermic process of MO adsorption. The magnetic adsorbent can be easily separated and recycled from solution, so the DCS/Fe₃O₄-Cu still maintained a large removal efficiency of 93.5% after five regeneration cycles. This work constructed a magnetic DCS/Fe₃O₄-Cu adsorbent with high adsorption capability and convenient recyclability, which provided the potential application for wastewater treatment.