One-step hydrothermal synthesis of hydrophilic Fe₃O₄/carbon composites and their application in removing toxic chemicals

Abstract

Hydrophilic Fe₃O₄/C nanocomposites are prepared by a simple one-step hydrothermal reaction route using FeCl₃ and glucose as raw materials, and sodium acetate as an alkali source. The phase structure, morphology, and composition are characterized by X-ray diffraction, Raman spectroscopy, transmission electron microscopy, mass spectrometry, Fourier transform infrared spectroscopy, and X-ray photoelectron spectroscopy, revealing that the glucose plays dual roles in the formation of final products: (I) the reduction of partial Fe³⁺ to Fe²⁺ ions and the formation of size-controlled Fe₃O₄ particles; (II) the stabilization and further functionalization of Fe₃O₄ nanoparticles by coating carbonous layers. The influence of crucial reaction factors such as the glucose concentration, pH value, reaction temperature and time on the aimed products is also investigated. The prepared Fe₃O₄/C samples demonstrate typical ferromagnetic behaviors and high removal capacities in removing the toxic Cr(VI) ions and organic pollutant Rhodamine B from wastewater, together with facile magnetic separability and good recyclability.

---

1. Introduction

The synthesis of nanomaterials has attracted extensive research interest over the past few decades due to their superior physical, chemical, and biological properties to bulk materials.^1–3 Magnetic iron oxide Fe₃O₄ nanoparticles are a kind of novel functional nanomaterial, which have been widely used in catalysis, environmental protection, sensors, magnetic storage media, and clinical diagnosis and treatment, based on their advantages such as magnetic properties, chemical stability, biocompatibility, and low toxicity.^4–8 Currently, various strategies have been developed to prepare magnetite nanoparticles with the desired physical and chemical properties, including the coprecipitation of ferric and ferrous ions with a 1 : 2 molar ratio in a basic system,⁹ thermal decomposition of the metal organic precursor (iron(III)acetylacetonate) at high temperature,^10,11 hydrothermal synthesis,^12,13 nonaqueous route,^14,15 sonochemical synthesis¹⁶ and microemulsion.¹⁷ However, in these previous studies, the process usually involves in the nitrogen or argon protection to avoid the oxidation of Fe²⁺ and ensure the formation of Fe₃O₄ nanoparticles, inevitably undergoing the complicated and harsh synthetic condition. In addition, the used metal organic precursors in the thermal decomposition technique are high-cost and the conventionally used organic solutions and reducing agents, such as ethylene glycol and diethylene glycol, are inferior to the aqueous media in view of the cost and the potential environmental or biological risks. In order to keep free from the trivial, expensive, and toxic fabrication process, a simple and green synthesis route to prepare magnetic nanoparticles at low cost is in demand.

Size and surface functionality have been proved to be two crucial factors in determining the applicable performance of magnetite materials.^18,19 However, the aggregation of magnetite nanoparticles, caused by the crystal-face attraction, electrostatic and dipolar fields, usually results in the large-sized particles.⁵ Moreover, the naked magnetite nanoparticles are easily oxidized in air due to the high chemical activity, leading to the loss of magnetism. The dispersibility and stability of magnetic nanoparticles could be resolved by encapsulating the nanoparticles in polymeric or inorganic matrixes.^20–22 Especially, the water-dispersible magnetic particles coated with the hydrophilic layer are highly in demand with respect to the biomedical and adsorbent application, since the magnetic particles are usually hydrophobic. For example, Zhang et al.⁹ have prepared the β-cyclodextrin coated Fe₃O₄ using a co-precipitation of FeCl₃·6H₂O and FeCl₂·4H₂O in NH₃·H₂O solution containing the hydrophilic small organic molecule of β-cyclodextrin under the protection of nitrogen. Sun et al.²³ have firstly synthesized the oleic acid coated iron oxide nanoparticles by thermal decomposition of Fe(acac)₃ (acac = acetylacetonate) in the presence of oleic acid and oleylamine, and then used the ligand-exchange strategies to offer them hydrophilic surface and aqueous dispersibility. Ge and co-workers¹⁵ have synthesized highly water-dispersible magnetite by using the surfactant of poly(acrylic acid) as the coated layer. However, some non-ignorable problems confined their practical application, such as the toxic and non-biodegradable coated agents, or high cost, and multiple, complicated procedures. Direct synthesis of water-soluble Fe₃O₄ nanoparticles coated with the low-cost and eco-friendly agent is very desirable for potential applications.

Heavy metal ions and organic pollutants in water system are great threats to the environment and human health. Among various water treatment technologies, such as the chemical precipitation, ion-exchange, membrane filtration, adsorption, and coagulation–flocculation, adsorption has been a preferred one because of its simplicity, high efficiency, and availability of a large number of adsorbents.^24,25 As one of the potential adsorbents, the magnetic iron oxide nanoparticles and nanocomposites have received much attention due to the high removal capacity for heavy metal ions (As(V), Cr(VI), etc.) and organic pollutants, as well as the convenient separation caused by their unique magnetic properties.^25–28 Moreover, the water-soluble magnetic nanocomposites with the functional groups and porous coating structure could improve dispersion and adsorption behaviors, providing a solid basis for the wastewater treatment.

In this work, our efforts were focused on the design of simple, green, and low-cost synthesis route for the preparation of carbon-coated magnetite nanoparticles. The one-step synthesis of carbon-coated magnetite nanoparticles using the glucose assisted hydrothermal reaction method was performed. The reaction parameters including the glucose concentration, pH value, reaction temperature and time were systematically studied. Based on the optimized synthesis strategy, the particle sizes of magnetite nanoparticles were well controlled in the range of 120–140 nm due to the confinement of carbon layer. Furthermore, the gluconic acid and glucose were anchored on the surface of magnetite/carbon composite particles, offering a large number of surface hydroxyl groups, and thus the composites showed good water solubility and an excellent removal capacity for the heavy metal ion Cr(VI) and organic pollutant Rhodamine B (RhB).

2. Experimental

2.1 Sample preparation

Ferric chloride (FeCl₃), glucose and sodium acetate, obtained from Tianjin Guangfu Chemical Co., were used as chemical reagents for the preparation of magnetite/carbon nanoparticles. All the chemicals were of analytical grade and used as received without further purification.

In a typical synthesis, 0.3, 0.6 or 1.1 g of glucose was firstly dissolved in 80 mL of deionized water under vigorous stirring, followed by the addition of 1.6 g of FeCl₃. Subsequently, 4.0 g of NaOAc was introduced to adjust the pH value of the solution to ca. 5. After stirring for 30 min, the final mixture was transferred to a Teflon-lined autoclave (150 mL capacity) and heated at 200 °C for 10 h. The resulting product was filtrated, washed with water, and dried at 60 °C overnight, which was labeled as Fe₃O₄/C-a, -b, or -c for the carbon-coated Fe₃O₄ composite sample obtained with 0.3, 0.6 or 1.1 g of glucose, respectively. For comparison, a control experiment without the addition of glucose was also carried out. Moreover, various factors in affecting the aimed products were also explored in the same synthesis procedure, including the glucose concentration, pH value, reaction temperature and time.

2.2 Sample purification

The prepared sample was stirred for 1 h, with the pure carbon spheres floating on the surface of the aqueous solution due to the lower density than those of Fe₃O₄ and Fe₃O₄/C. Then, the Fe₃O₄ and Fe₃O₄/C were collected by a magnet from the aqueous solution and put into the deionized water again, which continually stirred for 30 min and remained still for 10 min. Since the bigger particle size and larger density of the uncovered Fe₃O₄ compared to those of Fe₃O₄/C, the sediments in the bottom were abandoned and the upper suspension were obtained. In the end, the upper suspension was centrifuged at 8000 rpm for 10 min to obtain the final purified Fe₃O₄/C.

2.3 Characterization

Transmission electron microscopy (TEM) was performed on a JEOL JEM 2010F at 200 kV. All the samples subjected to TEM measurements were ultrasonically dispersed in ethanol and drop-casted onto copper grids covered with carbon film. X-ray diffraction (XRD) patterns were recorded on a Bruker D8 Focus diffractometer with Cu K_α radiation (λ = 1.5406 Å) operated at 40 kV and 40 mA. The diffraction data were collected over the 2θ angle of 10° to 80° at a scan rate of 6° min⁻¹. Raman spectroscopy was performed on a Renishaw-1000 Raman spectrometer. Thermogravimetric analysis (TGA) of the samples was conducted on a TA SDT Q600 analyzer in nitrogen atmosphere with a heating rate of 10 °C min⁻¹. N₂ adsorption–desorption isotherms were recorded on a Quantachrome NOVA 2000e sorption analyzer at liquid nitrogen temperature (77 K). The specific surface area was obtained by the Brunauer–Emmett–Teller (BET) method, and the pore size distribution was derived from the adsorption branches of isotherms using the DFT method. Mass spectrometry (MS) measurement was carried out on a Shimadzu LCMS-2020 mass spectrometer under the following conditions: interface voltage, 4.5 kV; DL temperature, 250 °C; heat block temperature, 200 °C. Fourier transform infrared (FT-IR) spectrum was recorded on a Bruker Vector 22 FT-IR spectrophotometer using a KBr pellet. X-ray photoelectron spectroscopy (XPS) measurement was performed on a Kratos Axis Ultra DLD (delay line detector) spectrometer equipped with a monochromatic Al K_α X-ray source (1486.6 eV). The XPS survey spectra were recorded with a pass energy of 160 eV, and high-resolution spectra with a pass energy of 40 eV. Binding energies were calibrated by using the containment carbon (C 1s 284.6 eV). Magnetization studies were performed at room temperature using a LDJ 9600 VSM magnetometer.

2.4 Adsorption measurements

For Cr(VI) adsorption: the aqueous solutions with different Cr(VI) concentrations (20, 40, 60, 80 and 100 mg L⁻¹) were prepared by dissolving different amounts of K₂Cr₂O₇ in deionized water. The pH values of the solution were adjusted to 4 by hydrochloric acid. 20 mg of Fe₃O₄/C-b was added into 100 mL of Cr(VI) solution with different concentrations. The mixture was continually stirred for 2.5 h in dark, followed by sampling the solution at defined time intervals. For analysis of Cr(VI), 5 mL of obtained clear solution was diluted with different amount of deionized water to ensure that the total Cr(VI) mass in the 50 mL colorimetric cylinder was less than 50 μg. Then 5 mL of diluted Cr(VI) solution and 2 mL of diphenylcarbazide were successively added into the colorimetric cylinder, diluted to 50 mL with deionized water, and adjusted pH to 1–2 (by adding H₂SO₄ (1 : 5, v/v) under the test of the pH-meter). The adsorbed capacity of Cr(VI) was monitored by measuring the UV absorption at λ_max = 540 nm of the initial and final solutions based on the Cr(VI) standard curve formula of C = 8.25092A − 0.00214, where C represents the Cr(VI) concentration, and A is the adsorption intensity.

In order to investigate the recycling potential, the Cr(VI)-loaded Fe₃O₄/C-b was separated from the solution by a conventional magnet, and ultrasonically cleaned in deionized water for 30 min, and subsequently soaked in 0.25 mol L⁻¹ NaOH for 3 h. The resulted Fe₃O₄/C-b sample was rinsed with 0.05 mol L⁻¹ NaOH and distilled water in succession, and such washing procedures were repeated until the rinsed water becomes colorless. Finally, the regenerated product was dried at 60 °C overnight before reuse as adsorbent.

For RhB adsorption: 20 mg of the Fe₃O₄/C-b sample was placed into a tubular quartz reactor filled with 100 mL of RhB aqueous solution (10⁻⁵ mol L⁻¹). The mixture was stirred for 2.5 h in dark at room temperature (30 °C). A magnet of about 0.002 T was used for collecting the magnetite nanoparticles in the solution. Finally, the content of RhB in the solution was measured by a SP-752 spectrophotometer at λ_max = 554 nm.

3. Results and discussion

3.1 Glucose assisted hydrothermal synthesis of Fe₃O₄/C nanocomposites

Fig. 1 shows the XRD patterns of the prepared samples obtained from the hydrothermal method with and without the use of glucose. The diffraction peaks of the Fe₃O₄/C-b prepared with the assistance of glucose are well consistent with the characteristic of the cubic Fe₃O₄ (a = b = c = 8.396 Å, Fd3m (227), JCPDS 19-0629) or γ-Fe₂O₃ (maghemite, JCPDS 39-1346), while the sample prepared in the absence of glucose exhibits the typical feature of the α-Fe₂O₃. This indicates the crucial role of glucose in the formation of Fe₃O₄ or γ-Fe₂O₃. However, the specific structural phase of the Fe₃O₄/C-b sample can still not be identified between Fe₃O₄ and γ-Fe₂O₃ only from the XRD data since they have similar XRD patterns. Raman spectroscopy is thus performed to further distinguish the structural phase of Fe₃O₄/C-b, as the two main bands of Fe₃O₄ centered at 668 (A_1g) and 535 (T_2g) cm⁻¹, and the ones of γ-Fe₂O₃ are located at 350, 500, and 720 cm⁻¹.¹² As shown in Fig. 2, the two main peaks at 663 and 535 cm⁻¹ are observed, assignable to the typical feature of Fe₃O₄, suggesting the presence of iron oxide in the form of Fe₃O₄ nanoparticles in Fe₃O₄/C-b. In addition, two broad peaks in the range of 1500–1650 cm⁻¹ and 1350–1450 cm⁻¹ are also observed in the Raman spectrum (Fig. 2 (inset)), which can be assigned to the crystalline graphitized carbon (G-band) and disordered amorphous carbon (D-band), respectively.^29,30 The much stronger peak intensity of G-band than that of D-band implies that glucose, as the carbon source, is partially pyrolyzed to form graphitized carbon layer after the hydrothermal treatment.

Fig. 1 XRD patterns of as-prepared Fe₃O₄/C-b and α-Fe₂O₃.

Fig. 2 Raman spectra of Fe₃O₄/C-b nanocomposites.

Fig. 3 shows TEM images and SAED pattern of the prepared Fe₃O₄/C-b. As shown in Fig. 3A and B, the dark Fe₃O₄ particles were encapsulated into the carbon spheres and the mean diameters of composite particles are in the range of 120–140 nm. The carbon layer is ca. 20 nm in thickness and the coated Fe₃O₄ particles exhibits quasi-spheric morphology with the size of 100–120 nm. The ordered lattice fringes were clearly observed in the high-resolution TEM image of Fe₃O₄/C-b (Fig. 3C), with the fringe separation of 0.48 nm corresponding to (111) plane of Fe₃O₄. The selected-area electron diffraction (SAED) in Fig. 3D also confirms the formation of Fe₃O₄.

Fig. 3 TEM images (A–C) and SAED pattern (D) of Fe₃O₄/C-b.

Fig. 4 shows the FT-IR spectra of the synthesized Fe₃O₄/C-b. An obvious absorption peak at 571 cm⁻¹ is regarded as the vibration of Fe–O bond, representing the magnetite phase of product.³¹ The broad absorption band in the range of 3160–3660 cm⁻¹ is assigned to O–H stretching vibration, arising from hydroxyl groups on nanoparticles, the adsorbed glucose and gluconic acid.³⁰ The two obvious peaks at the 1427 and 1593 cm⁻¹ correspond to the asymmetric and symmetric stretching vibrations of COO⁻, further confirming the existence of the gluconic acid on the surface of Fe₃O₄.³² While the band at 1050 cm⁻¹ of C–O stretching almost comes from the residue glucose, and the peaks around 2849 and 2927 cm⁻¹, assignable to asymmetric and symmetric vibrations of C–H in –CH₂–, is related to the carbon layer due to the carbonization of glucose.³³ The FT-IR analysis confirms that Fe₃O₄ nanoparticles not only be synthesized by the reduction of partial Fe³⁺ ions to Fe²⁺ ions with the glucose, but further stabilized and modified by the carbon layer, along with the residual glucose and gluconic acid. Moreover, the large amount of OH and COO⁻ groups covalently bonded to the carbon frameworks offers the hydrophilic surface of Fe₃O₄ nanoparticles and the Fe₃O₄ particles show the good water dispersibility, superior to the sample obtained by the traditional method (Fig. 4 (inset)).³⁴

Fig. 4 FT-IR spectrum of as-synthesized Fe₃O₄/C-b, and the photograph of aqueous solution of Fe₃O₄/C-b and the one synthesized according to the literature³² (inset).

XPS analysis was performed to reveal the compositions and chemical structures of the prepared Fe₃O₄/C-b (Fig. 5). The Fe 2p and O 1s peaks arise from the contribution of Fe₃O₄ nanoparticles, and the C 1s and O 1s peaks point to the carbon layer, glucose and gluconic acid which are coated on the surface of Fe₃O₄. The high-resolution Fe 2p spectrum shows that the Fe 2p_3/2 and Fe 2p_1/2 are located at the binding energy of 711.6 and 725.2 eV, respectively. The broadening of these two peaks and the disappearance of charge transferring satellites of Fe 2p_3/2 around 719 eV indicate the appearance of dual iron oxidation states (Fe²⁺ and Fe³⁺), well consistent with the previous reports.^35–37 This also provides another piece of proof that the structural phase of the prepared sample is Fe₃O₄.

Fig. 5 XPS survey scans (a) and high-resolution Fe 2p spectrum (b) of the synthesized Fe₃O₄/C-b.

As discussed above, the Fe₃O₄ particles have been successfully prepared from the FeCl₃ and glucose in the sodium acetate solution. In this hydrothermal synthesis route, the introduction of glucose is a crucial factor for the preparation of Fe₃O₄, since the structural phase of the product is pure α-Fe₂O₃ without the addition of glucose. It is suggested that the partial Fe³⁺ ions have been reduced to Fe²⁺ ions for the formation of Fe₃O₄ in the glucose-assisted synthesis route. As no external reductant was presented in the solution, the glucose is assumed to be responsible for the reduction reaction. The mass spectrometry in the ESI (Fig. S1†) further provides a solid evidence for the hydrothermal reaction process, by analyzing the composition of the initial and final reaction solution, respectively. Only the pristine glucose is present in the solution at the initial stage (Fig. S1a†), while the gluconic acid is detected in the reaction solution after hydrothermal treatment (Fig. S1b†), suggesting that the glucose has been oxidized to the gluconic acid due to the reduction of partial Fe³⁺ ions. Herein, the glucose plays the reduction role in reducing the partial Fe³⁺ ions to Fe²⁺ ions for the composition of Fe₃O₄.

Another observation is that the carbon layer derived from the carbonization of glucose is coated on the surface of Fe₃O₄ particles, as evidenced by the TEM images and Raman spectroscopy. A possible explanation is that the glucose would firstly reduce partial Fe³⁺ ions to Fe²⁺ ions to form the Fe₃O₄ crystalline nucleus and grow up, and the gluconic acid arising from the oxidation of glucose would be anchored on the surface of magnetite particles.³⁸ Meanwhile, the glucose as the carbon source also involves other chemical reaction in the hydrothermal process (Fig. 6): glucose is gradually polymerized into polysaccharides by the intermolecular dehydration and then carbonized into the carbon nanosphere, and such process would prefer to be conducted around the magnetite particles due to their catalytic carbonization role.^32,33 As a result, the carbon layer along with residual gluconic acid and glucose are completely or incompletely coated on the surface of Fe₃O₄ particles, thus stabilizing and modifying the surface properties of Fe₃O₄ particles.

Fig. 6 Schematic illustration of the glucose-assisted hydrothermal synthesis of the Fe₃O₄/C nanocomposites.

3.2 Factors influencing the formation of Fe₃O₄/C nanoparticles

Based on above powerful tools, such as XRD, TEM, mass spectrometry, Raman and FT-IR spectroscopy, the hydrothermal reaction route has been confirmed as an effective strategy to prepare carbon-coated magnetic nanoparticles. Moreover, the synthesis route is nontoxic and low-cost, without involving the organic solvents, initiators, or surfactants which are commonly used for the preparation of functionalized Fe₃O₄ particles. For the novel synthesis approach, systematic studies have revealed that many factors influence the successful formation of magnetite particles, such as the glucose concentration, pH value, reaction temperature and time, which should be discussed in detail.

3.2.1 Effect of the glucose concentration. The glucose plays the decisive role in the preparation of Fe₃O₄ nanoparticles in consideration of its reduction of partial Fe³⁺ ions to Fe²⁺ ions and further functionality of the as-prepared product. Fig. 7 shows the XRD patterns of magnetite nanoparticles obtained by changing the glucose amount from 0.3 to 1.1 g while keeping the other reaction conditions at the same level. The XRD patterns of all the samples agree well with the characteristic peaks of Fe₃O₄. A significant decrease in the relative intensity of diffraction peaks was observed with the increased glucose amount from 0.3 g of sample Fe₃O₄/C-a to 0.6 g of sample Fe₃O₄/C-b, accompanied with the broadening of Fe₃O₄ diffractions peaks, indicating a dramatic reduction of the crystal particle size. However, when the amount of added glucose increased to 1.1 g for sample Fe₃O₄/C-c, the diffraction peaks become sharper and higher again, suggesting that the Fe₃O₄ particles grew into larger crystallite size in the higher glucose concentration. The smallest particles were achieved when the appropriate amount of glucose was added (0.6 g), with the average size of 126 nm calculated from the Scherrer's equation (D = 0.89k/β cos θ).¹⁰

Fig. 7 XRD patterns of magnetite nanoparticles obtained by using different glucose concentration.

Fig. 8 displays TEM images of magnetite nanoparticles prepared by using different glucose concentration, which provide more detailed morphology information and size distribution of the carbon-coated Fe₃O₄. TEM analysis shows that the morphology and size of the prepared Fe₃O₄ particles are concentration-dependent. In a low glucose concentration, the Fe₃O₄ have a cubic shape with the average particle size of approximately 1.25 μm (Fig. 8A and B), and the coating layer of the carbon, glucose and gluconic acid with the thickness of ca. 10 nm are coated on the surface of Fe₃O₄ particles (Fig. 8B). When the glucose concentration increased, the Fe₃O₄ particles show the quasi-spheric structure and the reduced particle size (ca. 130 nm) (Fig. 8C and D). However, the particle size of Fe₃O₄/C-c centered on ca. 350 nm, bigger than those of Fe₃O₄/C-b, implying particles grew larger with further increment of glucose concentration. Consequently, the carbon layer may be too thin to restrict the particle size of Fe₃O₄ in low glucose concentration, while the glucose tends to form the large-sized carbon sphere in high concentration, which all leads to the undesired particle size and morphology of carbon-coated Fe₃O₄. The optimum particle size was achieved for the Fe₃O₄/C-b by controlling the glucose concentration, which also agrees well with the analysis of XRD data. The results demonstrate that the glucose concentration played an important role in controlling the morphology of the Fe₃O₄/C. The mechanism may be explained as follows: in the low glucose concentration, the thin carbon layer (shell) would tend to coat on the surface of crystallized Fe₃O₄ particles (core), forming the typical core–shell structure. The unique structural feature could restrict the growth of Fe₃O₄ particles and control their cubic morphology. However, once the glucose concentration increased to a certain degree, the glucose would be inclined to polymerize into carbon microspheres and then serve as the nucleating agent. Subsequently, the Fe³⁺ diffused into the carbon microspheres and crystallized to form the spherical morphology of Fe₃O₄/C particles.

Fig. 8 TEM images of Fe₃O₄/C nanoparticles prepared by using different glucose concentration: (A and B) Fe₃O₄/C-a, (C and D) Fe₃O₄/C-b, and (E and F) Fe₃O₄/C-c.

The coating agents in the synthesized Fe₃O₄/C samples were analyzed by the TGA measurements. All the samples were heated to 800 °C in nitrogen atmosphere to avoid the oxidation of Fe²⁺ ions, during which the weight loss of the samples is only attributed to the thermal decomposition of glucose, gluconic acid and glucose polymers. As shown in Fig. S2,† the TGA curves of synthesized magnetite nanoparticles present three weight loss processes. The first weight loss between 25 °C and 200 °C should be ascribed to the dehydration of the samples due to the loss of surface adsorbed water and crystal water. The second weight loss in the temperature range of 200–500 °C corresponds to the desorption and subsequent evaporation of the coating agents, including glucose, gluconic acid and glucose polymers, and the last one at 500–800 °C is due to the decomposition and carbonization of the glucose, gluconic acid and glucose polymers. In our experiment, the higher initial glucose concentration resulted in the higher weight loss, implying the larger amount of residual glucose and gluconic acid.

Fig. 9 shows the nitrogen adsorption–desorption isotherms and the corresponding pore size distribution curves of samples Fe₃O₄/C-b and Fe₃O₄/C-c. The nitrogen isotherms of Fe₃O₄/C-b and Fe₃O₄/C-c are correspondingly close to type-II and IV curves, with an increased nitrogen adsorbed volume at relative pressure P/P₀ of 0.8–1.0.³⁹ The BET surface areas of Fe₃O₄/C-b and Fe₃O₄/C-c are 75 and 84 m² g⁻¹, respectively. The pore diameters of Fe₃O₄/C-b and Fe₃O₄/C-c are both distributed in the range of 1–14 nm, accompanied with the pore volume of 0.125 and 0.171 cm³ g⁻¹, respectively. The sample Fe₃O₄/C-a obtained by low glucose concentration solution has almost no surface area because of the limited coating agents, which is not discussed here in detail.

Fig. 9 N₂ adsorption–desorption isotherms and the corresponding DFT pore size distribution curves (inset) of samples Fe₃O₄/C-b and Fe₃O₄/C-c.

3.2.2 Effect of the sodium acetate (NaOAc) amount. The role of NaOAc was also found to be very critical in this synthesis strategy. In a control experiment, different amounts of NaOAc (1.0–8.0 g) were added in the solution. As shown in Fig. S3 (ESI†), when the amount of NaOAc was more than 6.0 g or less than 3.0 g in this synthesis route, the XRD patterns of the prepared samples can only show the characteristics of α-Fe₂O₃. It is noteworthy that the NaOAc amount of 4.0 g resulted in the high-crystalline Fe₃O₄ particles. Generally, NaOAc is used to provide a steady OH⁻ ion supply through its hydrolysis. In our experiment, FeCl₃ hydrolyzes to form ferric hydroxide and releases H⁺ ions, leading to the byproducts of HCl. However, the accumulation of HCl would inhibit the further reduction of partial Fe³⁺ and thus affect the composition of the iron oxide samples.⁵ The appropriate addition of NaOAc would provide OH⁻ and neutralize the HCl, thus allowing the completion of reduction reaction and the formation of Fe₃O₄ particles.

3.2.3 Effect of the hydrothermal reaction temperature. Fig. S4† displays the XRD patterns of iron oxide samples obtained by different hydrothermal temperature. Notably, when the hydrothermal temperature is fixed on the 160 °C, no apparent Fe₃O₄ particles can be achieved. While when the temperature increased to 180 °C, the typical diffraction peaks of Fe₃O₄ could be observed. Moreover, the degree of crystallinity of Fe₃O₄ was higher along with the further increased temperature to 200 °C, which was taken as the optimum hydrothermal reaction temperature.

3.2.4 Effect of the hydrothermal treatment time. Since the high-quality Fe₃O₄ particles can be prepared when the glucose amount is 0.6 g, the NaOAc amount is 4.0 g, and hydrothermal temperature is 200 °C, the investigation of the effect of hydrothermal treatment time on the structural phase of the product was carried out based on the optimum composition and reaction temperature. Fig. S5† shows the XRD patterns of iron oxide samples prepared with different hydrothermal treatment time. It can be observed that the products are presented in the form of α-Fe₂O₃ when the hydrothermal treatment time was less than 4 hours, whereas the α-Fe₂O₃ gradually transformed into the Fe₃O₄ along with the reaction time exceeding 10 hours, as evidenced by the analysis of Fe₃O₄/C-b. No obvious change had been observed for the products as the reaction time ranges from 10 to 70 hours.

3.3 The magnetic properties of the prepared Fe₃O₄/C particles

Fig. 10 shows the magnetic hysteresis curves of the Fe₃O₄/C samples. All the samples with the small hysteresis loops exhibited diamagnetic behaviors, which could be reflected by B_r and H_c values in Table 1.³⁵ The phenomenon is attributed to the large particle size (>130 nm) of magnetite particles in this work, since the Fe₃O₄ nanoparticles with a particle size <30 nm can have superparamagnetic properties.⁴⁰ Nevertheless, when the magnetic field was 6000 Oe, the samples Fe₃O₄/C-a, Fe₃O₄/C-b and Fe₃O₄/C-c show the B_s values of 26.26, 63.48 and 57.05 emu g⁻¹, respectively, suggesting the prepared samples had strong magnetic response to the external magnetic field, especially the sample Fe₃O₄/C-b. The higher purity of Fe₃O₄/C-b than that of Fe₃O₄/C-a is mainly responsible for the larger magnetism. However, the excessive glucose concentration for the Fe₃O₄/C-c resulted in the relatively low mass percentage of Fe₃O₄ in the sample, which decreased the magnetism of Fe₃O₄/C-c. The highest saturation magnetization would render the Fe₃O₄/C-b a convenient magnetic separation process in the practical application.

Fig. 10 Magnetic hysteresis curves taken at room temperature for the prepared Fe₃O₄/C samples (inset: the enlarged view).

Table 1 Magnetic parameters of the prepared Fe₃O₄/C samples

Sample	Bs^a/emu g^−1	Hc^b/Oe	Br^c/emu g^−1
a Saturation magnetization.b Coercive force.c Remanent magnetization.
Fe3O4/C-a	26.26	40.37	1.096
Fe3O4/C-b	63.48	46.91	10.03
Fe3O4/C-c	57.05	62.67	3.729

---

3.4 The adsorption performance of Fe₃O₄/C particles in removing toxic chemicals

The heavy metal ions and organic chemicals are highly toxic environmental pollutants existing in industrial wastewater, which are very deleterious to humans, even in trace amount. The adsorption of these toxic environmental pollutants is one of the effective strategies to resolve the environmental problems involving wastewater treatment. Technologically, the constructed Fe₃O₄/C-b particles with high specific surface areas, strong magnetic properties, and excellent water dispersibility, could be used as an effective adsorbent for the removal of some toxic chemicals, such as Cr(VI) and Rhodamine B (RhB). In our experiment, the prepared Fe₃O₄/C-b particles were used to remove the Cr(VI) and RhB from wastewater. The effects of adsorption time, the reclaimed sample's performance and its equilibrium adsorption isotherms, and adsorption kinetics were investigated.

3.4.1 The adsorption performance of Fe₃O₄/C-b particles in removing Cr(VI).
(I) Adsorption isotherms of Fe₃O₄/C-b particles. The adsorption isotherm plays an important role in understanding sorption systems since the maximum capacity of adsorption could be determined for the adsorbent.⁴¹ Fig. 11 shows the adsorption performance of the prepared Fe₃O₄/C-b adsorbent for Cr(VI). As shown in Fig. 11a, the adsorption of Cr(VI) on Fe₃O₄/C-b adsorbent at different Cr(VI) concentration can all achieve the absorption equilibrium within 360 min. As a result, the equilibrium concentration of Cr(VI) solution was obtained. The amount of adsorbed Cr(VI) at equilibrium q_e was calculated from eqn (1):⁴²where C₀ (mg L⁻¹) is the initial Cr(VI) concentration, C_e (mg L⁻¹) is the equilibrium concentration of Cr(VI) solution, V (L) is the volume of the solution, and m (g) is the mass of adsorbent added to the solution.

Fig. 11 The removed percentage of Cr(VI) on as-prepared Fe₃O₄/C-b particles in different Cr(VI) concentration solution (a), the plot of q_e versus C_e for the non-linear adsorption isotherm of Cr(VI) onto Fe₃O₄/C-b (b), and adsorption isotherms of the Langmuir (c) and Freundlich (d) model.

The Langmuir and Freundlich models are two commonly used to investigate the adsorption process and mechanism of the adsorbent. The Langmuir model is primarily applied to the monolayer adsorption on a homogeneous surface without any interaction between the adsorbed molecules, and the Langmuir equation is commonly presented as eqn (2):^21,40

C_e/q_e = C_e/q_max + 1/K_Lq_max (2)

where q_max (mg g⁻¹) and C_e (mg L⁻¹) represent the maximum capacity and the equilibrium concentration of adsorbate, respectively. K_L (L mg⁻¹) is the Langmuir constant related to adsorption capacity.

The Freundlich model is widely used for the adsorption in a multilayer manner on an energetically heterogeneous surface, which can be expressed as eqn (3):⁴³

log q_e = log K_F + 1/n log C_e (3)

where K_F and n are the Freundlich constants that are related to adsorption capacity and intensity, respectively.

Fig. 11b shows the variation of q_e with respect to C_e for the adsorption of Cr(VI). The adsorption isotherms of Fe₃O₄/C–Cr(VI) adsorption system were built according to the two models (Fig. 11c and d), and the corresponding coefficients and non-linear R² values were shown in Table 2. It is shown that the R² value of the Langmuir isotherm model is higher than that of the Freundlich model, suggesting that a monolayer adsorption of Cr(VI) on the surface of Fe₃O₄/C-b occurs, which may originate from the inverse electrostatic adsorption and irreversible redox reaction on the Fe₃O₄ particles.⁴⁴ The maximum adsorption capacity (q_max) of Fe₃O₄/C-b for Cr(VI) is 61.69 ± 0.5 mg g⁻¹, much higher than previously reported values of other Cr(VI)-removal adsorbents, such as pristine Fe₃O₄ nanospheres (2.95 mg g⁻¹)⁴⁵ and Fe₃O₄ particles (4.38 mg g⁻¹),⁵ and comparable to the ethylenediamine-functionalized Fe₃O₄ magnetic polymers (61.35 mg g⁻¹).²⁷ The suitable coating agents (proper functional groups) and high specific surface area of the Fe₃O₄/C-b particles may result in such good removal performance to heavy metal ion Cr(VI).

Table 2 The kinetics parameters for the adsorption of Cr(VI) and RhB on the Fe₃O₄/C-b adsorbent

	C0 (mg L^−1)	qe,exp (mg g^−1)	Pseudo-first order			Pseudo-second order
			k1 (min^−1)	R^2	qe (mg g^−1)	k2 (g mg^−1 min^−1)	R^2	qe (mg g^−1)
Cr(VI)	20	26.50 ± 0.2	0.004	0.959	21.17 ± 0.2	0.000689	0.995	29.68 ± 0.2
	40	23.27 ± 0.2	0.003	0.914	18.01 ± 0.2	0.000723	0.982	25.62 ± 0.2
	60	18.25 ± 0.2	0.004	0.905	12.38 ± 0.2	0.001322	0.992	19.94 ± 0.2
	80	14.88 ± 0.2	0.005	0.948	9.06 ± 0.25	0.002683	0.999	15.70 ± 0.2
	100	12.00 ± 0.3	0.006	0.846	7.46 ± 0.3	0.002948	0.998	12.89 ± 0.25
RhB	4.79	27.39 ± 0.2	0.011	0.785	25.83 ± 0.2	0.000596	0.956	27.55 ± 0.2

---

(II) Kinetic adsorption rate equation of Fe₃O₄/C-b particles. The pseudo-first order equation and pseudo-second order equation are used to understand the adsorption kinetics of Fe₃O₄/C-b in the adsorption of Cr(VI), which could provide useful information about the adsorption type and the crucial factors involved in the adsorption. The two kinetic models can be expressed by eqn (4) and (5), respectively:^27,40

log(q_e − q_t) = log q_e − k₁t (4)

t/q_t = 1/(k₂q_e²) + t/q_e (5)

where q_e (mg g⁻¹) and q_t (mg g⁻¹) are the adsorbed amounts of Cr(VI) on Fe₃O₄/C-b at equilibrium and time t, respectively, and k₁ (min⁻¹) and k₂ (g mg⁻¹ min⁻¹) are the rate constants of the pseudo-first-order adsorption and pseudo-second-order adsorption, respectively.

The kinetic adsorption data for the Fe₃O₄/C–Cr(VI) adsorption system were simulated using the pseudo-first order rate equation and the pseudo-second order rate equation, as shown in Fig. 12. The fitted kinetic parameters are summarized in Table 2. The correlation coefficients (>0.982) for the pseudo-second order equation are higher than those (<0.959) for the pseudo-first-order rate equation, and the calculated q_e from the pseudo-second order equation was closer to the experimental value. Therefore, it can be concluded that the adsorption of Cr(VI) on the Fe₃O₄/C-b particles is a second-order reaction rather than a first-order one, which may involve the surface complexation between the Fe₃O₄/C surface and the toxic Cr(VI) ions in the aqueous solution.^5,43

Fig. 12 The pseudo-first order adsorption rate equation (a) and pseudo-second order adsorption rate equation (b).

(III) Cycling performance of Fe₃O₄/C-b particles. In order to investigate the recyclability of Fe₃O₄/C-b, recycling experiments were carried out. Firstly, the Cr(VI)-loaded Fe₃O₄/C adsorbent is conveniently separated from the solution by the conventional magnet, and further regenerated by dilute basic solution. The results of recycling experiments are shown in Fig. 13. It is found that, compared to the removal efficiency (26.5 ± 0.5%) of pristine Fe₃O₄/C-b within 6 h, the regenerated Fe₃O₄/C-b with a high saturation magnetization of 52.54 emu g⁻¹ shows the slightly inferior adsorption performance for the Cr(VI), but the 22 ± 0.5% of Cr-removal percentage can still be achieved after 3 cycles. The easy separation from the solution, the good recycling adsorption performance for Cr(VI) and the high remained magnetism suggest that Fe₃O₄/C particles can be utilized as an effective adsorbent for the wastewater treatment.

Fig. 13 The Cr(VI) removal percentage by using the Fe₃O₄/C-b composite in recycle runs (a) and the magnetism variation with the recycle runs (b).

3.4.2 The adsorption performance of Fe₃O₄/C particles in removing RhB. Fig. S6† shows the adsorption performance and the theoretical analysis of the RhB adsorption process for the Fe₃O₄/C-b absorbent. As shown in Fig. S6a,† the removal percentage of RhB can be up to 83.64 ± 0.5% at the initial concentration of 10 μmol L⁻¹, displaying a high adsorption capacity for the dye molecule. In order to investigate the maximum adsorption capacity, similarly, the Langmuir and Freundlich equations are applied to fit the adsorption data derived from the Fig. S6b.† In comparison of the values of the correlation coefficients (R²) in Fig. S6c and d,† the adsorption process is closer to the Langmuir model. This indicates that the adsorption of RhB is a monolayer adsorption process, and the calculated q_max is 43.86 ± 0.5 mg g⁻¹, higher than those of previously reported adsorbents, such as polystyrene/Fe₃O₄/graphene oxide (13.8 mg g⁻¹),⁴⁶ activated carbon (16.1 mg g⁻¹),⁴⁷ and sodium montmorillonite (42.2 mg g⁻¹).⁴⁸

The kinetic adsorption process for the Fe₃O₄/C-b absorbent was also investigated by the pseudo-first order and pseudo-second order models, as shown in Fig. S6e–g.† The fitting of adsorption data suggests that the pseudo-second order model with the larger correlation coefficient fits better with the experiment data than the pseudo-first order model. Thus, the adsorption of RhB on the Fe₃O₄/C-b is likely to be controlled by chemisorption,⁴⁹ which may be attributed to the hydrogen bonding interaction of RhB and functional groups of glucose and gluconic acid on the surface of Fe₃O₄/C-b.⁴⁶

4. Conclusions

The Fe₃O₄/C nanoparticles have been synthesized via a one-step glucose-assisted hydrothermal reaction approach, in which the Fe³⁺ ions were partially reduced to the Fe²⁺ by the glucose, and moreover, the carbon layer with the retention of functional groups, such as glucose and gluconic acid, were also coated on the surface of Fe₃O₄ nanoparticles. For this novel synthesis route, some crucial factors, such as the glucose concentration, pH value, reaction temperature and time, influence the formation of coated Fe₃O₄ nanoparticles. When applied as the adsorbent, the coated Fe₃O₄ nanoparticles with carbon layer, glucose and gluconic acid exhibited good hydrophilicity, superior magnetic properties, and excellent adsorption capacity of toxic chemicals of Cr(VI) and RhB. The Fe₃O₄ nanoparticles prepared from the facile and low-cost synthesis route may also be extended to the other fields, such as high-density magnetic recording media and catalysis.

Acknowledgements

This research was financially supported by the Foundation of Henan Educational Committee (No. 16A150023), Nanhu Scholars Program for Young Scholars of XYNU, the Doctoral Start-up Research Fund of Xinyang Normal University (15006) and the College of Chemistry and Chemical Engineering of Xinyang Normal University.

Notes and references

1. M. E. F. Brollo, R. López-Ruiz, D. Muraca, S. J. A. Figueroa, K. R. Pirota and M. Knobel, Sci. Rep., 2014, 4, 6839–6844.
2. N. B. Saleh, N. Aich, J. Plazas-Tuttle, J. R. Lead and G. V. Lowry, Environ. Sci.: Nano, 2015, 2, 11–18.
3. N. Saleh, K. Sirk, Y. Liu, T. Phenrat, B. Dufour, K. Matyjaszewski, R. D. Tilton and G. V. Lowry, Environ. Eng. Sci., 2007, 24, 45–57.
4. K. Naka, A. Narita, H. Tanaka, Y. Chujo, M. Morita, T. Inubushi, I. Nishimura, J. Hiruta, H. Shibayama, M. Koga, S. Ishibashi, J. Seki, S. Kizaka-Kondoh and M. Hiraoka, Polym. Adv. Technol., 2008, 19, 1421–1429.
5. L. S. Zhong, J. S. Hu, H. P. Liang, A. M. Cao, W. G. Song and L. J. Wan, Adv. Mater., 2006, 18, 2426–2431.
6. Y. H. Deng, Y. Cai, Z. K. Sun, J. Liu, C. Liu, J. Wei, W. Li, C. Liu, Y. Wang and D. Y. Zhao, J. Am. Chem. Soc., 2010, 132, 8466–8473.
7. G. Shi, B. Sun, Z. Jin, J. Liu and M. Li, Sens. Actuators, B, 2012, 171–172, 699–704.
8. T. P. Almeida, T. Kasama, A. R. Muxworthy, W. Williams, L. Nagy, T. W. Hansen, P. D. Brown and R. E. Dunin-Borkowski, Nat. Commun., 2014, 5, 5154–5159.
9. X. Zhang, Y. Wang and S. Yang, Carbohydr. Polym., 2014, 114, 521–529.
10. S. Sun and H. Zeng, J. Am. Chem. Soc., 2002, 124, 8204–8205.
11. S. Sun, H. Zeng, D. B. Robinson, S. Raoux, P. M. Rice, S. X. Wang and G. Li, J. Am. Chem. Soc., 2004, 126, 273–279.
12. T. J. Daou, G. Pourroy, S. Bégin-Colin, J. M. Grenèche, C. Ulhaq-Bouillet, P. Legarè, P. Bernhardt, C. Leuvrey and G. Rogez, Chem. Mater., 2006, 18, 4399–4404.
13. J. Wang, J. Sun, Q. Sun and Q. Chen, Mater. Res. Bull., 2003, 38, 1113–1118.
14. J. Liu, Z. Sun, Y. Deng, Y. Zou, C. Li, X. Guo, L. Xiong, Y. Gao, F. Li and D. Zhao, Angew. Chem., Int. Ed., 2009, 48, 5875–5879.
15. J. Ge, Y. Hu, M. Biasini, W. P. Beyermann and Y. Yin, Angew. Chem., Int. Ed., 2007, 46, 4342–4345.
16. A.-L. Morel, S. I. Nikitenko, K. Gionnet, A. Wattiaux, J. Lai-Kee-Him, C. Labrugere, B. Chevalier, G. Deleris, C. Petibois, A. Brisson and M. Simonoff, ACS Nano, 2008, 2, 847–856.
17. T. Lu, J. Wang, J. Yin, A. Wang, X. Wang and T. Zhang, Colloids Surf., A, 2013, 436, 675–683.
18. W. Wu, Q. He and C. Jiang, Nanoscale Res. Lett., 2008, 3, 397–415.
19. D. Wang and D. Astruc, Chem. Rev., 2014, 114, 6949–6985.
20. J. Deng, C. He, Y. Peng, J. Wang, X. Long, P. Li and A. S. C. Chan, Synth. Met., 2003, 139, 295–301.
21. M. Bhaumik, A. Maity, V. V. Srinivasu and M. S. Onyango, J. Hazard. Mater., 2011, 190, 381–390.
22. N. R. Jana, C. Earhart and J. Y. Ying, Chem. Mater., 2007, 19, 5074–5082.
23. S. I. C. J. Palma, M. Marciello, A. Carvalho, S. Veintemillas-Verdaguer, M. d. P. Morales and A. C. A. Roque, J. Colloid Interface Sci., 2015, 437, 147–155.
24. F. Fu and Q. Wang, J. Environ. Manage., 2011, 92, 407–418.
25. J.-S. Hu, L.-S. Zhong, W.-G. Song and L.-J. Wan, Adv. Mater., 2008, 20, 2977–2982.
26. H. Mittal and S. B. Mishra, Carbohydr. Polym., 2014, 101, 1255–1264.
27. Y.-G. Zhao, H.-Y. Shen, S.-D. Pan and M.-Q. Hu, J. Hazard. Mater., 2010, 182, 295–302.
28. C. Chen, P. Gunawan and R. Xu, J. Mater. Chem., 2011, 21, 1218–1225.
29. Y. Liu, Z. Ren, Y. Wei, B. Jiang, S. Feng, L. Zhang, W. Zhang and H. Fu, J. Mater. Chem., 2010, 20, 4802–4808.
30. S. Xuan, L. Hao, W. Jiang, X. Gong, Y. Hu and Z. Chen, Nanotechnology, 2007, 18, 035602.
31. Y. Wang, S. Wang, H. Niu, Y. Ma, T. Zeng, Y. Cai and Z. Meng, J. Chromatogr. A, 2013, 1283, 20–26.
32. X. Sun and Y. Li, Angew. Chem., Int. Ed., 2004, 43, 597–601.
33. Z. Wang, H. Guo, Y. Yu and N. He, J. Magn. Magn. Mater., 2006, 302, 397–404.
34. I. Martínez-Mera, M. E. Espinosa-Pesqueira, R. Pérez-Hernández and J. Arenas-Alatorre, Mater. Lett., 2007, 61, 4447–4451.
35. S. Gao, Y. Shi, S. Zhang, K. Jiang, S. Yang, Z. Li and E. Takayama-Muromachi, J. Phys. Chem. C, 2008, 112, 10398–10401.
36. Y. Lai, W. Yin, J. Liu, R. Xi and J. Zhan, Nanoscale Res. Lett., 2009, 5, 302–307.
37. L. Zhou, H. Deng, J. Wan, J. Shi and T. Su, Appl. Surf. Sci., 2013, 283, 1024–1031.
38. H. Hosseini-Monfared, F. Parchegani and S. Alavi, J. Colloid Interface Sci., 2015, 437, 1–9.
39. Y.-P. Zhu, J. Li, T.-Y. Ma, Y.-P. Liu, G. Du and Z.-Y. Yuan, J. Mater. Chem. A, 2014, 2, 1093–1101.
40. F.-H. Wang, W. Jiang, Y. Fang and C.-W. Cheng, Chem. Eng. J., 2015, 259, 827–836.
41. T. Hao, X. Rao, Z. Li, C. Niu, J. Wang and X. Su, J. Alloys Compd., 2014, 617, 76–80.
42. Y. Gao, D. Zhong, D. Zhang, X. Pu, X. Shao, C. Su, X. Yao and S. Li, J. Chem. Technol. Biotechnol., 2014, 89, 1859–1865.
43. Y. Li, S. Zhu, Q. Liu, Z. Chen, J. Gu, C. Zhu, T. Lu, D. Zhang and J. Ma, Water Res., 2013, 47, 4188–4197.
44. J. Hu, I. M. Lo and G. Chen, Water Sci. Technol., 2004, 50, 139–146.
45. J. Wang, B. Tang, T. Tsuzuki, Q. Liu, X. Hou and L. Sun, Chem. Eng. J., 2012, 204–206, 258–263.
46. P. Yuan, D. Liu, M. Fan, D. Yang, R. Zhu, F. Ge, J. Zhu and H. He, J. Hazard. Mater., 2010, 173, 614–621.
47. K. Kadirvelu, C. Karthika, N. Vennilamani and S. Pattabhi, Chemosphere, 2005, 60, 1009–1017.
48. P. P. Selvam, S. Preethi, P. Basakaralingam, N. Thinakaran, A. Sivasamy and S. Sivanesan, J. Hazard. Mater., 2008, 155, 39–44.
49. Y. S. Ho and G. McKay, Process Biochem., 1999, 34, 451–465.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra01408a

---