1 Introduction

Magnetic nanoparticles (MNPs) have attracted more attention not only for their peculiar characteristics, but also for their potential applications in various fields including magnetic recording, magnetic fluid, and biomedical materials [1]. Magnetite (Fe3O4), maghemite (γ-Fe2O3), and hematite (α-Fe2O3) are three main iron oxides [2]. Iron oxides are generally prepared by the coprecipitation method [35], microemulsion [6], hydrothermal [7], solvothermal [8], thermal decomposition of iron organicsalts and complex compounds, and laser pyrolysis technique [9, 10]. Among these methods, the coprecipitation is used more than any other method and has many advantages including; low cost, high yield, easy synthesis, and advanced laboratory facilities does not require [1]. MNPs with less than a critical size exhibit a unique magnetic property that is known as the superparamagnetism phase [11]. There is not a domain wall in these nanoparticles (NPs), therefore, the saturated magnetic response attains at relatively low external magnetic fields. The tuning size of MNPs is still one of the most challenging tasks to achieve the superparamagnetic phase. The stability of MNPs can be greatly enhanced when they are coated by external layers as the core–shell form. Compared to conventional organic or inorganic layers, carbon was found to be one of the most superior protecting materials; carbon being a light material. Also, it has low magnetic susceptibility [12], exhibits much higher stability in various chemical and physical environments such as acid or base media, as well as at high temperatures and pressures [13], and prevents MNPs core from agglomeration. Functional groups, such as −OH and C=O are present on the surface of the carbon shells, which will facilitate the linkage of functional groups or catalytic species to the surfaces in future applications [14].

Carbon encapsulated NPs were observed for the first time in 1993 during studies on fullerene formation in the carbon arc discharge process [15]. Various synthesis routes of carbon encapsulate have been reported, such as high temperature techniques including carbon arc discharge [16], flowing thermal plasma jet [17], magnetron and ion-beam co-sputtering [18], high-temperature annealing of the mixtures of carbon-based materials and metal precursors, catalytic chemical vapor deposition, pyrolysis of organometallic compounds, catalytic decomposition of methane [13], and a low temperature route by the hydrothermal method [14]. Within the last 10 years, much effort was put to develop a method that can fully supplant the high temperature routes. In 2006, Wei [13] and Wang [19] have successfully prepared carbon-encapsulated MNPs by using FeNi and Fe3O4 NPs as the nuclei in the hydrothermal reaction, respectively. In 2007, Xuan et al. [20] first successfully fabricated carbon-encapsulated Fe3O4 core/shell composites via a one-step hydrothermal reaction. In his work, glucose is both the carbon source for formation of carbon shells and the reducing agent for the reduction of Fe3C. However, they had no success in trying to regulate the particle diameters. Wang et al. [21] synthesized magnetite/carbon NPs with an average size of about 190 nm via one-step solvothermal process at 240 C. We understand there is no report about the preparation of Fe3O4/C core–shell NPs with diameters below 20 nm [22]. J. Zhang et al. [23] report the synthesis of the Fe3O4/C core–shell by using the one-step method. Carbon-encapsulated MNPs have received considerable attention in recent years due to their wide promising applications, including high density magnetic data storage [14], magnetofluid, magnetic resonance imaging contrast agents [24], biomedical techniques [25], and waste water [26, 27].

There are two methods for synthesis of core–shell NPs:

  1. (i)

    First method involves one-step processes.

    In this route, core and shell NPs were synthesized simultaneously [20].

  2. (ii)

    The second method involves two-step processes in which first, core NPs were synthesized, and then a thin layer as a shell encapsulating core NPs [20, 27].

    In our experiments, first Fe3O4 NPs were synthesized by coprecipitation method with two different conditions. The effects of these conditions on the size and crystalline strain of synthesized magnetite NPs was investigated. Subsequently, we used a very facile two-step process for synthesis of Fe3O4/C core–shell NPs by the hydrothermal process. The three advantages of the two-step method here are as follows:

    1. (1)

      Control of size, morphology, and crystal phase of carbon coated magnetite are easier in the two-step procedure.

    2. (2)

      The two-step process is facile, versatile, and suitable for the coating of iron oxide and other transition metals with carbon.

    3. (3)

      The carbonization reaction takes place at a relatively low temperature (160–180 C) in water and no reagents are added [13].

      FeCl3⋅6H2O and FeCl2⋅4H2O are used as an iron source and glucose as a carbon source. By a hydrothermal process, the aqueous glucose solution is maintained at 160–180 C in a Teflon-sealed autoclave. Very fine superparamagnetic Fe3O4/C core–shell NPs (below 11 nm) was prepared. Also, we have investigated the effect of the carbon shell on the crystalline strain and magnetic response of Fe3O4 NPs.

2 Experimental Details

2.1 Materials and Methods

All chemical materials were obtained in high purity and were used as received, including ferrous chloride tetrahydrate (FeCl2⋅4H2O), ferric chloride hexahydrate (FeCl3⋅6H2O), sodium hydroxide (NaOH) and glucose (C6H12O6) were purchased from the Merck Chemical Corporation.

2.2 Synthesis of Fe3O4 NPs

Fe3O4 NPs were prepared via the precipitation of Fe+3 and Fe+2 ions with a molar ratio (2/1) in the presence of NaOH as a precipitator agent.

2.2.1 Method A

In our experiment, the NaOH solution was dropped into the iron ions (Fe+3 and Fe+2) solution as following: FeCl2⋅4H2O and FeCl3⋅6H2O (with 1:2 molar ratios) were dissolved in 150 ml DI water. After 30 min of magnetic stirring and heating in 70 C, 10 ml solution of NaOH (4 M) was dropped into a salts solution. The black precipitations were composed. The generated precipitations were collected with a magnet and washed three times with DI water and ethanol, and then oven dried at 40 C. This sample was labeled Fe3O4 (A).

2.2.2 Method B

In this method, iron ions (Fe+3 and Fe+2) were dropped in to the NaOH solution as following: FeCl2⋅4H2O and FeCl3⋅6H2O (with 1:2 molar ratio) were dissolved in 25 mL of DI water. Then the resulting solution was added dropwise into 200 ml of 1.5 mol/L NaOH solution under vigorous stirring. The generated black precipitations were collected by magnetic separation and washed three times with DI water and ethanol, then oven-dried at 40 C for 24 h. This sample was labeled Fe3O4 (B).

2.3 Synthesis of Fe3O4/C Core–Shell (MNPs)

The Fe3O4/C core shell NPs were prepared via hydrothermal method as follows: Fe3O4 NPs were prepared by method B dispersed into glucose solution and the mixture was sonicated for 10 min. Next, the mixture was transferred to a Teflon-sealed autoclave and maintained at 180 C for 4 h. After the reaction was completed, the autoclave cooled naturally at room temperature and the suspension was isolated with a bar magnet. The generated precipitations were washed five times with DI water. Finally, the resulting Fe3O4/C NPs were oven-dried at 40 C for 12 h.

In the synthesis of iron oxide NPs by the coprecipitation process, two stages are involved: a short burst of nucleation occurs when the concentration of the species reaches to critical supersaturation, and then there is a slow growth of the nuclei by diffusion of the solutes to the surface of the crystal. To produce monodisperse iron oxide NPs, these two stages (nucleation and growth) should be separated, i.e., nucleation should be stopped during the growth period. Considering the LaMer diagram and Ostwald ripening [1, 28, 29], the process of monodisperse synthesis should be controlled. In a supersaturated solution when the nuclei form at the same time, subsequent growth illustrated the formation of monodispersed nano and microparticles.

In order to control size of monodispersed particles, nucleation must normally be performed during the very short period, because the final particle number is determined by the end of the nucleation, and it does not change during particle growth [30].

Difference between method A and B:

In method A, hydroxide ions are added to the environment filled with iron ions. By adding sodium hydroxide gradually (hydroxide ions), the pH increases slowly and iron ions have more time for extra nucleation, whereas in method B, iron ions are added to an environment full of hydroxide ions, which are surrounded by hydroxide ions immediately. So, iron ions have no time for extra nucleation and the nucleation process stopped.

Hence, the nucleation time in method B is less than method A, and prepared NPs in method B is smaller than method A.

In the two-step carbon encapsulation, the glucose molecules has major role as follows:

  1. (i)

    The glucose acts as a precursor, which leads to the formation of the carbon shell.

  2. (ii)

    The adsorption of carbon onto the surfaces of Fe3O4 NPs prevents their further aggregation.

The glucose can also decrease the interfacial tension between the crystallizing phase and the surrounding solution, therefore, the aggregation process performs slower, which leads to formation of smaller core–shell NPs. The main possible chemical reaction in our experiment to fabricate Fe3O4 and Fe3O4/C core shell NPs may be written as Eqs. (1) and (2) [27].

In the typical reaction, we heated the reactant mixture of glucose and Fe3O4 NPs at 180 C for 4 h. It should be noted that the hydrothermal temperature is critical for the formation of carbon. As the temperature was below 140 C, the carbonization did not start. If the temperature was higher than 190 C, the carbonization occurs in an unfavorite way, which leads to the agglomeration of the product. Therefore, the temperature of 180 C was preferred as the reaction temperature here [19]. Based on the results of our experiments, the growth of Fe3O4/C seems as follows:

  • Presumably, as the autoclave is heated to a certain temperature (higher than 160 C), the adjacent glucose molecules in the solution starts to dehydrate and polymerize together (polycondensation reaction) and form linear or branch-like oligosaccharides or other macromolecules.

  • The carbonization may arise from these compounds with either the Fe3O4 NPs or themselves as the nucleus. The resulting nucleus with carbon shell then grew uniformly and isotropically by the diffusion of solutes toward the particles’ surfaces until Fe3O4/C NPs were obtained [13, 31].

    (1)
    (2)

2.4 Characterization Methods

TEM images of the samples were recorded with a Philips CM10 Transmission Electron Microscopy (Philips, Netherland) operated at 100 KV. The samples were prepared by dispersing about 0.2 g powder with 15 ml ethanol by ultrasound for 30 min. A drop of the prepared sample was dispersed on a holey carbon TEM support grid. Excess solution was blotted off using a filter paper. XRD analysis was undertaken using a PW 1800 (Philips, Netherland) and Cu-K α radiation (λ=0.154 nm). The specimen were prepared for XRD analysis by sedimentation some of the dispersion on a zero-background slide and analyzed in ambient air. Continuous scan from 4 to 90 2θ were obtained at a scan rate of about 1 2θ min−1. FTIR spectra of KBr pellets were taken with a FTIR spectrometer (Nicolet magna, IR 560, USA). The concentration of the sample in KBr was 0.2–1 %. The magnetization of sample was measured in the varied magnetic field between −10 and 10 kOe at room temperature by using a Vibrating Sample Magnetometer (Meghnatis Daghigh Kavir Co, Iran).

3 Results and Discussion

3.1 X-Ray Diffraction (XRD)

The crystalline nature of the samples and its purity were characterized by X-ray diffraction (XRD). A representative XRD pattern of the as-prepared product is shown in Fig. 1. The diffraction peaks appeared at Bragg angles 2θ∼30.2, 35.5, 43.0, 53.1, 57.0, 62.5, and 74.2 corresponding to Miller indices (220), (311), (400), (422), (511), (440), and (533) could be attributed to Fe3O4 face centered cubic structure [JCPDS card No. 19-0629]. A preferred growth orientation shows along the [311] crystallographic direction. Here, it is worth mentioning that Fe3O4 has an inverse spinal structure with a phase centered cubic unit cell of 32 O2− ions with a lattice constant of 0.839 nm in which 1/3 of iron ions occupy the tetrahedral sites (all Fe3+) while 2/3 of the iron ions occupy both the octahedral and tetrahedral sites [32].

Fig. 1
figure 1

XRD patterns of the iron oxide prepared with method A, B, and Fe3O4/C core–shell

In Fig. 1 Fe3O4 (A), in addition to the main characteristic peaks of magnetite, there are a number of low-intensity peaks, which are related to impurity phases. It can be related to gradual change in pH and forming impurity phases such as iron hydroxide and acaganeite.

In Fig. 1 Fe3O4 (B), the characteristic peaks are higher in intensity, which indicates that the products have a good crystalline nature. No peaks corresponding to impurities are detected, showing that the final product is pure Fe3O4.

In the XRD pattern of Fe3O4/C, no additional peaks were observed except that of pure magnetite. The peaks broadening can be attributed to amorphous carbon, which coated as a shell on the magnetite [33].

The decreasing peak intensity can arise from the shielding effect of this amorphous surface layer.

The average crystallite size was estimated by the X-ray line broadening approach using from the Debye–Scherrer equation [34, 35]:

$$ \beta _{\mathrm{Observed}} = \frac{k\lambda}{D\cos \theta} $$
(3)

where D is crystallite size in nanometers, k is a constant equal to 0.9, λ is wavelength of X-rays Cu-K α radiation (λ=0.154 nm), β is FWHM (in radian), and θ is the diffraction peak angle. The results show that the size of Fe3O4 core particles, which have been synthesized using method A, are 13 nm and using method B are 6.6 nm. The size of particles was synthesized with method B, after coating with carbon, increase to 10.5 nm.

The XRD pattern reveals that the sharpness of peaks was increased after coating with carbon and the size of particles were grown which confirmed by Debye–Scherrer’s calculations. The increasing size of core may be due to a higher temperature, in coating carbon shell process (180 C).

Also, we have estimated and compared the crystallite size and strains in as-prepared Fe3O4 and Fe3O4/C NPs using the Williamson–Hall equation [36]. According to W–H equation, the observed line broadening is

(4)
(5)

By rearranging the above equation, we get

$$ \beta _{\mathrm{Observed}}\cos \theta = \frac{k\lambda}{D} + 4\varepsilon \sin \theta $$
(6)

where ε is the strain associated with the NPs. In Eq. (6), we have considered the homogeneous isotropic nature of crystal and, therefore, strain is independent from direction. The plot of Eq. (6) represents a straight line between 4sinθ (along the X-axis) and βcosθ (along the Y-axis). The slope of line gives the strain (ε) and intercept (/D) of this line on the Y-axis gives crystallite size (D).

Figures 2 and 3 depicted the W–H plots of Fe3O4 and Fe3O4/C samples by using best linear fit to data in ORIGIN 8 software. Results of W–H calculation show that the crystallite size and strain magnitude of Fe3O4 and Fe3O4/C are 7.6, 9.8 nm, 0.00142, and −0.00026, respectively, that are consistent with Debye–Scherrer’s calculations. The plots showed a negative strain for the Fe3O4/C NPs. This strain may be attributed due to the lattice shrinkage of core shell NPs.

Fig. 2
figure 2

The W–H analysis of Fe3O4 NPs. The strain is extracted from the slope and the crystalline size is extracted from the y-intercept of the fit

Fig. 3
figure 3

The W–H analysis of Fe3O4/C NPs. The strain is extracted from the slope and the crystalline size is extracted from the y-intercept of the fit

3.2 FT–IR Spectra

The FT–IR spectrum was used for identifying the functional groups formed on the surface of the nanocomposite. Compared with the standard spectrum, the following absorption bands can be observed:

The prepared Fe3O4 and Fe3O4/C NPs exhibit the chemical interaction between Fe3O4 and carbon. Figure 4 shows the FTIR spectra of the as-prepared Fe3O4 NPs and has a strong absorption peak at 576.11 cm−1, which is attributed to the Fe–O stretching vibration of Fe3O4 [26]. This peak (magnetite characteristic peak) reveals that the spinel structure of Fe3O4 was formed. Please note it has a good agreement with XRD results. The broad band region at 3400–3850 cm−1 is due to the symmetry v s (O–H) and asymmetry stretching vibrations v as (O–H) of H2O molecules [22]. The peak at 1340.47 cm−1 represents the bending vibrations of OH, which are assigned to OH absorbed by Fe3O4 NPs [20, 37].

Fig. 4
figure 4

FTIR spectra of the as-prepared Fe3O4

Figure 5 shows FTIR spectrum of the as-prepared Fe3O4/C NPs. In Fig. 5, the peak at 584.71 cm−1 is characteristic of the Fe–O vibrations. The broad band region at 3400–3850 cm−1 results from the stretching vibration of –OH [22]. The peak at 1400.89 cm−1 is the bending vibration, which is associated to the OH junction. The peaks at 1703 and 1619.74 cm−1 are attributed to C=O and C=C stretching vibrations, respectively, which supported the concept of aromatization of glucose during hydrothermal reaction [13, 31, 38]. The strong bonding of carbon on nanoparticle surface was checked by noting C=O and C–C vibration in FTIR spectra. The presence of carboxyl and hydroxyl groups endow Fe3O4/C NPs with hydrophilic surface, thus these NPs can disperse stably in a solution for practical application and which will facilitate the linkage of catalytic species or bimolecular to the surface in future application. For instance, they could be covalently bonded to biomacromolecules and used as a hydrophilic drug-delivery system [31].

Fig. 5
figure 5

FTIR spectra of the Fe3O4 /C core–shell

3.3 Particle Size and Morphology

The TEM micrograph of the Fe3O4/C core shell NPs is presented in Fig. 6. Most of the particles are quasi-spherical, and also the average size of most particles is nearly 10 nm as shown in the size distribution diagram (Fig. 7). Clearly, it is seen that the NPs have core–shell structures. In addition, the dark parts of NPs are the Fe3O4 core and the light parts are a carbon shell.

Fig. 6
figure 6

Transmission electron microscopy image of Fe3O4/C core–shell nanoparticles

Fig. 7
figure 7

Size distribution of Fe3O4/C nanoparticles

We understand there is no report about the preparation of Fe3O4/C NPs with diameters below 20 nm. Also, the particle size was estimated from TEM and was consistent with XRD results.

The results obtained from the Debye–Scherrer method, W–H calculation, and TEM are summarized in Table 1. The values of the average crystallite size of the Fe3O4 and Fe3O4/C NPs estimated from the different models more or less are similar, implying that the inclusion of strain in various forms has a very small effect on the average crystallite size of NPs.

Table 1 The average value of crystallite size of the Fe3O4 and Fe3O4/C NPs obtained from the different models

3.4 Magnetic Properties

For MNPs, the magnetic properties, especially the saturation magnetization M s and coercive force H c , have been found to be dependent on the size and shape of the particles as well as their chemical composition [2, 11]. Figure 8 presents the experimental hysteresis curve of as-prepared Fe3O4 and Fe3O4/C core–shell NPs, respectively, which obtained by vibrating sample magnetometer (VSM) instrument. The remanence and coercivity are zero, which indicates that the prepared Fe3O4 and Fe3O4/C NPs have a superparamagnetic phase [39].

Fig. 8
figure 8

Room-temperature magnetic hysteresis loops for the Fe3O4 and Fe3O4/C NPs

Small magnetic NPs often consist of only one domain and exhibit a phenomenon known as the supeparamagnetism phase. In this case, the magnetization direction may change under thermal excitation in the absence of a field. There are two critical sizes. The first critical radius, R SD, marks a region where in the particle changes from being a multidomain to a single-domain particle. There is not a domain wall in this radius because the increase of energy due to the domain wall formation dominates over the decrease in energy attributed to the formation of the domain. Thus, less than a critical size of particles, domain walls, will not exist. There is a second critical radius less than R SD that is called a superparamagnetic radius, R SP. Superparamagnetic particles do not possess long range magnetic order [2, 40]. The value of R SP is normally a few nanometers and different for many nanoparticles materials. According to a report by Frenkel and Dorfman, R SP is <15 nm for the common materials [14, 41]. R SP for Fe3O4 NPs is about 10 nm. In comparison, the saturation magnetization per unit mass of Fe3O4 NPs is 49 emu/g and for Fe3O4/C NPs is equal to 40 emu/g, which indicates that Fe3O4/C possesses a weaker response to the external magnetic field in comparison to the Fe3O4 sample. The magnetization of Fe3O4 was reduced after coating it with carbon. The reductions of saturation magnetization per unit mass attributed to the reduction of magnetic moment on the interface of core–shell due to interaction of iron oxide with carbon shell and shielding effect of the carbon shell, which depends on the amount of carbon and, therefore, thickness of the shell. Also, magnetization depends to the size and shape of particles and with an increase in size, it will increase. The saturation magnetization per unit mass for bulk Fe3O4 is about 92.8 emu/g. As the superparamagnetic Fe3O4 NPs are located under the external magnetic field; their magnetic momentum will align with an external field. Amorphous carbon is a diamagnetic material. There is not an unpaired electron in last orbital of carbon, therefore, when located in vicinity of the external magnetic field, it alters the orbital velocity of electrons around their nuclei, thus the magnetic dipole moment was changed. According to Lenz’s law, the field of these electrons will oppose the magnetic field changes provided by the applied field.

4 Conclusions

Carbon-encapsulated iron oxides (Fe3O4/C) NPs with core/shell structure have been successfully synthesized by simple hydrothermal method at low temperature, using the two-step process. The core–shell structure was confirmed by FTIR analysis. Size of core–shell particles were estimated by nearly 10 nm by the Debye–Scherer’s equation and Williamson–Hall methods, and results from these methods are well consistent with the estimated size from the TEM image. Prepared NPs are superparamagnetic and have no coercivity and remanence. The magnetization of Fe3O4 NPs was reduced after coating with carbon. The reduction of saturation magnetization per unit mass, M s , should be attributed to the interaction between the iron oxide core and the amorphous carbon shell at the interface and shielding effect of carbon shell, which depends on the amount of carbon and, therefore, the thickness of the shell. To the best of our knowledge, there is no report about the preparation of Fe3O4/C core–shell NPs with diameters below 20 nm by a two-step process.