Preparation of γ-Fe2O3/Ni2O3/FeCl3(FeCl2) Composite Nanoparticles by Hydrothermal Process Useful for Ferrofluids
Using a hydrothermal process in FeCl2 solution, γ-Fe2O3/Ni2O3/FeCl3(FeCl2) composite nanoparticles were obtained from the FeOOH/Ni(OH)2 precursor prepared by coprecipitation. The precursor and the as-prepared nanoparticles were investigated by vibrating sample magnetometer (VSM), X-ray diffraction (XRD), energy disperse X-ray spectroscopy (EDX), X-ray photoelectron spectroscopy (XPS), and transmission electron microscopy (TEM). The experimental results showed that the paramagnetic amorphous precursor, in which Ni(OH)2 is formed outside FeOOH, is transformed to ferrimagnetic γ-Fe2O3/Ni2O3 composite when it is processed in FeCl2 solution (0.25, 0.50, 1.00 M) in an autoclave at 100°C for 1 hr. In addition, the dismutation reaction of FeCl2 produces FeCl3 and Fe. Some FeCl3 and little FeCl2 can be absorbed to form γ-Fe2O3/Ni2O3/FeCl3(FeCl2) composite nanoparticles in which Ni2O3 forms outside the γ-Fe2O3 core and the outermost layer is FeCl3 (FeCl2). The content of FeCl3 (FeCl2) in the particles increased, and the magnetization of the particles decreased with the concentration of FeCl2 solution increasing in the hydrothermal process. The FeCl3 (FeCl2) surface is chemically passive and nonmagnetic (paramagnetic). Accordingly, the composite nanoparticles are chemically stable, and their aggregation is prevented. The specific saturation magnetization of such composite nanoparticles can get to 57.4–62.2 emu/g and could be very suitable for synthesizing ferrofluids.
Magnetic nanoparticles with diameters less than 100 nm have attracted increasing interest in the fields of basic science and technology [1–3]. Studies of magnetic nanoparticles have focused on the development of novel synthetic technology . A nanocomposite is a material composed of two or more phases where the combination of different physical or chemical properties may lead to completely novel materials . Magnetic nanocomposites have applications ranging from ferrofluids to separation science and technology . Synthesis of nanoparticles of several kinds of materials can be achieved by the coprecipitation method. Generally, the precursor, synthesized by chemical coprecipitation, needs to be further processed (by drying, calcinations, etc.) to form oxide nanoparticles . During the chemical reaction, followed by calcination or annealing, a new phase is formed. However, in addition to the transition from amorphous to crystalline, the particle size increases with the calcination temperature and also aggregation of crystallites occurs . Therefore, for the synthesis of ferrofluids, nanoparticles prepared by calcination could be unsuitable.
The hydrothermal process is a successful way to grow crystals of many different materials . In the present work, composite nanoparticles based on γ-Fe2O3, Ni2O3, and FeCl3 (FeCl2) were prepared by processing the amorphous FeOOH/Ni(OH)2 precursor in FeCl2 solution in an autoclave at 100°C. The effect of the concentration of FeCl2 solution on the product’s magnetization, crystal structure, chemical composition, and morphology was investigated. A new route for the preparation of magnetic composite nanoparticles is proposed.
The preparation of the nanoparticles can be divided into two steps. First, the precursor was synthesized using the coprecipitation method. An aqueous mixture of FeCl3 (40 mL, 1 M) and Ni(NO3)2 (10 mL, 2 M in HCl 0.05 mol), in which the ratio of Fe3+ to Ni2+ was 2 : 1, was added to NaOH solution (500 mL, 0.7 M). Then, the solution was heated to boiling for 5 minutes with stirring. After the heating was stopped, the black precursor gradually precipitated. The precursor was washed with a low concentration of HNO3 solution (0.01 M) to pH = 7~8. The second step was to obtain the composite nanoparticles. The precursor was added to FeCl2 solution, whose concentration was 0.25 M, 0.50 M, or 1.00 M, to form a mixture solution (200 mL). This was then poured into a stainless steel autoclave with a Teflon liner (whose volume is 400 mL), heated to 100°C for 1hr and allowed to cool naturally to room temperature. Finally, the particles were washed with acetone and allowed to dry naturally.
The magnetization, crystal structure, chemical composition, and morphology of both the precursor and processed samples were analyzed using vibrating sample magnetometry (VSM, HH-15), X-ray diffraction (XRD, XD-2), energy dispersive X-ray spectroscopy (EDX, Norton 8000), X-ray photoelectron spectroscopy (XPS, Thermo ESCA 250), and transmission electron microscopy (TEM, PHLIPS TECNAI 10).
3. Results and Analysis
Figure 1 shows the magnetization curves at room temperature for the precursor and the three processed samples. Clearly, the precursor appears to be paramagnetic, and the processed samples are ferromagnetic. It can be deduced from Figure 1 that the magnetization of the processed samples weakens as the concentration of FeCl2 solution increases. The specific saturation magnetization of ferromagnetic materials can be estimated from the linear relationship of versus 1/H at high field . The particles prepared by the hydrothermal process in 0.25 M, 0.50 M, and 1.00 M FeCl2 solution give values of 62.2, 59.2, and 57.4 emu/g, respectively.
Figure 2 displays the XRD patterns for the precursor and the three processed samples. The results indicate that the precursor is amorphous, and the three processed samples are crystalline, mainly containing γ-Fe2O3. According to Scherrer’s formula, the γ-Fe2O3 grain sizes of the particles may be calculated from the width of the (311) diffraction peak. This gives nearly the same size, about 8 nm, for all samples. For the processed samples in 1.00 M FeCl2 solution, in addition to the diffraction peaks of the γ-Fe2O3 phase, a few new peaks can be observed. Three new peaks correspond to crystal face spacings d of approximately 0.507, 0.480, and 0.304 nm, as indicated by arrows A, B, and C, respectively, in Figure 2(d). These values are close to the , 0.450, and 0.298 nm of the strongest peaks of 2FeCl3·5H2O. This demonstrates that 2FeCl3·5H2O has formed after the hydrothermal process, and its presence increases with the concentration of FeCl2. In addition, little peaks at 2θ = 44.8°and 65.2° may be seen in Figure 2(d) as indicated by arrows A′ and B′, and they belong to (110) and (200) diffraction peaks of Fe, respectively.
EDX results show that the precursor was constituted from Ni and Fe, although the processed samples contained some Cl as well. Quantitative analysis indicated that the mole ratio of Fe to Ni in the precursor was about 2 : 1, which is the same as that of the starting reagents, but the ratios in the processed samples are far larger than that of the starting reagents or the precursor. The detailed data are listed in Table 1. The XPS results indicated that the precursor consisted of Fe, Ni, and O, and the processed samples contained some Cl in addition. It was found by quantitative analysis that the ratio of Fe to Ni in the precursor is less than 2 : 1, but in the processed samples are greater than 2 : 1. In addition, the concentration of Cl in the processed samples increased with increasing concentration of FeCl2 solution in the hydrothermal process. These data are listed in Table 2. The values of the binding energy of the measured samples showed that the precursor consisted of both FeOOH and Ni(OH)2, and the processed samples consisted of Fe2O3, Ni2O3, and FeCl3 (FeCl2). The complete data of the binding energies are listed in Table 3.
Typical TEM photographs of the precursor and the processed samples are shown in Figure 3. It can be seen from the TEM observations that the precursor is loosely aggregated and the processed samples contain quasispherical particles with nearly uniform size of a few nanometers (as indicated by arrow A) in addition to some aggregates (as indicated by arrow B).
The experimental results show that the precursor is an amorphous composite of FeOOH and Ni(OH)2. Since the ratio of Fe to Ni measured by XPS is less than that of the starting reagents and the EDX measurement is in agreement, it can be concluded that the Ni(OH)2 species has formed outside the FeOOH species in the precursor because the XPS information comes from the surface layer, less than 3 nm thick, but the EDX information comes from a depth of about 1 μm. Similarly, it can also be concluded that the FeCl3 (FeCl2) is formed outside the outermost layer of the particles.
From the experimental results, it can be seen that when the amorphous precursor, consisting of FeOOH/Ni(OH)2, was hydrothermally processed in FeCl2 solution, a transition process took place to form the Fe-Ni oxide composite. The schematic main reactions in the hydrothermal process can be described as In the starting reagents, the content of Ni(NO3)2 is 0.02 mol, so that the one of Ni(OH)2 is so. If the Ni transforms completely into Ni2O3 following the schematic reaction (2), the reaction needs 0.005 mol O2 which is about 112 mL O2, that is, about 530 mL air. Obviously, the stainless steel autoclave is not as bigger enough to contain so much air. This means that only partial nickel hydroxide (Ni(OH)2) species transformed into Ni2O3, and other Ni(OH)2 species dissolved in water. This agrees with the measured results of both EDX and XPS in which the ration of Fe to Ni is much larger than 2 : 1 for the processed samples. The dissolution of Ni(OH)2 can enhance the pH value of the solution, which would assist precipitation. In addition, it is noted that for the processed samples, the concentrations of both Ni and Cl measured by XPS are larger than the ones measured by EDX. From the difference between the XPS and EDX data, it can be determined that the particle core consists of γ-Fe2O3 and Ni2O3 is formed outside the γ-Fe2O3, a result which is in agreement with the precursor structure consisting of both FeOOH and Ni(OH)2.
The FeCl3 could result from the dismutation reaction of some FeCl The Fe could form into iron nanoparticles. And, since pure FeCl3 is very easy to dissolve in water, the FeCl3 would be absorbed on γ-Fe2O3/Ni2O3 particles to form the outermost layer of the composite nanoparticles. XRD measurements of the processed samples in which there are new peaks of both FeCl3 and Fe, support the opinion.
In the FeCl2 solution process, since the Ni(OH)2 partially dissolved, the FeCl3 was absorbed, the Fe nanoparticles are formed, and the ratio of Fe to Ni in derived nanoparticles is larger than in the precursor for both EDX and XPS. In combination with the XRD results, it is further concluded that the amount of FeCl3 will increase with the concentration of FeCl2 solution in the hydrothermal process. The reaction (3) shows that the increasing amount of Fe is only a half of that of FeCl3, so that the affection of the FeCl3 increasing could be more obvious than the Fe increasing with the concentration of FeCl2 solution. This is in agreement with the magnetization measured by VSM since FeCl3 is paramagnetic, and its magnetization is less than the ferrimagnetic γ-Fe2O3 or weakly magnetic Ni2O3 (see the inset in Figure 1).
According to the data in Table 2(d), it can be known that if all the Cl is from FeCl3, the content of Fe should be larger than 72.51 which is three times that of Cl. But, in fact the content of Fe is only 65.37 as shown in Table 2(d). Thus, it is judged that FeCl2 could exist in the outer layer and increase with the concentration of the processing solution. Therefore, for the surface layer outside the magnetic γ-Fe2O3/Ni2O3, the exact description should be as FeCl3 (FeCl2). The FeCl3 (FeCl2) surface is a chemically passive and nonmagnetic (paramagnetic) coating. Thus, the chemical stability of the particles is improved and their aggregation is prevented, a phenomenon similar to the effect of silica coating on the surface of iron oxide nanoparticles [10–12]. And the content of FeCl3 (FeCl2) in the composite particles increased with the concentration of FeCl2 solution in the hydrothermal process, so the magnetization of the particles lessened accordingly. The saturation magnetizations of the as-prepared composite nanoparticles (57.4~62.2 emu/g) are close to the ones of the nanocrystalline γ-Fe2O3 particles prepared by generally hydrothermal synthesis (52.78~72.87 emu/g) . Using the particles prepared in 1 M FeCl2 solution, ionic ferrofluids  have been synthesized and the ferrofluid particles observed by TEM, as shown in Figure 4. It can be seen that these ferrofluid particles were well dispersed. Also, experiments have shown that such ferrofluids based on γ-Fe2O3/Ni2O3/FeCl3(FeCl2) nanoparticles exhibited greater optical transmission than the ones based on γ-Fe2O3 nanoparticles. This could be very interesting for the field of magneto-optical effects and will be investigated further.
Using a coprecipitation method, a precursor of amorphous composite in which FeOOH is coated with Ni(OH)2, has been produced. Using a hydrothermal process in FeCl2 solution, a transition is induced to form γ-Fe2O3/Ni2O3/FeCl3(FeCl2) composite nanoparticles about 10 nm in diameter besides little amount of Fe nanoparticles. The magnetizations of γ-Fe2O3, Ni2O3, and FeCl3 (FeCl2) are different. Accordingly, these composite nanoparticles show a gradient of magnetization from core to shell. In many cases, a passivating coating can be applied to avoid agglomeration (important for ferrofluids) . Therefore, such composite nanoparticles could be particularly suitable for the synthesis of ferrofluids. In the hydrothermal process, the formation of the inert FeCl3 (FeCl2) surface immediately follows the formation of the particles themselves, in situ and in the identical FeCl2 medium. This contrasts with the coating method, in which the coating is added to as-prepared particles and the possibility of aggregation of the particles exists before coating, that is, the so-called preaggregation . Therefore, the degree of dispersion of the particles prepared by this hydrothermal process in FeCl2 solution may be better than for particles coated using inert materials. The transition role from hydroxyl composite to oxide composite in FeCl2 solution could provide to a route to prepare oxide composite nanoparticles if any species in the precursor is insoluble or only slightly soluble in water.
Financial support for this work was provided by the National Nature Science Foundation Project of China (no. 11074205).
D. Y. Szabó and D. Vollath, “Nanocomposites from coated nanoparticles,” Advanced Materials, vol. 11, no. 15, pp. 1313–1316, 1999.View at: Google Scholar
Q. Liu, Z. Xu, J. A. Finch, and R. Egerton, “A novel two-step silica-coating process for engineering magnetic nanocomposites,” Chemistry of Materials, vol. 10, no. 12, pp. 3936–3940, 1998.View at: Google Scholar
L. N. Donselaar, A. P. Philipse, and J. Suurmond, “Concentration-dependent sedimentation of dilute magnetic fluids and magnetic silica dispersions,” Langmuir, vol. 13, no. 23, pp. 6018–6025, 1997.View at: Google Scholar
M. A. Correa-Duarte, M. Giersig, N. A. Kotov, and L. M. Liz-Marzán, “Control of packing order of self-assembled monolayers of magnetite nanoparticles with and without SiO coating by microwave irradiation,” Langmuir, vol. 14, no. 22, pp. 6430–6435, 1998.View at: Google Scholar
P. C. Scoholten, “The origin of magnetic birefringence and dichroism in magnetic fluids,” IEEE Transactions on Magnetics, vol. 16, p. 221, 1980.View at: Google Scholar