Abstract

During the synthesis of the γ-Fe2O3 nanoparticles via the chemically induced transition method, Cu(I) modification has been attempted by adding CuCl/NaOH to the treatment solution. The experimental results showed that, under the condition of a NaOH content equal to 0.04 moles, when the content of CuCl is as low as or moles, the products are single γ-Fe2O3/Cu(I)FeO2/FeCl3·6H2O composite nanoparticles, whereas when the content of CuCl is higher, moles, the product is a mixture consisting of γ-Fe2O3/Cu(I)FeO2/FeCl3·6H2O nanoparticles and Cu(II)(OH)Cl nanoparticles. For the γ-Fe2O3/Cu(I)FeO2/FeCl3·6H2O composite nanoparticles, the Cu(I)FeO2 interface layer is not thick enough to form one unit cell, but it can modify the formation of a FeCl3·6H2O surface layer and the effective magnetization of the γ-Fe2O3 core.

1. Introduction

By definition, nanomaterials have one or more dimensions in the nanometer scale range (<100 nm) and consequently show novel properties when compared to bulk materials [1]. A nanocomposite is a material composed of two or more phases. Nanoparticles are typically defined as solids measuring less than 100 nm in all the three dimensions; composite nanoparticles are generally coatings in which the combination of different physical and chemical properties may lead to completely novel materials with modified properties [1, 2]. Significant research effort has shown that the surface modification of the particles can be easily accomplished in postsynthesis steps or during the synthesis, thereby providing the nanoparticles with additional functionalities [3]. Magnetic nanoparticles constitute an important class of functional materials and can be categorized, based on single or multiple materials, into simple and core/shell—or composite—nanoparticles, which are gaining increasing interest because of their novel properties and the numerous applications in many diverse fields [35]. Composite magnetic iron oxide nanoparticles have applications ranging from ferrofluids to separation science and technology [68].

Studies on nanoparticles have focused on the development of simple and effective methods for fabricating nanomaterials with controlled size and morphology and hence tailoring their properties [9]. Liquid-phase synthesis is often used to prepare inorganic nanoparticles [10]; the conventional aqueous synthesis of the γ-Fe2O3 particles involves three or more steps [11, 12]. We have proposed a method to synthesize γ-Fe2O3 nanoparticles by thermally treating the FeOOH/Mg(OH)2 precursor in FeCl2 treating solution [13, 14]. Through this method, known as chemically induced transition (CIT) method, FeOOH species were transformed into γ-Fe2O3 nanocrystallites by dehydration and Mg(OH)2 was dissolved to assist the precipitation of the nanoparticles. Besides, the Fe2+ ions in the FeCl2 solution were oxidized to Fe3+ to form a FeCl3·6H2O coating on the γ-Fe2O3 nanocrystallites [14]. During the synthesis of the γ-Fe2O3 nanoparticles, the surface modification was performed by adding a salt solution composed of metal ions with a valency of two, such as Zn(II)Cl2 [15, 16] and Co(II)(NO3)2 [17], to fabricate Zn(II)Fe2O4 and Co(II)Fe2O4 epitaxial layers, respectively, on the γ-Fe2O3 crystallites. Additional NaOH can enhance these modifications to produce more ZnFe2O4 or CoFe2O4. In this work, a surface modification using metal ions with a valency of one is attempted by adding Cu(I)Cl/NaOH to the FeCl2 treating solution, and the as-prepared products were characterized using multiple techniques. Accordingly, the features of the Cu(I)-modified γ-Fe2O3 nanoparticles were investigated.

2. Materials and Methods

2.1. Chemicals

Ferric chloride (FeCl3), magnesium nitrate (Mg(NO3)2), ferrous chloride (FeCl2), cuprous chloride (CuCl), and sodium hydroxide (NaOH) were of analytical grade, and all other chemicals were used as received without further purification. Distilled water was used throughout the experiments.

2.2. Preparation

The preparation of Cu(I)-modified γ-Fe2O3 nanoparticles by the so-called chemically induced transition method can be divided into two steps. First, the FeOOH/Mg(OH)2 precursor was synthesized by coprecipitation, as described in detail elsewhere [13]. Second, 5 g of the precursor was added to a boiling FeCl2 treating solution (0.25 M, 400 mL) and kept boiling under reflux for 20 min. Then, both CuCl solution (50 mL), with a varying concentration, and NaOH solution (2 M, 20 mL) were added simultaneously to the solution, and the resulting mixture was boiled continuously for 10 min. After cooling naturally to room temperature, the products precipitated from the solution; subsequently, they were washed with acetone and allowed to dry. The concentrations of the CuCl solutions were 0.025, 0.050, and 1.000 M, corresponding to samples (1), (2), and (3), respectively. For comparison, unmodified particles were also prepared by adding the precursor to the FeCl2 solution and boiling for 30 min, producing sample (0).

2.3. Characterization

The bulk chemical compositions of the as-prepared products were obtained by energy dispersive X-ray spectroscopy (EDS, Quanta-200). The morphologies were observed by transmission electron microscopy (TEM, G20ST). The crystal properties were analyzed by high-resolution TEM (HRTEM, JEM-2100F) and X-ray diffractometry (XRD, D/Max-RC). The surface chemical compositions were determined by X-ray photoelectron spectroscopy (XPS, ESCALAB250Xi). The specific magnetization curves were measured at room temperature by vibrating sample magnetometry (VSM, HH-15).

3. Results and Discussion

3.1. Results

The EDS measurements revealed that samples (1), (2), and (3) contained not only O, Fe, and Cl and no Mg or Na—as sample (0)—but also Cu. Figure 1 shows the EDS spectra. The atomic percentages of the Fe, Cl, and Cu elements ( values), listed in Table 1, show that the amount of Cu increases from sample (1) to sample (3).

Typical TEM images are shown in Figure 2. All the samples are made of approximately spherical nanoparticles involving hexagonal particles. The size of the nanoparticles for both samples (1) and (2) is clearly larger than that of the unmodified sample (0). Additionally, sample (3) contained two kinds of particles, larger size and smaller size, as the arrows marked with “A” and “B,” respectively, indicate.

The HRTEM revealed these particles to be small crystallites. Figure 3 is a typical HRTEM image of sample (2), which shows that the interlaced lattice fringes of two sets of planes have the same spacing of approximately 0.29 nm. As shown by the XRD spectra in Figure 4, the modified samples can be divided into two types, with samples (1) and (2) belonging to the same type and sample (3) showing notable differences. Similar to the unmodified sample (0), samples (1) and (2) predominantly possess a ferrite-like spinel structure with the features of γ-Fe2O3 (JCPDS card number 39-1346). Sample (3) clearly contained γ-Fe2O3 and Cu(II)(OH)Cl (JCPDS card number 23-1063).

For samples (1), (2), and (3), the XPS results confirmed the presence of the same chemical elements as determined by EDS. The quantitative results for the Fe, Cl, and Cu elements are listed in Table 1. Figure 5 shows the O 1s, Fe 2p3/2, Cl 2p, and Cu 2p spectra for samples (1) to (3). Similar to the XRD results, the XPS spectra can also be divided into two categories and are analyzed as follows.

Both samples (1) and (2) have the same XPS spectral structure. Their O 1s spectra exhibited two peaks: the P1 peak at approximately 529.5 eV can be attributed to Fe2O3, and the P2 peak at approximately 532.2 eV is due to molecular H2O [19]. The Cl 2p3/2 peak at approximately 198.7 eV corresponds to Cl 2p3/2 in FeCl3 (199 eV), which is in good agreement with the probable presence of FeCl3·6H2O, as unmodified sample [14]. Thus, the Fe 2p3/2 peak at ~711.2 eV resulted from both γ-Fe2O3 and FeCl3. The Cu 2p3/2 peak at ~932.0 eV clearly corresponds to Cu(I)FeO2 (Cu 2p3/2 peaks at 932.6 eV), rather than Cu(II)Fe2O4 (Cu 2p3/2 peaks at 933.7 eV) [20]. The binding energy data of both samples (1) and (2) are listed in Table 2(a). For sample (3), the O 1s and Fe 2p3/2 spectra are similar to those of samples (1) and (2). The Cl 2p spectrum exhibited two peaks, and the P1 peak (198.59 eV) could correspond to FeCl3. Therefore, the Fe-containing compounds in sample (3) also consisted of both γ-Fe2O3 and FeCl3·6H2O. The spectrum of Cu 2p of sample (3) has a more complicated structure than the spectra of both samples (1) and (2). It is known that the filled 3d shell of Cu+ could prevent the ligand-metal charge transfer shake-up transition from occurring, so that one major difference between Cu(I) oxide and Cu(II) oxide is the satellite structure on the high energy side of the copper core lined in Cu(II) oxide [20]. The XRD analysis clearly determined that sample (3) contained Cu(OH)Cl; thus, besides the P0 peak corresponding to the Cu peaks present in samples (1) and (2) and attributed to Cu 2p3/2 in CuFeO2, the P0′ peak could originate from Cu 2p1/2 in CuFeO2 and other peaks could result from Cu(OH)Cl. The Cu spectrum of sample (3) shows a pronounced shake-up satellite (SAT) structure, which is similar to the structure of Cu 2p in CuCl2 [18]. Furthermore, as the Cl 2p spectrum of CuCl2 has two distinguishable Cl 2p1/2 and Cl 2p3/2 peaks [18], the spectrum of Cu(OH)Cl could contain a contribution from both Cl 2p1/2 and Cl 2p3/2, explaining the two peaks exhibited by the Cl spectrum of sample (3). The binding energy data of sample (3) are listed in Table 2(b). As a comparison, the data of CuCl2 are also listed in Table 2(b).

The specific magnetization curves of samples (0), (1), (2), and (3) in Figure 6 show that the samples exhibit a ferromagnetic behavior. Thus, the specific saturation magnetization, , can be estimated from a plot of versus at high field [21]. For samples (0), (1), (2), and (3), the are 57.71, 66.59, 71.80, and 38.37 emu/g, respectively.

3.2. Discussion

According to the experimental results, every sample modified by CuCl/NaOH contained same γ-Fe2O3 phase as the unmodified sample (0). Also, the modified samples have Cu-containing compounds. This result indicates that the precursor first transformed into γ-Fe2O3 crystallites in the FeCl2 treating solution and, then, following the addition of CuCl/NaOH, the Cu-containing compounds were formed. Under an additional specific content of NaOH equal to moles (20 mL, 2 M), the composition of the as-prepared products depended on the content of additional CuCl. For lower contents of CuCl— moles (50 mL, 0.025 M) and moles (50 mL, 0.050 M), that is, samples (1) and (2), respectively—the Cu-containing compound is Cu(I)FeO2, whereas when the content of CuCl is higher— moles (50 mL, 1.000 M), that is, sample (3)—the product exhibits two different Cu-containing compounds: Cu(I)FeO2 and Cu(II)(OH)Cl. These results reveal that, by adding CuCl/Na(OH) to the treating solution, the oxidation of the Cu+ ions and the formation of the Cu-containing compounds could depend on the concentration of CuCl under a certain content of NaOH. At low CuCl concentration, all the Cu+ ions formed Cu(I)FeO2 and no Cu(II) compound was created, whereas, at higher CuCl concentration, part of the Cu+ ions formed Cu(I)FeO2, and the remaining Cu+ ions were oxidized to Cu2+ ions to produce Cu(II)(OH)Cl. The formation of Cu(OH)Cl rather than CuFe2O4 could be due to a reason similar to that behind the formation of Cu2(OH)3NO3, where Cu2+ does not coprecipitate with Fe3+ as the pH of the reaction is too low [22].

For the unmodified sample (0), it is known that FeCl3·6H2O forms on the γ-Fe2O3 crystallites [14]. XPS results revealed the presence of FeCl3·6H2O in the modified samples (1), (2), and (3). For both samples (1) and (2), based on the γ-Fe2O3, Cu(I)FeO2, and FeCl3·6H2O phases, the nanoparticles contained γ-Fe2O3 and FeCl3·6H2O, as confirmed by the lattice fringes of the two sets of planes observed in the HRTEM images, which precisely correspond to the (220) plane of γ-Fe2O3 (spacing of 0.2953 nm) and (002) plane of FeCl3·6H2O (spacing of 0.2927 nm), respectively. The Cu(I)FeO2 could be an intermediate layer formed between the γ-Fe2O3 core and the FeCl3·6H2O surface layer, as the ratio of Cu to Cl obtained from the XPS measurement is lower than that obtained from the EDS measurement. As a consequence, a schematic model of the particle structure of both samples (1) and (2) is shown in Figure 7. This inference can be explained as follows.

Notably, the collection depth of the signal in the EDS analysis largely exceeds the dimensions of the nanoparticles, whereas the signal collection depth in the XPS experiment is ~3λ, where λ = 1.24 nm and 1.51 nm for Fe 2p and Cu 2p electrons, respectively [23, 24]. After statistical analysis [25], the TEM results revealed that the size of the particles of both samples (1) and (2) followed a lognormal distribution similar to that of sample (0). Their median diameter and standard deviation are listed in Table 3. Consequently, the average size of the particles is calculated by the formula [25], and the results show the size of samples (1) and (2) larger than 11 nm, as also listed in Table 3. Thus, for the nanoparticles in samples (1) and (2), the EDS results reflected the average ratio of the elements in the nanoparticles, whereas the XPS results reflected the ratio of the elements near the nanoparticle surface. Therefore, for the nanoparticles having a multiple layer structure, as shown in Figure 7, as the depth of the XPS detection is smaller than the radius of the particles (as ), the ratio of the elements in the internal layer to the elements in the outer layer, from XPS measurement, is lower than the average ratio in the total nanoparticle measured by EDS. For samples (1) and (2), the experimental results (see Table 1) show that the ratio of Cu to Cl from XPS measurements is far lower than that obtained from EDS measurements. Accordingly, for samples (1) and (2), the Cu(I)FeO2 grows between the γ-Fe2O3 core and the FeCl3·6H2O surface layer to form γ-Fe2O3/Cu(I)FeO2/FeCl3·6H2O composite nanoparticles.

Furthermore, sample (3) exhibited strong diffraction peaks of Cu(II)(OH)Cl in the XRD spectra. Therefore, sample (3) is a mixture of γ-Fe2O3/Cu(I)FeO2/FeCl3·6H2O nanoparticles and Cu(II)(OH)Cl nanoparticles, which may correspond to the larger and smaller particles, respectively, observed in the TEM image.

For sample (0), based on γ-Fe2O3/FeCl3·6H2O nanoparticles, and samples (1) and (2), based on γ-Fe2O3/Cu(I)FeO2/FeCl3·6H2O nanoparticles, the molar percentages of the γ-Fe2O3 phase, , Cu(I)FeO2 phase, , and FeCl3·6H2O phase, , can be estimated by where , , and are the atomic percentages of Fe, Cu, and Cl, respectively, and is equal to zero for sample (0). Thus, the molar percentages of every phase in samples (0), (1), and (2) can be obtained from the values of , , and as measured by EDS (see Table 1). As a consequence, the mass percentages of these phases can be derived from where is the molar percentage and is the molar mass of the phase. Accordingly, the mass percentages of each phase in samples (0), (1), and (2) were calculated from the values of and the molar masses of γ-Fe2O3, Cu(I)FeO2, and FeCl3·6H2O. The values of both and are listed in Table 4.

For samples (1) and (2), the specific magnetization can be described by where , , and are the specific magnetization of the γ-Fe2O3, Cu(I)FeO2, and FeCl3·6H2O phase, respectively, and is the mass fraction of the phase. As γ-Fe2O3 and Cu(I)FeO2 are ferrimagnetic and FeCl3·6H2O is paramagnetic, the specific magnetization of both samples (1) and (2) results mainly from the mass percentages of the ferrimagnetic phase, . From the results listed in Table 4, for sample (1) and for sample (2). Thus, the specific magnetization of sample (1) is slightly less than that of sample (2). In addition, sample (3) contained Cu(II)(OH)Cl particles, so the specific magnetization of sample (3) is lower than that of both samples (1) and (2).

The magnetization (moment per unit volume) is an important parameter used to characterize magnetic materials. For a particle system, the magnetization can be obtained generally from , where is the specific magnetization (moment per unit mass) and is the density of the material. For composite nanoparticles containing many phases with different densities, the density of the particles should be taken as the average density , which can be derived from where is volume percentage and is density of the phase. The can be described as Accordingly, the volume percentages of the γ-Fe2O3, Cu(I)FeO2, and FeCl3·6H2O phases in samples (0), (1), and (2) can be calculated, and the average density for every sample can be derived. Therefore, the saturation magnetization can be obtained. These results, , , and for samples (0), (1), and (2), are listed in Table 5.

For samples (1) and (2), the amount of Cu(I)FeO2 listed in Table 5 is so small that the thickness of the Cu(I)FeO2 in the γ-Fe2O3/Cu(I)FeO2/FeCl3·6H2O nanoparticles is no more than one unit cell, and Cu(I) is not in the same state as in bulk Cu(I)FeO2; this case is similar to that of γ-Fe2O3 nanoparticles having CoFe2O4 layer less thick than one unit cell [26]. The presence of Cu(I)FeO2 could modify the formation of FeCl3·6H2O and make the FeCl3·6H2O layer assume a three-dimensional (3D) oriented arrangement [27], so as to have a certain crystallized orientation relatively to the γ-Fe2O3 crystallites. Therefore, both γ-Fe2O3 and FeCl3·6H2O are identified in the HRTEM stripe images showing the two sets of planes. Also, the saturation magnetization of both the modified samples (1) and (2) is higher than that of the unmodified sample (0), although the ferrite volume percentages of the formers are lower than that of the latter. It is known that the spins close to the surface to be pinned by surfactant molecules, which cause anomalously large magnetic anisotropy, would result in the less apparent saturation magnetization of nanoparticles than that of the bulk [28]. Accordingly, it is judged that the Cu(I)FeO2 thin layer in the γ-Fe2O3/Cu(I)FeO2/FeCl3·6H2O nanoparticles may modify the magnetically silent “dead layer” [29], which existed at the interface between the γ-Fe2O3 and FeCl3·6H2O phases of the γ-Fe2O3/FeCl3·6H2O nanoparticles of the unmodified sample (0) and did not provide any contribution to the effective magnetization.

4. Conclusions

When the FeOOH/Mg(OH)2 precursor was thermally treated in FeCl2 solution, the Mg(OH)2 dissolved, FeOOH transformed into γ-Fe2O3 nanocrystallites, and Fe2+ in the FeCl2 treating solution was simultaneously oxidized to Fe3+. The nanocrystallites absorbed Fe3+ and Cl to form γ-Fe2O3/FeCl3·6H2O nanoparticles, in which the γ-Fe2O3 core was coated with the FeCl3·6H2O layer. By adding Cu(I)Cl/NaOH to the FeCl2 solution during the synthesis, the compositions of the as-prepared products can be modified. For a certain content of NaOH, 0.04 moles, using a low content of Cu(I)Cl ( moles or moles), single Cu(I) modified composite nanoparticles can be prepared. The structure of such composite nanoparticles can be described as γ-Fe2O3/Cu(I)FeO2/FeCl3·6H2O and consisted of three parts as follows: γ-Fe2O3 core, a Cu(I)FeO2 intermediate layer, and an outermost FeCl3·6H2O layer. For a higher content of CuCl, moles, Cu+ was partially oxidized to Cu2+, and the as-prepared product was a mixture of γ-Fe2O3/Cu(I)FeO2/FeCl3·6H2O nanoparticles and Cu(II)(OH)Cl nanoparticles. The Cl and Cu spectra of Cu(OH)Cl measured by XPS have the same structure as those of CuCl2, and the binding energies of Cl 2p3/2, Cl 2p1/2, Cu 2p3/2, and Cu 2p1/2 for the Cu(OH)Cl compound are approximately 198.6, 200.1, 934.6, and 954.5 eV, respectively.

For the γ-Fe2O3/Cu(I)FeO2/FeCl3·6H2O composite nanoparticles, the average thickness of the Cu(I)FeO2 layer is not enough to form one unit cell, and Cu(I) is not in the same state in bulk Cu(I)FeO2. The experimental results show that the Cu(I)FeO2 intermediate layer could modify the formation of the FeCl3·6H2O layer to stimulate a 3D oriented attachment of the layer relatively to the γ-Fe2O3 crystallites; it may also modify the magnetic “dead layer” between the γ-Fe2O3 core and FeCl3·6H2O surface layer to enhance the effective magnetization. Besides having high magnetization, such nanoparticles have an inert FeCl3·6H2O surface; therefore they could possess a relatively good chemical stability and can be used directly to synthesize ionic ferrofluids without ferric nitric treatment as the Massart method [30].

Conflict of Interests

The authors declare that there is no conflict of interests regarding the publication of this paper.

Acknowledgment

Financial support for this work was provided by the National Science Foundation of China (Grant no. 11274527, 5375039).