1 Introduction

Ni and Fe have important applications as catalysts for Fischer–Tropsch (FT) conversion of synthesis gases to liquids and as carriers for chemical looping combustion systems. In the FT application, Ni and Fe catalysts have been used in both supported and powdered forms. Two drawbacks of powdered catalysts are low surface area and the lack of uniformity of individual particles. To overcome these challenges, the past decade has seen increased interest in synthesis of monodispersed nanoparticles for applications in catalysis. Initial efforts in synthesis of nanoparticles were aimed at control of nanoparticle size and shape [1–3]. Recently, efforts have turned towards synthesis of alloy nanoparticles [4–6]. Alloy nanoparticles can be more effective catalysts or have better magnetic properties than pure component nanoparticles [1, 7, 8]. Performance of alloy nanoparticles in applications such as FT depends on control of both composition and particle size [9–13]. However, one disadvantage of subjecting small metallic nanoparticles to the high temperatures required in catalytic reactors is that they tend to sinter or aggregate into larger particles. To prevent aggregation, there has been significant interest and success in the preparation of metal nanoparticles coated with thin shells of silica and other ceramic materials [8, 14–17]. A core–shell multi-component structure can be catalytically active if the ceramic shell is sufficiently porous to allow transport of gases to and from the surface of the metal alloy core [18]. At the same time, the shell prevents or retards sintering of the metallic alloy cores, thus allowing their use over an extended range of temperatures. The synthesis of such core–shell nanoparticles structures requires a fundamental understanding of the factors that influence their growth, structure, morphology and porosity.

In this work, we report our one-pot preparation of bimetallic FeNi alloy nanoparticles with porous SiO2 shells (NiFe2O4@SiO2). The preparation delivers nanoparticles with well-controlled shapes, sizes and compositions. We demonstrate the porosity of the silica shell by N2 sorption methods, and use X-ray diffraction (XRD) and temperature programmed techniques to compare the nanoparticles’ stability in extreme environments to those of commercially available catalytic materials.

2 Experimental

Ni(II) nitrate hexahydrate (>99.99 %), Ni (II) chloride, Brij 56 (polyethylene glycol hexadecyl ether, average M n  ~ 683), Fe(II) chloride, ammonium hydroxide solution (33 % solution), propanol, tetraethyl orthosilicate (TEOS, 99 %), and cylcohexene were purchased from Sigma Aldrich and were used without purification. NiFe2O4@SiO2 nanoparticles were synthesized using a one-pot technique. Parameters used for synthesis of spherical and rod-shaped NiFe2O4@SiO2 are summarized in Table 1. In a typical preparation, Brij 56 surfactant was dissolved in 100 mL of cyclohexene and stirred in a three-neck glass flask for 30 min at ~342 K, forming a clear solution. Next, 4 mL of an aqueous solution of Fe(II) chloride solution was added; the mixture turned yellow. After stirring for 10 min, 4 mL of an aqueous solution of Ni(II) nitrate hexahydrate was added to the reaction mixture. Next, 30 % aqueous NH4OH solution was added; the solution turned blue, then grey, then black. The reaction mixture was heated for another 10 min before drop-wise addition of a solution of TEOS in methanol. The final reaction mixture was heated and stirred for another 15 min–1 h. Product NiFe2O4@SiO2 nanoparticles were centrifuged and washed with polar solvent to remove excess surfactant.

Table 1 Parameters for preparation of NiFe2O4@SiO2
Full size table

Relative solution amounts and concentrations were varied, as shown in Table 1, to change the geometry of the product nanoparticles. The main factor that determines nanoparticle shape is the concentration of surfactant (Brij 56). The surfactant concentration used to prepare nanorods is twice that used for nanospheres [19]. Except for the TEM images, all characterizations reported in this work are for nanosphere samples.

2.1 X-ray Diffraction (XRD) and X-ray Fluorescence (XRF)

XRD patterns were acquired on a Panalytical X’Pert Pro X-ray diffractometer. A Cu radiation source was used with a polycapillary lens and a Ni filter on the incident beam. Nanoparticle samples were prepared for XRD by drying overnight at 373 K in a vacuum oven, followed by annealing in air at temperatures from 773 to 1,173 K in a box-furnace. XRF measurements were performed on a Panalytical MiniPal 4 XRF spectrometer. Annealed samples were placed in a plastic cup with a bottom X-ray window made of polyvinylidene chloride (Saran Wrap). A rhodium target X-ray source was used with 15 kV and 0.28 mA power settings. Scans were run for 30 s in air with no filter.

2.2 Transmission Electron Microscopy (TEM)

TEM studies were conducted using a JEOL JEM-2000 EX II microscope operating at 200 keV with a Gatan Camera. The high-resolution TEM studies were performed using a Technai F20 FEG/HRTEM/STEM with a Gatan imaging filter and an energy dispersive X-ray spectroscopy system. TEM samples were prepared by evaporation of 20 μL of the colloidal NiFe2O4@SiO2 nanoparticle solution in propanol onto a carbon-coated, 200 mesh, Cu TEM grid.

2.3 Characterization of Shell Porosity by N2 Sorption

The pore size distribution and surface area of the nanoparticles were measured by N2 sorption methods using a Quantachrome Nova 2200e instrument. Surface areas were calculated using the BET method; pore size distributions were determined from desorption isotherms. Nanoparticle samples were prepared for characterization by drying overnight at 373 K in a vacuum oven, followed by annealing in air at 773 K for 1 h in a box furnace.

2.4 Temperature Programmed Reduction (TPR) and Temperature Programmed Desorption (TPD) Measurements

TPR and H2-TPD experiments were performed using a Micromeritics AutoChem HP 2950 instrument. Before characterization, samples of NiFe2O4@SiO2 nanoparticles and an uncoated reference were annealed in air for 1 h in a box furnace at temperatures between 773 and 1,173 K. Approximately 0.065 g of NiFe2O4@SiO2, or 0.013 g of uncoated NiFe2O4 sample, was loaded into the instrument’s stainless steel reactor. Once in the reactor, the sample was dried at 623 K in flowing He for 30 min. Then, after cooling to 323 K, the gas flow was changed to 10 % H2 in Ar and the sample was heated to 773 K at 10 K/min. The sample was held at 773 K for 60 min. During both heating and the hold at 773 K, hydrogen consumption by the sample was measured by comparing the thermal conductivities of the gas streams on the inlet and outlet sides of the reactor. The sample was then cooled to 193 K while still under H2/Ar flow. After 30 min at 193 K in H2/Ar, the flow was changed to pure Ar. After 30 min in Ar at 193 K, the sample was heated at 10 K/min to 773 K. The sample was held at 773 K for 30 min. Desorption of hydrogen from the sample during both heating and the hold at 773 K was monitored by mass spectrometry (signal at m/e = 2). A NiO/SiO2–Al2O3 catalyst (66 wt% Ni) and Fe2O3 powder (5 μ, 99 %) were obtained from Alfa-Aesar for use as single metal references for the TPR and TPD experiments.

3 Results and Discussion

The size and morphology of the NiFe2O4@SiO2 nanoparticles were examined using TEM. Figure 1 shows typical examples of bright field TEM images of NiFe2O4@SiO2 nanoparticles synthesized using the one-pot synthesis. By adjusting preparation parameters, either spherical (Fig. 1a) or rod-shaped (Fig. 1b) NiFe2O4@SiO2 nanoparticles can be prepared. Within the silica shell, the NiFe2O4 nanoparticle cores are either irregular or rod-shaped.

Fig. 1
figure 1

Bright field TEM images of NiFe2O4@SiO2 nanoparticles packed to form nanospheres (a) or nanorods (b). The silica shell is 10–20 nm thick

Full size image

NiFe2O4@SiO2 nanosphere samples were annealed in air before structural characterization by XRD. Figure 2a shows the XRD pattern of NiFe2O4@SiO2 nanoparticles annealed at 773 K for 1 h. Diffraction peaks are observed at 2θ = 30.29, 35.69, 37.31, 43.36, 53.80, 57.35, 62.91 and 74.6°. This pattern is characteristic of NiFe2O4, suggesting formation of nickel ferrite nanoparticles with inverse spinel structure [20, 21]. The sharpness of the diffraction features indicates that the sample is highly crystalline. The atomic ratio of Fe:Ni in the sample, measured by XRF, is 1.3:1; this ratio is consistent with the ratio of the precursors used in preparation of the sample. Fe:Ni for the spinel is expected to be 2:1, thus it is possible that this sample contains X-ray amorphous Ni oxides. Figure 2b shows the diffraction pattern of the same sample after annealing at 1,173 K. New features appear at 24.00, 33.15, 40.85, and 65.47°, indicating the formation of an independent hematite (Fe2O3) phase that coexists with the spinel [22, 23].

Fig. 2
figure 2

XRD patterns of NiFe2O4@SiO2 nanoparticles annealed at 773 K (a) and 1,173 K (b). Asterisk indicates diffraction features of NiFe2O4. Plus indicates diffraction features of hematite (Fe2O3)

Full size image

The diffraction patterns of silica coated NiFe2O4 nanoparticles were compared to those of spherical NiFe2O4 nanoparticles without silica coating. Figure 3 shows the XRD pattern of NiFe2O4 nanoparticles without silica coatings annealed at 973 K in air for 1 h. The pattern is characteristic of nickel ferrite, with no evidence of hematite. The ratio of Fe:Ni in the uncoated sample, as measured by XRF, is 0.8:1. As was the case for the SiO2 coated nanoparticles, this ratio suggests Ni content in excess of that expected for a stoichiometric spinel; X-ray amorphous Ni oxides may also be present in this sample. Interestingly, the relative intensities of the uncoated NiFe2O4 diffraction features differ from those of the NiFe2O4@SiO2 (Fig. 2). For example, the (400) reflection dominates the pattern of uncoated NiFe2O4, but it is a minor feature in the pattern of the SiO2-coated material. This difference suggests that preparation parameters—including the presence of the coating—can be leveraged to direct the direction of nanoparticle crystal growth.

Fig. 3
figure 3

XRD pattern of NiFe2O4 nanoparticles without silica coating annealed at 973 K

Full size image

To be useful for catalysis, the NiFe2O4 cores must be accessible by reactant molecules, thus the silica coating must be porous. The porosity of the silica coating was characterized by N2 sorption methods. Figure 4 shows the pore size distribution of NiFe2O4@SiO2 nanoparticles annealed at 773 K. The pore size distribution displays a sharp maximum at ~3.7 nm; this is the characteristic diameter of the pores in the silica shells [24]. The surface area of this sample, determined by BET analysis, is 72 m2/g. These results are comparable to those we previously reported for SiO2 coatings on Ni nanorods [25].

Fig. 4
figure 4

Pore size distribution, measured by N2 desorption, of NiFe2O4@SiO2 nanoparticles annealed at 773 K

Full size image

Figure 5 compares the TPR spectra of uncoated NiFe2O4 nanospheres annealed in air at 773 K and NiFe2O4@SiO2 nanospheres annealed at 773, 973, and 1,173 K. TPR spectra for commercially available NiO/SiO2–Al2O3 and Fe2O3, both annealed at 773 K, are included as references. The NiFe2O4 samples generally display two features; in the uncoated sample, the first feature appears at 648 K and the second as a shoulder between ~680 and 770 K. TPR features in the range 600–700 K have been associated with reduction of NiO to Ni and Fe2O3 to Fe3O4 [26–31]. The single component Fe2O3 sample shows a small feature at 640 K, likely due to reduction of Fe2O3–Fe3O4; a large feature at higher temperature indicates further reduction of Fe [30, 31]. The NiO-only sample presents a pair of peaks, which are generally explained by reduction of NiO that is stabilized to different extents by its support [26, 32].

Fig. 5
figure 5

Temperature programmed reduction of NiFe2O4@SiO2 (annealed at different temperatures) and NiFe2O4 nanoparticles; and NiO and Fe2O3 reference samples

Full size image

The reduction profile of the uncoated NiFe2O4 sample more closely resembles that of the NiO reference than that of the Fe2O3 reference. The predominant low-temperature feature of uncoated NiFe2O4 likely corresponds to reduction to Ni + Fe2O3. As shown in Fig. 6, the XRD pattern of this sample after TPR/D confirms formation of Fe2O3. The appearance of this reduction feature at a higher temperature than in the Ni reference suggests that NiFe2O4 is more stable than NiO. Coating the NiFe2O4 with silica further stabilizes the oxide and delays its reduction to higher temperatures. As NiFe2O4@SiO2 is annealed at higher temperatures, both reduction features move to higher temperatures. When annealed at 1,173 K, relatively little reduction of NiFe2O4@SiO2 is observed; its reduction profile now resembles that of the Fe2O3 reference. This result is consistent with the XRD pattern of SiO2@NiFe2O4 annealed at 1,173 K, which displayed features of both NiFe2O4 and Fe2O3 (Fig. 3). Fe2O3 formed by aggressive oxidation of NiFe2O4@SiO2 may segregate to the surface of the nanoparticle and delay reduction of underlying Ni-containing species.

Fig. 6
figure 6

Post TPR/D XRD pattern of NiFe2O4 nanoparticles without silica. Asterisk indicates diffraction features of NiFe2O4. Plus indicates diffraction features of hematite (Fe2O3)

Full size image

H2-TPD spectra for the same six samples are compared in Fig. 7. The Fe2O3 reference sample adsorbs little hydrogen in this experiment. The NiO reference, on the other hand, shows three features, which have been ascribed to H in three different surface regions of Ni: sub-surface (α), “second-layer” (β), and top-surface (γ) [33]. NiFe2O4@SiO2 annealed at 773 and 973 K display the α and γ desorption features; β appears as a separate feature in the 973 K sample. Clearly, H2 adsorbs onto Ni sites in these samples. However, when NiFe2O4@SiO2 is annealed at 1,173 K, only a small α feature remains; the γ peak, characteristic of top-surface Ni, is no longer present. The uncoated NiFe2O4 displays none of the features characteristic of desorption of hydrogen from Ni; instead, it bears a strong resemblance to the TPD spectrum of the Fe2O3 reference.

Fig. 7
figure 7

H2 temperature programmed desorption spectra of NiFe2O4@SiO2 (annealed at different temperatures) and NiFe2O4 nanoparticles; and NiO and Fe2O3 reference samples. Samples were exposed to H2 at 193 K

Full size image

Our results suggest that the SiO2 shell stabilizes Ni at the nanoparticle surface; after reduction at 773 K, NiFe2O4@SiO2 annealed at temperatures through 973 K show clear evidence of surface Ni in their H2-TPD spectra. Oxidation at 1,173 K induces conversion to Fe2O3, which, as noted earlier, may segregate to the nanoparticle surface and obscure Ni atoms. Without the shell, reduction at 773 K can draw Fe to the surface, and eliminate surface Ni sites.

4 Conclusion

We demonstrate a one-pot synthesis for NiFe2O4@SiO2 nanoparticles. Nitrogen desorption experiments show that the silica coatings are porous, with pore diameter ~3.7 nm. The coating stabilizes the NiFe2O4 cores against reduction to hematite (Fe2O3) and preserves surface Ni sites.