1 Introduction

The perovskite-type oxides, which have a general formula of ABO3, have been the focus of extensive research due to their outstanding properties ranging from high-temperature superconductivity to colossal magnetoresistance [1]. Among these oxides, strontium titanate (STO) has exhibited several promising features for application of this material in new electronic devices [2–9]. Its innumerable applications in many fields are due to its large nonlinear optical properties, extremely high and electric-field tuneable dielectric constant, energy storage ability, high thermal stability and photocatalytic properties [10–12].

However, the properties of STO are dependent on its chemical composition as well as its structure, shape and size [13, 14]. Other important characteristics of these materials include the partial substitution of the cations at both the A and B sites of ABO3 and the ability to maintain the stability of unusual mixed oxidation states in the crystal structure. To obtain a high dielectric constant and semiconducting behaviour of devices based on STO, it is necessary to modify or tailor the crystal structure. This modification is typically accomplished by chemical substitution with transition metals and the promotion of defects (i.e. vacancy in the structure) [15, 16].

Fe-doped strontium titanate, SrTi(1−x)O3:Fe x , is a continuous solid-state solution between strontium ferrite (SrFeO3) and strontium titanate. Due to the difference in the oxidation states between the titanium (Ti4+) and iron atoms (Fe3+), oxygen vacancies are commonly observed in the Fe-doped strontium titanate compound, STO:Fe. In fact, the formation of oxygen vacancies maintains the charge balance [17–19]. This structural characteristic can be used to reduce the electrical resistance of the undoped STO and to engineer high-sensitivity oxygen sensors [20].

STO:Fe has been studied for its redox behaviour [15], photocatalytic activity [22] and nonlinear optical properties [18], as well as for use as a gas sensor material [21] and phase shifter in communication systems [23]. However, the STO:Fe compound is not well understood, and more research is needed.

To achieve a comprehensive understanding of the structure–property relationships in multiferroic materials, the study of phonon dynamics is crucial. For example, Raman scattering spectroscopy is an invaluable tool for studying the lattice dynamics of these systems due to its sensitivity to short-range distortions arising from microstructural defects.

Recently, a Raman study reported by Van Minh et al. [24] on STO:Fe indicated that a strong reduction in the broad structure characteristic of SrTiO3 in the 200–500 cm−1 region was observed as the iron concentration increased. In addition, the activation of a single peak at approximately 690 cm−1, which is absent in undoped STO, was observed due to doping with Fe+4 ions [25].

In this study, the preparation of STO and SrTi(1−x)O3:Fe x (x = 0, 1, 3 and 5 mol% Fe) using a sol–gel hydrothermal method is reported. Structural and surface chemistry analyses indicate that amorphisation occurs as the Fe content increases.

2 Experimental

2.1 Synthesis

SrTi(1−x)O3:Fe x powders with x = 0, 1, 3 and 5 mol% Fe were synthesised by a sol–gel hydrothermal process using titanium tetrachloride (TiCl4, 1 M, Aldrich), strontium chloride (SrCl2·6H2O, 99.999 %, Aldrich) and iron oxide (Fe2O3, Aldrich) as starting materials.

In a typical procedure, three different solutions were prepared as follows: solution (A) containing 7 mL of TiCl4 was diluted with 2.3 mL of 2 M HCl to form a predominantly yellow solution; solution (B) was prepared by dissolving 2.4 g of SrCl2·6H2O in 4 mL of deionised water; and solution (C) was prepared by dissolving Fe2O3 in 2 mL of HCl.

The Fe-doped STO precursor was prepared by the dropwise addition of (A) and (B) into (C). Under stirring and N2 bubbling, 13 mL of 6 M NaOH was added to the strontium iron titanium solution and a white homogeneous colloidal strontium iron titanium slurry was formed.

The mixed solution was transferred to a 100-mL Teflon-lined stainless steel reactor, which was sealed and heated at 180 °C for 24 h under an oxygen partial pressure of 35 bar. Finally, the autoclave was cooled to room temperature under environmental conditions. The as-synthesised white powder attached to the walls of the Teflon container was collected, centrifuged, washed with distilled water and ethanol to remove the remaining ions, and dried at 60 °C for 6 h under reduced pressure.

2.2 Characterisation

The surface chemistry of the STO:Fe samples was characterised by X-ray photoelectron spectroscopy (XPS; Physical Electronics system model 1257) using Al Kα emission. The binding energies and oxidation states were obtained from high-resolution scans. The energy scale was calibrated by assigning 284.8 eV to the C 1 s peak. X-ray diffraction (XRD) data were acquired using a Siemens Advanced D-8 diffractometer with CuKα radiation at 40 kV and 30 mA. The Raman spectra were recorded with a T64000 Raman spectrometer manufactured by HORIBA Jobin–Yvon using the single grating mode (2400 lines) with a spectral resolution better than 0.4 cm−1. The spectra were excited by a 2.5-mW laser with a wavelength of 514.5 nm. The temperature-dependent measurements were performed using an Oxford constant flow cryostat, which allowed for a variation of temperature from 20 to 300 K with ±1 K control. Transmission electron microscopy (TEM) studies were performed on a Jeol JEM-2011 operated at 200 kV in diffraction and phase-contrast modes.

3 Results

3.1 X-ray diffraction

Figure 1 shows the X-ray diffraction patterns of the STO and STO:Fe samples. These samples were not subjected to annealing. As shown in Fig. 1, the sharp and well-defined peaks indicated the high crystallinity of both the STO and Fe-doped STO samples at lower concentrations (1 mol%). The XRD Bragg reflection was assigned to the cubic perovskite structure of STO. In fact, all of the main peaks for the STO samples in Fig. 1 were stabilised in the cubic lattice (space group Pm3m), and the calculated lattice constants were consistent with the tabulated values (a = 3.912 Å; JCPDS card no. 73-0661). No diffraction peaks related to SrCO3 or iron oxide phases were observed, which indicates that the iron ions were homogeneously incorporated in the STO lattice.

Fig. 1
figure 1

XRD patterns of STO doped with different content of Fe: (a) 0, (b) 1, (c) 3 and (d) 5 mol%, respectively. The inset shows the 2θ region of 31° and 33°

Full size image

However, in comparison with pure STO, the relative peak intensity of the perovskite peaks decreased with increasing Fe concentrations, suggesting that the samples become less crystalline. The samples with higher Fe concentrations were primarily amorphous with a small degree of crystallisation.

The inset in Fig. 1 shows the XRD patterns of all of the samples in the 31°–33° 2θ range. A single-peak diffraction between 31° and 33° was observed corresponding to the (1 1 0) Bragg reflection. The effect of an increase in the Fe substitution on the STO structure was detected as a shift in the (1 1 0) reflection towards a higher diffraction angle. No double peaks were observed in the 45°–48° 2θ range, which is characteristic of the STO cubic lattice.

The XRD peaks shift to a higher diffraction angle with increasing Fe concentrations, which confirms the increasing substitutional doping of Fe in STO. The ionic radius of Fe3+ (0.585 Å) is smaller than that of Ti4+ (0.605 Å). Therefore, doping of Fe leads to decreased lattice parameters and interplanar spacing in STO. The Rietveld refinements of the XRD data were carried out using the perovskite structure (space group Pm3m) and JCPDS card no 91062-ICSD to measure the lattice parameter. The calculated values of the lattice parameter decreased from 3.916 Å (STO) to 3.899 Å (STO:Fe 3 mol%) due to smaller ionic radius of Fe compared with that of Ti. However, this continuous shift in the XRD spectrum disappears at STO:Fe 5 mol%. This behaviour may be due to the mixture of Fe3+ and Fe4+ atoms or the strong amorphisation of the structure [26].

3.2 XPS analysis

The STO:Fe samples were analysed by XPS to elucidate their surface chemical characteristics. The oxidation states of the Fe, Sr, Ti and O atoms are shown in Fig. 2. High-resolution core-level scans that were acquired from the XPS spectra for the Sr 3d, Ti 2p, Fe 3d and O 1s levels of the STO:Fe 5 mol% sample are shown in Fig. 2a–d. The binding energy positions of the Sr 3d-doublet lines at 132.7 and 134.4 eV are due to Sr 3d 5/2 and Sr 3d 3/2, respectively (Fig. 2a), which these values correspond to α-Sr in strontium [27]. Here, the α-component corresponds to Sr atoms in the ‘perovskite-like superconductor phase’ of the STO:Fe structure.

Fig. 2
figure 2

XPS spectra of the elements for 5 mol% Fe-doped STO sample. a Sr 3d, b Ti 2p, c O 1s and d Fe 3d

Full size image

The binding energies of Ti 2p 3/2 and Ti 2p 1/2 for the STO:Fe samples are 458.1 and 464.0 eV, as shown in Fig. 2b, and are close to the reported values of α-Ti atoms for Ti atoms in the perovskite structure of STO and correspond to Ti with an oxidation state of 4+ [27]. The binding energies of the weak peaks at 456.5 (2p 3/2) and 459.1 eV (2p 1/2) were due to Ti3+ ions. These ions correspond to a Ti2O3 (Ti3+) compound, which is related to a non-perovskite structure of the STO:Fe structure.

As shown in Fig. 2c, the binding energy positions of the O 1s spectra yield two peaks. The stronger peak is associated with O2− ions at 529.6 eV, which is related to oxygen in the perovskite structure of STO:Fe, indicating that the oxygen ions remain coordinated in TiO6 octahedra [27]. The peak at 531.3 eV was due to an intermediate oxidation state for oxygen [i.e. O −x (0 < x<2)] and may be related to carbonate compounds as well as to defect, such as oxygen vacancies and chemical adsorption [28]. The O−x oxidation state is due to oxygen in the non-perovskite structure of STO:Fe [27].

The Fe 2p XPS region, which contains a doublet of 2p 3/2 and 2p 1/2, is shown in Fig. 2d. The peaks are located at 710.2 eV (Fe 2p 3/2) and 723.5 eV (Fe 2p 1/2), which are similar to those reported in a recent study of Fe 2p corresponding to the presence of Fe3+ [29]. The second doublet was located at 713.3 eV (Fe 2p 3/2) and 727.2 eV (Fe 2p 1/2) and assigned to Fe4+ [30, 31]. Therefore, Fe3+ and Fe4+ ions substitute for Ti in the structure [32].

Bocquet et al. [30] investigated the electronic structure of the Fe4+ perovskite oxide SrFeO3 by XPS and concluded that the large increase in charge at the Fe site led to a chemical shift to a higher binding energy. In an XPS study, Ghaffari et al. reported that iron in the STO:Fe perovskite structure participated as a mixture of Fe3+ and Fe4+ (SrTi(1−x)[Fe3+, Fe4+](x)O(3−δ)). In an separate report, an increasing Fe content resulted in a significant increase in the amount of Fe3+ and Fe4+, resulting in the formation of oxygen vacancies [31].

The presence of the O−x state in the sample is consistent with the XRD results because an increasing Fe concentration in the samples results in poorer crystallinity (i.e. increased amorphisation), which may be due to different defects in the network, such as oxygen vacancies, and the chemical adsorption process.

3.3 Transmission electron microscopy

The morphology of the STO:Fe sample was investigated by low-magnification TEM and EDX, as shown in Fig. 3a–d. The Fe-doped ST nanoparticles are nearly spherical with an average diameter of 100, 200 and 600 nm at 1, 3 and 5 mol% STO:Fe, respectively. 1 mol% Fe results provide better definition of the nanoparticles with spherical shape and defined contours. In addition, these nanoparticles do not appear to agglomerate, as shown in Fig. 3a.

Fig. 3
figure 3

TEM micrographs and SAED patterns of STO doped with different content of Fe: a 1, b 3, c 5 mol% and d corresponding to EDX for the 5 mol% sample. The colour rings in each SAED patterns are for guides to the eye (Color figure online)

Full size image

With an increasing concentration of the dopant ion, the definition of the nanoparticles becomes worse and the nanoparticles exhibit irregular agglutination and become difficult to differentiate (Fig. 3b, c). The average diameter of samples increased due to the emergence of the amorphous phase at higher Fe concentrations.

The EDX result shown in Fig. 3d indicates that the 5 mol% sample only contains the Sr, Ti, O and Fe elements in the sample. The results from the TEM analysis are in agreement with those obtained from the XRD analysis, where at higher Fe concentrations, the amorphous phase increases results in less crystallinity. The insets in Fig. 3a, b, c show the SAED patterns for 1, 3 and 5 mol% STO:Fe, respectively. The sharp and decreasing intensity of peaks was observed when the doped Fe content are increased, which indicates that the crystallinity of the samples was lost as the iron content increased,

3.4 Raman spectroscopy

At room temperature, STO single crystals have an ideal cubic perovskite structure (space group: Pm3m). Due to the odd symmetry of all of the zone-centre optical phonons, no first-order Raman activity is expected based on group symmetry analysis. Therefore, the room temperature spectrum should be dominated by second-order scattering [33]. However, Raman scattering studies of STO have shown that first-order Raman scattering can actually be observed when the central symmetry is broken due to various factors, such as strain effects, the presence of impurities and defects and oxygen vacancies [34]. For the STO nanoparticles, the origin of symmetry breaking comes from frozen surface dipoles or from local tetragonal arrangement as well as the presence of nanoscopic polar domains due to the incorporated impurities [35].

The room temperature Raman spectra of the STO:Fe nanoparticles with a spectral resolution higher than 0.4 cm−1 are shown in Fig. 4a. By comparing these Raman spectra with the published data [24, 33–39], three first-order phonon modes were identified at approximately 171, 540 and 800 cm−1 corresponding to the TO2, TO4 and LO4 modes, respectively. The frequencies of the TO2 and TO4/LO4 modes have been determined by fitting with Fano and Lorentzian line shapes, respectively.

Fig. 4
figure 4

a Room temperature Raman spectra of Fe-doped STO nanoparticles showing peaks associated with first-order scattering processes involving TO2, TO4 and LO4 phonons modes, peaks marked with star correspond to the second-order phonons, and peaks marked with triangle correspond to traces of SrCO3. b Linewidth (red circles) and frequency (black circles) of TO4 phonon mode as a function of Fe content. The position and the linewidth were determined by fitting with Lorentzian line shape. The black and the red dotted lines are for guides to the eye (Color figure online)

Full size image

The pronounced Fano-like asymmetric line shape of the TO2 polar mode suggests that the activation of this first-order process is most likely due to the presence of defect-induced ferroelectric polar regions in the nanoparticles [35]. This Fano-like behaviour has been observed in other perovskite nanoparticles [35, 36] and thin films [34, 40–42].

The weak peaks marked with a star at 120 cm−1 and between 200 and 400 and 600–800 cm−1 were identified as second-order scattering involving two-phonon processes. The peaks marked with a black triangle at approximately 148 and 181 cm−1 correspond to trace amounts of SrCO3, which is typically not detected by XRD [36].

Figure 4a shows that with an increasing Fe content, the Raman scattering peaks exhibit a strong decrease in intensity and become broader. Similar results were reported by Van Minh et al. for STO:Fe nanoparticles with Fe concentrations higher than 10 % [24]. In addition, a strong decay in the Raman intensities and overall modification of the Raman spectra was observed as the Fe concentration increased. Van Minh et al. attributed this effect to disorder and compositional fluctuations induced by the random distribution of Ti and Fe atoms in the B-site in the ABO3 sublattices. This effect will destroy the translational invariance, leading to symmetry breaking and corresponding changes in the Raman spectrum.

The effect of the Fe concentration on the TO4 phonon modes is shown in Fig. 4b. A marked shift to lower frequencies and a broadening of the linewidth (full width at half maximum, FWHM) were observed as the Fe concentration increased. Based on the difference in the atomic weights of the Ti (47.88) and Fe (55.85) atoms, the lower frequency shift can be explained by the simple mass effect [phonon frequency ~(k/m)1/2, where k is the force constant and m the weight] and the weaker bonds due to the amorphisation with increasing the Fe content.

The Raman spectra and a close up of the TO4 mode for STO:Fe 1 mol% as a function of temperature from 20 to 300 K are shown in Fig. 5. The results in Figs. 5b and 6 indicate that as the temperature decreased, a strong blue shift in the frequency and sharpening of the FWHM of the TO4 mode were observed. These two related effects arise because an increase in the temperature (decrease) displaces the atoms from their equilibrium positions, resulting in an overall volume expansion or contraction of the lattice and a change in the interatomic forces due to the anharmonicity of the bonds. This change in the interatomic force is reflected in the Raman frequencies, which shift to lower frequencies as the temperature increases. Similarly, the linewidth becomes broader with increasing temperature due to the temperature dependence of the phonon lifetime. This temperature dependence of the Raman signal can be used as a fingerprint to probe the thermal properties of the material [43, 44].

Fig. 5
figure 5

a Temperature evolution of the Raman spectra of STO:Fe 1 mol%. The peaks with star correspond to the second-order processes. b Temperature dependence of the TO4 phonon mode. The Raman spectra are denoted by blue dots, and the Lorentzian fits are shown in red solid lines. All spectra are plotted in the same scale with a small offset for clarity (Color figure online)

Full size image
Fig. 6
figure 6

Temperature dependence of the position TO4 phonon frequency (black dots and dashed line) and linewidth (red dots and dashed line) of STO:Fe 1 % mol. Both parameters were determined by a Lorentzian fit (as shown by the red lines in Fig. 5b) (Color figure online)

Full size image

The temperature dependence of the Fano asymmetry parameter (q) of the TO2 mode is shown in Fig. 7. The Fano-like shape was fitted using:

$$I(\omega ) = \frac{{A(q + E(\omega ))^{2} }}{{1 + E(\omega )^{2} }} + C$$
(1)

where E(ω) is the reduced frequency given by

$$E(\omega ) = 2(\omega - \omega_{0} )/\varGamma$$
(2)

and I(ω) is the intensity, A is the amplitude, ω 0 is the phonon frequency, q is the Fano asymmetry parameter, Γ is the FWHM, and C is a constant that accounts for the background signal. Positive values of q indicate that the scattering intensity is greater on the high-energy side of the peak. In contrast, negative values of q indicate the interference with the electronic continuum [35]. The q parameter did not exhibit significant temperature dependence, confirming the results reported by Rabuffetti et al. [36] for STO nanocubes. However, this result is different from the results reported by Banerjee et al. [35], who reported an oscillating temperature dependence of q in similar nanostructures. This controversial behaviour remains unclear, and more research is needed.

Fig. 7
figure 7

Temperature dependence of the asymmetry parameter q for STO:Fe 1 % mol (blue dots), the values were determined by fitting the Fano line shape to Eq. (1). The blue dashed line is used as guide to the eye. Inset Fano line shape of the TO2 phonon mode of STO:Fe 1 % mol at 20 K (blue dots), the red line corresponds to the Fano profile fitted by using Eq. (1) (Color figure online)

Full size image

4 Conclusions

In summary, nanocrystalline SrTi(1−x)O3:Fe x powders with x = 0, 1, 3 and 5 mol% Fe were synthesised by a sol–gel hydrothermal process to study doping effects on the structure using X-ray photoelectron and Raman scattering spectroscopy.

The XRD and Raman results indicate that the Fe-doped STO nanoparticles have a cubic perovskite structure for all of the Fe-doping contents with no evidence of a secondary phase. However, XRD indicated that at higher Fe concentrations, the materials primarily exhibit an amorphous phase with a very small degree of crystallisation, which is in agreement with the TEM images. In fact, with an increasing percentage of dopant ions, the crystalline definition of the nanoparticles is lost and the nanoparticles exhibit agglutination. The XPS results indicated that Sr ions were in the 2+ oxidation state, and in the perovskite structure, Fe participated as a mixture of Fe3+ and Fe4+, which significantly increased the formation of oxygen vacancies as the Fe content increased.

The presence of a Fano-like asymmetric line shape in the TO2 phonon mode suggested that the activation of the first-order Raman active modes at room temperature may be due to the presence of a defect-induced ferroelectric polar region in the nanoparticles. The Fe content dependence of the TO4 phonon mode suggests that this mode can be used as a fingerprint of the Fe content in slightly doped STO nanoparticles. Similarly, the temperature dependence of the TO4 phonon mode of the 1 mol% Fe-doped sample also suggests that it could be used as an ideal fingerprint to probe the thermal properties of the sample.