Background

Interest in multiferroics has been recently revived, since coexistence and interactions of ferroelectric, ferromagnetic, and ferroelastic orderings in multiferroics[1–6] could be applied potentially to a range of novel multifunctional devices[6, 7]. As one of the special multiferroic materials, EuTiO3 was found that in the bulk exhibits a G-type antiferromagnetic ordering below 5.3 K[8, 9], and its epitaxial films transform into ferromagnetic under large enough lattice strain[10–13].

A variety of techniques are available to grow fine epitaxial perovskite films, such as pulsed laser deposition[11], molecular beam epitaxy[12], radio-frequency magnetron sputtering[14], and metal-organic chemical vapor deposition[15]. These methods share a common feature that high growth temperatures (>500°C) and costly equipments are usually necessary. In contrast, an attractive alternative technique for preparing epitaxial perovskite films is hydrothermal epitaxy[16–20], which allows direct deposition crystalline films using mild aqueous solutions at temperatures as low as 150°C[16, 18] and avoids the research dependence on the costly aforementioned epitaxial growth equipments. In consideration of the merits of the hydrothermal epitaxy, however, nothing is currently known about the hydrothermal growth of epitaxial EuTiO3 films and their properties.

In this paper, we report the hydrothermal epitaxy of EuTiO3 films on SrTiO3(001) substrate at 150°C and the properties of the films. We find that the as-grown epitaxial EuTiO3 films show an out-of-plane lattice shrinkage and room-temperature ferromagnetism. Postannealing at 1,000°C evidences that this lattice shrinkage relates to the instabilities of Eu oxidation state in the films.

Methods

The heteroepitaxial EuTiO3 films investigated were grown on SrTiO3(001) substrate by hydrothermal method. Prior to growth, a solution of KOH (10 M, 15 mL) was added into a suspension which was composed of TiO2 (0.2 g), Eu(NO3)3 · x H2O (1.0 g) and H2O (50 mL) with a subsequent constant stirring for 30 min. The resulting solution was then introduced into a 100-mL Teflon-lined stainless autoclave with a fill factor of 65%, where the SrTiO3(001) substrate was fixed inside. The autoclave was shifted to a preheated oven holding at 150°C. After 24 h of growth, the sample was removed from the autoclave, cleaned by deionized water, and then dried ready in the air for the subsequent measurements. The phase structure of the films was assessed by high-resolution X-ray diffractometry (HRXRD; Bede D1, Durham, UK). HRXRD longitudinal ω- 2θ scans were recorded with an analyzer composed of Ge channel-cut crystals, while a pole figure was taken in skew geometry and with open detector. To assess the morphology and microstructure of the films, the samples were cleaved into smaller pieces for investigation by scanning electron microscopy (SEM; Hitachi S-4800, Chiyoda-ku, Tokyo, Japan) and transmission electron microscopy (TEM; TecnaiTMG2F30, FEI, Hillsboro, OR, USA), the latter through the standard mechanical thinning and ion-milling processes. The elemental composition of the films was analyzed by X-ray photoelectron spectroscopy (XPS; Kratos AXIS UltraDLD, Manchester, UK). The absence of water or hydroxyl in the films was evidenced by Fourier transform infrared spectroscopy (FTIR; Nexus870, Nicolet, Madison, WI, USA). The magnetic properties of the as-grown and annealed samples were measured in a superconducting quantum interference device magnetometry (SQUID). All magnetization data presented here are corrected for the diamagnetic background of the substrate. Postannealing of the as-grown sample was carried out in an Ar ambient for 10 h at 1,000°C.

Results and discussion

Most remarkable is the peculiar morphology observed by SEM from which a sequential growth of the films is proposed. Figure1a,b,c,d,e displays the SEM images taken in the top view of one sample with surface graded from edge to center and the typical morphology of epitaxially overgrown and coalesced EuTiO3 films. Note that the surface area of the SrTiO3(001) substrate we used for growth is 5 × 5 mm2. We may indirectly visualize the growth evolution of the EuTiO3 films from the spacial morphological nonuniformity. As shown in Figure1a, the existence of side facets observed at the top of micro-crystals reveals an initial nucleation growth in cross-like shape. The nucleation then processes from cross-shaped into tetragonal and after that into cuboidal. Accompanying the coalescence of cuboid in the first layer, nucleation on the second layer starts and develops, as shown in Figure1b. Figure1c,d clearly reveals the coalescence process of the micro-crystals on the second layer. A crisscross consisting of dense crosses shown in Figure1c forms to coalesce the side facets of conjoined micro-crystals. Figure1d shows coalescence of the crisscross on top of layers. The complete coalescence of the crisscross results in a great smooth surface of the films shown in Figure1e. Interestingly, the crosses and the micron-sized tetragon develop regularly and orient highly, which reveals that the films are highly oriented and suggests a tetragonal structure of the film. This indication is evidenced by the following TEM and HRXRD results. Figure1f shows a cross-sectional SEM image taken on an arbitrary portion of the sample. A layer with a uniform thickness of about 600 nm is clearly observed.

Figure 1
figure 1

Top-view and side-view SEM images. Bird’s-eye view from the (a) edge, (b) near-edge, (c) middle-of-edge-and-center, (d) near-center, and (e) center of one sample surface. Note that the surface area of the SrTiO3(001) substrate is 5 × 5 mm2. (f) Cross-sectional SEM image taken in an arbitrary portion of the sample.

Full size image

To directly investigate this peculiar epitaxial growth of the EuTiO3/SrTiO3(001) structure, the interface of the structure was examined by TEM. Figure2a shows a cross-sectional high-resolution transmission electron micrograph of the EuTiO3/SrTiO3(001) interface along the SrTiO3[ 1 ̄ 00] zone axis. The lattice planes of the EuTiO3 film are clearly resolved and are found to be well ordered. Consecutive lattice planes at the interface between the film and the substrate is clear, which precisely and directly evidences a well epitaxial relationship between the deposited film and the substrate, although there might be few dislocations in the interface to release the internal stress due to slight lattice mismatch. The insets in Figure2a show the high-resolution micrographs of the EuTiO3 films and SrTiO3 substrate taken in focus, respectively. Selected area electron diffraction (SAED) patterns of the films and substrate were also taken and are shown in Figure2b,c, respectively. Both SAED patterns depict the identical crystallographic structure and indicate their epitaxial orientations with small lattice misfit and a highly oriented tetragonal structure of the film, which leads to a tetragonal surface morphology generally presented in nucleation developing stage, as shown in the aforementioned SEM images.

Figure 2
figure 2

High-resolution transmission electron micrographs and selected area electron diffraction patterns. (a) Cross-sectional high-resolution transmission electron micrograph of the EuTiO3/SrTiO3(001) interface along the SrTiO3[ 1 ̄ 00] zone axis. The insets show the high-resolution micrographs of the EuTiO3 films and SrTiO3 substrate taken in focus, respectively. Selected area electron diffraction patterns of (b) EuTiO3 and (c) SrTiO3, respectively.

Full size image

To investigate the crystallographic uniformity of this epitaxial growth, the EuTiO3/SrTiO3(001) structure was assessed by HRXRD. Both EuTiO3 and SrTiO3 were reported to have the cubic perovskite crystal structure at room temperature and have a lattice constant of 0.3905 nm[21], indicating zero lattice mismatch between EuTiO3 and SrTiO3. Figure3a shows symmetric HRXRD longitudinal ω- 2θ scans taken within a 2θ range from 10° to 110° for the as-grown and postannealed samples. Apart from the (00l) (l = 1, 2, 3, and 4) reflections of SrTiO3, the (00l) reflections of EuTiO3 for the as-grown sample can be identified and no reflections pertinent to a secondary phase can be found, indicating that the epitaxial growth of EuTiO3 is oriented along the c-axis. The out-of-plane lattice constant of the as-grown films calculated from the (001), (002), and (004) peaks are 0.3789, 0.3821, and 0.3831 nm, respectively. They are much smaller than the reported value of 0.3905 nm for bulk EuTiO3[22, 23] and show an out-of-plane lattice shrinkage of 2.9%, 2.1%, and 1.9%, respectively. The average shrinkage is 2.3%, which means that the out-of-plane lattice shrinks by about 2.3% along the c-axis. The in-plane epitaxial relationship between the films and the substrate was measured by azimuthal scans in skew geometry. Figure3b shows an XRD {211} pole figure of the as-grown sample measured by setting 2θ = 57.92°. The reflections from EuTiO3 and SrTiO3 overlap in every streak measured by an azimuthal and sample-tilting angular scans. The in-plane fourfold symmetry of the EuTiO3/SrTiO3 orientation relationship is revealed by the four streaks in the pole figure, which shows an in-plane orientation relationship of EuTiO3〈100〉∥SrTiO3〈100〉. Evidently, the pole figure provides the same qualitative information as the SAED patterns, in that it reveals a fourfold symmetry and an excellent in-plane alignment of the EuTiO3 films and SrTiO3 substrate. Postannealing of the as-grown sample was carried out in an Ar ambient for 10 h at 1,000°C in order to compare the result with the report where the epitaxial EuTiO3 films were prepared by pulsed laser deposition[11]. Upon postannealing, symmetric HRXRD longitudinal ω- 2θ scans display that the EuTiO3 peaks shift toward lower angles and are superimposed on the SrTiO3 peaks without yielding any impurity phases, as shown in Figure3a. It means that the out-of-plane lattice shrinkage of the as-grown films was relaxed by postannealing, possibly corresponding to the changes of oxidation state in Eu surroundings. It is reported that valence instabilities are an interesting and general phenomenon for rare earth ions in their compounds, for example, mixed valences, valence fluctuations, and surface valence transitions[24–27]. Our present work provides an opportunity to study further valence instabilities of Eu in EuTiO3 and their resultant properties.

Figure 3
figure 3

HRXRD longitudinal scans and XRD pole figure. (a) Symmetric HRXRD longitudinal ω- 2θ scans of the as-grown and postannealed EuTiO3 films on SrTiO3(001) substrate. (b) XRD {211} pole figure of the as-grown sample.

Full size image

The elemental composition of the films was then analyzed by XPS, which was taken within a binding energy scan range from 0 to 1,300 eV. No signals pertinent to K+ cation can be found, indicating that the films have no incorporation of K from the solvent. The Eu 3d and Ti 2p core-level XPS spectra of the as-grown sample are shown in Figure4a,b, respectively. The results clearly exhibit that the as-grown sample consists of mixed Eu2+, Eu3+, and Ti4+ cations, in agreement with the peak positions of the cations shown in the XPS spectra from other studies[25–29]. The presence of Eu3+ indicates the necessity of anion excess in the as-grown films for charge balance and may affect the crystal lattice and magnetic properties of the films, which will be discussed later on. The Eu 3d core-level XPS spectra of the annealed sample are shown in Figure4a, which reveals a reduction of Eu3+ quantity. The Ti 2p core-level XPS spectra of the annealed sample not only are dominated by the Ti4+ contribution but also plausibly exhibit the Ti3+ shoulders, as shown in Figure4b. These results reflect a necessity to lose part of the ionic charge during the annealing process for charge compensation. Further investigations are necessary to understand the chemical details of the films and annealing process.

Figure 4
figure 4

XPS spectra of the as-grown and postannealed samples. (a) A comparison of the Eu 3d core-level XPS spectra between the as-grown and postannealed samples. (b) Ti 2p core-level XPS spectra of the as-grown and postannealed samples.

Full size image

It is important to realize the possible inclusion of water or hydroxyl in the as-grown films. Such issues have been reported in various perovskites prepared hydrothermally[30–32]. These impurities can contribute to charge balance in the as-prepared perovskites and be removed by annealing to produce defects, which when coupled with a metal can account for charge compensation[30, 31]. Thus, our films were studied by FTIR. Figure5 shows the FTIR spectra of the as-grown and postannealed samples for a comparison. No peaks pertinent to water or hydroxyl can be seen and resolved from the spectra; hence, the presence of water or hydroxyl and their resultant charge balance/compensation mechanisms are excluded in our films. The charge balance (compensation) in our as-grown (annealed) films is possibly made by oxygen excess (loss).

Figure 5
figure 5

FTIR spectra of the as-grown and postannealed samples. The peak at 2,360.39 cm-1 is due to contributions from CO2 present in air.

Full size image

Finally, we are interested in the magnetic properties of these films. The in-plane hysteresis loops for the as-grown films shown in Figure6a were measured by SQUID with the magnetic field (H) parallel to the EuTiO3[100] direction at 300 K. The as-grown EuTiO3 films exhibit a ferromagnetic-like behavior. To quantitatively show the impact of the postannealing on its magnetic properties, the same piece of the sample after annealing was measured by SQUID to avoid errors from sample volume measurements. A striking decrease of MS and a negligible ferromagnetic behavior for the annealed films are found and shown in Figure6a. These results indicate that the oxidation states of Eu in the as-grown films provides magnetic moments and contributes to the magnetization. In order to get more information about the magnetism in these films, the in-plane hysteresis loops for the as-grown and annealed films were measured at 2 K. It can be seen from the loops shown in Figure6a that both films exhibit a ferromagnetic behavior and an increase of MS at 2 K. Surprisingly, the MS value of the annealed films is much larger than that of the as-grown films at 2 K. It means that Eu2+ in the annealed films has magnetic contribution to magnetization at low temperature and implies that Eu3+ ion probably possesses less magnetic moment than Eu2+. Temperature dependence of the magnetization curves shown in Figure6b compares the magnetic properties between the as-grown and annealed films in more detail. It clearly shows that the annealed films convert to ferromagnetic behavior as external magnetic field applied to the films is raised, implying the presence of a ferromagnetic phase transition in the annealed films at low temperature. Evidently, a thermal treatment of the as-grown films demonstrates obvious impact on their magnetic properties. Combining this result with that from XPS investigations, we can obtain that the valence instabilities of Eu in EuTiO3 films could result in the material being ferromagnetic at room temperature, which may extend the range and potential of this material for application.

Figure 6
figure 6

Hysteresis loops and temperature dependence of magnetization. (a) Hysteresis loops obtained at 300 and 2 K for the as-grown and annealed films with external field applied parallel to EuTiO3[100] direction. The inset magnifies the low magnetic field range. (b) Temperature dependence of the magnetization curves of the as-grown and annealed films at 1,000 Oe and 20 kOe external fields applied parallel to EuTiO3[100] direction.

Full size image

Conclusions

To summarize and conclude, using a hydrothermal method, EuTiO3 films with high crystalline quality were successfully grown on SrTiO3(001) substrate at a temperature of 150°C. The films show highly oriented and regularly shaped morphologies with graded spacial distribution, which reflects a sequential growth process of the films. Using this growth technique, EuTiO3 films grown on SrTiO3 substrate exhibit an out-of-plane lattice shrinkage, which could be relaxed by postannealing. Valence instabilities of Eu were found in the sample and result in the EuTiO3 films being ferromagnetic at room temperature, which provides an opportunity to study further their properties and potential applications.