Epitaxial growth and thermostability of cubic and hexagonal SrMnO{}_{\text{3}} films on SrTiO{}_{\text{3}}(111)

Abstract

The growth of SrMnO₃ films on SrTiO₃(111) substrates by pulsed laser deposition was studied and found to produce cubic and hexagonal (4H) structures in the SrMnO₃ films. By adjusting the substrate temperature and oxygen pressure, the stability of the two phases was fine-tuned, resulting in the growth of cubic-SrMnO₃(111) or 4H-SrMnO₃(0001) film, with the 4H phase being the more stable at room temperature and ambient pressure in the bulk form. The growth temperature of the cubic phase was also further lowered relative to the bulk thermodynamics by strain at the heterointerface, and once obtained, it was stable at temperatures of up to 800 ^∘C.

Graphical abstract

Keywords

oxide film / pulsed laser deposition / heteroepitaxy / SrMnO₃

1 Introduction

Thin films of [111]-oriented perovskite oxides ABO₃ (where A and B are metal cations) have attracted extensive attention of late because of their peculiar physical properties [1–3], such as the fact that the six-fold symmetry of their honeycomb structure can result in novel topology states at their surfaces or interfaces [4, 5]. In addition, [111]-oriented superlattices allow for the design and construction of long-range ordering of cation dopants. One example of this is the double perovskite La₂FeCrO₆ structure with B site cations (Fe and Cr) distributed alternatively along the ⟨001⟩ direction that has been obtained by growing a LaFeO₃/LaCrO₃(111) superlattice [6, 7], which was found to have magnetic properties distinctly different from a disordered solid solution of LaFeO₃ and LaCrO₃. An exchange bias has also been unexpectedly discovered [8] in the [111]-oriented superlattice of paramagnetic LaNiO₃ and ferromagnetic LaMnO₃. Meanwhile, in the perovskite AMnO₃ manganite, A-site doping has been proven effective in tuning the properties of the material by inducing a rich phase diagram that includes colossal magnetoresistance and ferromagnetism [9, 10]. Normally, these cation dopants are randomly distributed and produce a localization effect that reduces the metallicity of electron transport in low-dimensional manganite structures; this is one of the possible reasons for the so-called “deadlayer” problem in La_2/3Sr_1/3MnO₃ ultrathin films [11]. Thus, a [111]-oriented LaMnO₃/SrMnO₃ superlattice in which La and Sr are distributed alternatively with long-range ordering along all ⟨100⟩ directions offers a realistic potential to solve this problem by suppressing the localization effect originating from A-site disorder.

It is quite challenging to grow perovskite (111) thin films due to their strong surface polarity and the multiple competing phases in the lattice structure. For example, SrMnO₃ is commonly synthesized in either cubic (C) or hexagonal (4H) structures (Fig. 1), both of which are antiferromagnetic insulators with respective Neel temperatures (T_{\mathrm{N}}) of 240 and 280 K [12–14]. The 4H-SrMnO₃ structure is made up of alternating face-sharing and corner-sharing MnO₆ molecules along the [0001] orientation [Fig. 1(b)], which contribute to long-range antiferromagnetic ordering [13–15]. The lattice structure of C-SrMnO₃, on the other hand, is quite simple [Fig. 1(a)], though first-principles calculations have shown that strain could produce a multiferroic phase due to its strong spin–phonon coupling [16]. It is important to control the phase of SrMnO₃ during its growth, as otherwise, its properties can be quite varied. The stability of the perovskite structure can generally be described by the tolerance factor: t=\frac{r_{\mathrm{A}}+r_{\mathrm{O}}}{\sqrt{2}(r_{\mathrm{B}}+r_{\mathrm{O}})}, where r represents the ionic radii of A and B cations, and O atoms. For SrMnO₃ the value of t is 1.04, meaning that the 4H phase is normally stable at room temperature and ambient pressure [17], whereas the cubic phase can only be obtained in a metastable state [13, 15]. In 1970, Negas and Roth [18] presented a phase diagram for bulk SrMnO_3-x that demonstrated 4H-SrMnO₃ is stable in air up to about 1035 ^∘C. Higher temperatures were shown to induce the formation of oxygen vacancies (V_{\mathrm{O}}\mathrm{s}), resulting in a deformation of the 4H phase that causes the crystal to transform into a perovskite-like structure at ¿1400 ^∘C. By annealing in air at 300 ^∘C for several hours, this perovskite-like SrMnO₃ was oxidized to form the cubic phase [13–15, 19].

For thin film growth, 1400 ^∘C is an impractically high temperature and is likely to deteriorate the desired [111]-oriented superlattice by causing serious interdiffusion across the interfaces. With a perovskite (001) substrate, the growth of a 4H-SrMnO₃ film can be suppressed by the symmetry mismatch, allowing the cubic phase to be obtained by pulsed laser deposition (PLD) at a reduced temperature of 800 ^∘C [20]. Conversely, with a perovskite (111) surface, the competition between 4H (0001) and cubic (111) cannot be tuned by symmetry alone. This paper therefore looks at introducing an appropriate strain to the heteroepitaxial growth of SrMnO₃ films by using SrTiO₃(111) as the substrate. These single-crystalline SrMnO₃ films are grown by PLD as either a 4H or cubic phase depending on the growth parameters used. In this way, a C-SrMnO₃(111) film can be obtained at a temperature as low as 875 ^∘C, which represents a significant reduction when compared to bulk synthesis. This interface epitaxial stabilization also ensures that a C-SrMnO₃ film grown on SrTiO₃(111) has a good thermostability.

2 Experiments

Films of SrMnO₃ were deposited on Nb-doped (0.7 wt.%) SrTiO₃(111) single-crystalline substrates (3×12×0.5 mm{}_{}{}^3) by ultra-high vacuum PLD using a 248 nm KrF excimer laser (CompexPro201F). Prior to deposition, the substrates were treated by Ar⁺ ion sputtering (10 μA/1 KV) for 10 minutes followed by annealing at 900 ^∘C under 3×10^-6 mbar of oxygen pressure for 1 hour. This resulted in a 4×4 surface reconstruction with a slight Sr accumulation, which effectively stabilizes the SrTiO₃(111) polar surface [21]. The samples were then heated by passing a direct current through the substrates, during which time the temperature was monitored by an infrared pyrometer. All depositions were carried out using a laser repetition rate of 2 Hz and an energy density of 1.5 J/cm². To control the phase of the SrMnO₃ film, the substrate temperature (T_{\mathrm{S}}) and oxygen pressure (P_{\mathrm{O}}) were adjusted to within 850–1000 ^∘C and 2×10^-6–5×10^-3 mbar, respectively. The film thickness was \sim60 nm for all samples, based on which it is considered reasonable to exclude any effect of substrate surface reconstruction on the film growth. The samples were post-annealed in a tube furnace at 300 ^∘C under an oxygen flow of \sim1 bar to minimize the oxygen vacancies formed. The crystalline structure of the film was characterized by X-ray diffraction (XRD) using an UltimaIV X-ray diffractometer with Cu K radiation (\lambda =1.5405 Å). The DC magnetic susceptibilities were measured between 400 and 10 K by a commercial SQUID-vibrating sample magnetometer (Model S-VSM of Quantum Design) with zero-field cooling.

3 Results and discussion

The structure of the SrMnO₃ films produced on SrTiO₃(111) substrates was controlled to be either a 4H or cubic phase by changing the T_{\mathrm{S}} and P_{\mathrm{O}} during growth. Figure 2 shows the XRD spectra of samples deposited at 920 ^∘C/3.3×10^-5 mbar and 950 ^∘C/4.0×10^-3 mbar, in which the their different phases can be clearly distinguished. That is, if the diffraction peaks of the substrate are ignored, then one sample shows sharp peaks at 2\theta ={41.06}^{\circ } and 89.03^∘ that correspond to the (111) and (222) lattice plane interspacings of C-SrMnO₃ (JCPDS Card No. 231413) [22] respectively, while the other shows peaks at 2\theta ={19.57}^{\circ } and 61.37^∘ that can be attributed to the (0002) and (0006) interspacings of the 4H phase (JCPDS Card No. 841612) [13], respectively. No other diffractions are detected, indicating that monophased single-crystalline C- or 4H-SrMnO₃ films were obtained.

Table 1 summarizes the XRD data of samples grown under different T_{\mathrm{S}} and P_{\mathrm{O}} conditions, and indicates that a high T_{\mathrm{S}} and low P_{\mathrm{O}} favor the formation of the cubic phase, while a low T_{\mathrm{S}} and high P_{\mathrm{O}} lead to a 4H phase. This can be understood through thermodynamic analysis by considering the competition between the 4H and cubic phases to be a chemical process of

\begin{array}{}\text{(1)}& (SrMnO{}_3{}^{}){}_{4\mathrm{H}}\leftrightarrow (\text{SrMnO}{}_{3-x}{}^{}){}_{\mathrm{cubic}}+\frac{x}{2}{\text{O}}_{\text{2}}.\end{array}

Oxygen vacancies are involved here since previous study of bulk SrMnO₃ has shown them to play a key role in the deformation of 4H-SrMnO₃ at temperatures above 1400 ^∘C [18]. In the present study, the V_{\mathrm{O}}s density during film growth is controlled by adjusting P_{\mathrm{O}}, and the T_{\mathrm{S}} and P_{\mathrm{O}} required for transformation from 4H- to C-SrMnO₃ should follow the relation [23]:

\begin{array}{}\text{(2)}& \ln {\displaystyle \frac{P_{\mathrm{O}}}{p_0}}=-{\displaystyle \frac{2\mathrm{\Delta }H}{nR{T}_{\mathrm{S}}}}+C_0,\end{array}

where p_0=1 bar is the standard pressure, \mathrm{\Delta }H=H_{\mathrm{cubic}}-H_{4\mathrm{H}} is the standard enthalpy change (independent of T_{\mathrm{S}}), R is the ideal gas constant, n is the total particle number, and C₀ is a constant. The boundary between C- and 4H-SrMnO₃ should therefore appear as a straight line in a phase diagram plot of 1/T_{\mathrm{S}} vs. lgP_{\mathrm{O}} [24], and the slope of this line should be negative given that \mathrm{\Delta }H>0 [25, 26]. As shown schematically in Fig. 3, reducing the growth temperature of the C-SrMnO₃ film should be accompanied by a decrease in oxygen pressure; the minimum value being limited by the oxidation requirements of oxide growth. Putting the experimentally determined values of T_{\mathrm{S}} and P_{\mathrm{O}} needed to grow C- and 4H-SrMnO₃ films into the 1/T_{\mathrm{S}} vs. \lg P_{\mathrm{O}} plot, we find that these two phases are indeed separated by a straight line that is consistent with Eq. (2). It is also worth noting here that as the 4H phase is thermodynamically stable, it is reasonable to assume that the growth parameters of the cubic phase would be close to this boundary.

The temperature and P_{\mathrm{O}} required for the transformation between cubic and 4H phases in the bulk material (1400 ^∘C / 0.2 bar) [18] are also labeled in Fig. 3; and as can be seen, both these parameters fall well below the line for film growth. This indicates that the growth of C-SrMnO₃(111) film is possible at lower temperatures than those prescribed by thermodynamics. In the SrMnO₃/SrTiO₃ heteroepitaxial structure, any interface strain induced by the lattice mismatch must also be considered. The in-plane lattice mismatch between C-SrMnO₃ (space group Pm\overline{3}m, a=3.804 Å) and the SrTiO₃ substrate (space group Pm\overline{3}m, a=3.905 Å) is 2.655%, but drops to 1.443% when 4H-SrMnO₃(0001) (space group P{6}_3/mmc, a=5.4434 Å, c=9.0704 Å) is grown on SrTiO₃(111) (see Fig. 1). When formed on SrTiO₃(111), either phase of SrMnO₃ will experience a compressive strain along its growth direction, but this is expected to be much larger in the case of the cubic phase due to its greater in-plane lattice mismatch. The experimentally observed vertical strain; which is defined as \mathrm{\Delta }=\frac{d-d_{\mathrm{JCPDS}}}{d_{\mathrm{JCPDS}}}, where d is the lattice interspacing determined by XRD and d_{\mathrm{JCPDS}} is the standard JCPDS card value (No. 231413 [22] and No. 841612 [13]) of the according interspacing; does not, however, show any significant difference between the two phases (Table 1). This may be attributed to the different lattice response of the [111] direction to distortion along [\overline{1}01] in the cubic phase when compared to the response of the [0001] direction to distortion along [11\overline{2}0] in the 4H phase. Consequently, the stability of the cubic phase is enhanced at the heterointerfaces, allowing film growth at lower temperatures.

Figure 4 shows the susceptibility χ of the cubic and 4H phase films at temperatures of between 10 and 400 K. Both curves exhibit antiferromagnetic phase transitions, with a T_{\mathrm{N}} at 38 K and 41 K for the C- and 4H-SrMnO₃ films, respectively. The existence of defects or domain boundaries in the films might be the reason for this being lower than their respective bulk materials [27].

The thermostability of the C-SrMnO₃ film was evaluated by annealing it at 800 ^∘C for 8 hours under an oxygen flow of \sim1 bar. As shown in Fig. 5, the XRD peaks corresponding to the cubic lattice structure showed no decay or displacement after annealing, indicating that no impurities were formed and the transformation to a 4H phase was inhibited. It is clear from this that thanks to the stabilizing effect of the heteroepitaxial interface cubic-SrMnO₃ films should be capable of surviving conditions typical of oxide film growth. They therefore have the potential to provide a reliable template for the design and construction of artificial perovskite superlattices.

4 Conclusion

High-quality SrMnO₃ films have been successfully grown on SrTiO₃(111) substrates as either a cubic or 4H phase, the resulting structure being controlled by the substrate temperature and oxygen pressure during PLD growth. Following the principal of thermodynamics, the growth temperature of the cubic SrMnO₃(111) film is significantly lowered by the strain that is induced by the lattice mismatch at the heterointerfaces. This enhances the thermostability of the cubic-SrMnO₃ film, giving it the potential to be used for consistent growth of artificial perovskite superlattices.

References

[1] J. Heber, Materials science: Enter the oxides, Nature 459(7243), 28 (2009)

[2] D. G. Schlom, L.-Q. Chen, X. Pan, A. Schmehl, and M. A. Zurbuchen, A thin film approach to engineering functionality into oxides, J. Am. Ceram. Soc. 91(8), 2429 (2008)

[3] M. Johnsson and P. Lemmens, Perovskites and thin films-crystallography and chemistry, J. Phys.: Condens. Matter 20(26), 264001 (2008)

[4] D. Xiao, W. Zhu, Y. Ran, N. Nagaosa, and S. Okamoto, Interface engineering of quantum Hall effects in digital transition metal oxide heterostructures, Nat. Commun. 2, 596 (2011)

[5] Y. Wang, Z. Wang, Z. Fang, and X. Dai, Interaction-induced quantum anomalous Hall phase in (111) bilayer of LaCoO₃, arXiv: 1409.6797 (2014).

[6] B. Gray, H. N. Lee, J. Liu, J. Chakhalian, and J. W. Freeland, Local electronic and magnetic studies of an artificial La₂FeCrO₆ double perovskite, Appl. Phys. Lett. 97(1), 013105 (2010)

[7] K. Ueda, H. Tabata, and T. Kawai, Ferromagnetism in LaFeO₃–LaCrO₃ superlattices, Science 280(5366), 1064 (1998)

[8] M. Gibert, P. Zubko, R. Scherwitzl, J. Iñiguez, and J. M. Triscone, Exchange bias in LaNiO₃–LaMnO₃ superlattices, Nat Mater 11(3), 195 (2012)

[9] E. Dagotto, T. Hotta, and A. Moreo, Colossal magnetoresistant materials: The key role of phase separation, Phys. Rep. 344(1), 1 (2001)

[10] A. M. Haghiri-Gosnet and J. P. Renard, CMR manganites: Physics, thin films and devices, J. Phys. D: Appl. Phys. 36(8), R127 (2003)

[11] M. Huijben, L. W. Martin, Y. H. Chu, M. B. Holcomb, P. Yu, G. Rijnders, D. H. A. Blank, and R. Ramesh, Critical thickness and orbital ordering in ultrathin La_0.7Sr_0.3MnO₃ films, Phys. Rev. B 78(9), 094413 (2008)

[12] R. Søndenå, P. Ravindran, S. Stølen, T. Grande, and M. Hanfland, Electronic structure and magnetic properties of cubic and hexagonal SrMnO₃, Phys. Rev. B 74(14), 144102 (2006)

[13] P. D. Battle, T. C. Gibb, and C. W. Jones, The structural and magnetic properties of SrMnO₃: A reinvestigation, J. Solid State Chem. 74(1), 60 (1988)

[14] A. A. Belik, Y. Matsushita, Y. Katsuya, M. Tanaka, T. Kolodiazhnyi, M. Isobe, and E. Takayama-Muromachi, Crystal structure and magnetic properties of 6H–SrMnO₃, Phys. Rev. B 84(9), 094438 (2011)

[15] I. N. González-Jiménez, A. Torres-Pardo, A. E. Sánchez-Peláez, Á. Gutiérrez, M. García-Hernández, J. M. González-Calbet, M. Parras, and Á. Varela, Synthesis of 4H–SrMnO_3.0 nanoparticles from a molecular precursor and their topotactic reduction pathway identified at atomic scale, Chem. Mater. 26(7), 2256 (2014)

[16] J. H. Lee and K. M. Rabe, Epitaxial-strain-induced multiferroicity in SrMnO₃ from first principles, Phys. Rev. Lett. 104(20), 207204 (2010)

[17] V. M. Goldschmidt, Die Gesetze der Krystallochemie, Naturwissenschaften 14(21), 477 (1926)

[18] T. Negas and R. S. Roth, The system SrMnO_3-x, J. Solid State Chem. 1(3), 409 (1970)

[19] A. Sacchetti, M. Baldini, P. Postorino, C. Martin, and A. Maignan, Raman spectroscopy on cubic and hexagonal SrMnO₃, Journal of Raman Spectroscopy 37(5), 591 (2006)

[20] S. Kobayashi, Y. Tokuda, T. Ohnishi, T. Mizoguchi, N. Shibata, Y. Sato, Y. Ikuhara, and T. Yamamoto, Cation off-stoichiometric SrMnO₃ thin film grown by pulsed laser deposition, J. Mater. Sci. 46(12), 4354 (2010)

[21] J. Feng, X. Zhu, and J. Guo, Reconstructions on SrTiO₃(111) surface tuned by Ti/Sr deposition, Surf. Sci. 614, 38 (2013)

[22] National Bureau of Standards Monograph 25, Section 10–Data for 84 Substances, 58 (1972)

[23] R. A. Alberty and R. J. Silbey, Physical Chemistry, 3rd Ed., John Wiley &Sons, Inc., 2001

[24] For similicity, the horizonal axis of Fig. 3 is plotted in lgPO. The linear dependence of lnPO on 1/TS as described in Eq. (2) remains in the plot since lgPO = lge⋅lnPO.

[25] M. B. Nielsen, D. Ceresoli, P. Parisiades, V. B. Prakapenka, T. Yu, Y. Wang, and M. Bremholm, Phase stability of the SrMnO₃ hexagonal perovskite system at high pressure and temperature, Phys. Rev. B 90(21), 214101 (2014)

[26] L. Rørmark, A. B. Mørch, K. Wiik, S. Stølen, and T. Grande, Enthalpies of oxidation of CaMnO_3-δ, Ca₂MnO_4-δ and SrMnO_3-δ deduced redox properties, Chem. Mater. 13(11), 4005 (2001)

[27] S. Farokhipoor, C. Magén, S. Venkatesan, J. Íñiguez, C. J. M. Daumont, D. Rubi, E. Snoeck, M. Mostovoy, C. de Graaf, A. Müller, M. Döblinger, C. Scheu, and B. Noheda, Artificial chemical and magnetic structure at the domain walls of an epitaxial oxide, Nature 515(7527), 379 (2014)

Rights & permissions

The Author(s) 2015.