In recent years, the Ruddlesden–Popper iridium oxides (n = 1, 2, …) have attracted broad interest because of their novel quantum ground states, which arise from the large spin–orbit coupling of the heavy iridium element (Z = 77).^[1–3] The n = 1 Sr₂IrO₄ is a representative spin–orbit-coupled Mott insulator with J_eff = 1/2.^[4] Due to the competition between the strong spin–orbit coupling and Coulomb repulsion (∼0.5 eV), a narrow gap is formed between the half-filled and half-empty J_eff = 1/2 bands.^[2,5,6] The electrical properties and lattice structure of Sr₂IrO₄ are both similar to those of the parent compound of high-temperature superconducting cuprate La₂CuO₄.^[6,7] Previous theoretical studies suggest the possibility of unconventional superconductivity in electron doped Sr₂IrO₄.^[8,9] To study the physical properties and perhaps to reveal exotic phenomena such as unconventional superconductivity, techniques to fabricate high-quality single-crystalline Sr₂IrO₄ thin films are in urgent need.^[10] Pulsed laser deposition (PLD) assisted with reflective high-energy electron diffraction (RHEED) is a typical physical vapor-deposition technique that has been widely used to prepare high quality complex oxide thin films.^[11–13] One can accurately control the composition of the thin films via PLD.^[5] Though the growth of simple orthorhombic SrIrO₃ films is relatively straight-forward, only a handful of groups reported the successful growth of Sr₂IrO₄ single-crystalline films.^[2,10,14] The Sr₂IrO₄ films usually contain various impurity phases such as SrIrO₃ and Sr₃Ir₂O₇, and have poor crystalline structures even with a stoichiometric Sr₂IrO₄ target.^[5,10] Besides, very high growth temperatures are usually needed to prepare Sr₂IrO₄ films.^[14] Unfortunately, the high growth temperature may exceed the temperature range of the heating plate of the PLD chamber and long-term experiments at those temperatures may damage the heating plate. Therefore, the capability to grow high quality epitaxial Sr₂IrO₄ thin films at lower temperatures is urgent required. According to previous works, an important method to prepare high-quality layered-perovskite films is to reduce the interface energy between the substrates and deposited films.^[15–17] At present, the commonly used method to reduce the interface energy is to grow a conductive layer; i.e., a high-quality buffer layer between the films and substrates.^[15] Whereas, the insertion of buffer layers can cause new problems such as ion-diffusion between the film and buffer layers that can lead to the formation of impurities at high temperatures.

In our study, we find that the optimal growth temperature for epitaxial Sr₂IrO₄ films on SrTiO₃:Nb (001) conductive substrates is ∼90 °C lower than that on insulating SrTiO₃ (001) substrates using PLD assisted with RHEED. Electrical transport measurements of the Sr₂IrO₄ films on SrTiO₃:Nb substrates are performed, which show a three-dimensional Mott variable-range hopping mechanism, and the films exhibit weak ferromagnetism below ∼240 K with a saturation magnetization of /Ir at 5 K. Both the above electrical transport and magnetization characters match those of the high quality Sr₂IrO₄ films grown on insulating SrTiO₃ substrates at the higher temperature. This study demonstrates that conductive substrates reduce the optimal growth temperature for preparing high quality Sr₂IrO₄ thin films, which may facilitate future studies of exotic phenomena in the Sr₂IrO₄ family.

The Sr₂IrO₄ thin films (with thickness of about 25 nm) were grown on (001) oriented SrTiO₃ and SrTiO₃: 0.7%wt. Nb single-crystalline substrates using PLD assisted with RHEED. A commercial stoichiometric Sr₂IrO₄ target was used for all film growths in this study. The background pressure of the PLD chamber was about 10⁻⁶ Pa. A KrF exciter laser (λ = 248 nm) was used with different fluence energies at the frequency of 1 Hz. During the deposition, oxygen pressures between 0.01 Pa to 20 Pa and substrate temperatures ranging from 670 °C to 870 °C for SrTiO₃:Nb and 770 °C to 860 °C for SrTiO₃ were used. A high-quality LaNiO₃ buffer layer of 40 nm was deposited at 660 °C and 30 Pa to make a comparison with the growth directly on SrTiO₃:Nb substrates.^[15] The actual growth temperatures were measured using an external infrared detector for every deposition condition. The possibility of the observed temperature difference being from different emissivities was rule out using Au/Ti back-coated SrTiO₃ substrates and simultaneous heating of both insulating and conductive substrates. X-ray diffraction (XRD) characterization was performed using a Rigaku SmartLab high-resolution diffractometer with wavelength of 0.15406 nm from Cu radiation. The I–V measurements were done using the 4-wire method with Ti/Au top electrodes. The magnetic and electrical properties of the Sr₂IrO₄ films were characterized by a Quantum Design physical property measurement system.

Figure 1(a) displays the XRD patterns of the Sr₂IrO₄ films grown on insulating SrTiO₃ substrates with an energy density of 0.5 J/cm² and oxygen pressures of 0.2 Pa and 0.4 Pa, respectively. All diffraction peaks of the film grown at 850 °C can be indexed to the (0 0 L) reflections of the Sr₂IrO₄ single crystal (see Fig. 1(a)), thus indicating the Ruddlesden–Popper n = 1 crystal structure. In addition, there are no impurity phase related peaks. The film thickness oscillation is observed, which suggests the film interface and surface being very smooth. The c-axis lattice constant of the film is about 2.5072 nm. When the growth temperature is 800 °C, the (0 0 L) diffraction peaks of the Sr₂IrO₄ film are weak and even the (0 0 8) peaks are almost invisible, which can be summarized as phase II range in Fig. 1(c). An extra peak at 2θ = 20.92° (marked with a red star) appears, which matches very well with the (0 0 1) diffraction of the SrIrO₃ phase. As the deposition temperature is decreased below 700 °C, the diffraction peaks of SrIrO₃ with much smaller c-axis lattice constant appear at 2θ = 21.8° and 44.3°. The SrIrO₃ phase is very stable and no Sr₂IrO₄ phase is observed. This corresponds to phase I in Fig. 1(c). Unfortunately, the heating temperature of our PLD instrument cannot be higher than 870 °C. Therefore, Figure 1(c) shows no XRD data of samples on insulating substrates with higher growth temperatures in the phase IV region.

Fig. 1.

	Figure Option ViewDownloadNew Window
	Fig. 1. The XRD patterns of (a) Sr2IrO4/SrTiO3 (001) films and (b) Sr2IrO4/SrTiO3:Nb (001) samples grown at different substrate temperatures. The black stars mark the impurity phase peaks. (c) Temperature dependence of the structure of different SrO layer number phases. Temperatures in the phase III range indicate the substrate temperature for Sr2IrO4 epitaxial film growth.

Figure 1(b) shows that the optimized deposition temperature of Sr₂IrO₄ films on SrTiO₃:Nb conductive substrates is about 780 °C. Both (0 0 12) peaks of the Sr₂IrO₄ samples on SrTiO₃ and SrTiO₃:Nb substrates are at 42.2° (c = 2.572 nm). The film in phase IV, which was grown at 840 °C, exhibits the impurity phases of iridium and SrIrO₃. The inconspicuous appearance of an extra peak at 2θ = 44.4° matches very well with the (0 0 2) diffraction of the SrIrO₃ phase in Fig. 1(b). The XRD pattern of the film deposited at 725 °C shows poor crystallinity of the Sr₂IrO₄ phase and a clear perovskite peak of SrIrO₃ without film-thickness oscillation in phase II. It is remarkable that the optimal deposition temperature of the Sr₂IrO₄ films on SrTiO₃:Nb conductive substrates is about 90 °C lower than that directly deposited on insulating SrTiO₃ substrates. The Sr₂IrO₄ films are coherently strained to the substrates, as confirmed by the x-ray reciprocal space mapping measurements and RHEED patterns obtained before and after the deposition. Therefore, the substrate induced strain cannot explain the large drop of the optimal deposition temperatures. The sharp drop in the growth temperature of Sr₂IrO₄ films on conductive substrates may suggest that the interface between the Sr₂IrO₄ film and SrTiO₃:Nb substrate has a much smaller interface energy than that between Sr₂IrO₄ and the SrTiO₃ insulating substrate.

To elucidate the specific phase compositions of the films, the samples deposited at diverse substrate temperatures are separated into I, II, III, and IV regions in Fig. 1(c), where the horizontal axis represents the substrate temperature and the vertical axis represents the oxygen pressure. Figure 1(c) also shows the crystal structure of SrIrO₃ as a major impurity in addition to the Sr₂IrO₄ main phase. Below the optimal temperature range (phases I and II), the films exhibit either a single SrIrO₃ phase or a Sr₂IrO₄ phase with SrIrO₃ impurities. There is a wide window for the stabilization of pseudo-cubic SrIrO₃ phase for all grown at substrate temperatures below 740 °C. In the region of phase III, a single Sr₂IrO₄ phase with good crystallinity is found. Above this range (i.e., in phase IV), the films exhibit a layered Sr₂IrO₄ phase and impurity phases of SrIrO₃ and metallic iridium. This may indicate that the layered Sr₂IrO₄ films are partially converted to the perovskite SrIrO₃. During exploring the substrate temperature dependence for Sr₂IrO₄ films on both types of substrates, the oxygen pressure was also optimized in a range between 0.01 Pa and 20 Pa. Therefore, our results indicate that the optimal deposition temperature for Sr₂IrO₄ films on SrTiO₃:Nb conductive substrates is approximately 90 °C lower than that grown on SrTiO₃ substrates.

To further confirm the results, Sr₂IrO₄ epitaxial films were grown on SrTiO₃ substrates with LaNiO₃ buffer layers as well as on SrTiO₃ substrates with Ti (50 nm) or Au (50 nm) layers sequentially sputtered on the back. The RHEED was used to monitor the in-situ layer-by-layer epitaxial growth or three-dimensional islands-growth of the thin films at low oxygen pressure. The typical XRD patterns of the Sr₂IrO₄ films grown on different substrates under optimal conditions are shown in Fig. 2(a). The sharp (0 0 L) diffraction peaks of the Sr₂IrO₄ films suggest that the samples are of high crystallinity, and there are no observable impurities, such as SrIrO₃. Figure 2(b) shows the intensity oscillations of RHEED for Sr₂IrO₄ samples deposited on diverse substrates. The reflected beam-intensity oscillations, with a certain intensity for the films deposited on conductive substrate SrTiO₃:Nb or LaNiO₃ buffer layer, demonstrate a two-dimensional layer-by-layer growth mode. The growth temperature for the Sr₂IrO₄ films on SrTiO₃ substrates with the LaNiO₃ buffer layer is also reduced by nearly 40 °C compared to that directly grown on SrTiO₃ substrates. On the other hand, the optimal temperature on the SrTiO₃ substrates with Ti and Au metal sputtered on the back is not significantly reduced compared with that of the films grown on bare SrTiO₃. Therefore, the possibility that the decrease of growth temperature on SrTiO₃:Nb due to impurities at the substrate and film interface may be ruled out. Thus, it is speculated that the interface energy between the Sr₂IrO₄ films and SrTiO₃:Nb conductive substrates is reduced similar to the interface energy decrease between the Sr₂IrO₄ films and LaNiO₃ buffer layers. In a previous study, our group has achieved great enhancement of crystalline quality of Aurivillius films by applying a conductive LaNiO₃ buffer layer to screen the electric dipole field in the film. However, the Sr₂IrO₄/SrTiO₃:Nb interface does not exhibit the polar catastrophe as in previous Aurivillius films and the mechanism might be different.^[15] A simple model is used to explain the difference in optimal growth temperatures on different substrates. When the atoms arrive at the substrate, they are adsorbed by the surface to form a stable phase of the oxide. So far, this adsorption is a physical adsorption and not a chemical bond. To form a heterostructure, these atoms need to overcome an energy barrier to position at the correct place for forming a stable chemical bond, mostly by thermal diffusion. One possible explanation for the reduced optimal growth temperature is that the bonding energy of Sr₂IrO₄ on SrTiO₃:Nb is much lower than that of Sr₂IrO₄ on SrTiO₃.

Fig. 2.

	Figure Option ViewDownloadNew Window
	Fig. 2. (a) XRD patterns of Sr2IrO4 samples grown on different substrates. (b)–(e) RHEED intensity oscillations for Sr2IrO4 films grown under optimization condition on (b) (001)-oriented SrTiO3 substrate without buffer layer, (c) SrTiO3 substrates with titanium and gold sequentially sputtered on the back of the substrates, (d) SrTiO3 substrate with buffer layer LaNiO3, (e) SrTiO3:Nb substrate.

Figure 3(a) and 3(b) show that the growth of Sr₂IrO₄ films is affected by the oxygen pressure (P) and the deposition temperature (T) on both SrTiO₃:Nb (see Fig. 3(a)) and SrTiO₃ substrates (see Fig. 3(b)). A large number of samples were grown at oxygen pressures ranging from 0.01 Pa to 20 Pa and deposition temperatures from 670 °C to 830 °C on SrTiO₃:Nb and from 770 °C to 860 °C on SrTiO₃. The different patterns represent different identified phases in the films for different growth conditions. Figure 3(b) shows that the deposition-condition range for Sr₂IrO₄ films grown on SrTiO₃ (001) substrates is very narrow, and the temperature needs to exceed 840 °C. It is clear that high-quality Sr₂IrO₄ (n = 1) films on conductive SrTiO₃:Nb substrates have lower growth temperatures as well as a wider window for the substrate temperature (ranging from 750 °C to high temperatures shown in Fig. 3(a)). By comparing the two figures, the growth temperature for Sr₂IrO₄ thin films on SrTiO₃:Nb substrates is about 90 °C lower than that directly grown on SrTiO₃ substrates. To explain the phase diagram, the laser plume is assumed to contain multiple substances, such as SrO, IrO₃, SrIrO₃, Sr₂IrO₄, IrO₂, IrO₆, etc. As the oxygen pressure increases, a pure SrIrO₃ film is obtained. If a large amount of SrIrO₃ is contained in the plume when the oxygen pressure increases, then the reaction will generate more Sr₂IrO₄.^[1] As it is not the case, this possibility can be excluded. However, it can explain why SrIrO₃ is also produced under the low oxygen pressure. The high oxygen pressure leads to the pure phase SrIrO₃ via the reactions and , where Ir may be derived from enriched Ir in the target.^[10] According to the literature, IrO₃ is a stable gaseous compound, and is more stable than IrO₂, which therefore excludes IrO₂ as the main component of the plume. Another explanation is that the lighter SrO species is likely to be backscattered by oxygen atoms under high oxygen pressure. Recent studies found that the light species SrO has a wider distribution than the heavy species at a high oxygen pressure, according to the x-ray photoelectron spectroscopy measurement. Hence, the epitaxial films have an Ir rich formation like SrIrO₃.^[5]

Fig. 3.

	Figure Option ViewDownloadNew Window
	Fig. 3. The phase diagram of Sr2IrO4 grown on (a) SrTiO3:Nb and (b) SrTiO3 substrates.

Figure 4(a) displays that the full-width at half-maximum of the (0 0 12) rocking curve is 0.04° for the Sr₂IrO₄ films on SrTiO₃:Nb (001), consistent with phase III in Fig. 1(c). The inset of Fig. 4(b) shows the resistivity as a function of temperature between 50 K and 360 K for the high-quality Sr₂IrO₄ samples grown on SrTiO₃ and SrTiO₃:Nb (001) substrates, respectively. The conduction mechanism responsible for the electrical transport in Sr₂IrO₄ films is illustrated in Fig. 4(b).^[2] In Fig. 4(b), the illustration of lnρ versus T^−1/4 for the Sr₂IrO₄ films shows two temperature ranges. The three-dimensional Mott variable-range hopping mechanism is used to explain the transport characteristics of the Sr₂IrO₄ film. The values of T_M, the average hopping distance R_M, the localization length a, and the Mott hopping energy of the sample on an insulating substrate were calculated.^[2] The corresponding results are listed in Table 1. By comparison, the resistivity ρ of the film on insulating substrate is slightly higher. While the sample on insulating substrate also shows insulating behavior and two temperature ranges with good linear relationship of – , which are consistent with the film on conductive substrate. The zero-field cooling and field cooling magnetization results for the film grown on SrTiO₃:Nb (001) are shown in Fig. 4(c). The transition temperature T_C) of the weak ferromagnetism is about 240 K and the saturation magnetization is /Ir (see inset in Fig. 4(c)).

Fig. 4.

Figure Option ViewDownloadNew Window

Fig. 4. (a) The crystal quality of epitaxial Sr2IrO4 film on SrTiO3:Nb (001) is shown by the full-width at half-maximum of (0 0 12) x-ray rocking curve. (b) Temperature dependence of the resistivity of Sr2IrO4 epitaxial films on SrTiO3 and SrTiO3:Nb substrates, the inset shows versus T−1/4. (c) The field cool and zero-field cool curves and the inset shows the hysteresis loops of epitaxial Sr2IrO4 film at 5 K and 20 K, respectively.

Table 1.

Table 1.

Table 1. Fitting parameters to the three-dimensional Mott variable range hopping model for Sr2IrO4 thin film on SrTiO3 substrate. . Temperature TM/106 K a/Å (RM/a)/K−1/4 EM/meV High-T 1.5 1.9 12.9/T1/4 0.71/T3/4 Low-T 0.6 2.6 10.1/T3/4 0.58/T3/4

Table 1. Fitting parameters to the three-dimensional Mott variable range hopping model for Sr2IrO4 thin film on SrTiO3 substrate. .

We have grown high-quality Sr₂IrO₄ thin films on both conductive SrTiO₃:Nb and insulating SrTiO₃ substrates using PLD assisted with RHEED. The results indicate that the suitable growth temperature for Sr₂IrO₄ films on SrTiO₃:Nb substrate is at least 90 °C lower than that on SrTiO₃, suggesting that the interface energy barrier between the Sr₂IrO₄ films and SrTiO₃:Nb substrates is lower than that between Sr₂IrO₄ and the SrTiO₃ substrates. The electrical transport of high-quality Sr₂IrO₄ films grown on SrTiO₃ substrates is discussed, which agrees to the three-dimensional Mott variable-range hopping conduction rule. The magnetization measurements show that the films exhibit weak ferromagnetism with T_C of about 240 K and the saturation magnetization of /Ir at 5 K. The study proves the crucial role of conductive substrates (such as SrTiO₃:Nb) to reduce the optimal growth temperature in depositing high-quality iridium oxide thin films.