Employing High-temperature-grown SrZrO$_3$ Buffer to Enhance the Electron Mobility in La:BaSnO$_3$-based Heterostructures

TL;DR

This paper demonstrates a method to significantly improve electron mobility in La:BaSnO$_3$ heterostructures by using a high-temperature-grown SrZrO$_3$ buffer layer, enabling high-quality films on various substrates.

Contribution

The study introduces a novel high-temperature pulsed laser deposition process for SrZrO$_3$ buffers that reduces dislocation density, enhancing electron mobility in La:BaSnO$_3$ heterostructures.

Findings

Achieved room-temperature mobility of up to 157 cm$^2$V$^{-1}$s$^{-1}$.

Dislocation density was reduced below $1.0\times 10^{10}$cm$^{-2}$.

Effective on multiple oxide substrates with lattice mismatch.

Abstract

We report a synthetic route to achieve high electron mobility at room temperature in epitaxial La:BaSnO$_3$/SrZrO$_3$ heterostructures prepared on several oxide substrates. Room-temperature mobilities of 157, 145, and 143 cm$^2$V$^{-1}$s$^{-1}$ are achieved for heterostructures grown on DyScO$_3$ (110), MgO (001), and TbScO$_3$ (110) crystalline substrates, respectively. This is realized by first employing pulsed laser deposition to grow at very high temperature the SrZrO$_3$ buffer layer to reduce dislocation density in the active layer, then followed by the epitaxial growth of an overlaying La:BaSnO$_3$ active layer by molecular-beam epitaxy. Structural properties of these heterostructures are investigated, and the extracted upper limit of threading dislocations is well below $1.0\times 10^{10}$cm$^{-2}$ for buffered films on DyScO$_3$, MgO, and TbScO$_3$ substrates. The present results provide a promising route towards achieving high mobility in buffered La:BaSnO$_3$ films prepared on most, if not all, oxide substrates with large compressive or tensile lattice mismatches to the film.

Tables

Table 1. Table 1: Electronic transport characteristics (carrier density and mobility) of the samples discussed in this study.

Sample name	Sample layout	Carrier density	Carrier mobility
		($\times {\text{10}}^{\text{20}}$ cm${}^{-\text{3}}$)	(cm${}^{\text{2}}$ V${}^{-\text{1}}$ s${}^{-\text{1}}$)
A	La:BaSnO${}_{\text{3}}$ (25 nm)/SrZrO${}_{\text{3}}$ (100 nm)/DyScO${}_{\text{3}}$	$\text{1.27}\pm \text{0.05}$	$\text{157}\pm \text{2}$
B	La:BaSnO${}_{\text{3}}$ (25 nm)/SrZrO${}_{\text{3}}$ (100 nm)/TbScO${}_{\text{3}}$	$\text{1.57}\pm \text{0.05}$	$\text{143}\pm \text{2}$
C	La:BaSnO${}_{\text{3}}$ (25 nm)/SrZrO${}_{\text{3}}$ (100 nm)/MgO	$\text{1.13}\pm \text{0.05}$	$\text{145}\pm \text{2}$
D	La:BaSnO${}_{\text{3}}$ (25 nm)/SrZrO${}_{\text{3}}$ (100 nm)/DyScO${}_{\text{3}}$	$\text{1.47}\pm \text{0.05}$	$\text{148}\pm \text{2}$
E	La:BaSnO${}_{\text{3}}$ (25 nm)/SrZrO${}_{\text{3}}$ (100 nm)/TbScO${}_{\text{3}}$	$\text{1.56}\pm \text{0.05}$	$\text{143}\pm \text{2}$
F	La:BaSnO${}_{\text{3}}$ (25 nm)/SrZrO${}_{\text{3}}$ (100 nm)/MgO	$\text{1.06}\pm \text{0.05}$	$\text{131}\pm \text{2}$

Full text

PLD, MBE, transparent conducting oxides

The perovskite alkaline earth stannate, La-doped BaSnO3 (La:BaSO3), is an attractive transparent semiconductor that exhibits outstanding room-temperature electron mobility (RT $\mu_{e}$) of $320\text{ cm}^{2}\text{ V}^{-1}\text{s}^{-1}$ at a carrier density of $n\simeq 8.0\times 10^{19}\text{ cm}^{-3}$ in bulk single crystals [1, 2]. Besides its wide bandgap (3.1 eV) and unique optical properties, La:BaSO3 is highly stable at high temperatures and it exhibits unique electronic properties. This makes La:BaSO3 an enticing material for the exploration of device physics in transparent high RT $\mu_{e}$ field-effect transistors (FETs) and a suitable candidate material for integration in thermally stable capacitors and power electronic devices [3, 4, 5, 2, 6, 7, 8, 9, 10, 11].

The potential of La:BaSO3 for oxide electronics and fundamental realization of 2-dimensional electron gases with high RT $\mu_{e}$ in transparent semiconductors have triggered considerable interests in thin films and heterostructures [12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28]. Unfortunately due in part to a lack of lattice-matched substrates, La:BaSnO3 films suffer from a high density of structural defects, stacking faults and point defects, which limit their electron mobility. Noteworthy structural defects in epitaxial La:BaSO3 films are threading dislocations (TDs), the density of which are often in the order of $1.5\times 10^{11}\text{ cm}^{-2}$ and higher for these films [12, 20, 17, 27, 15, 29]. Such TDs are due to the large lattice mismatch between La:BaSO3 films and commercially available substrates. As shown in Fig. 1(a), the commercially available substrate with the closest lattice match is the scandate material PrScO3, which presents compressive lattice of -2.3% [30]. Other usual perovskite oxide substrates such as SrTiO3, (LaAlO3)0.3(Sr2AlTaO6)0.7 (LSAT) and LaAlO3 have compressive lattice mismatches of -5.4%, -6.4%, and -8.6%, respectively. Although the rock-salt substrates like MgO offer moderate tensile mismatch of +2.3%, however they have a structure and symmetry mismatch to the La:BaSnO3 perovskite structure.

Furthermore, other factors such as complex point defects, Ba/Sn antisites and Ruddlesden–Popper shear faults that form during epitaxial growth are also known to act as extra electron traps or scattering sites, which limit electron mobility in films [20, 22, 13, 31, 27, 32]. Thus, as compared to bulk single crystals, the reported RT $\mu_{e}$ in epitaxial La:BaSnO3 films have only reached a maximum value of $183\text{ cm}^{2}\text{ V}^{-1}\text{s}^{-1}$ ($n\simeq 1.2\times 10^{20}\text{ cm}^{-3}$) for films prepared by molecular beam epitaxy (MBE) [20]. Other film deposition techniques achieved maximum RT $\mu_{e}$ of $140\text{ cm}^{2}\text{ V}^{-1}\text{s}^{-1}$ ($n\simeq 5.2\times 10^{20}\text{ cm}^{-3}$) for pulsed laser deposition (PLD) [22], $121\text{ cm}^{2}\text{ V}^{-1}\text{s}^{-1}$ ($n\simeq 4.0\times 10^{20}\text{ cm}^{-3}$) for high-pressure magnetron sputtering [25], $96\text{ cm}^{2}\text{ V}^{-1}\text{s}^{-1}$ ($n\simeq 2.6\times 10^{20}\text{ cm}^{-3}$) for vacuum-annealed films grown by high-pressure-oxygen sputter deposition [15], and $53\text{ cm}^{2}\text{ V}^{-1}\text{s}^{-1}$ ($n\simeq 2.0\times 10^{20}\text{ cm}^{-3}$) for the chemical solution deposition technique [23].

To reduce the defect densities in epitaxial films, various approaches have been explored. These include, for example, the use of insulating buffer layers (e.g., (Sr,Ba)SnO3 and BaSnO3) inserted between the substrate and the active La:BaSnO3 top layers to sever the large lattice mismatch [20, 17, 27, 33, 34]; a very high-temperature-grown insulating buffer layer to reduce the density of TDs [22]; the use of flux-grown undoped BaSnO3 (001) single crystals as substrates [35]; as well as post-growth annealing processes [19, 36, 35] and adsorption-controlled MBE growth for improved stoichiometry control [11, 32, 24, 13, 20, 37, 28]. These prior approaches suggest that there is still room for exploring other strategies to boost $\mu_{e}$ in epitaxial La:BaSnO3 films.

By the combination of PLD and MBE, we report an effective synthetic route that employs a high-temperature-grown buffer layer to boost RT $\mu_{e}$ in epitaxial La:BaSnO${}_{\text{3}}$ films. The density of TDs is reduced considerably by first growing an insulating buffer layer of SrZrO${}_{\text{3}}$ at a very high temperature using PLD, followed by the epitaxial growth of an overlaying La:BaSnO3 active layer by MBE. Besides the demonstration of the enhancement of electron mobility in epitaxial La:BaSnO${}_{\text{3}}$/SrZrO3 heterostructures prepared on TbScO3 [22], the current study reveals that the insertion of a very high-temperature-grown SrZrO3 epitaxial layer between the film and the substrate is an effective synthetic route for minimizing the density of defects and boosting the transport properties of La:BaSnO3 films prepared on most oxide substrates. We demonstrate that this synthesis approach is applicable to many oxide substrates that induce large compressive or tensile strains to the films, which is a significant contribution for addressing the challenge of lack of a commercially available lattice-matched substrates close to the BaSnO${}_{\text{3}}$ cubic lattice parameter (4.116Å). The effectiveness of the synthesis approach has been explored by preparing La:BaSnO${}_{\text{3}}$/SrZrO3 heterostructures on scandate DyScO3 and TbScO3 substrates, and also on rock-salt MgO substrates. Surface and structural characterization demonstrates smooth surface morphologies and high crystalline quality of the films. Electronic transport measurements revealed RT $\mu_{e}$ as high as $157\text{ cm}^{2}\text{ V}^{-1}\text{s}^{-1}$ ($n\simeq 1.27\times 10^{20}\text{ cm}^{-3}$), $145\text{ cm}^{2}\text{ V}^{-1}\text{s}^{-1}$ ($n\simeq 1.13\times 10^{20}\text{ cm}^{-3}$), and $143\text{ cm}^{2}\text{ V}^{-1}\text{s}^{-1}$ ($n\simeq 1.57\times 10^{20}\text{ cm}^{-3}$) for heterostructures grown on DyScO3, MgO, and TbScO3 crystalline substrates, respectively. As compared to prior reports, these RT $\mu_{e}$ are the second-highest mobilities achieved in epitaxial La:BaSnO3 films; and so far, the highest RT $\mu_{e}$ obtained for La:BaSnO3 films of small thickness ($\leq 25$ nm) prepared using non-BaSnO3 buffer layers.

Figure 1(b) depicts a schematic view of the sample types investigated. Our approach to minimizing dislocation density starts by growing at very high temperature (1300°C) an insulating SrZrO3 buffer layer using PLD. The SrZrO3 layers were deposited on several (001)-oriented MgO, (110)-oriented DyScO3, and TbScO3 crystalline substrates ($5\times 5\times 1\text{ mm}^{3}$). Prior to deposition, all the substrates were terminated in situ at very high temperatures using a CO2 laser substrate heating system, as described in Ref. [38]. To grow SrZrO3 buffer layers by PLD ($\lambda=248$ nm), we used a laser fluence of $2\text{ J cm}^{-2}$ at $1.4\times 10^{-2}\text{ mbar}$ of O2. The buffer layers were deposited at 4 Hz to a thickness of 100 nm. SrZrO3 is chosen because it has a low vapor pressure and can therefore be grown at high temperatures [22]. Also, SrZrO3 has a psuedocubic lattice parameter value (4.101 Å) that is very close to that of La:BaSnO3 [Fig. 1(a)]. Ideally, an undoped BaSnO${}_{\text{3}}$ buffer layer grown at higher substrate temperatures could also be used to lower dislocation densities and improve $\mu_{e}$ further, but due to the significant volatility of tin oxide at substrate temperatures above $~{}850$°C, this is not a viable option. Details about the PLD growth of the SrZrO3 buffer films are provided in Ref. [22].

Epitaxial La:BaSnO3 (25 nm) films were grown on top of the SrZrO3 buffer layers using a Veeco GEN10 MBE system. Separate effusion cells containing lanthanum (99.996% purity, Ames Lab), barium (99.99% purity, Sigma-Aldrich), and SnO2 (99.996% purity, Alfa Aesar) were heated. The fluxes of the resulting molecular-beams emanating from the effusion cells were measured by a quartz crystal microbalance before growth. The La:BaSnO3 films were grown in an adsorption-controlled regime by supplying an excess SnO${}_{\text{x}}$-flux. The background pressure of the oxidant, $10\%\text{ O}_{3}+90\%\text{ O}_{2}$, was held at a constant ion gauge pressure of $1.33\times 10^{-6}\text{ mbar}$. The substrate temperature was maintained between 830 and 850°C, as measured by an optical pyrometer. Details on the growth of La:BaSnO3 films by MBE are provided in Ref. [20].

For transport measurements, we used a Nanometrics Hall measurement system to characterize the resistivity, carrier concentration, $n$, and the electron mobility, $\mu_{e}$, of the La:BaSnO3 films, using four spring-loaded tips (Au/Ir) arranged in a Van der Pauw geometry.

We first characterize the buffer layer. The high-temperature growth of the SrZrO3 buffer layers on different substrates was in situ monitored by reflection high-energy electron diffraction (RHEED) [Fig. S1(a)-(c)]. RHEED oscillations and sharp, diffracted and specular RHEED patterns were observed throughout the deposition of the SrZrO3 layers on DyScO3, TbScO3 and MgO substrates, indicating a smooth film surface [Fig. 2(a)]. The time-dependent RHEED intensity oscillations observed for SrZrO3 buffer layers prepared on DyScO3 and TbScO3 substrates [Fig. S1(a)-(b)] are suggestive of a layer-by-layer growth mode for these buffer layers prepared at very high-temperature. Based on these, we estimate the thickness of the SrZrO3 intermediate layer to 100 nm, in consistency with the scanning transmission electron microscopy cross sections. The surface morphology of a SrZrO3 layers was investigated using atomic force microscopy (AFM). Figure 2(b) depicts typical AFM images for representative 100 nm thick SrZrO3 grown at 1300°C on DyScO3, TbScO3, and MgO substrates. From a surface morphology point of view on a small scale, all samples exhibit a relatively smooth surface. We observe slight variations of the surface morphology for samples grown on different substrates; in particular, the SrZrO3 layers grown on MgO exhibit some island growth. SrZrO3 films grown by PLD are known to exhibit a significant surface roughness [39]. The existence of the small islands in these SrZrO3 buffer layers may be an indication of nucleation sites caused by interfacial strain energy originating from the lattice mismatch between the film and substrates, which results in a slightly increased surface roughness. For a lateral scan size of $\sim 2\times 2\,\mu\text{m}^{2}$, the extracted surface roughness is around 4 nm for 100-nm-thick SrZrO3 layers grown on scandate substrates (DyScO3 and TbScO3); and it increases to $\sim 8$ nm for SrZrO3 films grown on MgO. [Fig. 2(c)]. From $\theta-2\theta$ x-ray diffraction (XRD) scans of the buffer layers, we extracted out-of-plane lattice parameters of c $=4.11\pm 0.02$ Å for the 100 nm-thick SrZrO3 films grown on DyScO3, TbScO3, and MgO substrates. These values are within experimental error of the fully relaxed psuedocubic lattice constant of SrZrO3, $4.101$ Å. Fully relaxed films are expected given the 100 nm thickness of the SrZrO3 buffer layer, the high growth temperature, and the significant (2.3% to 4.1%) lattice mismatch between SrZrO3 and these substrates.

Next we used adsorption-controlled MBE to deposit 25-nm-thick La:BaSnO3 films on top of PLD-grown SrZrO3 prepared on different oxide substrates. Results discussed below demonstrate that our approach to use very high-temperature-grown SrZrO3 buffer layers to reduce the density of TDs is applicable not only for epitaxial heterostructures prepared in the same deposition chamber without exposing samples to ambient conditions as reported in Ref. [22] but also for SrZrO3 buffer layers exposed to air for days prior to the subsequent epitaxial growth of La:BaSnO3 active layers [40].

After MBE growth, the crystalline quality and phase purity of La:BaSnO3/SrZrO3 heterostructures were characterized by XRD. Figure 3(a) shows the representative $\theta-2\theta$ scans for the La:BaSnO3/SrZrO3/DyScO3 (sample A), La:BaSnO3/SrZrO3/TbScO3(sample B), and La:BaSnO3/SrZrO3/MgO (sample C) heterostructures. Only the substrate peaks and phase-pure 0 0 $l$ family of the film diffraction peaks are resolved, indicating a high crystallinity; and also, verifying that the heterostructures were aligned along the c-axis. The extracted out-of-plane lattice parameter of all three films is c$=4.13\pm 0.01$ Å. This value is close to the bulk lattice constant of BaSnO3, and consistent with the out-of-plane lattice constants reported previously in La:BaSnO3 films [22, 41, 20, 35]. Figure S3 shows a closeup view of the $\theta-2\theta$ scan around the $002$ diffraction peak for the La:BaSnO3/SrZrO3 heterostructures grown on the three substrates. The asymmetry in the peaks highlight the presence of the La:BaSnO3 and SrZrO3 layers in these heterostructures. In particular, the 002 peaks exhibited by heterostructures grown on DyScO3 and TbScO3 substrates reveal noticeable thickness fringes. The observation of Laue thickness fringes and solely $00l$ peaks is an indication of phase purity and smooth growth.

Figures 3(b) and 3(c) show reciprocal space maps (RSM) around the asymmetric $\bar{1}03_{p}$ reflection peaks of the heterostructures (samples A and B) prepared on scandate substrates (DyScO3 and TbScO3). Figure 3(d) is for the $204_{p}$ reflection peak of the film (sample C) on MgO. From all three RSM maps, it is evident that the La:BaSnO3 layer is commensurately strained to the SrZrO3 buffer layer, but in all cases that the commensurate La:BaSnO3/SrZrO3 bilayer is relaxed from the underyling substrate. This makes sense given the excellent lattice match and structural match between La:BaSnO3 and the SrZrO3 buffer layer. This result is also consistent with the literature for epitaxial La:BaSnO3/SrZrO3 heterostructures [22].

Now we turn to the electronic transport and microstructural data. Figure 4 presents the electron mobility at 300 K as a function of carrier density for samples A, B, and C. The highest RT $\mu_{e}$ of $157\text{ cm}^{2}\text{ V}^{-1}\text{s}^{-1}$ with a carrier concentration of $n\simeq 1.27\times 10^{20}\text{ cm}^{-3}$ is achieved for sample A. This RT $\mu_{e}$ is $\sim 10\%$ higher than the previously reported mobility ($140\text{ cm}^{2}\text{ V}^{-1}\text{s}^{-1}$) achieved by inserting a high-temperature-grown SrZrO3 buffer layer between the La:BaSnO3 film and the substrate [22]. It is the second-highest reported RT $\mu_{e}$ achieved in epitaxial La:BaSnO3 films and the highest attained for thin ($\leq 25$ nm) epitaxial La:BaSnO3 films. For samples B and C, we achieve RT $\mu_{e}$ of $143\text{ cm}^{2}\text{ V}^{-1}\text{s}^{-1}$ ($n\simeq 1.57\times 10^{20}\text{ cm}^{-3}$) and $145\text{ cm}^{2}\text{ V}^{-1}\text{s}^{-1}$ ($n\simeq 1.13\times 10^{20}\text{ cm}^{-3}$), respectively. These RT $\mu_{e}$ are reproducible in different structurally comparable heterostructures prepared in similar conditions [Table 1]. The observed slight carrier mobility difference in structurally comparable heterostructures (samples A and D, and sample C and F) is attributed to experimental fluctuations. Also, as the thickness of the active layer is thin, the interface defect density or reconstruction effects may vary in structurally comparable samples; thus, causing the observed slight carrier mobility difference. The achieved improvements in RT $\mu_{e}$ of La:BaSnO3 films on different oxide substrates are attributed to the use of the high-temperature-grown buffer layer, which is know to minimize the density of defects, and thus results in an increase in carrier mobility.

To investigate the defect density and provide complementary real-space structural characterization of these films, cross-sectional transmission electron microscopy (TEM) imaging was performed. Figure 5(a),  5(b), and  5(c) depict bright-field TEM images of the entire film thickness for representative (25 nm) La:BaSnO3/(100 nm) SrZrO3 heterostructures prepared on DyScO3, TbScO3, and MgO substrates, respectively. We observe misfit dislocations along the interface between the films and substrates. As expected for high-temperature-grown SrZrO3 buffer layers [22], TDs were barely observed in all the three representative samples [Fig. 5(a)-(c)]. STEM investigations over wide areas showed hardly any TDs in La:BaSnO3/SrZrO3 heterostructures, and electron energy-loss spectroscopy (EELS) map analyses indicate expected elemental composition in the films [see Fig. S2(a)-(c) of the supplementary material]. For films prepared on DyScO3, TbScO3, and MgO substrates, the extract upper limit of TD density is well below $1.0\times 10^{10}\text{ cm}^{-2}$, in agreement with previous report [22].

The low density of TDs in these samples is attributed to the very high temperature (1300 °C) used for the growth of the SrZrO3 buffer layer. This approach helps to eliminate most of TDs that would act as scattering centers and trap electrons. As SrZrO3 has an excellent lattice match to La:BaSnO3, inserting a high-temperature-grown SrZrO3 layer between the La:BaSnO3 film and substrate minimizes the TD density [22]. At the high substrate temperature used for the growth of the SrZrO3 buffer layer, which has a significant lattice mismatch to all of the underlying commercial substrates used, the TDs are able to more readily move and react with each other. The result is misfit dislocation segments that relieve the misfit strain and a lower TD density than would be the case had the SrZrO3 buffer layer been grown at lower temperature where the TDs are less mobile. The lowered TD density of the SrZrO3 directly benefits the subsequently grown La:BaSnO3 layer as it is not only well lattice matched to the SrZrO3 buffer layer, but inherits relatively few TDs from it. Notably for our growth approach, the subsequent growth of an overlying La:BaSnO3 by adsorption-controlled MBE helps in achieving better stoichiometry control, thus allowing to enhance electron mobility in these films. Our results clearly demonstrate that the high-temperature-grown SrZrO3 epilayer is a suitable template for subsequent growth of high mobility La:BaSnO3 films with fewer TDs not only on TbScO3 [22], but also on other oxide substrates (DyScO3 and MgO). It is envisaged that this synthesis approach of high mobility La:BaSnO3 films could also be extended to most, if not all, oxide substrates such as SrTiO3, (LaAlO3)0.3(Sr2AlTaO6)0.7 (LSAT), LaAlO3, MgAl2O4, and LaLuO3 that present large compressive or tensile lattice mismatches to La:BaSnO3.

Although our improved synthesis approach of epitaxial La:BaSnO3/SrZrO3 heterostructures reduces TDs and increases RT mobilities, we are not able to achieve reported mobilites as high as those reported in MBE grown thick (60 nm) La:BaSnO3/ (330 nm) BaSnO3 films [20]. As the La:BaSnO3 active layer is thin (25 nm) in our heterostructures, it could be that not only TDs are trapping electrons, but also effects such as surface scattering or interface traps are lowering the density of mobile carriers. These contributions are expected to be less pronounced in thick La:BaSnO3/BaSnO3 heterostructures as the buffer layer and the active layer consist of almost the same materials.

In summary, we have explored an approach to enhance room-temperature electron mobility in La:BaSnO3/SrZrO3 heterostructures. For MBE-grown La:BaSnO3 films prepared on PLD-grown SrZrO3 buffer layers that were grown at 1300°C, we achieve RT mobilities of 157, 145, and 143 cm2V*-1s-1* for films prepared on DyScO3, MgO and TbScO3 substrates, respectively. The density of TDs are very low in these films with an upper limit well below $1.0\times 10^{10}\text{ cm}^{-2}$ for all films prepared on these oxide substrates, thus verifying the efficacy of our synthesis approach. Our work provides an effective approach for the growth of high mobility La:BaSnO3 epitaxial films on most, if not all, oxide substrates that present large compressive or tensile lattice mismatches to La:BaSnO3, which is an essential step in tackling the challenges caused by the lack of commercially available substrates with lattice parameters matching the BaSnO3 unit cell. Also, we note that achieving high RT $\mu_{e}$ at low thickness and relatively low carrier concentrations in these heterostructures provides an opportunity to fabricate La:BaSnO3-based FETs on various oxide substrates in which channels may be fully depleted. Based on these results, future directions are expected to focus on exploring the physics of La:BaSnO3-based devices for their potential practical applications in oxide electronics.

See supplementary material for additional surface and microstructural characterizations (RHEED, TEM and XRD) of the La:BaSnO3/SrZrO3 heterostructures.

P. Ngabonziza acknowledges startup funding from the College of Science and the Department of Physics & Astronomy at Louisiana State University.

W. Sigle and P. van Aken acknowledge funding from the European Union’s Horizon2020 research and innovation program under Grant Agreement No.823717-ESTEEM3.

D. G. Schlom and J. Park acknowledge support by the Air Force Office of Scientific Research under Award No. FA9550-16-1-0192 and gratefully acknowledge Professors Grace Xing and Debdeep Jena for use of their Nanometrics Hall measurement system.

The authors declare no conflict of interest.

The data that support the findings of this study are available from the corresponding authors upon reasonable request.