Microsc. Microanal., page 1 of 7 doi:10.1017/S1431927614000750
© MICROSCOPY SOCIETY OF AMERICA 2014
Oxygen Octahedral Distortions in LaMO3/SrTiO3 Superlattices Gabriel Sanchez-Santolino,1,2 Mariona Cabero,1,2 Maria Varela,1,2,3,* Javier Garcia-Barriocanal,1 Carlos Leon,1 Stephen J. Pennycook,4 and Jacobo Santamaria1 1
GFMC, Departamento de Fisica Aplicada III, Universidad Complutense de Madrid, 28040 Madrid, Spain Instituto Pluridisciplinar, Universidad Complutense de Madrid, 28040 Madrid, Spain 3 Materials Science and Technology Division, Oak Ridge National Laboratory, Oak Ridge, TN 37831, USA 4 Department of Materials Science and Engineering, The University of Tennessee, Knoxville, TN 37996, USA 2
Abstract: In this work we study the interfaces between the Mott insulator LaMnO3 (LMO) and the band insulator SrTiO3 (STO) in epitaxially grown superlattices with different thickness ratios and different transport and magnetic behaviors. Using atomic resolution electron energy-loss spectral imaging, we analyze simultaneously the structural and chemical properties of these interfaces. We find changes in the oxygen octahedral tilts within the LaMnO3 layers when the thickness ratio between the manganite and the titanate layers is varied. Superlattices with thick LMO and ultrathin STO layers present unexpected octahedral tilts in the STO, along with a small amount of oxygen vacancies. On the other hand, thick STO layers exhibit undistorted octahedra while the LMO layers present reduced O octahedral distortions near the interfaces. These findings are discussed in view of the transport and magnetic differences found in previous studies. Key words: scanning transmission electron microscopy, electron energy-loss spectroscopy, complex oxides, spectrum imaging, manganites, oxygen octahedral tilts, oxygen vacancies
I NTRODUCTION Complex oxide heterostructures exhibit exciting new physics such as the 2D metallic state found in insulating/insulating LaAlO3/SrTiO3 interfaces (Ohtomo & Hwang, 2004; Zubko et al., 2011; Tsymbal et al., 2012). The physical behavior of these systems is directly related to the crystal and electronic structures present at the interfaces (Reyren et al., 2007; GarciaBarriocanal et al., 2008), and the possibility of tailoring their electronic properties has generated great interest (Thiel et al., 2006; Caviglia et al., 2008; Cen et al., 2009; Rivera-Calzada et al., 2011). Subtle distortions from the perfect cubic ABO3 perovskite structure change the behavior of these materials (Wang et al., 2003), more specifically: deformations in the BO6 oxygen octahedron around the cations and collective tilts of the octahedral network play an important role in the electronic functionalities of perovskite materials (Jia et al., 2009; Berger et al., 2011; Kim et al., 2013). Aberration corrected scanning transmission electron microscopy (STEM) combined with electron energy-loss spectroscopy (EELS) is a powerful tool to study both the electronic and the structural properties of these systems (Varela et al., 2012). Thanks to the high spatial resolution achieved with the convergent scanning probe, it is possible to obtain atomic resolution spectrum images (Bosman et al., 2007) and to simultaneously study chemical and structural information in real space with atomic resolution (Varela et al., 2009). Received September 29, 2013; accepted March 21, 2014 *Corresponding author. [email protected]
STEM-EELS holds the key to understanding a number of further aspects of oxide interfaces that remain unclear. Among others, the recent study of the interface between the Mott insulator LaMnO3 (LMO) and the band insulator SrTiO3 (STO) (Garcia-Barriocanal et al., 2010). The physical properties of the system are related to structural changes driven by epitaxial mismatch, which is controlled by the relative thickness ratio between these layers in a superlattice heterostructure. In our previous work, we found that by varying the STO spacer thickness, and hence the epitaxial strain, the electronic properties of the manganite layers can be tuned. When STO/LMO superlattices with ultrathin STO are grown, the LMO is relaxed and displays a ferromagnetic and conducting state. In contrast, LMO layers grown on top of thicker STO are partially strained, insulating, and display depressed ferromagnetism. Intriguingly, the oxidation state of the Mn, as measured by EELS, is in all cases close to + 3. This finding suggests that the crystal structure changes due to epitaxial strain in LaMnO3, affecting the Jahn-Teller distortion or the oxygen octahedral rotations, may be coupled with magnetic ordering (Lee et al., 2013). These structural variations may explain the electronic and magnetic properties of these samples. Here we will examine an important aspect of the structure of these interfaces: what is the O sublattice behavior? Since the transition metal-to-oxygen bonds control the electronic properties of these oxides, we aim to shed some light on how the details of the structural reconstruction induced by epitaxial strain at the interface can have an effect on electronic coupling. To do so, we use aberration-corrected STEM and EELS to study the samples
2
Gabriel Sanchez-Santolino et al.
at the atomic level, and we focus on the structural differences at the interfaces between LMO and STO layers in superlattices with very different thickness ratios.
200 kV. The cross-section specimens were prepared by conventional mechanical grinding and polishing and Ar ion milling.
M ATERIALS AND M ETHODS
RESULTS
The samples are epitaxial LMO/STO superlattices grown on top of STO (100) substrates with a high pressure, hightemperature (810°C) sputtering deposition system (Varela et al., 1999). Two superlattices with different nominal thickness ratios were studied: LMO17/STO2 and LMO17/ STO12 unit cells (u.c.), eight bilayer repetitions each. Aberration-corrected STEM Z-contrast images and spectrum images were obtained using a Nion UltraSTEM100 (Krivanek et al., 2008) equipped with a 5th-order corrector and a Gatan Enfina EELS spectrometer and operated at 60 and 100 kV. The probe forming aperture was ~30 mrad while the EELS collection angle was 35 mrad. For spectral imaging, the electron beam was scanned along the region of interest and an EEL spectrum was acquired for every pixel, along with the simultaneous annular dark field (ADF) signal. Random noise in the EEL spectrum images was removed using principal-component analysis (Bosman et al., 2006). Annular bright field (ABF) images were obtained in an aberration-corrected Nion UltraSTEM 200 operated at
In order to image the oxygen sublattice under the electron microscope, cross-sectional samples were prepared in the pseudo-cubic [110] orientation (we will use the pseudo-cubic notation from now on), where the oxygen “ripple” due to the octahedral rotations is easily resolved. Figure 1 shows both low and high magnification Z-contrast images of the superlattices studied. Figure 1a shows a (LMO17u.c./STO12u.c.) × 8 sample and Figure 1b shows a (LMO17u.c./STO2u.c.) × 8 superlattice. The samples are epitaxial and have a very high structural quality. The interfaces are coherent and defectfree, but the contrast in the images looks somewhat smeared across the interfaces, which could be due to beam broadening through the sample thickness but also to major chemical interdiffusion. In order to examine this possibility, Ti, Mn, and La elemental maps obtained from atomic resolution EEL spectrum images across a STO (left)/LMO (right) interface are displayed in Figure 2. An upper estimate for the length scale of any chemical intermixing can be obtained by looking at the width of the 75–25% drop of the respective signals
AND
DISCUSSION
Figure 1. a: Low and high magnification high angle annular dark field (HAADF) images of a (LMO17 u.c./STO12 u.c.) × 8 superlattice grown on STO(001), obtained at 100 kV. b: Low and high resolution HAADF images of a (LMO17 u.c./STO2 u.c.) × 8 superlattice grown on STO(001) at 100 kV. STO layers exhibit a darker contrast, while LMO layers appear brighter. Yellow arrows mark the interface planes.
O Octahedral Distortions in LMO/STO Superlattices
Figure 2. Atomic resolution EEL spectrum image of a LMO (right)/STO (left) interface. A vertical yellow dashed line marks the interface position. The top panel shows the simultaneously acquired Z-contrast signal. Some drift is visible. Ti (blue), Mn (green), and La (red) elemental maps have been produced by fitting the raw EELS data to reference spectra using a multiple linear least squares fit (Longo et al., 2012). The bottom panel shows the laterally averaged La, Mn, and Ti signals (same color code), showing an interface width of one perovskite block (marked with a yellow rectangle for clarity). Data acquired at 100 kV.
across the interface (highlighted with a yellow rectangle for the Ti L2,3 profile in the bottom panel) (Browning et al., 1993). For all Ti, Mn, and La profiles the interface width is approximately one perovskite block. This finding points to a relatively sharp interface, which is consistent with our previous work (Garcia-Barriocanal et al., 2010). Both STO and LMO layers are flat and continuous over long lateral distances. Figure 3a shows an ABF image of a crushed bulk LMO sample down the pseudo-cubic [110] direction. Figure 3b shows a model of the LMO structure from neutron scattering data (Norby et al., 1995). We have labeled the O atoms in the LaO plane as O1, while O2 atoms will be the O atoms in the MnO2 plane (both planes are marked with black dashed lines). For clarity a yellow dashed line highlights the ripple within the O2 plane due to the octahedral tilts. The arrows placed in the model highlight the noticeable difference in the distances along the vertical (y) direction between O2 atoms in adjacent planes, due to the octahedral rotations. The ripple can be quantified from the images by assigning a pair of coordinates (xi, yi) to each O column within every O2 plane.
3
To obtain these values, we map the oxygen atomic positions from the ABF image with an iterative process locating the center-of-mass of every column. Through this procedure, adjacent O atoms within a given O2 plane exhibit a relative vertical offset of Δy = (yi + 1 − yi). The upper inset in Figure 3a shows the calculated Δy differences for an area of the ABF image. Each pixel corresponds to 1 u.c., and the crystal structure is drawn on the Δy image for clarity. The results exhibit a familiar “checkerboard” pattern reported in other oxides (Borisevich et al., 2010), which illustrates the antiphase behavior of the in-plane tilts in LMO. In the model used this difference is close to 0.6 Å, similar to the values obtained from the bulk LMO ABF image (around 0.5 ± 0.1 Å). Having used ABF imaging as a reference to show the ripple in the oxygen columns, we now use EELS atomic resolution imaging, which is a more powerful technique since it allows simultaneous studies of chemistry and electronic properties through fine structure analysis. EEL spectrum images of bulk LMO samples in the [110] direction from previous studies already showed this “ripple” distortion in the oxygen sub-lattice (Pennycook & Varela, 2011). Figure 4a shows a high angle annular dark field (HAADF) image of a superlattice with 17 u.c. of LMO and 2 u.c. of STO, both ferromagnetic and metallic (Garcia-Barriocanal et al., 2010). The layers are fully epitaxial and the interfaces are flat and coherent, as reported previously by GarciaBarriocanal et al. The yellow square marks the area where an EEL spectrum image was acquired with an exposure time of 0.05 s/pixel. In Figure 4b, we show the O K edge integrated signal, where the oxygen atomic lattice from the LMO/STO/ LMO interfaces region is resolved. Some spatial drift is present. It is worth mentioning that only the pure O columns in the MnO2 plane are visible. O atoms on the heavy La-O or Sr-O atomic columns are invisible due to dynamical diffraction (Varela et al., 2009). To measure the oxygen octahedral tilts, we mapped the atomic positions from the white square region in the O K edge map with the same process used in the ABF image from Figure 3a. For an accurate calibration of the spectrum images, we used the ADF image acquired simultaneously with our EEL spectra. Figure 4c shows the map for the oxygen sublattice ripple within the O2 plane, the values measured fluctuate slightly around 0.35–0.55 Å with maximum absolute values close to 0.6 Å in the LMO layers, in good agreement with the ABF image analysis and the structure deduced from neutron scattering (Norby et al., 1995) in bulk LMO samples. We have averaged the absolute values from Figure 4c for each atomic plane in Figure 4d (error bars correspond to the standard deviation within each atomic plane) to show more clearly the behavior along the different layers: in both top and bottom LMO layers, the averaged ripple value (around 0.4 Å), is close to the bulk value. Interestingly, in the STO layer (region marked in red) instead of finding the flatness expected for a cubic material, we observe that the oxygen octahedra are also tilted, with an average ripple value of 0.2 Å. Next, we studied a superlattice with a quite distinct thickness ratio: 17 LMO u.c. and 12 STO u.c., which is still ferromagnetic but is also an insulating sample. In Figure 5 we
4
Gabriel Sanchez-Santolino et al.
Figure 3. a: ABF image of a bulk LaMnO3 crystal down the [110] direction. The upper inset reflects the oxygen octahedral tilt map. Each pixel corresponds to the difference in vertical coordinates (yi + 1 − yi) from a given O atom and the adjacent O first neighbor within the O2 plane, as shown in the magnified view in the lower inset. b: LaMnO3 crystal structure from neutron scattering data (Norby et al., 1995) showing La (green), Mn (blue), and O (red) atom positions. Black dashed lines show the different oxygen planes (labeled O1 and O2). A yellow line highlights the ripple in the O2 plane, for visual clarity. Data acquired at 200 kV.
show the analysis from a STO/LMO interface region. In the HAADF image of Figure 5a, STO and LMO layers are clearly distinguished thanks to the Z-contrast, where the STO layer appears darker and the LMO layers are brighter; the yellow square indicates the acquisition region for the EEL spectrum image. Figure 5b shows the O K edge integrated intensity map from the STO/LMO interface, which is marked with a red dashed line. The spectrum image was acquired with 0.05 s/pixel acquisition time and some spatial drift is observed. In Figure 5c we show the calculated ripple values from the oxygen map in Figure 5b. In the STO layer, there is no appreciable contrast between different unit cells (pixels). However, the “checkerboard” pattern, associated with tilted octahedra, is recovered again in the LMO layer. The average ripple (resulting from the octahedral tilts) is plotted in Figure 5d. This figure confirms that towards the middle of the LMO layer (top) the structure is bulk-like, with values similar to those in the LMO17/STO2 superlattice. However, the O ripple values decrease to around 0.2 Å within the four LMO planes right on top of the thick STO layer (shadowed in red), indicating a depression of the octahedral tilts. This change could be related to the difference in transport and magnetic properties of these samples. No distortion is found in the STO layer (within the noise), as one would expect in a cubic system. It is worth noting that EELS imaging allows quantification of the O concentration relative to the 3d metals (Mn, Ti) from the EELS data using the routines available in the Gatan
Digital Micrograph software. These routines may not be completely accurate in the presence of dechanneling, but we can comment on general trends and qualitative changes observed in relation to the structural distortions measured. The quantification of O concentration (relative to Ti/Mn), in LMO/STO superlattices with thick STO layers, results in a relatively flat O content across the interfaces (Fig. 5b). If anything, a slightly higher O concentration can be detected within the STO layers (of 1–2%, barely above the noise level). This finding is an artifact due to the reduced scattering to high angles in the lighter STO material (ensuing in a higher intensity going forward into the spectrometer than when the electron beam is on the heavier LMO layers). Interestingly, when the same quantification procedure is carried out on the ultrathin STO layers of the LMO17/STO2 superlattice (Fig. 4b), we detect a slight O depletion around 3% in the atomic plane in the middle of the STO layer. This finding is unlikely to be an artifact, since a reduced O signal is not to be expected in the lighter STO layer (actually, quite the contrary). The O depletion is very likely to be real and, furthermore, its value is probably underestimated here (again due to the reduced scattering and ADF signal on the STO layers). This deficit points to the presence of a small amount of O vacancies. These vacancies may help the system accommodate the relatively large octahedral tilts imposed by the proximity of LMO. Indeed, it has been reported that oxygen vacancies may induce oxygen octahedral rotations in STO (Choi et al., 2013); this result also agrees with the
O Octahedral Distortions in LMO/STO Superlattices
5
Figure 4. a: High angle annular dark field (HAADF) image of a LMO17/STO2 superlattice. The yellow square shows the region where the spectrum image was acquired. b: Oxygen K edge map obtained from the area marked in (a) showing the STO thin layer, obtained through integration of the edge signal after background subtraction using a power law. The interfaces are marked with red lines. Some spatial drift is present. A structural model is shown (Mn columns in blue, La in green, O atoms in red). The oxygen relative composition profile is presented on the right end, in a matching scale. This concentration has been calculated relative to the 3d metals (Mn and Ti) using the Digital Micrograph quantification routine; hence, it increases on the La/Sr-O planes (where the 3d metal signal decreases). c: Oxygen octahedral tilt map in Ångströms, from the area marked in (b). Each pixel corresponds to the difference in vertical displacements from a given O atom and the adjacent O atom within the O2 plane. The red shadowed region corresponds to the STO layer. d: Averaged oxygen distortion values from (c) along the LMO/STO/LMO layers. A red rectangle marks the location of the STO layer. Data acquired at 60 kV. PCA was employed to remove random noise. A spectral-spatial scaling was applied. Thirteen principal components were used for reconstruction.
findings from Garcia-Barriocanal et al. (2010) where a reduction of the Ti oxidation state is measured in the LMO/ STO superlattices with thinner STO spacers, which could imply the presence of oxygen vacancies. We can also discuss the different oxygen octahedral tilt maps obtained by EELS in our samples in connection with their electronic properties. From the results in GarciaBarriocanal et al. (2010), the samples have completely different magnetic and electrical properties: the superlattice with the thinner STO spacer (2 u.c.) is conducting and strongly ferromagnetic, while the sample with thicker STO layers (12 u.c.) is insulating and shows a reduced magnetic moment of 0.5 μB/Mn atom. Conductivity in the superlattices with ultrathin STO layers may be related to O vacancies and also the presence of “charge leakage” in these symmetric systems (Garcia-Barriocanal et al., 2010).
Results from the EEL spectrum imaging analysis also point to a possible relation between the oxygen octahedral rotations and the strain in the layers. In the LMO17/STO2 superlattice, the LMO layers (with a relaxed structural state) exhibit pronounced, bulk-like octahedral tilts. On the other hand, in the LMO17/STO12 superlattice, the LMO atomic planes closer to the interface show smaller octahedral tilts, suggesting strain related structural changes induced by the thicker STO spacers. As discussed in previous studies, the relaxed LMO thin films are found to be metallic and ferromagnetic (Smadici et al., 2007; May et al., 2009), while strain may drive a phase transition to an insulating state with residual ferromagnetism. However, a complete explanation of the link between the observed structural changes and the macroscopic physical properties would need complex theoretical calculations and will be the object of future work.
6
Gabriel Sanchez-Santolino et al.
Figure 5. a: High angle annular dark field (HAADF) image of an interface in a LMO17/STO12 superlattice. The yellow square shows the region where the spectrum image was acquired. b: Oxygen K edge map obtained from the area marked in (a) showing the STO/LMO interface, marked with a red line. A structural model is shown (Mn columns in blue, La in green, O atoms in red). The right panel shows the O concentration relative to the 3d metal content averaged across the image, in a matching scale. c: Oxygen octahedral tilt map in Ångströms extracted from the O K edge map in (b). Each pixel corresponds to the difference in vertical displacements between any given O atom and the adjacent O atom within the O2 plane. d: Averaged oxygen distortion values from (c) along the STO/LMO interface showing a decreased octahedral rotation in the first four planes of the LMO layer (marked in red). Data acquired at 60 kV. PCA was employed to remove random noise. A spectral-spatial scaling was applied. Twelve principal components were used for reconstruction.
CONCLUSIONS By means of aberration-corrected STEM-EELS we have studied two LMO/STO superlattices with different thickness ratios (17/2 and 17/12), which present different transport and magnetic properties. Analysis of spectrum images reveals a clear difference in lattice distortions and oxygen octahedral tilt behavior in both samples. The thin STO superlattice, which is conducting, presents “bulk-like” octahedral tilts in the LMO layers and also a noticeable ripple in the thin STO layer, which could be related to the presence of a small amount (≥3%) of oxygen vacancies. On the other hand, the insulating sample with thicker STO spacers shows no oxygen ripple in the STO layers but we detect an interfacial region of depressed octahedral tilts in the LMO layers, which may play a role in the different magnetic and transport behaviors in these samples.
ACKNOWLEDGMENTS The authors thank Masashi Watanabe for the Digital Micrograph PCA plug-in, Andrew Lupini for the atom
mapping scripts, and J. Luck for help with specimen preparation. Research at ORNL (SJP and MV) was supported by the US Department of Energy, Basic Energy Sciences, Materials Sciences and Engineering Division. Research at UCM was supported by Spanish MICINN through grants MAT2011-27470-C02 and Consolider Ingenio 2010 CSD2009-00013 (Imagine), by CAM through grant S2009/ MAT-1756 (Phama) and by the ERC starting Investigator Award, grant #239739 STEMOX
REFERENCES BERGER, R.F., FENNIE, C.J. & NEATON, J.B. (2011). Band gap and edge engineering via ferroic distortion and anisotropic strain: The case of SrTiO3. Phys Rev Lett 107, 146804. BORISEVICH, A.Y., CHANG, H.J., HUIJBEN, M., OXLEY, M.P., OKAMOTO, S., NIRANJAN, M.K., BURTON, J.D., TSYMBAL, E.Y., CHU, Y.H., YU, P., RAMESH, R., KALININ, S.V. & PENNYCOOK, S.J. (2010). Suppression of octahedral tilts and associated changes in electronic properties at epitaxial oxide heterostructure interfaces. Phys Rev Lett 105, 087204.
O Octahedral Distortions in LMO/STO Superlattices BOSMAN, M., KEAST, V., GARCÍA-MUÑOZ, J.L., D’ALFONSO, A.J., FINDLAY, S.D. & ALLEN, L.J. (2007). Two-dimensional mapping of chemical information at atomic resolution. Phys Rev Lett 99, 086102. BOSMAN, M., WATANABE, M., ALEXANDER, D.T.L. & KEAST, V.J. (2006). Mapping chemical and bonding information using multivariate analysis of electron energy-loss spectrum images. Ultramicroscopy 106, 1024–1032. BROWNING, N.D., CHISHOLM, M.C. & PENNYCOOK, S.J. (1993). Atomic resolution chemical analysis using a scanning transmission electron microscope. Nature 366, 143–146. CAVIGLIA, A.D., GARIGLIO, S., REYREN, N., JACCARD, D., SCHNEIDER, T., GABAY, M., THIEL, S., HAMMERL, G., MANNHART, J. & TRISCONE, J.-M. (2008). Electric field control of the LaAlO3/SrTiO3 interface ground state. Nature 456, 624–627. CEN, C., THIEL, S., MANNHART, J. & LEVY, J. (2009). Oxide nanoelectronics on demand. Science 323, 1026–1030. CHOI, M., OBA, F., KUMAGAI, Y. & TANAKA, I. (2013). Anti‐ ferrodistortive‐Like oxygen‐octahedron rotation induced by the oxygen vacancy in cubic SrTiO3. Adv Mater 25, 86–90. GARCIA-BARRIOCANAL, J., BRUNO, F.Y., RIVERA-CALZADA, A., SEFRIOUI, Z., NEMES, N.M., GARCIA-HERNÁNDEZ, M., RUBIO-ZUAZO, J., CASTRO, G. R., VARELA, M., PENNYCOOK, S.J., LEON, C. & SANTAMARIA, J. (2010). “Charge leakage” at LaMnO3/SrTiO3 interfaces. Adv Mater 22, 627–632. GARCIA-BARRIOCANAL, J., RIVERA-CALZADA, A., VARELA, M., SEFRIOUI, Z., IBORRA, E., LEON, C., PENNYCOOK, S.J. & SANTAMARIA, J. (2008). Colossal ionic conductivity at interfaces of epitaxial ZrO2:Y2O3/ SrTiO3 heterostructures. Science 321, 676–680. JIA, C.L., MI, S.B., FALEY, M., POPPE, U., SCHUBERT, J. & URBAN, K. (2009). Oxygen octahedron reconstruction in the SrTiO3/ LaAlO3 heterointerfaces investigated using aberration-corrected ultrahigh-resolution transmission electron microscopy. Phys Rev B 79, 081405. KIM, Y.-M., KUMAR, A., HATT, A., MOROZOVSKA, A.N., TSELEV, A., BIEGALSKI, M.D., IVANOV, I., ELISEEV, E.A., PENNYCOOK, S.J., RONDINELLI, J.M., KALININ, S.V. & BORISEVICH, A.Y. (2013). Interplay of octahedral tilts and polar order in BiFeO3 films. Adv Mater 25, 2497–2504. KRIVANEK, O.L., CORBIN, G.J., DELLBY, N., ELSTON, B.F., KEYSE, R.J., MURFITT, M.F., OWN, C.S., SZILAGYI, Z.S. & WOODRUFF, J.W. (2008). An electron microscope for the aberration-corrected era. Ultramicroscopy 108, 179–195. LEE, J.H., DELANEY, K.T., BOUSQUET, E., SPALDIN, N.A. & RABE, K.M. (2013). Strong coupling of Jahn-Teller distortion to oxygenoctahedron rotation and functional properties in epitaxiallystrained orthorhombic LaMnO3. arXiv:1307.3347 [cond-mat. str-el], 1–7. LONGO, P., THOMAS, P.J. & TWESTEN, R.D. (2012). Atomic-level EELS mapping using high-energy edges in DualEELS mode. Microsc Today 20, 30–36. http://www.gatan.com/files/PDF/ Use_of_MLLS_fitting_EELS_FL.pdf
7
MAY, S.J., RYAN, P.J., ROBERTSON, J.L., KIM, J.-W., SANTOS, T.S., KARAPETROVA, E., ZARESTKY, J.L., ZHAI, X., TE VELTHUIS, S.G.E., ECKSTEIN, J.N., BADER, S.D. & BHATTACHARYA, A. (2009). Enhanced ordering temperatures in antiferromagnetic manganite superlattices. Nat Mater 8, 892–897. NORBY, P., KROGH ANDERSEN, I.G., KROGH ANDERSEN, E. & ANDERSEN, N.H. (1995). The crystal structure of lanthanum manganate(lll), LaMnO3 at room temperature and at 1273 K under N2. J Solid State Chem 119, 191–196. OHTOMO, A. & HWANG, H.Y. (2004). A high-mobility electron gas at the LaAlO3/SrTiO3 heterointerface. Nature 427, 423–426. PENNYCOOK, S.J. & VARELA, M. (2011). New views of materials through aberration-corrected scanning transmission electron microscopy. J Electron Microsc 60(Suppl 1), S213–S223. REYREN, N., THIEL, S., CAVIGLIA, A.D., KOURKOUTIS, L.F., HAMMERL, G., RICHTER, C., SCHNEIDER, C. W., W., KOPP, T., RÜETSCHI, A.-S., JACCARD, D., GABAY, M., MULLER, D.A., TRISCONE, J.-M. & MANNHART, J. (2007). Superconducting interfaces between insulating oxides. Science 317, 1196–1199. RIVERA-CALZADA, A., DIAZ-GUILLEN, M.R., DURA, O.J., SANCHEZSANTOLINO, G., PENNYCOOK, T.J., SCHMIDT, R., BRUNO, F.Y., GARCIA-BARRIOCANAL, J., SEFRIOUI, Z., NEMES, N.M., GARCIAHERNANDEZ, M., VARELA, M., LEON, C., PANTELIDES, S.T., PENNYCOOK, S.J. & SANTAMARIA, J. (2011). Tailoring interface structure in highly strained YSZ/STO heterostructures. Adv Mater 23, 5268–5274. SMADICI, Ş., ABBAMONTE, P., BHATTACHARYA, A., ZHAI, X., JIANG, B., RUSYDI, A., ECKSTEIN, J., BADER, S. & ZUO, J.M. (2007). Electronic reconstruction at SrMnO3-LaMnO3 superlattice interfaces. Phys Rev Lett 99, 196404. THIEL, S., HAMMERL, G., SCHMEHL, A., SCHNEIDER, C.W. & MANNHART, J. (2006). Tunable quasi-two-dimensional electron gases in oxide heterostructures. Science 313, 1942–1945. TSYMBAL, E.Y., DAGOTTO, E.R., EOM, C.-B. & RAMESH, R. (2012). Multifunctional Oxide Heterostructures. Oxford, UK: Oxford University Press. VARELA, M., GAZQUEZ, J. & PENNYCOOK, S.J. (2012). STEM-EELS imaging of complex oxides and interfaces. MRS Bull 37, 29–35. VARELA, M., OXLEY, M., LUO, W., TAO, J., WATANABE, M., LUPINI, A., PANTELIDES, S. & PENNYCOOK, S. (2009). Atomic-resolution imaging of oxidation states in manganites. Phys Rev B 79, 085117. VARELA, M., SEFRIOUI, Z. & ARIAS, D. (1999). Intracell changes in epitaxially strained YBa2Cu3O7-x ultrathin layers in YBa2Cu3O7-x/ PrBa2Cu3O7 superlattices. Phys Rev Lett 83, 3936–3939. WANG, J., NEATON, J.B., ZHENG, H., NAGARAJAN, V., OGALE, S.B., LIU, B., VIEHLAND, D., VAITHYANATHAN, V., SCHLOM, D.G., WAGHMARE, U.V., SPALDIN, N.A., RABE, K.M., WUTTIG, M. & RAMESH, R. (2003). Epitaxial BiFeO3 multiferroic thin film heterostructures. Science 299, 1719–1722. ZUBKO, P., GARIGLIO, S., GABAY, M., GHOSEZ, P. & TRISCONE, J.-M. (2011). Interface physics in complex oxide heterostructures. Ann Rev Cond Mat Phys 2, 141–165.
© MICROSCOPY SOCIETY OF AMERICA 2014
Oxygen Octahedral Distortions in LaMO3/SrTiO3 Superlattices Gabriel Sanchez-Santolino,1,2 Mariona Cabero,1,2 Maria Varela,1,2,3,* Javier Garcia-Barriocanal,1 Carlos Leon,1 Stephen J. Pennycook,4 and Jacobo Santamaria1 1
GFMC, Departamento de Fisica Aplicada III, Universidad Complutense de Madrid, 28040 Madrid, Spain Instituto Pluridisciplinar, Universidad Complutense de Madrid, 28040 Madrid, Spain 3 Materials Science and Technology Division, Oak Ridge National Laboratory, Oak Ridge, TN 37831, USA 4 Department of Materials Science and Engineering, The University of Tennessee, Knoxville, TN 37996, USA 2
Abstract: In this work we study the interfaces between the Mott insulator LaMnO3 (LMO) and the band insulator SrTiO3 (STO) in epitaxially grown superlattices with different thickness ratios and different transport and magnetic behaviors. Using atomic resolution electron energy-loss spectral imaging, we analyze simultaneously the structural and chemical properties of these interfaces. We find changes in the oxygen octahedral tilts within the LaMnO3 layers when the thickness ratio between the manganite and the titanate layers is varied. Superlattices with thick LMO and ultrathin STO layers present unexpected octahedral tilts in the STO, along with a small amount of oxygen vacancies. On the other hand, thick STO layers exhibit undistorted octahedra while the LMO layers present reduced O octahedral distortions near the interfaces. These findings are discussed in view of the transport and magnetic differences found in previous studies. Key words: scanning transmission electron microscopy, electron energy-loss spectroscopy, complex oxides, spectrum imaging, manganites, oxygen octahedral tilts, oxygen vacancies
I NTRODUCTION Complex oxide heterostructures exhibit exciting new physics such as the 2D metallic state found in insulating/insulating LaAlO3/SrTiO3 interfaces (Ohtomo & Hwang, 2004; Zubko et al., 2011; Tsymbal et al., 2012). The physical behavior of these systems is directly related to the crystal and electronic structures present at the interfaces (Reyren et al., 2007; GarciaBarriocanal et al., 2008), and the possibility of tailoring their electronic properties has generated great interest (Thiel et al., 2006; Caviglia et al., 2008; Cen et al., 2009; Rivera-Calzada et al., 2011). Subtle distortions from the perfect cubic ABO3 perovskite structure change the behavior of these materials (Wang et al., 2003), more specifically: deformations in the BO6 oxygen octahedron around the cations and collective tilts of the octahedral network play an important role in the electronic functionalities of perovskite materials (Jia et al., 2009; Berger et al., 2011; Kim et al., 2013). Aberration corrected scanning transmission electron microscopy (STEM) combined with electron energy-loss spectroscopy (EELS) is a powerful tool to study both the electronic and the structural properties of these systems (Varela et al., 2012). Thanks to the high spatial resolution achieved with the convergent scanning probe, it is possible to obtain atomic resolution spectrum images (Bosman et al., 2007) and to simultaneously study chemical and structural information in real space with atomic resolution (Varela et al., 2009). Received September 29, 2013; accepted March 21, 2014 *Corresponding author. [email protected]
STEM-EELS holds the key to understanding a number of further aspects of oxide interfaces that remain unclear. Among others, the recent study of the interface between the Mott insulator LaMnO3 (LMO) and the band insulator SrTiO3 (STO) (Garcia-Barriocanal et al., 2010). The physical properties of the system are related to structural changes driven by epitaxial mismatch, which is controlled by the relative thickness ratio between these layers in a superlattice heterostructure. In our previous work, we found that by varying the STO spacer thickness, and hence the epitaxial strain, the electronic properties of the manganite layers can be tuned. When STO/LMO superlattices with ultrathin STO are grown, the LMO is relaxed and displays a ferromagnetic and conducting state. In contrast, LMO layers grown on top of thicker STO are partially strained, insulating, and display depressed ferromagnetism. Intriguingly, the oxidation state of the Mn, as measured by EELS, is in all cases close to + 3. This finding suggests that the crystal structure changes due to epitaxial strain in LaMnO3, affecting the Jahn-Teller distortion or the oxygen octahedral rotations, may be coupled with magnetic ordering (Lee et al., 2013). These structural variations may explain the electronic and magnetic properties of these samples. Here we will examine an important aspect of the structure of these interfaces: what is the O sublattice behavior? Since the transition metal-to-oxygen bonds control the electronic properties of these oxides, we aim to shed some light on how the details of the structural reconstruction induced by epitaxial strain at the interface can have an effect on electronic coupling. To do so, we use aberration-corrected STEM and EELS to study the samples
2
Gabriel Sanchez-Santolino et al.
at the atomic level, and we focus on the structural differences at the interfaces between LMO and STO layers in superlattices with very different thickness ratios.
200 kV. The cross-section specimens were prepared by conventional mechanical grinding and polishing and Ar ion milling.
M ATERIALS AND M ETHODS
RESULTS
The samples are epitaxial LMO/STO superlattices grown on top of STO (100) substrates with a high pressure, hightemperature (810°C) sputtering deposition system (Varela et al., 1999). Two superlattices with different nominal thickness ratios were studied: LMO17/STO2 and LMO17/ STO12 unit cells (u.c.), eight bilayer repetitions each. Aberration-corrected STEM Z-contrast images and spectrum images were obtained using a Nion UltraSTEM100 (Krivanek et al., 2008) equipped with a 5th-order corrector and a Gatan Enfina EELS spectrometer and operated at 60 and 100 kV. The probe forming aperture was ~30 mrad while the EELS collection angle was 35 mrad. For spectral imaging, the electron beam was scanned along the region of interest and an EEL spectrum was acquired for every pixel, along with the simultaneous annular dark field (ADF) signal. Random noise in the EEL spectrum images was removed using principal-component analysis (Bosman et al., 2006). Annular bright field (ABF) images were obtained in an aberration-corrected Nion UltraSTEM 200 operated at
In order to image the oxygen sublattice under the electron microscope, cross-sectional samples were prepared in the pseudo-cubic [110] orientation (we will use the pseudo-cubic notation from now on), where the oxygen “ripple” due to the octahedral rotations is easily resolved. Figure 1 shows both low and high magnification Z-contrast images of the superlattices studied. Figure 1a shows a (LMO17u.c./STO12u.c.) × 8 sample and Figure 1b shows a (LMO17u.c./STO2u.c.) × 8 superlattice. The samples are epitaxial and have a very high structural quality. The interfaces are coherent and defectfree, but the contrast in the images looks somewhat smeared across the interfaces, which could be due to beam broadening through the sample thickness but also to major chemical interdiffusion. In order to examine this possibility, Ti, Mn, and La elemental maps obtained from atomic resolution EEL spectrum images across a STO (left)/LMO (right) interface are displayed in Figure 2. An upper estimate for the length scale of any chemical intermixing can be obtained by looking at the width of the 75–25% drop of the respective signals
AND
DISCUSSION
Figure 1. a: Low and high magnification high angle annular dark field (HAADF) images of a (LMO17 u.c./STO12 u.c.) × 8 superlattice grown on STO(001), obtained at 100 kV. b: Low and high resolution HAADF images of a (LMO17 u.c./STO2 u.c.) × 8 superlattice grown on STO(001) at 100 kV. STO layers exhibit a darker contrast, while LMO layers appear brighter. Yellow arrows mark the interface planes.
O Octahedral Distortions in LMO/STO Superlattices
Figure 2. Atomic resolution EEL spectrum image of a LMO (right)/STO (left) interface. A vertical yellow dashed line marks the interface position. The top panel shows the simultaneously acquired Z-contrast signal. Some drift is visible. Ti (blue), Mn (green), and La (red) elemental maps have been produced by fitting the raw EELS data to reference spectra using a multiple linear least squares fit (Longo et al., 2012). The bottom panel shows the laterally averaged La, Mn, and Ti signals (same color code), showing an interface width of one perovskite block (marked with a yellow rectangle for clarity). Data acquired at 100 kV.
across the interface (highlighted with a yellow rectangle for the Ti L2,3 profile in the bottom panel) (Browning et al., 1993). For all Ti, Mn, and La profiles the interface width is approximately one perovskite block. This finding points to a relatively sharp interface, which is consistent with our previous work (Garcia-Barriocanal et al., 2010). Both STO and LMO layers are flat and continuous over long lateral distances. Figure 3a shows an ABF image of a crushed bulk LMO sample down the pseudo-cubic [110] direction. Figure 3b shows a model of the LMO structure from neutron scattering data (Norby et al., 1995). We have labeled the O atoms in the LaO plane as O1, while O2 atoms will be the O atoms in the MnO2 plane (both planes are marked with black dashed lines). For clarity a yellow dashed line highlights the ripple within the O2 plane due to the octahedral tilts. The arrows placed in the model highlight the noticeable difference in the distances along the vertical (y) direction between O2 atoms in adjacent planes, due to the octahedral rotations. The ripple can be quantified from the images by assigning a pair of coordinates (xi, yi) to each O column within every O2 plane.
3
To obtain these values, we map the oxygen atomic positions from the ABF image with an iterative process locating the center-of-mass of every column. Through this procedure, adjacent O atoms within a given O2 plane exhibit a relative vertical offset of Δy = (yi + 1 − yi). The upper inset in Figure 3a shows the calculated Δy differences for an area of the ABF image. Each pixel corresponds to 1 u.c., and the crystal structure is drawn on the Δy image for clarity. The results exhibit a familiar “checkerboard” pattern reported in other oxides (Borisevich et al., 2010), which illustrates the antiphase behavior of the in-plane tilts in LMO. In the model used this difference is close to 0.6 Å, similar to the values obtained from the bulk LMO ABF image (around 0.5 ± 0.1 Å). Having used ABF imaging as a reference to show the ripple in the oxygen columns, we now use EELS atomic resolution imaging, which is a more powerful technique since it allows simultaneous studies of chemistry and electronic properties through fine structure analysis. EEL spectrum images of bulk LMO samples in the [110] direction from previous studies already showed this “ripple” distortion in the oxygen sub-lattice (Pennycook & Varela, 2011). Figure 4a shows a high angle annular dark field (HAADF) image of a superlattice with 17 u.c. of LMO and 2 u.c. of STO, both ferromagnetic and metallic (Garcia-Barriocanal et al., 2010). The layers are fully epitaxial and the interfaces are flat and coherent, as reported previously by GarciaBarriocanal et al. The yellow square marks the area where an EEL spectrum image was acquired with an exposure time of 0.05 s/pixel. In Figure 4b, we show the O K edge integrated signal, where the oxygen atomic lattice from the LMO/STO/ LMO interfaces region is resolved. Some spatial drift is present. It is worth mentioning that only the pure O columns in the MnO2 plane are visible. O atoms on the heavy La-O or Sr-O atomic columns are invisible due to dynamical diffraction (Varela et al., 2009). To measure the oxygen octahedral tilts, we mapped the atomic positions from the white square region in the O K edge map with the same process used in the ABF image from Figure 3a. For an accurate calibration of the spectrum images, we used the ADF image acquired simultaneously with our EEL spectra. Figure 4c shows the map for the oxygen sublattice ripple within the O2 plane, the values measured fluctuate slightly around 0.35–0.55 Å with maximum absolute values close to 0.6 Å in the LMO layers, in good agreement with the ABF image analysis and the structure deduced from neutron scattering (Norby et al., 1995) in bulk LMO samples. We have averaged the absolute values from Figure 4c for each atomic plane in Figure 4d (error bars correspond to the standard deviation within each atomic plane) to show more clearly the behavior along the different layers: in both top and bottom LMO layers, the averaged ripple value (around 0.4 Å), is close to the bulk value. Interestingly, in the STO layer (region marked in red) instead of finding the flatness expected for a cubic material, we observe that the oxygen octahedra are also tilted, with an average ripple value of 0.2 Å. Next, we studied a superlattice with a quite distinct thickness ratio: 17 LMO u.c. and 12 STO u.c., which is still ferromagnetic but is also an insulating sample. In Figure 5 we
4
Gabriel Sanchez-Santolino et al.
Figure 3. a: ABF image of a bulk LaMnO3 crystal down the [110] direction. The upper inset reflects the oxygen octahedral tilt map. Each pixel corresponds to the difference in vertical coordinates (yi + 1 − yi) from a given O atom and the adjacent O first neighbor within the O2 plane, as shown in the magnified view in the lower inset. b: LaMnO3 crystal structure from neutron scattering data (Norby et al., 1995) showing La (green), Mn (blue), and O (red) atom positions. Black dashed lines show the different oxygen planes (labeled O1 and O2). A yellow line highlights the ripple in the O2 plane, for visual clarity. Data acquired at 200 kV.
show the analysis from a STO/LMO interface region. In the HAADF image of Figure 5a, STO and LMO layers are clearly distinguished thanks to the Z-contrast, where the STO layer appears darker and the LMO layers are brighter; the yellow square indicates the acquisition region for the EEL spectrum image. Figure 5b shows the O K edge integrated intensity map from the STO/LMO interface, which is marked with a red dashed line. The spectrum image was acquired with 0.05 s/pixel acquisition time and some spatial drift is observed. In Figure 5c we show the calculated ripple values from the oxygen map in Figure 5b. In the STO layer, there is no appreciable contrast between different unit cells (pixels). However, the “checkerboard” pattern, associated with tilted octahedra, is recovered again in the LMO layer. The average ripple (resulting from the octahedral tilts) is plotted in Figure 5d. This figure confirms that towards the middle of the LMO layer (top) the structure is bulk-like, with values similar to those in the LMO17/STO2 superlattice. However, the O ripple values decrease to around 0.2 Å within the four LMO planes right on top of the thick STO layer (shadowed in red), indicating a depression of the octahedral tilts. This change could be related to the difference in transport and magnetic properties of these samples. No distortion is found in the STO layer (within the noise), as one would expect in a cubic system. It is worth noting that EELS imaging allows quantification of the O concentration relative to the 3d metals (Mn, Ti) from the EELS data using the routines available in the Gatan
Digital Micrograph software. These routines may not be completely accurate in the presence of dechanneling, but we can comment on general trends and qualitative changes observed in relation to the structural distortions measured. The quantification of O concentration (relative to Ti/Mn), in LMO/STO superlattices with thick STO layers, results in a relatively flat O content across the interfaces (Fig. 5b). If anything, a slightly higher O concentration can be detected within the STO layers (of 1–2%, barely above the noise level). This finding is an artifact due to the reduced scattering to high angles in the lighter STO material (ensuing in a higher intensity going forward into the spectrometer than when the electron beam is on the heavier LMO layers). Interestingly, when the same quantification procedure is carried out on the ultrathin STO layers of the LMO17/STO2 superlattice (Fig. 4b), we detect a slight O depletion around 3% in the atomic plane in the middle of the STO layer. This finding is unlikely to be an artifact, since a reduced O signal is not to be expected in the lighter STO layer (actually, quite the contrary). The O depletion is very likely to be real and, furthermore, its value is probably underestimated here (again due to the reduced scattering and ADF signal on the STO layers). This deficit points to the presence of a small amount of O vacancies. These vacancies may help the system accommodate the relatively large octahedral tilts imposed by the proximity of LMO. Indeed, it has been reported that oxygen vacancies may induce oxygen octahedral rotations in STO (Choi et al., 2013); this result also agrees with the
O Octahedral Distortions in LMO/STO Superlattices
5
Figure 4. a: High angle annular dark field (HAADF) image of a LMO17/STO2 superlattice. The yellow square shows the region where the spectrum image was acquired. b: Oxygen K edge map obtained from the area marked in (a) showing the STO thin layer, obtained through integration of the edge signal after background subtraction using a power law. The interfaces are marked with red lines. Some spatial drift is present. A structural model is shown (Mn columns in blue, La in green, O atoms in red). The oxygen relative composition profile is presented on the right end, in a matching scale. This concentration has been calculated relative to the 3d metals (Mn and Ti) using the Digital Micrograph quantification routine; hence, it increases on the La/Sr-O planes (where the 3d metal signal decreases). c: Oxygen octahedral tilt map in Ångströms, from the area marked in (b). Each pixel corresponds to the difference in vertical displacements from a given O atom and the adjacent O atom within the O2 plane. The red shadowed region corresponds to the STO layer. d: Averaged oxygen distortion values from (c) along the LMO/STO/LMO layers. A red rectangle marks the location of the STO layer. Data acquired at 60 kV. PCA was employed to remove random noise. A spectral-spatial scaling was applied. Thirteen principal components were used for reconstruction.
findings from Garcia-Barriocanal et al. (2010) where a reduction of the Ti oxidation state is measured in the LMO/ STO superlattices with thinner STO spacers, which could imply the presence of oxygen vacancies. We can also discuss the different oxygen octahedral tilt maps obtained by EELS in our samples in connection with their electronic properties. From the results in GarciaBarriocanal et al. (2010), the samples have completely different magnetic and electrical properties: the superlattice with the thinner STO spacer (2 u.c.) is conducting and strongly ferromagnetic, while the sample with thicker STO layers (12 u.c.) is insulating and shows a reduced magnetic moment of 0.5 μB/Mn atom. Conductivity in the superlattices with ultrathin STO layers may be related to O vacancies and also the presence of “charge leakage” in these symmetric systems (Garcia-Barriocanal et al., 2010).
Results from the EEL spectrum imaging analysis also point to a possible relation between the oxygen octahedral rotations and the strain in the layers. In the LMO17/STO2 superlattice, the LMO layers (with a relaxed structural state) exhibit pronounced, bulk-like octahedral tilts. On the other hand, in the LMO17/STO12 superlattice, the LMO atomic planes closer to the interface show smaller octahedral tilts, suggesting strain related structural changes induced by the thicker STO spacers. As discussed in previous studies, the relaxed LMO thin films are found to be metallic and ferromagnetic (Smadici et al., 2007; May et al., 2009), while strain may drive a phase transition to an insulating state with residual ferromagnetism. However, a complete explanation of the link between the observed structural changes and the macroscopic physical properties would need complex theoretical calculations and will be the object of future work.
6
Gabriel Sanchez-Santolino et al.
Figure 5. a: High angle annular dark field (HAADF) image of an interface in a LMO17/STO12 superlattice. The yellow square shows the region where the spectrum image was acquired. b: Oxygen K edge map obtained from the area marked in (a) showing the STO/LMO interface, marked with a red line. A structural model is shown (Mn columns in blue, La in green, O atoms in red). The right panel shows the O concentration relative to the 3d metal content averaged across the image, in a matching scale. c: Oxygen octahedral tilt map in Ångströms extracted from the O K edge map in (b). Each pixel corresponds to the difference in vertical displacements between any given O atom and the adjacent O atom within the O2 plane. d: Averaged oxygen distortion values from (c) along the STO/LMO interface showing a decreased octahedral rotation in the first four planes of the LMO layer (marked in red). Data acquired at 60 kV. PCA was employed to remove random noise. A spectral-spatial scaling was applied. Twelve principal components were used for reconstruction.
CONCLUSIONS By means of aberration-corrected STEM-EELS we have studied two LMO/STO superlattices with different thickness ratios (17/2 and 17/12), which present different transport and magnetic properties. Analysis of spectrum images reveals a clear difference in lattice distortions and oxygen octahedral tilt behavior in both samples. The thin STO superlattice, which is conducting, presents “bulk-like” octahedral tilts in the LMO layers and also a noticeable ripple in the thin STO layer, which could be related to the presence of a small amount (≥3%) of oxygen vacancies. On the other hand, the insulating sample with thicker STO spacers shows no oxygen ripple in the STO layers but we detect an interfacial region of depressed octahedral tilts in the LMO layers, which may play a role in the different magnetic and transport behaviors in these samples.
ACKNOWLEDGMENTS The authors thank Masashi Watanabe for the Digital Micrograph PCA plug-in, Andrew Lupini for the atom
mapping scripts, and J. Luck for help with specimen preparation. Research at ORNL (SJP and MV) was supported by the US Department of Energy, Basic Energy Sciences, Materials Sciences and Engineering Division. Research at UCM was supported by Spanish MICINN through grants MAT2011-27470-C02 and Consolider Ingenio 2010 CSD2009-00013 (Imagine), by CAM through grant S2009/ MAT-1756 (Phama) and by the ERC starting Investigator Award, grant #239739 STEMOX
REFERENCES BERGER, R.F., FENNIE, C.J. & NEATON, J.B. (2011). Band gap and edge engineering via ferroic distortion and anisotropic strain: The case of SrTiO3. Phys Rev Lett 107, 146804. BORISEVICH, A.Y., CHANG, H.J., HUIJBEN, M., OXLEY, M.P., OKAMOTO, S., NIRANJAN, M.K., BURTON, J.D., TSYMBAL, E.Y., CHU, Y.H., YU, P., RAMESH, R., KALININ, S.V. & PENNYCOOK, S.J. (2010). Suppression of octahedral tilts and associated changes in electronic properties at epitaxial oxide heterostructure interfaces. Phys Rev Lett 105, 087204.
O Octahedral Distortions in LMO/STO Superlattices BOSMAN, M., KEAST, V., GARCÍA-MUÑOZ, J.L., D’ALFONSO, A.J., FINDLAY, S.D. & ALLEN, L.J. (2007). Two-dimensional mapping of chemical information at atomic resolution. Phys Rev Lett 99, 086102. BOSMAN, M., WATANABE, M., ALEXANDER, D.T.L. & KEAST, V.J. (2006). Mapping chemical and bonding information using multivariate analysis of electron energy-loss spectrum images. Ultramicroscopy 106, 1024–1032. BROWNING, N.D., CHISHOLM, M.C. & PENNYCOOK, S.J. (1993). Atomic resolution chemical analysis using a scanning transmission electron microscope. Nature 366, 143–146. CAVIGLIA, A.D., GARIGLIO, S., REYREN, N., JACCARD, D., SCHNEIDER, T., GABAY, M., THIEL, S., HAMMERL, G., MANNHART, J. & TRISCONE, J.-M. (2008). Electric field control of the LaAlO3/SrTiO3 interface ground state. Nature 456, 624–627. CEN, C., THIEL, S., MANNHART, J. & LEVY, J. (2009). Oxide nanoelectronics on demand. Science 323, 1026–1030. CHOI, M., OBA, F., KUMAGAI, Y. & TANAKA, I. (2013). Anti‐ ferrodistortive‐Like oxygen‐octahedron rotation induced by the oxygen vacancy in cubic SrTiO3. Adv Mater 25, 86–90. GARCIA-BARRIOCANAL, J., BRUNO, F.Y., RIVERA-CALZADA, A., SEFRIOUI, Z., NEMES, N.M., GARCIA-HERNÁNDEZ, M., RUBIO-ZUAZO, J., CASTRO, G. R., VARELA, M., PENNYCOOK, S.J., LEON, C. & SANTAMARIA, J. (2010). “Charge leakage” at LaMnO3/SrTiO3 interfaces. Adv Mater 22, 627–632. GARCIA-BARRIOCANAL, J., RIVERA-CALZADA, A., VARELA, M., SEFRIOUI, Z., IBORRA, E., LEON, C., PENNYCOOK, S.J. & SANTAMARIA, J. (2008). Colossal ionic conductivity at interfaces of epitaxial ZrO2:Y2O3/ SrTiO3 heterostructures. Science 321, 676–680. JIA, C.L., MI, S.B., FALEY, M., POPPE, U., SCHUBERT, J. & URBAN, K. (2009). Oxygen octahedron reconstruction in the SrTiO3/ LaAlO3 heterointerfaces investigated using aberration-corrected ultrahigh-resolution transmission electron microscopy. Phys Rev B 79, 081405. KIM, Y.-M., KUMAR, A., HATT, A., MOROZOVSKA, A.N., TSELEV, A., BIEGALSKI, M.D., IVANOV, I., ELISEEV, E.A., PENNYCOOK, S.J., RONDINELLI, J.M., KALININ, S.V. & BORISEVICH, A.Y. (2013). Interplay of octahedral tilts and polar order in BiFeO3 films. Adv Mater 25, 2497–2504. KRIVANEK, O.L., CORBIN, G.J., DELLBY, N., ELSTON, B.F., KEYSE, R.J., MURFITT, M.F., OWN, C.S., SZILAGYI, Z.S. & WOODRUFF, J.W. (2008). An electron microscope for the aberration-corrected era. Ultramicroscopy 108, 179–195. LEE, J.H., DELANEY, K.T., BOUSQUET, E., SPALDIN, N.A. & RABE, K.M. (2013). Strong coupling of Jahn-Teller distortion to oxygenoctahedron rotation and functional properties in epitaxiallystrained orthorhombic LaMnO3. arXiv:1307.3347 [cond-mat. str-el], 1–7. LONGO, P., THOMAS, P.J. & TWESTEN, R.D. (2012). Atomic-level EELS mapping using high-energy edges in DualEELS mode. Microsc Today 20, 30–36. http://www.gatan.com/files/PDF/ Use_of_MLLS_fitting_EELS_FL.pdf
7
MAY, S.J., RYAN, P.J., ROBERTSON, J.L., KIM, J.-W., SANTOS, T.S., KARAPETROVA, E., ZARESTKY, J.L., ZHAI, X., TE VELTHUIS, S.G.E., ECKSTEIN, J.N., BADER, S.D. & BHATTACHARYA, A. (2009). Enhanced ordering temperatures in antiferromagnetic manganite superlattices. Nat Mater 8, 892–897. NORBY, P., KROGH ANDERSEN, I.G., KROGH ANDERSEN, E. & ANDERSEN, N.H. (1995). The crystal structure of lanthanum manganate(lll), LaMnO3 at room temperature and at 1273 K under N2. J Solid State Chem 119, 191–196. OHTOMO, A. & HWANG, H.Y. (2004). A high-mobility electron gas at the LaAlO3/SrTiO3 heterointerface. Nature 427, 423–426. PENNYCOOK, S.J. & VARELA, M. (2011). New views of materials through aberration-corrected scanning transmission electron microscopy. J Electron Microsc 60(Suppl 1), S213–S223. REYREN, N., THIEL, S., CAVIGLIA, A.D., KOURKOUTIS, L.F., HAMMERL, G., RICHTER, C., SCHNEIDER, C. W., W., KOPP, T., RÜETSCHI, A.-S., JACCARD, D., GABAY, M., MULLER, D.A., TRISCONE, J.-M. & MANNHART, J. (2007). Superconducting interfaces between insulating oxides. Science 317, 1196–1199. RIVERA-CALZADA, A., DIAZ-GUILLEN, M.R., DURA, O.J., SANCHEZSANTOLINO, G., PENNYCOOK, T.J., SCHMIDT, R., BRUNO, F.Y., GARCIA-BARRIOCANAL, J., SEFRIOUI, Z., NEMES, N.M., GARCIAHERNANDEZ, M., VARELA, M., LEON, C., PANTELIDES, S.T., PENNYCOOK, S.J. & SANTAMARIA, J. (2011). Tailoring interface structure in highly strained YSZ/STO heterostructures. Adv Mater 23, 5268–5274. SMADICI, Ş., ABBAMONTE, P., BHATTACHARYA, A., ZHAI, X., JIANG, B., RUSYDI, A., ECKSTEIN, J., BADER, S. & ZUO, J.M. (2007). Electronic reconstruction at SrMnO3-LaMnO3 superlattice interfaces. Phys Rev Lett 99, 196404. THIEL, S., HAMMERL, G., SCHMEHL, A., SCHNEIDER, C.W. & MANNHART, J. (2006). Tunable quasi-two-dimensional electron gases in oxide heterostructures. Science 313, 1942–1945. TSYMBAL, E.Y., DAGOTTO, E.R., EOM, C.-B. & RAMESH, R. (2012). Multifunctional Oxide Heterostructures. Oxford, UK: Oxford University Press. VARELA, M., GAZQUEZ, J. & PENNYCOOK, S.J. (2012). STEM-EELS imaging of complex oxides and interfaces. MRS Bull 37, 29–35. VARELA, M., OXLEY, M., LUO, W., TAO, J., WATANABE, M., LUPINI, A., PANTELIDES, S. & PENNYCOOK, S. (2009). Atomic-resolution imaging of oxidation states in manganites. Phys Rev B 79, 085117. VARELA, M., SEFRIOUI, Z. & ARIAS, D. (1999). Intracell changes in epitaxially strained YBa2Cu3O7-x ultrathin layers in YBa2Cu3O7-x/ PrBa2Cu3O7 superlattices. Phys Rev Lett 83, 3936–3939. WANG, J., NEATON, J.B., ZHENG, H., NAGARAJAN, V., OGALE, S.B., LIU, B., VIEHLAND, D., VAITHYANATHAN, V., SCHLOM, D.G., WAGHMARE, U.V., SPALDIN, N.A., RABE, K.M., WUTTIG, M. & RAMESH, R. (2003). Epitaxial BiFeO3 multiferroic thin film heterostructures. Science 299, 1719–1722. ZUBKO, P., GARIGLIO, S., GABAY, M., GHOSEZ, P. & TRISCONE, J.-M. (2011). Interface physics in complex oxide heterostructures. Ann Rev Cond Mat Phys 2, 141–165.
