The anomalous conductivity at the interface between nominally-insulating polar and non-polar materials, such as LaAlO3/SrTiO3 (LAO/STO), has sparked considerable debate within the oxide community and given rise to multiple hypotheses explaining both the conductivity and the resolution of the polar iscontinuity. A possibility that has been suggested is the unintentional doping of either the film or the substrate, as might be caused by imbalanced cation diffusion across the interface. For instance, there have been indications that the extent of he A- and B-site cation interdiffusion are uneven in the LAO/STO interfaces grown by some groups. The interface between similarly structured LaCrO3 (LCO) and STO is investigated to provide additional insight. Contrary to the case of LAO/STO, the LCO/STO system appears to be thoroughly insulating. The extent of the intermixing at the interface is examined using a combination of high resolution Rutherford backscattering (HRRBS) and cross-sectional electron microscopy, including aberration-corrected high resolution TEM, probe-corrected AADF-STEM, x-ray energy dispersive spectroscopy (EDS), electron energy loss spectroscopy (EELS), and chromaticaberration-corrected energy filtered TEM (EFTEM) imaging, both for a range of La/Cr stoichiometries and film thicknesses from 5-125 unit cells. Cross-sectional TEM/STEM provides a far more local measurement than RBS, illuminating the effects of interfacial non-uniformity, the presence of which appears to affect the average interface width, particularly for thicker LCO films. berrationcorrected HAADF-STEM enables the formation of a probe fine enough to sufficiently separate intensities from the A and B-sites, allowing a comparison of the diffusion widths for the two cation sites. LCO/STO is better suited to such analysis than, e.g., LAO/STO, as both the A- and B-site cations are heavier in the film than the substrate. HAADF-STEM analysis suggests roughly equal diffusion for both cation sites, with an interface width that generally increases with film thickness (or growth time). Several analysis methods are considered and,for instance, the full width at a tenth maximum (FWTM) for best-fit curves is for no film thickness less than a few unit cells A- and B-sites in. For all approaches, both the shapes and widths of the profiles are very similar, suggesting even diffusion of A- and B-sitecations, with EDS, EELS, and EFTEM confirming similar trends. The interface width trends are in rough qualitative agreement with HRRBS, but HRRBS is more sensitive to small concentrations La than HAADF-STEM, EELS, or EDS, and shows that there are long diffusion tails of the La into the STO. The results together suggest that while there is significant interdiffusion at this interface, there are nonetheless no conclusive indications that there is an imbalance in A- and B-site cation interdiffusion, as would be expected for insulating interfaces.
|Publication status||Published - 2012|
|Event||Materials Research Society 2012 Fall Meeting - Boston, MA, USA|
Duration: 25 Nov 2012 → 30 Nov 2012
|Conference||Materials Research Society 2012 Fall Meeting|
|City||Boston, MA, USA|
|Period||25/11/12 → 30/11/12|