The interplay between strain gradients and lattice defects at the phase boundaries in complex oxides

Abstract

Epitaxial growth of oxide thin films has lattice mismatches to the substrate. Unique physical phenomena or bulk properties emerge by lattice mismatch. The lattice mismatch is gradually alleviated as the oxide thin film becomes thicker. This phenomenon is called strain gradient. Strain gradient induces emergent physical properties of parent materials. Because the strain gradient induces internal fields, it affects the distribution of point defects. The rearrangement of point defects induced by the strain gradient causes changes in dielectric and conductivity properties. A strain gradient is a general physical property. Since the strain gradient is a physical phenomenon applicable to bent dielectrics, it is meaningful to contribute to the development of nanoelectronics by revealing the role of the strain gradient on point defects inside the dielectric. Despite the large lattice mismatch between bismuth ferrite oxide (BiFeO$_3$, $a$$_{\rm pc}$$\sim$4.014 $\textnormal{\AA}$) and lanthanum aluminum oxide substrate (LaAlO$_3$, $a_{\rm pc}$$\sim$3.789 $\textnormal{\AA}$), BiFeO$_3$ which possesses both ferroelectricity and antiferromagnetism at room temperature is successfully deposited on LaAlO$_3$ which stabilizes the structural phase of large elongation of $c$/$a$ ratio. Especially, if the BiFeO$_3$ deposition on LaAlO$_3$ is increased over 20 nm, mixed phases which have a large strain gradient and a mixture of more than two phases appear. If the lanthanum is substituted for the bismuth site, strain inhomogeneity is enhanced along with the generation of mixed phases, and point defects are aggregated in these areas, showing great changes in electrical properties. Impedance spectroscopy measurement, X-ray scattering, neutron scattering, and atomic force microscopy are exploited to characterize the cause of changes in electronic properties in La-substituted BiFeO$_3$. Also, strain gradient can be derived from the epitaxial growth of large lattice-mismatched materials. If we deposited strontium titanate oxide (SrTiO$_3$, $a_{\rm pc}$$\sim$3.905 $\textnormal{\AA}$) on dysprosium scandinate oxide substrate (DyScO$_3$, $a$$_{\rm pc}$$\sim$3.945 $\textnormal{\AA}$) which is known to have an anisotropic lattice parameter, SrTiO$_3$ exerts a tensile strain of 1.21\% along [100]$_{\rm O}$ and 1.09\% along [1\textnormal{$\bar{1}$}0]$_{\rm O}$ on DyScO$_3$ substrate. If the thickness increases without strain relaxation, DyScO$_3$ will not stand the tensile strain and form cracks along [1\textnormal{$\bar{1}$}0]$_{\rm O}$ direction. In the crack, defect accumulates near the crack structure which involves a large strain gradient. In this thesis, we lay the basis for the application of oxide thin films to nanoelectronics by investigating the emergence of unique physical phenomena through the control of the strain gradient of the oxide thin film.