UC Berkeley

It is historically proven that artificial heterostructures are of paramount importance for both fundamental research and technological application. One distinguishable example is superlattices and quantum-well heterostructures of conventional semiconductors (III-V). Several fundamental observations such as two-dimensional electron gas, quantum confinement effect, quantum Hall effect and fractional quantum Hall effect, were first realized in artificial heterostructures of conventional semiconductors. On the technological front, artificial heterostructures of conventional semiconductors were found to revolutionize optoelectronic and high-speed electronic industry. Given the capability of artificial heterostructures to enable new physical phenomena, with an appropriate choice of material system, these heterostructures hold the potential to uncover hidden physical phenomena and exotic phases, which are otherwise not observed in bulk systems. One choice of such a system could be a strongly correlated material where the presence of strong electronic correlations has been known to enable unique electronic and magnetic properties. Complex oxides, in particular, are known to exhibit a strong interaction between various degrees of freedom available such as spin, charge, orbital and lattice degree, and thus offer a suitable choice of material system. In literature, one can find plenty of examples from different areas of research such as superconductivity, magnetism, ferroelectricity and thermoelectricity, where artificial superlattices led to the observation of fundamentally different behavior, compared to bulk superlattice constituents. These artificial superlattices and heterostructures continue to be the most promising candidate for exploring new phenomena and enhancing physical properties in complex oxide material systems. In this dissertation, artificial superlattices of complex oxides were synthesized in a thin-film geometry to enable the observation of fundamentally new physical phenomena, compared to their bulk counterparts, in few selected areas of investigation.

First, I will present the experimental results on heat transport across superlattice structures composed of insulating perovskite oxides. The measured thermal conductivity of these artificial superlattices exhibited a unique phonon transport, i.e. coherent phonon transport, phenomenon that is extremely rare to observe in bulk form. The key element for enabling the coherent transport of phonon in these superlattice systems is the range of phonon wavelengths, which carry most of the heat in the material. The critical range being between 1-3 nanometers, the size of the system needs to reduce to this length scale in order to observe the effect of the wave nature of phonons on heat transport. Thus, superlattice structures offer an ideal candidate to search for this novel phenomenon. Several theoretical studies predicted the existence of these phenomena, but the experimental evidence of their existence remained largely absent or inconclusive. By synthesizing the superlattice of perovskite oxides, I’ve observed unambiguous evidence of the coherent transport of phonons at short-period superlattices of SrTiO3-CaTiO3 and BaTiO3-SrTiO3. In contrast to conventional heat conduction mechanisms, where phonon transport can be described by energy-carrying particles, the uniqueness of this phenomenon is highlighted by the fact that the wave aspect of phonons needs to be invoked to understand their transport behavior. After observing the wave nature of phonons in dictating heat conduction across periodic superlattices, other artificial heterostructures were studied to understand the nature of coherent phonon transport at the superlattice interfaces. Designed superlattice-like sequences, where the structural order can be controlled from a periodic sequence to a completely random sequence, were synthesized to further understand the role of the coherent scattering of phonons in thermal transport across superlattice structures.

Next, I will discuss the experimental observation of another unique phenomenon, which was enabled by artificial superlattices of complex oxides in a different subject of research, i.e. ferroelectricity. Several theoretical studies on nanostructured ferroelectric systems such as nanodisc, nanocomposites, superlattices etc., predicted the stabilization of novel ferroelectric ground states. A number of different topologies of electrical polarization such as vortices and skyrmions, were predicted in ferroelectric nanostructures, which showed a strong resemblance to spin topologies such as skyrmion, merons etc. found in magnetic systems. Experimental confirmation for the existence of these exotic polarization states, however, remained absent. By leveraging the competition among charge, orbital and lattice degrees of freedom in superlattices of a paraelectric (SrTiO3) and ferroelectric (PbTiO3) material, vortex-antivortex structures of electrical polarization were stabilized in ferroelectric (PbTiO3) layers of PbTiO3-SrTiO3 superlattice. Only for a narrow range of superlattice periods, the polarization vortices are stabilized with a balance between one gradient energy associated with the non-uniform polarization profile of the vortex structure and the other’s electrostatic and elastic energies associated with depolarization fields and epitaxial constraints from the substrate, respectively.

In the last section, I will describe unique and unusual phenomena associated with the existence of polarization vortices in paraelectric/ferroelectric superlattices. Superlattices of PbTiO3-SrTiO3 with varying periodicity showed a rich spectrum of characteristically different ferroelectric domains. Specifically, in the short-period regime, domain size was observed to evolve with a negative scaling coefficient, which is unusual for typical ferroelectrics. When the superlattice periodicity was increased, the average domain size decreased, suggesting an opposite behavior to the universal Kittel’s Law, where in the latter, an increase in ferroelectric thickness leads to an increase in the average domain size. Second, I will present intriguing results on the fundamental characteristics of polarization vortices as revealed from X-ray circular dichroism studies. The finite difference in absorption spectra of left-circular vs. right-circular polarized light from vortex structures, suggested that vortex-antivortex arrays are chiral in nature. The presence of chirality in polarization vortex structures is a characteristically different behavior compared to bulk ferroelectric systems where uniformly polarized regions are expected to exhibit a linear X-ray dichroism (i.e. a difference in the absorption spectra of linearly polarized light in a direction parallel and perpendicular to ferroelectric order). Lastly, a possibility of manipulating phonon dispersion using vortex ordering or vortex “lattice” is discussed, along with a hypothesis of inducing localization in propagative phonons.