Sandia staff members Jon Ihlefeld (in the Electronic, Optical, & Nanostructured Materials Dept.) and Stephen Lee (in the Semiconductor Material and Device Sciences Dept.) in collaboration with professors Patrick Hopkins (Univ. of Virginia), Bryan Huey (Univ. of Connecticut), and Darrell Schlom (Cornell Univ.) recently published “Effects of coherent ferroelastic domain walls on the thermal conductivity and Kapitza conductance in bismuth ferrite” in Applied Physics Letters. The team observed phonon scattering by coherent domain walls within epitaxial BiFeO3 thin films at room temperature—the first known observation of this effect at noncryogenic temperatures. They believe that the close spacing of the walls in the thin film embodiment allowed this to be observed.
Ferroelectric domain orientation maps for 4-domain (a) and 2-domain (b) specimens, with corresponding images of domain boundary types and charging in (c) and (d) revealing the substantial domain wall den-sity difference for distinct polarization variants. The 200 nm scale bars in (a) and (b) apply to all images for each specimen (columns).
Ferroelectric and ferroelastic domain structure has a profound effect on the piezoelectric, ferroelectric, and dielectric responses of ferroelectric materials. These responses drive the development of capacitors, nonvolatile memory, sensors, and tunnel junctions. Recently, progress in thin-film growth of materials using these ferroelectric effects promises novel nanoscale device applications. Much recent interest is focused on bismuth ferrite (BiFeO3), a rhombohedrally distorted perovskite exhibiting room-temperature coexistence of ferroelectric and antiferromagnetic orders, yielding many unusual properties.
Several studies show it is possible to engineer the domain structure of epitaxial BiFeO3 films using SrTiO3 single-crystalline substrates and by using symmetry breaking step edges resulting from intentional miscuts along high-symmetry crystallographic directions. Given the growth- and substrate-dependent domain structures that BiFeO3 exhibits, its properties and responses to various stimuli can be affected by the presence of the various domain walls. These properties must be considered and well characterized as BiFeO3 thin films are considered for use in various nanometer-scale devices.
Our team measured the thermal conductance of a series of BiFeO3 thin films with different domain variants. Piezo force microscopy (PFM) revealed a corresponding variation in the average domain wall density, based on polarization mapping that allows domain wall identification, local polarization rotation, and interfacial charging to be determined with 4.5 nm resolution. We measured the thermal conductance of the BiFeO3 films with time domain thermoreflectance (TDTR) from 100–400 K.
The effective thermal conductivities we observed in the BiFeO3 films varied based on the density of domain walls in the thin film, where the presence of more domains leads to a decrease in thermal conductivity, indicating the strong effect domain walls have on phonon scattering and thermal conductance even though domain walls are generally considered to be nearly perfect interfacial regions. The thermal boundary conductances across the coherent domain walls are lower than the thermal boundary conductances associated with grain boundaries in silicon, strontium titanate, and yttria-stabilized zirconia (YSZ). At higher temperatures, the effective thermal conductivity of the 4-domain variant BiFeO3 thin films is roughly equivalent to the thermal conductivity of silicon dioxide.
Future work in domain wall thermal engineering should focus on studying the influence of domain wall type, domain characteristics (charged or neutral), and population on thermal transport to optimize the performance of practical devices.
Read the abstract at Applied Physics Letters.