Zernike Seminar: Nagarajan Valanoor (University of New South Wales, Sydney, Australia) - "Ferroelastic Domain Wall Mobility in Ferroelectric Multilayers: Stretching the possibilities"

ABSTRACT
Over the past 10 years our group has been actively seeking innovative methods to enhance the electromechanical response of ferroelectric thin films. One such technique has been the use of bilayers- in 2009 we showed that a tetragonal (T) PbZrxTi1-xO3 (PZT)/rhombohedral (R) PbZrxTi1-xO3 (PZT) bilayer heterostructure shows remarkable ferroelastic mobility. The motion of these ferroelastic domain walls leads to giant electromechanical responses- some of the largest ever reported for thin integrated films. In my talk I want to cover the evolution of our thought process- from the first “poor” experiment using sol-gel samples to the current state-of-the-art using model epitaxial systems. Some of these results appear in Advanced Materials (2009), Phys. Rev. Lett (2010), Acta Materialia (2010) and Advanced Materials Interfaces (2015,2016)

The middle part of the talk will focus on the nanoscale origins of ferroelastic domain wall motion in ferroelectric multilayer thin films that lead to the aforementioned giant electromechanical responses. We present direct evidence for complex underpinning factors that result in ferroelastic domain wall mobility using a combination of atomic-level aberration corrected scanning transmission electron microscopy and phase-field simulations. The local electric dipole distribution is imaged on an atomic scale for a ferroelastic domain wall that nucleates in the R-layer and cuts through the composition breaking the T/R interface. Our studies reveal a highly complex polarization rotation domain structure that is nearly on the knife-edge at the vicinity of this wall. Induced phases, namely tetragonal-like and rhombohedral-like monoclinic were observed close to the interface, and exotic domain arrangements, such as a half-four-fold closure structure, are observed. Phase field simulations show this is due to the minimization of the excessive elastic and electrostatic energies driven by the enormous strain gradient present at the location of the ferroelastic domain walls. Thus, in response to an applied stimulus, such as an electric field, any polarization reorientation must minimize the elastic and electrostatic discontinuities due to this strain gradient, which would induce a dramatic rearrangement of the domain structure. This work appears in ACS Nano, 2016, 10 (11), pp 10126–10134.

Where does this all fit in for CogniGron? Our insight into the origins of ferroelastic domain wall motion has allowed us to better “craft” such multilayered ferroelectric systems with precisely tailored domain wall functionality and enhanced sensitivity. The last bit of my talk will have unpublished data where I show that we can no nucleate ferroelastic domains at site-specified locations. This is a massive leap forward which we plan to exploit to study synaptic plasticity using elastic motion of ferroic domains.