Charged domain wall phenomena and associated negative capacitance in boracites

  • Joseph Guy

Student thesis: Doctoral ThesisDoctor of Philosophy


Ferroelectric materials, defined by a switchable, spontaneous electrical polarisation, have made significant inroads into our lives under the guise of invaluable technologies such as data storage and nanoelectronics. These technologies typically rely on ferroelectric domains, that is, macroscopic regions of uniformly oriented dipole moments within the crystal. Over the past decade, however, attention has now turned to the entity that separates these domains, the domain wall, due to their widely reported enhanced functional properties distinct from the bulk material. Running parallel to this fertile branch of ferroelectrics research is the troubling realisation that we are approaching a bottleneck in nanoscale transistor design due to fundamental operational limits. To circumvent these issues, exotic physical concepts have been invoked. In particular, the phenomenon of negative capacitance has been proposed, and it is here that ferroelectrics, particularly their constituent domain walls, could play a crucial role in the continued success of transistor-based technologies; ephemeral domain wall-based transistors, for example, have been envisioned as superseding current transistor designs. In this thesis, I report the anomalous motion of specific domain boundaries in a sample of improper ferroelectric-ferroelastic copper-chlorine boracite Cu3B7O13Cl (Cu-Cl boracite) that is commensurate with a non-transient negative capacitance effect.
As a benchmark, and to ensure that I could make valid statements about the electric field-induced changes in polarisation observed in these crystals, a variety of optical and scanning probe techniques were employed to characterise the local domain states. Information regarding the ferroelastic domains, manifest as distinct sample surface corrugations, was used in conjunction with electrical-based scanning probe measurements such as conducting Atomic Force Microscopy (c-AFM) and Piezoresponse Force Microscopy (PFM) to determine that domain walls with an associated dc conductivity distinct from the bulk are caused by two types of polar discontinuity: 90° tail-tail conducting or 90° head-head insulating.

The same charged domain boundaries can be readily injected into the crystal upon application of point stress from a sharp preparation needle. I demonstrate that subsequent sweeping of the needle across the sample surface can uniformly switch considerable portions of the crystal into only a handful of distinct domain states delineated by large, rectilinear domain boundaries that are charged and span hundreds of micrometres in length. Site-specific injection of such boundaries subsequently paved the way for their precise control using externally applied electric fields.
The electric field-induced motion of 90° tail-tail conducting walls was found to be entirely conventional and thermodynamically concrete; domains aligned with the biasing field grow favourably at the expense of domains with a spontaneous polarisation anti-aligned with the field. Remarkably, however, the development of polarisation across 90° head-head insulating boundaries is out-of-phase with the applied field, and subsequent positional hysteresis measurements confirm this. The inferred polarisation-electric field hysteresis indicates that dP/dE is negative throughout the range of fields applied which, consequently, is proportional to the capacitance.

This study culminates with the direct measurement of the negative capacitance effect by examining the switching currents induced by the anomalous motion of charged insulating domain boundaries. To circumvent the issues associated with low field-driven domain wall propagation speeds, a relatively new scanning probe technique called Charge Gradient Microscopy (CGM) was utilised, which allowed me to resolve the small switching currents by recreating the same relative motion between the probe (electrode) and wall observed in typical bias-driven experiments, but at a significantly faster rate. The CGM results, in conjunction with domain wall velocity-voltage measurements, were used to infer the current-voltage (I-V) characteristics of the domain boundary and confirm the negative capacitance effect. On the basis of these measurements, certain mechanisms that, historically, have been attributed to negative capacitance, such as the magnetic storage of energy and local field electric field reversal at the domain boundary, are excluded. The exact mechanism generating the negative capacitance effect within the boracites is still presently unknown, but outstanding experiments, including consideration of long-range microstructural changes and investigation of charged domain wall dynamics in lamellar-sized geometries are considered.
Date of AwardDec 2021
Original languageEnglish
Awarding Institution
  • Queen's University Belfast
SupervisorMarty Gregg (Supervisor) & Amit Kumar (Supervisor)

Cite this