Comment on ``Elastic Stabilization of a Single-Domain Ferroelectric State in Nanoscale Capacitors and Tunnel Junctions" [N.A. Pertsev and H. Kohlstedt, Phys. Rev. Lett. 98, 257603 (2007).]

A. M. Bratkovsky,A. P. Levanyuk

DOI: https://doi.org/10.1103/PhysRevLett.100.149701

2007-09-28

Abstract:In a recent Letter [N.A. Pertsev and H. Kohlstedt, Phys. Rev. Lett. 98, 257603 (2007)] the authors claim that "even nanoscale capacitors and tunnel junctions may have out of plane polarization sufficient for memory applications." Here we show in an elementary way that this conclusion is not substantiated by their calculations and that they should have come to the opposite conclusion within their approximations.

Materials Science,Statistical Mechanics

What problem does this paper attempt to address?

The core problem that this paper attempts to solve is to evaluate the conclusion proposed by Pertsev and Kohlstedt (PK) in their previous research, that is, "even nano - scale capacitors and tunnel junctions may have sufficient out - of - plane polarization for memory applications." Authors Bratkovsky and Levanyuk believe that PK's conclusion is not supported by their calculations and is misleading in some aspects.

### Main problems
1. **Influence of the gradient coefficient**:
PK simplified the problem in their calculations by setting the gradient coefficient \(g_{ln}\) to zero. Physically, this means that the domain wall energy and width are set to zero. However, this simplification ignores the important influence of the gradient term on the stability of the system. In fact, when \(g_{ln} = 0\), the system can eliminate the depolarization field by forming domains, resulting in an unstable state.

2. **Role of elastic energy**:
PK believes that the elastic energy associated with the domain structure can prevent the formation of domains. However, in their approximation method, the polarization distribution is strictly rectangular \(P(x)=\pm p\), and at this time the elastic energy is zero. Therefore, the elastic energy has no effect in this approximation. In addition, elastic effects are indeed important when considering the transition from the asymmetric phase to the symmetric phase, but PK's discussion is not related to this.

3. **Stability and memory function**:
PK studied the stability under very small fluctuations, but this is not equivalent to the memory function in practical applications. Stability and memory function can be divided into absolute stability and relative stability (i.e., metastability). For metastability, the memory function depends on the time to escape from the metastable state. If this time is too short, effective storage function cannot be achieved. PK did not take this into account but focused on theoretical instability.

4. **Key theoretical issues**:
The key to calculating the "critical thickness of ferroelectric storage" is to determine the time to escape from the metastable state. It is necessary to find appropriate methods to solve the problem of ferroelectric storage, and PK's research failed to provide valuable insights and may mislead research in this field instead.

### Summary
The main purpose of this paper is to point out that PK's research has methodological flaws and to emphasize that when evaluating the memory function of nano - scale ferroelectric capacitors, more complex physical mechanisms such as gradient terms, elastic effects, and the time to escape from the metastable state must be considered. These factors are crucial for understanding ferroelectric storage in practical applications.