Electromechanical properties of the 180° domain wall in PbTiO3

We analyze the electromechanical response of the 180 degree ferroelectric domain wall in tetragonal PbTiO3 by combining first-principles calculations with a Landau-Ginzburg-Devonshire (LGD) description. Using regular multidomain structures with varying domain-wall density, we extract polarization profiles and lattice distortions and map them onto the continuum model to determine conventional (homogeneous) and gradient (inhomogeneous) electrostriction. Conventional electrostriction yields only a small negative length change of the sample, whereas gradient electrostriction–arising from the coupling between strain and polarization gradients–produces a positive contribution nearly an order of magnitude larger and localized at the wall core. Our results demonstrate that gradient electrostriction dominates the electromechanical response of 180 degree walls in PbTiO3, supporting its inclusion in LGD models that stabilize Bloch-type domain wall structures.

💡 Research Summary

This paper presents a comprehensive investigation of the electromechanical response of 180° ferroelectric domain walls in tetragonal PbTiO₃ by integrating first‑principles density‑functional theory (DFT) calculations with a Landau‑Ginzburg‑Devonshire (LGD) phenomenological framework. The authors construct periodic supercells containing two symmetry‑related domain walls and vary the wall density (ρ) by changing the number of unit cells (n) between walls. For each supercell, they fully relax atomic positions and lattice parameters under zero external stress, extracting layer‑resolved polarization components (P₁, P₂, P₃) and the associated lattice distortions (ΔL, Δb, Δc).

The LGD description is extended beyond the conventional homogeneous electrostriction term (Q tensor) to include a gradient electrostriction term (R tensor) that couples strain to polarization gradients. The strain‑polarization relation is written as
eᵢⱼ = fᵢⱼₖₗ∂Pₖ/∂xₗ + QᵢⱼₖₗPₖPₗ + Rᵢⱼₖₗₘₙ∂Pₖ/∂xₘ∂Pₗ/∂xₙ + Sᵢⱼₖₗσₖₗ,
where the flexoelectric term f is negligible for the present geometry. Because the wall normal is along x, only the longitudinal strain e₁ varies across the wall. By separating e₁ into a homogeneous electrostriction contribution (Δe₁,eh) and a gradient contribution (Δe₁,en), the total change in sample length becomes ΔL = ΔL_eh + ΔL_en. Analytic expressions (Eqs. 9‑13) relate ΔL to the spontaneous polarization P_S, the Bloch‑type polarization component P_B, the inverse wall thickness k, and two effective coefficients I_Q and I_R that combine the Q‑ and R‑tensor elements, respectively.

From DFT, bulk elastic compliances (S₁₁, S₁₂) and homogeneous electrostriction coefficients (Q₁₁, Q₁₂) are obtained, yielding I_Q = –5.181 × 10⁻³ m⁴ C⁻². The domain‑wall profiles are fitted to hyperbolic functions: P₃(x)=P_S tanh(kx) and, for Bloch walls, P₂(x)=P_B cosh⁻¹(kx). The extrapolation to zero wall density provides Ps≈0.86 C m⁻², k≈3.65 Å⁻¹, and for Bloch walls P_B≈0.26 C m⁻² (Ising walls have P_B≈0).

The calculated length changes for isolated walls are:

• Ising wall: ΔL_eh ≈ –0.42 × 10⁻¹¹ m (small contraction), ΔL_en ≈ +4.3 × 10⁻¹¹ m (expansion).
• Bloch wall: ΔL_eh ≈ –0.38 × 10⁻¹¹ m, ΔL_en ≈ +4.6 × 10⁻¹¹ m.

Thus, the gradient electrostriction contribution is an order of magnitude larger than the conventional electrostriction and has opposite sign, resulting in a net positive elongation of the crystal along the wall normal. The extracted gradient coefficient I_R = –6.019 × 10⁻²¹ m⁶ C⁻² confirms that the R‑tensor dominates the electromechanical response at the wall core.

The discussion emphasizes that conventional LGD models, which include only Q‑terms, predict an unstable Bloch configuration because the homogeneous electrostriction alone cannot offset the energy cost of the Bloch component. Incorporating the R‑term provides a positive strain localized at the wall, stabilizing the Bloch‑type polarization and reconciling theory with previous ab‑initio reports of Bloch walls in PbTiO₃. The authors also note that the wall‑induced transverse strains (Δb, Δc) vanish in the dilute limit, confirming that the observed elongation originates solely from the longitudinal strain.

In conclusion, the study quantitatively demonstrates that gradient electrostriction is the primary driver of the electromechanical response of 180° domain walls in PbTiO₃. This insight has significant implications for the design of ferroelectric devices that exploit domain‑wall functionality, such as non‑volatile memories, nanoscale actuators, and acoustic transducers, where precise control of wall‑induced strain is essential. The methodology—combining first‑principles extraction of microscopic profiles with a rigorously extended LGD model—offers a robust pathway for evaluating gradient‑coupling effects in other ferroic systems.