Ferroelectric Properties and Topological Textures in PbTiO 3 from a Second-Principles Open-Source interatomic potential

Louis Bastogne, Philippe Ghosez · 2025

Read Ferroelectric Properties and Topological Textures in PbTiO 3 from a Second-Principles Open-Source interatomic potential by Louis Bastogne, Philippe Ghosez (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2505-02624.txt.

Introduction

Louis Bastogne and Philippe Ghosez present an interatomic potential for lead titanate in two thousand twenty-five. Lead titanate is a benchmark ferroelectric with strong polarization. It shifts phase at high heat in ways labs prize. Density functional theory models it well but costs too much for big runs.

Their second principles potential is open source and fully atomistic. It matches the costly theory on key traits of the material. Domain wall motion and polar textures come out right in its runs. The model stays predictive across a wide span of settings. Big, accurate studies of memory grade uses are the target.

Catalog record

CREATORLouis Bastogne, Philippe Ghosez
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE37 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2505-02624

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Ferroelectric Properties and Topological Textures in PbTiO 3 from a Second-Principles Open-Source interatomic potential Louis Bastogne, Philippe Ghosez Abstract We introduce an open-source, fully atomistic second-principles interatomic potential for lead titanate (PbTiO 3 ), a benchmark ferroelectric material known for its strong polarization and high-temperature phase transitions. While density functional theory excels at capturing atomic-scale behavior, it remains computationally prohibitive for large-scale simulations required to explore complex phenomena. Our model addresses this limitation by accurately reproducing key properties of PbTiO 3 , including domain wall dynamics and skyrmion formation, which are known as key features to next-generation memory and energy-efficient technologies. Validated against DFT data, the model remains predictive across a wide range of conditions. It offers an accessible and efficient framework for high-accuracy large-scale simulations, allowing deeper insights into PbTiO 3 and its potential applications. I Introduction Lead titanate (PbTiO 3 ) is a cornerstone material in the ferroelectric (FE) community, well known for its large spontaneous polarization Nelmes and Kuhs 1985 and high-temperature ferroelectric-to-paraelectric phase transition Shirane and Hoshino 1951 . These attributes underpin its widespread use in applications such as transducers and actuators Zhang et al. 2013 . They also establish PbTiO 3 as a prototypical compound for studying fundamental phenomena such as domain formation Meyer and Vanderbilt 2002 , phase transitions, and piezoelectricity Ghosez and Junquera 2022 . Over the past decade, several unexpected discoveries in PbTiO 3 have significantly advanced our understanding of ferroelectric materials. Notably, distinctive ferroelectric domain walls (DWs) have emerged Wojdeł and Iniguez 2014 , redefining their role from passive interfaces to active, dynamic entities with novel functionalities such as heat flow regulation Langenberg et al. 2019 . This paradigm shift establishes DWs as key components for enhancing energy efficiency and reducing power consumption in future electronic devices. Another breakthrough was the prediction of polar skyrmions Pereira Gonçalves et al. 2019 , topologically non-trivial solitons analogous to their magnetic counterparts. These skyrmions are stable, electrically controllable, and can exhibit intriguing properties such as chirality and tunable negative capacitance Tikhonov et al. 2020 ; Zubko et al. 2016 ; Guo et al. 2025 , positioning them as promising candidates for ultra-dense data storage and advanced memory technologies Du et al. 2024 . Understanding such complex structures and phenomena requires computational methods that combine high accuracy with scalability. While density functional theory (DFT) offers reliable atomic-scale insights, its computational cost limits its applicability to large systems or long time scales. Second-principles (SP) interatomic potentials Wojdeł et al. 2013 ; Escorihuela-Sayalero et al. 2017 overcome these constraints by describing the potential energy surface via effective potentials fitted to DFT data Ghosez and Junquera 2022 . These models have shown predictive capabilities across various studies Graf and Íñiguez 2021 ; Bastogne et al. 2024 ; Murillo-Navarro et al. 2021 ; Zhang et al. 2023 ; Pereira Gonçalves et al. 2019 ; Zubko et al. 2016 ; Gómez-Ortiz et al. 2024 ; Gómez-Ortiz et al. 2025 , advancing the exploration, at the atomic scale, of ferroelectric phenomena of technological relevance. In this work, we introduce an open-source, fully atomistic second-principles model for PbTiO 3 , rigorously constructed and validated against DFT data. Our model faithfully captures structural, energetic, and dynamical properties – including phase transitions and finite-temperature phonon softening. We use it to investigate domain-walls and their interactions as well as topological textures such as skyrmions. Doing so, we confirm the usefulness of the approach to uncover previously unexplored polar configurations in PbTiO 3 . More broadly, its potential to explore previously inaccessible polar textures and complex behaviors in ferroelectric materials. II Methods II.1 First-principles We generated the ab-initio data for fitting the second-principles interatomic potential using density functional theory calculations with the Abinit software package Gonze et al. 2020 . We used the generalized gradient approximation (GGA) with the PBESol exchange-correlation functional and a planewave-pseudopotential approach, relying on optimized norm-conserving pseudopotentials from the PseudoDojo server Hamann 2013 ; Van Setten et al. 2018 . We treated Pb\penalty\ 5d^{10}6s^{2}6p^{2} , Ti\penalty\ 3s^{2}3p^{6}3d^{2}4s^{2} and O\penalty\ 2s^{2}2p^{4} electrons as valence states. We adopted a plane-wave energy cutoff of 65\ Ha and used a 8\times 8\times 8 \Gamma -centered Monkhorst-Pack Monkhorst and Pack 1976 k-point mesh for the five-atom cubic unit cell, while adapting the k-point sampling to the supercell size in order to maintain a similar high level of convergence. We computed the dynamical matrices for the relaxed cubic Pm\bar{3}m on a 4\times 4\times 4 q-point mesh. From DFPT, we also extracted essential properties, including the dielectric constant, Born effective charges, strain-phonon coupling, and elastic constants. II.2 Second-principles The second-principles model employed in this work is designed to accurately and efficiently describe the Born-Oppenheimer potential energy surface (PES) of crystalline materials. It is based on a Taylor expansion of the total energy in terms of structural degrees of freedom and around a reference structure—typically a stationary point where all forces and stresses vanish. This reference can represent a configuration relevant to competing phases or a stable structure suited for analyzing local anharmonicity. The model incorporates all atomic and strain degrees of freedom by systematically distinguishing between harmonic and anharmonic energy contributions. Harmonic interactions, including both short-range forces and long-range dipole-dipole couplings, are derived from density functional perturbation theory (DFPT). Anharmonic effects, which capture deviations from harmonic behavior, are introduced through a selection of higher-order terms whose parameters are fitted to first-principles data, such as energies, forces, and stresses from a representative set of atomic configurations. Assuming periodic boundary conditions, the model constructs the PES using material-specific parameters. Provided the bond topology remains unchanged, this approach effectively captures the behavior of systems with closely competing structural phases—such as perovskites—where small energy differences drive phase transitions We developed the model using the Multibinit Gonze et al. 2020 software package. In practice, the interatomic force constants (IFC), elastic constants, strain-phonon coupling, Born effective charges, and dielectric constant of the cubic phase Pm\bar{3}m were computed using DFPT calculations from Abinit Gonze et al. 2020 . A training set of 3644 DFT configurations was used to select and fit the anharmonic coefficients of the model, applying a cutoff radius of \sqrt{3}/2 times the cubic lattice parameter to generate symmetry-adapted terms (SAT) up to the eighth order, including strain-phonon coupling and anharmonic elastic constants. We selected 29 SATs via a goal-function (GF) minimization approach and added 66 additional terms to ensure boundness of the model energy. Finite-temperature simulations were conducted using a hybrid molecular dynamics Monte Carlo (HMC) method Duane et al. 1987 ; Betancourt 2017 , implemented in Abinit Gonze et al. 2020 , with a time step of 0.72 fs and 29 fs HMC sweeps. After 500 HMC sweeps of thermalization, simulations comprised 2000 HMC sweeps in general, increasing to 4000 near phase transitions. Structural relaxations used the Broyden-Fletcher-Goldfarb-Shanno (BFGS) algorithm in Abinit Gonze et al. 2020 , and energy barriers were explored using the nudged elastic band (NEB) method with the Quick-Min algorithm Henkelman and Jónsson 2000 ; Sheppard et al. 2008 . Phonon dispersions at 0 K were obtained with the SP potential in Abinit Gonze et al. 2020 , while finite-temperature phonons were calculated using TDEP Knoop et al. 2024 . This atomistic model has been successfully applied in previous

Read the complete text