The ground state of CuInP${}_2$S${}_6$ thin films: A study of the deep potential method

Abstract

The two-dimensional ferroelectric (FE) material CuInP${}_2$S${}_6$ (CIPS) has garnered considerable interest due to its out-of-plane ferroelectricity at room temperature. However, a notable discrepancy exists between experiments and density functional theory (DFT) calculations regarding the ground state of CIPS thin films: experiments suggest a state with net polarization, while DFT predicts an antiferroelectric (AFE) state as the lowest-energy state. Here, we investigate the stability of polarization states in CIPS thin films by combining first-principles calculations with the deep potential (DP) method. Our results reveal that for films thicker than the bilayer, an AFE state that has intralayer AFE ordering in the inner layers and intralayer FE ordering in the two surface layers has the lowest electronic energy. This state is significantly lower than the uniform FE state. In addition, we find that a ferrielectric (FiE) state with pure intralayer FE ordering is very close to the AFE state in energy. By using the DP model, we calculated the phonon free energy of CIPS thin films. For the monolayer, the intralayer FE ordering possesses a lower phonon free energy than the intralayer AFE ordering. This energetic preference for the intralayer FE ordering maintains as the thickness grows. Consequently, when the phonon-free energy is incorporated, the FiE state becomes energetically favorable over the AFE state for multilayers CIPS. Our findings demonstrate that the inclusion of vibrational entropy stabilizes the FiE state as the ground state in multilayers CIPS at finite temperatures, reconciling the previous discrepancy between experimental observations and DFT predictions. This insight is vital for understanding the FE properties of CIPS and its potential applications in devices.

arXiv comment10.2 pages, 9 figures

References in corpus (19)

• Gaussian Approximation Potentials: the accuracy of quantum mechanics, without the electrons
• Deep Potential Molecular Dynamics: a scalable model with the accuracy of quantum mechanics
• DeePMD-kit: A deep learning package for many-body potential energy representation and molecular dynamics
• Moment Tensor Potentials: a class of systematically improvable interatomic potentials
• Implementation strategies in phonopy and phono3py
• DP-GEN: A concurrent learning platform for the generation of reliable deep learning based potential energy models
• Out-of-plane Piezoelectricity and Ferroelectricity in Layered $\alpha$-In2Se3 Nano-flakes
• Colossal electroresistance in metal/ferroelectric/semiconductor tunnel diodes for resistive switching memories
• Robust Ferroelectricity in Monolayer Group-IV Monochalcogenides
• A Ferroelectric Semiconductor Field-Effect Transistor
• DeePMD-kit v2: A software package for Deep Potential models
• Active Learning of Uniformly Accurate Inter-atomic Potentials for Materials Simulation
• Thickness-dependent in-plane polarization and structural phase transition in van der Waals Ferroelectric CuInP2S6
• DeePMD-kit v3: A Multiple-Backend Framework for Machine Learning Potentials
• Deep Learning of Accurate Force Field of Ferroelectric HfO${}_2$
• Negative piezoelectricity in low--dimensional materials
• Structural Phase Transitions in SrTiO3 from Deep Potential Molecular Dynamics
• Unconventional ferroelectric domain switching dynamics in CuInP2S6 from first principles
• Interface engineering of ferroelectricity in thin films of thiophosphate ABP 2 X 6 (A = Cu, Ag; B = In, Bi, Cr, V; and X = S, Se)