Recent research has highlighted the potential of ferroelectricity in van der Waals bilayers in providing an unconventional route for improving device performance. Understanding the static and dynamic properties of domain wall (DW) is critical unlocking this potential, as key parameters such as switching field and speed heavily rely on them. In this article, we conduct a theoretical exploration of the fundamental properties of textures in stacking-engineered ferroelectrics using a machine-learning potential model. Our results demonstrate that competition between the switching barrier of stable ferroelectric states and in-plane lattice distortion leads to a DW width of ten nanometers. We also demonstrate that DW motion can drastically reduce the critical ferroelectric switching field of a monodomain by two orders of magnitude and enable domain switching on a picosecond timescale, suggesting the potential for ultrafast and energy-saving non-volatile memory devices. Moreover, twisting the bilayer into a stacking Moiré structure results in a super-paraelectric state, because the ferroelectric order is reversibly broken by DW motion already at ultralow electric fields. These findings offer valuable insights into the behavior and properties of stacking-engineered ferroelectrics, with significant implications for the development of next-generation electronic devices.

最近的研究强调了范德华双层铁电性在提供提高器件性能的非常规途径方面的潜力。了解畴壁 (DW) 的静态和动态特性对于释放这一潜力至关重要，因为开关场和速度等关键参数严重依赖于它们。在本文中，我们使用机器学习电位模型对堆叠工程铁电体中纹理的基本特性进行了理论探索。我们的结果表明，稳定铁电态的转换势垒与面内晶格畸变之间的竞争导致 DW 宽度为 10 纳米。我们还证明，DW 运动可以将单域的临界铁电切换场大幅降低两个数量级，并实现皮秒时间尺度的域切换，这表明超快且节能的非易失性存储设备的潜力。此外，将双层扭曲成堆叠莫尔结构会导致超顺电态，因为在超低电场下，铁电秩序已被 DW 运动可逆地破坏。这些发现为堆叠工程铁电体的行为和特性提供了宝贵的见解，对下一代电子设备的开发具有重大影响。