Robust Zeeman-type band splitting in sliding ferroelectrics
Abstract: Transition metal dichalcogenides exhibit giant spin-orbit coupling, and intriguing spin-valley effects, which can be harnessed through proximity in van der Waals (vdW) heterostructures. Remarkably, due to the prismatic crystal field, the Zeeman-type band splitting of valence bands reach values of several hundreds of meV. While this effect is suppressed in the commonly studied hexagonal (H)-stacked bilayers due to the presence of inversion symmetry, the recent discovery of sliding ferroelectricity in rhombohedral (R-)stacked MX$_2$ bilayers (M=Mo, W; X=S, Se) suggests that the Zeeman effect could be present in these non-centrosymmetric configurations, making it even more intriguing to investigate how the spin-resolved bands would evolve during the phase transition. Here, we perform density functional theory calculations complemented by symmetry analysis to unveil the evolution of ferroelectricity during sliding and the behavior of Zeeman splitting along the transition path. While the evolution of the out-of-plane component of the electric polarization vector resembles the conventional ferroelectric transition, we observe significant in-plane components parallel to the sliding direction, reaching their maximum at the intermediate state. Moreover, we demonstrate that the R-stacked bilayers exhibit persistent Zeeman-type band splitting throughout the transition path, allowed by the lack of inversion symmetry. Further analysis of different stacking configurations generated by sliding along various directions confirms that the Zeeman effect in MX$_2$, primarily arising from the polarity of prismatic ligand coordination of the metal atom, is remarkably robust and completely governs the spin polarization of bands, independently of the sliding direction. This resilience promises robust spin transport in vdW based MX$_2$ bilayers, opening new opportunities for ferroelectric spintronics.
Paper Prompts
Sign up for free to create and run prompts on this paper using GPT-5.
Top Community Prompts
Collections
Sign up for free to add this paper to one or more collections.