Johan MENTINK
Antiferromagnets host the fastest and smallest magnetic waves of all magnets. With their additional intrinsically small dissipation antiferromagnets are ideal candidates to challenge the limits for energy and speed in data storage and processing technologies. However, understanding the magnon spectrum at short wavelengths and high oscillation periods has been challenging even for the simplest model: the antiferromagnetic Heisenberg model in 2D [1]. Furthermore, studying the space-time dynamics of this model defines an intricate quantum many-body problem out of equilibrium, for which until recently no accurate methods were available.
Beyond the limitations of existing methods, we adopt a machine learning inspired ansatz [2] to simulate the dynamics of the 2D Heisenberg model [3-4]. By sudden perturbations of the exchange interaction, we directly trigger dynamics of short-range spin correlations that is often described as the dynamics of magnon-pairs. Interestingly, although the anisotropic pattern can be indeed qualitatively understood with magnon theory, the spreading at the smallest length and time scales is up to 40% faster than expected from the highest magnon velocity. We explain the enhanced propagation speed by magnon-magnon interactions, which become exceptionally strong in the two dimensions and in the deep quantum limit (S=1/2).
Beyond sudden perturbations of the exchange interaction, we consider parametric driving of magnon pairs and explore the potential for switching between two stable oscillation states [5]. Using a semi-classical theory, we predict that switching can occur at the femtosecond timescale with an energy dissipation down to a few zepto Joule. This result touches the thermodynamical bound of the Landauer principle and approaches the quantum speed limit up to 5 orders of magnitude closer than demonstrated with magnetic systems so far.
[2] G. Carleo and M. Troyer, Science 355, 602 (2017)
[3] G. Fabiani & J.H. Mentink, SciPost Phys 7, 004(2019)
[4] G. Fabiani, M.D. Bouman and J.H. Mentink, Phys. Rev. Lett. 127, 097202 (2021)
[5] G. Fabiani & J.H. Mentink, Appl. Phys. Lett. 120, 152402 (2022)