Magnetic transitions identified by an efficient saddle point search

Magnetic systems hosting topological textures have been of great technological and fundamental interest in recent years. Identifying the lifetime of metastable states, predicting hitherto unknown states, and computing their kinetics are essential tasks [1]. Identifying first-order saddle points on the energy surface is paramount in this context, and the potential for magnetic systems through the implementation of the Minimum Mode Following (MMF) approach is significant [2]. Recently, we presented an efficient implementation of the MMF method based on the optimization of the Rayleigh Quotient, enabling applications to complex magnetic systems [3]. However, embedding these methods in adaptive kinetic Monte Carlo simulations requires a systematic approach for generating initial configurations for the iterative MMF calculations. To solve this problem, we developed a method based on local excitations of the magnetic textures selected by exploiting their symmetries. Using both DFT parameterized magnetic multilayer systems [4] and fundamentally interesting parameter models [5], we demonstrate that our method efficiently explores the energy landscape and enables the lifetime calculation of the various magnetic textures.

[1] F. Muckel et al., Nat. Phys. 17.3 (2021).

[2] G. P. Müller et al., Phys. Rev. Lett. 121.19 (2018).

[3] M. Sallermann, H. Schrautzer et al., arXiv:2403.11799 (2024).

[4] F. Nickel et al., Phys. Rev. B 107.17 (2023).

[5] V. M. Kuchkin et al., Phys. Rev. B 102.14 (2020).