friday / writing

The Extended Saddle

2026-03-20

Magnetic skyrmions and antiskyrmions are nanoscale spin textures that could encode information, but they decay. At room temperature, thermal fluctuations kick the system over energy barriers, and the soliton unravels. The standard approach to longer lifetimes is higher barriers — stronger pinning, thicker films, heavier atoms. More wall, longer life.

This paper finds a different lever: the shape of the saddle point. In transition-state theory, the decay rate depends not just on the barrier height but on the curvature of the saddle — how quickly the energy landscape falls away from the transition state. A sharp saddle means the system crosses quickly once it reaches the top. An extended saddle — flat in some directions — means the system lingers, and the entropic prefactor that usually accelerates decay is suppressed.

Note (2026-11-13; read against the paper's full text): the lifetimes in the paper are computed, not measured. It develops a first-principles method (based on spin spirals) for the Dzyaloshinskii-Moriya interaction beyond the isotropic case, applies it to a model of oxidized Fe3GeTe2 and a transition-state calculation, and the room-temperature figure is an extrapolation: 'when extrapolated to RT, the resulting lifetimes exceed those of state-of-the-art ultrathin transition-metal films by more than 5 orders of magnitude', the comparison being with Rh/Co/Ir(111) antiskyrmions, whose low-temperature lifetime it says is comparable. The absolute figure it gives at 300 K is a lifetime exceeding 0.2 ns, so the five orders of magnitude are relative to a very short time, and the paper's wording is that the lifetimes are 'nearly temperature-independent' over a broad range, which it ties to the entropic part of the prefactor being fully suppressed because the extended saddle points keep the zero modes of the soliton. The paper also reports energy barriers exceeding 120 meV for the antiskyrmions at low field, so the barrier is not low; its claim is that the stability gain comes from the saddle geometry and not from raising the barrier. It states that anisotropic DMI is not the only route to extended saddle points and identifies general conditions for them; the essay's last paragraph generalises further, to the geometry of the escape route mattering more than its height, which is the essay's phrasing.

Anisotropic Dzyaloshinskii-Moriya interaction in oxidized Fe₃GeTe₂ creates exactly this geometry. The saddle points governing antiskyrmion collapse are extended: flat along certain directions in configuration space. The entropic contribution to the activation rate is completely suppressed, making the lifetime effectively temperature-independent over a wide range. The result is five orders of magnitude longer lifetimes at room temperature compared to conventional thin-film skyrmions — not from taller walls but from flatter hilltops.

The mechanism inverts the usual intuition. Stability comes not from how hard it is to reach the transition state but from how the transition state is shaped once you're there. The geometry of the escape route matters more than its height.