370
X. S. Wang and X. R. Wang
Fig. 14.8 Schematic diagrams for vortices of different chiralities and polarities
where v is the velocity of the texture; u = −μ B pJ/[eM s (1 + β
2
)] is a vector with
the dimension of a velocity that is proportional to the current density J, in which
p is the spin polarization and e is the electron charge; G is the gyrovector;
← →
D is
the dissipation tensor defined as D i j =
∂ i m · ∂ j mdV ; and T is the force on the
texture and its i
th component is T i = −
∂
F dV
∂ X i
−
∂m
∂x i
· (m × τ )dV , where F is
the free-energy functional and X is the position of the texture center. The components
of gyrovector G is defined as
G i =
1
2
ε i jk m i ·
∂ j m × ∂ k m
d
2 r,
(14.13)
where ε i jk is the Levi-Civita symbol and i, j, k = 1, 2, 3 stand for x, y, z. For a 2D
film in the x y plane, only G z is present, and it is easy to see G z = 2π Q. The term
G × v says that a finite Q induces a transverse motion to the velocity, similar to the
Lorentz force or Magnus force.
Due to the non-zero winding number, an isolated vortex cannot be a localized
object in an infinite film. However, they can exist locally in confined geometries
such as nanodisks [57–63]. The vortices in vortex domain walls (VDWs) could also
be used as information carriers in a racetrack memory designs [46]. In the presence
of STT only [T = 0 in (14.12)], if α = β, the vortices not only move along the
electron flow, but also move in the transverse direction, which hinders the practical
Précédent

- 385/587

Suivant