-
7
-
-
1842684062
-
-
M. Yamanouchi et al.: Nature 428 (2004) 539;
-
(2004)
Nature
, vol.428
, pp. 539
-
-
Yamanouchi, M.1
-
9
-
-
9144255659
-
-
E. Saitoh et al.: Nature 432 (2004) 203.
-
(2004)
Nature
, vol.432
, pp. 203
-
-
Saitoh, E.1
-
15
-
-
33748641485
-
-
L. Thomas et al.: Nature 443 (2006) 197.
-
(2006)
Nature
, vol.443
, pp. 197
-
-
Thomas, L.1
-
30
-
-
34547399594
-
-
We define n to be a unit vector in the direction of spin, which is opposite to the direction of magnetization
-
We define n to be a unit vector in the direction of spin, which is opposite to the direction of magnetization.
-
-
-
-
37
-
-
33847330320
-
-
H. Kohno, G. Tatara, and J. Shibata: J. Phys. Soc. Jpn. 75 (2006) 113706. Referred to as I in the text.
-
H. Kohno, G. Tatara, and J. Shibata: J. Phys. Soc. Jpn. 75 (2006) 113706. Referred to as I in the text.
-
-
-
-
39
-
-
34547464768
-
-
cond-mat/0703414
-
R. A. Duine et al.: cond-mat/0703414.
-
-
-
Duine, R.A.1
-
40
-
-
84858095885
-
-
This term renormalizes the magnitude of spin from the localized component (S) to Stot, S, δS, where δS is a contribution from s electrons. One can eliminate this (fourth) term by multiplying a factor S/Stot to all terms on the r.h.s. of eq, 1
-
tot to all terms on the r.h.s. of eq. (1).
-
-
-
-
41
-
-
84858089432
-
-
Please note that, in the literature (except for, e.g., ref. 17), the β-term is regarded as non-adiabatic. In this letter, we take a different view that adiabatic following is a consequence of the adiabatic theorem, rather than the definition of adiabaticity. We are grateful to Y. Tserkovnyak for sharing this viewpoint, to H. J. Skadsem for informing us of ref. 17, and to M. D. Stiles for his thoughtful comments on this issue.
-
Please note that, in the literature (except for, e.g., ref. 17), the β-term is regarded as non-adiabatic. In this letter, we take a different view that "adiabatic following" is a consequence of the adiabatic "theorem", rather than the definition of adiabaticity. We are grateful to Y. Tserkovnyak for sharing this viewpoint, to H. J. Skadsem for informing us of ref. 17, and to M. D. Stiles for his thoughtful comments on this issue.
-
-
-
-
42
-
-
84858095882
-
-
There is also a non-adiabatic torque, τna, due to electron reflection, which is spatially nonlocal and oscillatory. We do not study this torque in this letter; please refer to refs. 27, 35, and 36
-
na, due to electron reflection, which is spatially nonlocal and oscillatory. We do not study this torque in this letter; please refer to refs. 27, 35, and 36.
-
-
-
-
45
-
-
34547431658
-
-
cond-mat/0702020
-
M. D. Stiles et al.: cond-mat/0702020.
-
-
-
Stiles, M.D.1
-
47
-
-
34547439373
-
-
cond-mat/0602075
-
A. Sakuma: cond-mat/0602075.
-
-
-
Sakuma, A.1
-
48
-
-
34547486044
-
-
cond-mat/0611588
-
H. J. Skadsem et al.: cond-mat/0611588.
-
-
-
Skadsem, H.J.1
-
53
-
-
84858085778
-
-
We use the vector (bold italic) notation for the spin component. The space-time components are indicated by subscripts such as μ, ν (= 0, 1, 2, 3) or i; j (= 1, 2, 3).
-
We use the vector (bold italic) notation for the spin component. The space-time components are indicated by subscripts such as μ, ν (= 0, 1, 2, 3) or i; j (= 1, 2, 3).
-
-
-
-
54
-
-
84858087264
-
-
The subscript σ =↑, ↓ corresponds, respectively, to σ = +1, -1 in the formula (and to σ = ↑, ↓ or -1; +1), as in I.
-
The subscript σ =↑, ↓ corresponds, respectively, to σ = +1, -1 in the formula (and to σ = ↑, ↓ or -1; +1), as in I.
-
-
-
-
55
-
-
84858085779
-
-
⊥ already contains a time derivative.
-
⊥ already contains a time derivative.
-
-
-
-
57
-
-
84858089429
-
-
sreproduces eq. (2).
-
sreproduces eq. (2).
-
-
-
-
58
-
-
34547473696
-
-
This choice ensures the rotational symmetry of n and t el
-
el.
-
-
-
|