Page 305 - Electrical Properties of Materials

Spintronics 287

dependent, one of the spin species has a higher chance of getting scattered from
a 4s state into an empty 3d state. Because of its reduced mobility this spin will
contribute less to the total current.
The dependence of resistance on spin is the phenomenon that launched the
new discipline of spin electronics or briefly, spintronics.

11.11 Spintronics

With the discovery of giant magnetoresistance, spin entered the minds of en-
gineers. Until then conventional electronics always, without exception, ignored
spin. But all that changed in 1988. The effect of spin has been the subject of
intensive research ever since. Commercial applications are not far away as will
be discussed in Section 11.12 and in Appendix VI, devoted entirely to the phys-
ical principles on which memory elements have been working or might work
in the future.

11.11.1 Spin current

We know what current is. The substances may be different but the definition
is always the same. We measure current by counting the number of elements
(whatever they are) passing through a cross-section in unit time. The substance
could be water (we talk then of a stream, a brook, or a river), it can be oil or
gas (which flows in pipes), or electrons (which flow in conductors).
Spin current is not analogous to any of these. It arises when the two different
kinds of spin, up and down, are not in equilibrium. One kind is more numerous
(a)
than the other. To find out more about spin let us set up the experiment shown in
Fig. 11.24(a). There are two materials connected to each other, a ferromagnetic
material, F, and a non-magnetic metal, N, and there is a voltage applied between
the two materials. We can easily predict what would happen. Both materials are
conductors so electrons will flow from the ferromagnetic material into the non- F N
magnetic metal. That is obvious. There is now a current: an electron current.
To emphasize the difference between the flow of electrons and the flow of (b)
the electron spin we shall often refer to the former as charge current. What M
else is there to say? Thirty years ago we would have said: ‘That’s all. There
is nothing else we need to think about.’ But ever since the discovery of giant
magnetoresistance we have become spin-conscious. We should ask now: ‘What δM
about the spin?’
Let us start by saying that the magnetism in F is polarized. The electrons
(say) have spin up. Obeying the electric field that has been set up, many of Position
them will move into N. If the spin of the electron that moved over was up in
F, then surely its spin will still be up when it crosses the boundary into N. An Fig. 11.24
(a) Voltage applied to an F–N
electron is not so oblivious as to forget its past history. But, the next ques-
junction (between a ferromagnetic
tion is, how long can it keep its spin up while moving in the non-magnetic
and a non-ferromagnetic material).
metal? Clearly, those spin-up electrons find themselves in an entirely different
(b) The resulting spatial distribution
environment, one might even say a hostile environment. All the physical phe- of magnetization.
nomena taking place in the non-magnetic material conspire against them. They
will be knocked off their privileged position by the various scattering phenom-
ena. What can they do? They cannot vanish. Electrons will remain electrons. ∗ ∗ We are not concerned with particle
Thus all they can do is to turn into spin-down electrons. But if the number physics.