Blog entry

What is spin? (and why do we want to tranport it?)

By José Lado (INL) 

Today I would like to try to explain in a pictorial way what is the spin of the electrons and why it is technologically very appealing to electronic applications. I would like to warn you that I will oversimplify some aspects, so if want to get a more accurate view of any detail, checking Wikipedia is always a great idea. Lets get started then.  


As most of you maybe have already heard, the spin is like an intrinsic (i.e. a fundamental property) of the electrons, which (at the home level!) is usually understood as if the electron where a small ball rotating over itself. Although  this picture is sometimes useful, it can be rather misleading given that (up to now), the electron has zero volume, it is hard to understand how something that doesn't have a volume can rotate, and of course its true nature is not related at all with a usual rotation. However, from now on, I would refer to the spin of the electron as a color (of course completely different from the one of high energy physics!), assuming that in the universe we can have two types of electrons, “red” electrons and “blue” electrons. 


But wait a moment, why shall we even consider that the electron has something called spin? Even if you believe that there exist a particle which the unit electric charge which has a certain mass, why do you would need to suppose that there is something more? A quick answer to this could be to refer to the experiment of Stern-Gerlach (see Fig. 1), where they see that in a magnetic field there are two (and only two!) types of electrons, “red” and “blue”  (following the convention I anticipated before). Another clue of the existence of the spin would be the magnets that you might have sticked to the door of your fridge; in them, for “some reason”, the electrons choose to be most of them “red”. Anyway, lets assume that the electrons can have two colors.

Lets review a bit about electronics at the home level. All the computation science is based of being able to perform logic operations. By logic operation, I refer (roughly speaking!) to being able to give answers to simple logic questions like in human language would be something like: 

        Question: “February has at most 29 days and today is 31, am I in February? “

        Answer: “No”

If you ask it to a person, it looks kind of easy to answer, but the point is, how can you build a machine that does this in a systematic (and efficient!) way? The answer was given in the last century, when the first transistors where invented. This devices allow to perform this kind of operations

(more precisely called boolean) by controlling how many electrons (or more precisely how much charge) are passing. Once you have a small device that is capable of doing this you are done, putting a lot of them together and connecting them in an intelligent way you will (at very end) get a computer. Since their invention, the increasing computational power relied on the manufacturing of smaller and more efficient transistors, but this looks like cannot go much further (check the so called Moore's law).


So, a bunch of transistors are able perform (classical) logic operations by counting how much charge is passing, so you could ask yourself, why are you using charge and not other things, like for example mass? The answer would be that if you still try to use electrons, since they a very small mass, it would be kind of hard to measure their mass efficiently. If you try to use heavier stuff (like molecules), you might have a hard time paying the electricity bill, since you would have to pay a lot of energy just only to move them!


So, transporting charge works (reading this article is a proof of it!) but mass doesn't look like a good idea, but (and coming back to the beginning) why not try to transport spin? So, the goal is to rebuild all the electronic technology based of this intrinsic property of the electron, the spin, having in mind the great success achieved using the other intrinsic property, the charge. Of course the spin already plays a large important goal in our present computers, for example your hard disk drives hard based in a phenomena called giant magneto-resistance, in which the spin plays a central role. 

At this point, you might think why it should be tried  to do again with spin something that has been largely optimized with charge. One of the problems of transporting charge is that a lot of energy is lost just by heating (the electrons crash with different stuff and they move slower). This is one of the key problems (which wouldn't be a worry if there where superconductors at room temperature), but it could be avoidable if you choose to transport spin instead of charge. The reason for the last statement is that there exist some materials in which a spin current cannot stop, very much like in a superconductor a charge current doesn't stop (but their physical nature is completely different!). 

In one sentence, a topological insulator (see Fig. 2) is a material which is an insulator (i.e. electrons cannot move large distances) in the bulk of the material, but on the surface they are like in a metal (i.e. they can move freely). You might think that this is trivial, you could just take a piece of plastic (which is an insulator), and surround it by a bath of copper (which is a metal), and you would have something that matches the previous definition. Since this wouldn't be what it is referred by a topological insulator, lets just add to that definition that the system is made by just one material and that the metallic character  does not depend on whatever you do on the surface (you could have a very dirty material and it would still behave as the perfect one!!). 


Moreover, the special property of this metallic surface is that the electrons that move in their surface cannot slow down by crashing with something (in contrast with the usual metals) nor go backwards, the just go ahead forever! In can be understood as imaging the surface of the material as two road highway, one for red electrons and one for blue electrons, but this road is a one direction way (and opposite for for each color), so if you are a red electron you cannot go back unless you change your color. You already can figure out the potential application of this, if you create a transistor based on topological insulators, which relies on the flowing of spin current, you would not have energy losses by heat, and you would just annihilate one of the biggest problems in the current electronic technology.


Just to finish, just mention that in comparison with the charge currents, you can have a (non-adiabatic) spin current in an electrical insulator, so you could transmit information using spins by a material that would not allow to transmit information using charge currents, but this is another history.