Mixed Signals
Any sufficiently advanced information is indistinguishable from noise
Thursday 12 March 2009, 1:21 PM
A new spin on battery technology
The MJT is the top part of the illustration, and is roughly the same diameter as a human hair. Beneath that is a magnified image of the central part of the device: the white spots are atoms, and the circles contain the nano-magnets that store the power. Picture: Phan Nam Hai/University of Miami
I've yet to dig out the paper in Nature to find out how far this is from being useful as a power source - as the current device is a few hundred micrometers across, it's not going to be storing megawatts. But it's the sort of thing that could be created in vast arrays, like any semiconductor device, and if they're getting to the bottom of the underlying physics then the same effect could be used in many different configurations. What may be much more interesting than just power storage is the fact, mentioned almost in passing in the press release, that the current delivered by the MTJ is spin-polarised; the electrons are predominately spinning in one direction.
That's hot news for spintronics, which we've covered previously and which, together with graphene, has the most exciting potential for fundamentally new computational devices. Spin logic could work much faster at much lower power than even today's finest electronics, because it doesn't rely on currents flowing and the consequent unavoidable loss.
And, as the researchers say, if this discovery leads to new insights into basic magnetic theory, there are almost no limits to how profoundly it could affect modern life. Which is a bit over the top - just not that much.
PreviousComments on this post
This voltage that's '100 times too big', are we talking microvolts, millivolts, volts? How many hairs for a battery? Also, unless something moves between them (no current flow, you say), how would spintronics components inter-react with each other, surely a basic requirement for any kind of logic engine? Sorry if these are doh basic questions.
I haven't got the paper yet, so I can't say -- all they're saying is that the effect was much bigger than they expected, and that's huge news for physicists. I'd love to know about internal resistance, energy conversion efficiency and all of the other things that really matter in a battery, and I'm sure I'll find out - although the really interesting thing is the theoretical limits. Perhaps those aren't known yet.
This isn't a spintronics logic device. It's just a battery. It does, however, produce spin-polarised electrons, which are essential for spintronic logic devices and are remarkably difficult to produce in useful ways (at least, they were last time I looked!). It's one of the major problems, so if there's a new way to do it, that's got huge implications.
It's not that spin-polarised electrons don't flow: they do. They're just electrons. But they don't _need_ to flow to be useful in spintronics, so you can do lots of computation, in theory, without having to move the electrons at every step. You could have a spintronic NOR gate, for example, that has clockwise spin as 1 and anticlockwise as 0; if both inputs have anticlockwise electrons, then the output would be clockwise - and that output can be used as an input in a coupled gate.
Doesn't have to work that way, but it's possible.
This comment has been deleted at the users request
This comment has been deleted at the users request
