My last post jogged my brain cells a bit, and I got to thinking about current flow through a ferromagnetic material.
Now, we know that a magnetized ferromagnetic material has intrinsic spin to its field, thus electrons which have the same spin can easily flow, whereas electrons with opposite spin will undergo spin flip scattering (where the electron is reflected), leading to a higher electrical resistance. Usually the electron undergoes spin flip scattering, flips its spin, and then re-enters the ferromagnetic material, whereupon it can flow with little resistance.
So as I was doing more research, I came across a diagram for a SQUID (Superconducting Quantum Interference Device):
Look familiar?
So I got to wondering... what if we used a setup like above, but rather than putting just magnetism through the round part, we put actual current flow?
Follow me here, it gets a bit complicated. Rather than using a solid core as we traditionally do, we use a bifilar coil of ferromagnetic wire. The electron's entrance to the bifilar coil's windings would be within the influence of the magnetic field, whereas the exit of the bifilar coil's windings would be outside the influence of the magnetic field... this forces segregation of the electrons into spin-up and spin-down, each going into their respective coil winding, but once through the coil, the spin current flow doesn't recombine until it's outside the magnetic influence.
What does this get us? Well, the electrons with spin-up will easily go through one of the windings of the bifilar, whereas the electrons with spin-down will easily go through the other winding. So we've subdivided the current flow into only-spin-up and only-spin-down electron flows, otherwise known as spin polarization.
In my Understanding and Exploiting Physical Phenomena thread, I mentioned the Spin Hall Effect, which is exactly what this is.
What does this get us? Very low resistance.
http://www.rpi.edu/dept/phys/ScIT/InformationStorage/magres/magnetism_b.html
When the superlattice is placed in a magnetic field, however, the magnetization of all layers will align with the external field, creating the situation depicted below. Now only conduction electrons with spins toward the left of the page will experience the higher scattering rate. Thus the resistance of the material decreases in a magnetic field.
So we're using a bifilar coil as our 'superlattice', and we place it in a magnetic field. The bifilar coil is wound such that spin-up electrons and spin-down electrons are equally influenced by the applied magnetic field. Thus we get very little spin flip scattering once the electrons are in the bifilar coil (although at the entrance to the bifilar coil there will be spin flip scattering and electron segregation as spin-up and spin-down electrons flow into their respective windings of the bifilar coil).
Thus the conduction electrons of both spin-up and spin-down experience lower scattering rates inside the bifilar coil, and thus the resistance of the bifilar coil should be very low. This is the inverse of Giant Magnetoresistance (GM)... and we know GM can result in as much as 100x higher resistance than normally exhibited. So the resistance in the bifilar coil should be very low indeed.
Spin-up and spin-down currents circulate in opposite directions under a static magnetic field (and thus they usually cancel when within the same wire, which is why we don't get a current flow under a static magnetic field), so with proper winding of the bifilar coil, we can force them to flow in the same direction (toward the bifilar coil's exit wires) under influence of the magnetic field. Thus we get an additive spin current flow. We're taking electrons which normally would go in opposite directions under a static magnetic field, thus cancelling the voltage effect of that magnetic field, and we're getting them to go in the same direction, generating an additive voltage.