[This is a “literature impression“.]
Masahiro Hotta has a series of paper about what he calls “quantum energy teleportation (QET)”, modeled after the well-known notion of quantum teleportation (of information). Although it sounds like crazy crack pot stuff, and they contain the red-flag term “zero-point energy”, the basic physics of Hotta’s work are sound. But they don’t appear to have important consequences for energy transmission.
The idea is to exploit the fact that the ground state of the vacuum in QFT is, in principle, entangled over arbitrary distances. In a toy Alice and Bob model with respective systems and , you assume a Hamiltonian for which the ground state is unique and entangled. Then, Alice makes a local measurement on her system . Neither of the two conditional global states for the joint system — conditional on the outcome of the measurement — are eigenstates of the Hamiltonian, and so therefore the average energy must increase for the joint system. The source of this energy is the device Alice used to make the measurement. Now, if Bob were to independently make a measurement of his system, he would find that energy would also necessarily flow from his device into the joint system; this follows from the symmetry of the problem. But if he waits for Alice to transmit to him the outcome of her result, it turns out that he can apply a local unitary to his system and a subsequent local measurement that leads to a net average energy flow to his equipment. The fact that he must wait for the outcome of Alice’s measurement, which travels no faster than the speed of light, is what gives this the flavor of teleportation.
The first thing to note is that this protocol would work just as good for classically correlated states as for quantum mechanically entangled states. Think of a classically correlated pair of bits. Each has entropy, locally, which can be thought of as lacking the resource of negentropy. But if you measured one (which consumes negentropy) you can extract negentropy from the other (because you then know it’s pure state, and you can get useful work by allowing it to thermalize with any non-zero temperature).
So this isn’t really quantum, except for the important fact that in QFT the vacuum is an entangled state. In other words, correlated states are just lying around “out there”, and don’t need to be generated (using negentropy) like they would in the classical case. This is what might lead one to hope that we can use this naturally available entanglement for energy transmission.
However, the distance this can work over turns out to be very, very tiny. To send of energy, it only goes over the corresponding relativistic quantum distance scale, i.e. . Basically, for a particle with de Broglie wavelength determining a certain energy, you can’t send the energy further than the order of a single wavelength! This distance limitation can be fixed by starting with squeezed vacuum, but squeezed vacuums aren’t just available out there; you need to create them, which takes negentropy.
Therefore, this idea does not appear to be useful. It does make me wonder, though, whether you can use the vacuum state for “old fashioned” quantum teleportation of a quantum state. (Presumably you’d use a very low energy mode with wavelength comparable to the distance you want to operate over.) I assume this would never work, but I haven’t thought about it.
Added 2014-2-13: Apparently, Benni Reznik has written on distilling EPR pairs from the vacuum.
[Most of the insight in this post is due to Charlie Bennett.]