colegu said:
It's the second time I've posted this message, and my english simply sux, so excuse me if I will not be able to explain my misunderstanding quite clear.
Let's imagine a verry long metal cable, which is heated at one head. The quasi-free electrons at that head, as their energy increases, must change their quantum state, but there must be a way for them to *know* what the other electrons quantum states are, in order not to occupy the same state.
So how does a fermion *know* what states are not possible to ocupy?
Your question actually isn't that elementary to answer.
First of all, when we solve for the
ideal case of a conductor using such free-electron gas, we make the explicit assumption of "plane wave states" and "very large distances" for the boundary conditions. We then obtain the typical fermion distribution of states. Now, under most circumstances, this is OK because typically, the scale of system in question is large enough that, according to the electron, it is boundless and all it cares about are its immediate environment.
However, this is where your example is no longer applicable and has breeched the assumption made. In an ordinary metal, the electrons have no
long range coherence. This is because the scattering with the lattice ions and other electrons would destroy such coherence very quickly. So essentially, the Exclusion Principle is only a
localized principle, maybe up to an order of 10^3 lattice constants (this is a very rough estimate). Only when you get a superconducting state would there be such long-range coherence for the entire solid. In the latter case, all the electrons are already in a single state that is non-local.
Secondly, even after mentioning the above, you need to remember that the scattering into the high energy states isn't "static". Electrons continue to scatter in and out of those states due to thermal fluctuations. So cannot assume that only one electron will occupy a particular state and that's that. As with the Cooper Pair states in a superconductor, the electrons continue to scatter in and out of the states, like a game of musical chair. As soon as one scatters out, another one will occupy that state. At equilibrium, these scattering process reaches a statistical equilibrium.
Lastly, you need to remember that all the momentum and energy states are
continuous for the conduction band. This means that there are infinitesimally high or lower states that are part of the density of states. So it doesn't take that much to find an available state to scatter into.
Zz.