[kaypep (OP)]

Fast heat transfer in metals are claimed to be a result of electron diffusion, the movement of free mobile electrons that spread kinetic energy faster than vibrating metal cations can collide with one another.

Since heat causes the movement of electrons, wouldn't it cause a flow of electrons that is able to produce a current?

Thanks!

[graham.d]

It is more to do with the electron mobility rather than a flow. The mobility means that the electrons can pass on the energy quickly rather than there being any net movement.

[lyner]

There is no net charge transfer - the fast electrons move into the 'cooler' regions and slower electrons fall into the resulting positive charge of the region which has just lost some fast electrons. Remember, there are huge electric forces inside solids which keep them together and in equilibrium.

[kaypep (OP)]

It is more to do with the electron mobility rather than a flow. The mobility means that the electrons can pass on the energy quickly rather than there being any net movement.

Yeahh, but the electrons still move right? Electricity is about electron movement, isn't it?

There is no net charge transfer - the fast electrons move into the 'cooler' regions and slower electrons fall into the resulting positive charge of the region which has just lost some fast electrons.

but the same thing's happening when you pass electricity through a metal. electrons move from the negative to positive terminal, across the metal.
electrons move from the hotter to cooler region, across the metal.

[graham.d]

Electricity is about electron movement but even here individual electrons don't move very far, it is just that there is a net flow of electrons in one end and a equal net flow of electrons out the other end. They actually jiggle about a lot in all directions. With thermal conduction it is just their jiggling about that is passed on and there is no net flow. As you say, in a crystal lattice the atoms are fixed and it is a transfer of the vibrations that constitutes thermal flow. With a conductor it is the freely moving electrons that allows a more rapid thermal transfer. They are a conduit for transfer of an average vibrational energy that constitutes thermal energy.

[lyner]

individual electrons don't move very far

How very true. Only a matter of mm in a second, average for conduction electrons. I never thought about it before but the actual vibrational, random motion is quite a bit faster than that but over a small distance before it interacts with another charge. The Fermi speed of electrons in a metal is about 1.5million m/s!!!; that corresponds to the speed corresponding to the average kinetic energy of an electron at room temperature, but they only travel about 40billionths of a metre before bumping into something. Crazy numbers, don't you think?
See http://hyperphysics.phy-astr.gsu.edu/hbase/electric/ohmmic.html
Thermal conduction along a bit of copper must involve an awful lot of collisions; each one transferring a bit of energy in the 'cold' direction.

[chris]

Very interesting question, and the simple answer to whether heating a material can drive a current is YES!

This is the thermoelectric effect and is used to power space probes venturing to distant reaches of the solar system where solar power would be insufficient. "New Horizons", the probe bound for Pluto, is powered this way. Usually the heat supply is a radioactive source such as polonium or plutonium which decays sufficiently vigorously to produce a strong heating effect.

This heat source is coupled to a thermoelectric material, which is some kind of semi-conductor. The warmth shakes electrons free from one component of the semi-conductor, they are harvested by an electrode and sent around a circuit, returning to the cold-size of the semi-conductor to complete the circuit.

Recently scientists have developed a highly efficient thermoelectric material based on lead telluride doped with thallium that can harvest heat from car exhausts. The researchers say this could make the petrol in the tank got 10% further by building the system into a hybrid car:

http://www.thenakedscientists.com/HTML/content/news/news/1440/

Chris