Specifically, why do resistors give off heat?

[taylaron]

At the atomic and molecular level, why exactly does a resistor give off heat?
I'm trying to understand the fundamental science acting behind this device.
I read that when treated like water, forcing water through a bottleneck creates increased pressure on the smaller diameter pipe which translates to heat in the real world. But why??
From my perspecive at the atomic level the valence electrons should simply 'bunch up' like in a funnel or pipette waiting to reach the nozzle. Why must the water or current flow at the same 'rate' as in the larger diameter pipe?

Regards,
-Tay

 

[dlgoff]

In a large pipe, the molecules travel slower giving them more chance to strike the walls. Hence more pressure.

 

[waht]

In a perfect conductor, the electron are free do move. In a resistor, the electrons are constantly impeded by other atoms - as they drift. Heat is given off in a collision.

 

[taylaron]

So, can one take a non conductive polymer and impregnate it with x percent lengths of nano scale ballistic conductors (metallic single wall carbon nano tubes) and achieve partial conductivity but without the loss of a lot of heat? To my understanding, since the electrons can't travel through the polymer the only route to travel through is the carbon nano tubes which are near perfect conductors (hence the ballistic conduction).

Or, would the electrons traveling through the SWCNT's "collide" with the molecules of the non conductive polymer and create friction thus heat? (although I still don't understand how electricity would be able to flow at all in those areas because it is an open circuit thus no current and no friction).
I see it like a bolt of lightning where there are many many splits from the main discharge path, but they stop because they don't have a path to ground. Similar to the lengths of CNT's empregnated in the polymer woud act as the ionized air particles in a lightning bolt. the current would only flow through the sections of CNT's that are all touching one another creating a path for the electrons to flow.

Where the electrons flow through there would be a direct path to the other side of the polymer, there would be incredibly little resistance creating a "single position valve for electrons" without the loss of energy through collisions (heat). the conductivity can vary with the concentration of CNT's as long as the polymer is a homogeneous mixture.
Am I correct?

briefly proved here: http://www.entegris.com/Resources/Images/20257.pdf
and http://www.iop.org/EJ/abstract/0957-4484/20/40/405702