Re: How is heat transported by a Peltier Junction?

Fig.1 shows two metal strips and a battery connected together. We know that metal has some "binding" forces on the conductive electrons moving in it. When we pull out a electron from a metal we need to do some work, therefore a electron in a metal has some negative potential energy compared to a free electron. (If that seems difficult to understand, imagine a ball. If a ball is at a higher position, it has some positive energy because when it falls, this energy will convert to its kinetic energy. However, if it is at a lower position, we need to do some work to bring it up to our level, therefore it has a negative potential energy when it is at a lower position.) In fig.1 the "green metal" has a lower energy level than the "red metal". At the upper junction, when a electron moves from the green metal into the red metal, part of its kinetic energy will convert to potential energy and it will move slower. Since the temperature is just a representation of the average speed of the random movements of the particles, the temperature at the upper junction seems to decrease. Of course at the upper junction there are also electrons moving back from the red into the green, but a electric current means a net electron flux from the green to the red, therefore a net cooling effect. Similarly, At the lower junction, there is a net flux of electrons from the red to the green, therefore a net heating effect.

This explanation is rather crude. There is a very good article on the web about the Peltier effect in semiconductors [1], it can also help you understand the Peltier effect at the junction of two different metals.

References

[1]http://jchemed.chem.wisc.edu/Journal/Issues/1996/Oct/abs940.html.