Most thermocouples that I've seen being sold online consist of two separate metal wires connected at junctions. Many smaller thermoelectric generators consist of a combination of iron and copper or iron and aluminum. However, I'm a tad confused why two different materials are being used.

If I take a look at the image below, I see that an n-type and p-type semiconductor is attached to a conducting wire with a heat source and heat sink on opposite ends. Do the semiconductor materials need to be different? For instance, can't I simply use n-type and p-type doped germanium? Why do I need to use two distinct metals? Is there a benefit, for instance using silicon for one and germanium for the other?

Does the circuit wire need to be a special alloy, or just regular copper wire to transfer current?

Ultimately, I'm a bit confused regarding the Seebeck Effect. I've seen this diagram, but also models with different alloys. Any clarification will be great!

Science ABC, Venkatesh

(Image by Science ABC - Venkatesh Vaidyanathan)

  • 1
    $\begingroup$ Look up Peltier and Thompson effects. Thermocouple technology will also explain the principles. $\endgroup$ Jun 29, 2021 at 15:46
  • $\begingroup$ I don't know how Peltier and Thompson effects help me understand a Seebeck TEG. I have read about thermocouples, but I don't understand the need for two different alloys when it comes to current generation. I need some clarification whether two similar p and n type semiconductor might work. $\endgroup$
    – ARJ
    Jun 29, 2021 at 19:15

2 Answers 2


A conductor, hot on one side, and cold on the other will produce a voltage proportional to the Seebeck constant and temp gradient. In order to flow current, you need to connect the hot side to the cold side with a conductor. The conductor will now equalize its temperature at its respective connection points. If the conductor was made from the same material as the heated metal then it would now have the same voltage potential (due to the same seebeck coefficient) so no current would flow. This is why you need separate materials, with different seebeck constants



It is possible to make a thermoelectric generator from a single metal, but there needs to be some change in the Seebeck coefficient at the heated "junction". Forming the junction from two different metals is not the only way to do this, but it is the easiest and most scalable way.

As is discussed in the answer to Seebeck Effect in a homogenous material on Physics SE, a thermoelectric junction can be made by changing the thickness of the metal from larger to smaller than the electron mean free path. Single-Metal Nanoscale Thermocouples have been constructed based on this.

One can also expect differences in Seebeck coefficient in metals that are otherwise identical except for their microcrystalline structure, or in semiconductors that are doped differently on either side of the junction.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.