March 28, 2024, 06:43:06 PM
Forum Rules: Read This Before Posting


Topic: Current through an N-type semiconductor  (Read 1768 times)

0 Members and 1 Guest are viewing this topic.

Offline Irresponsible

  • Very New Member
  • *
  • Posts: 1
  • Mole Snacks: +0/-0
Current through an N-type semiconductor
« on: October 08, 2015, 09:09:38 PM »
Hi folks. I am new here and I have a simple question.

If I take a regular piece of Silicone and dope it with Arsen, I get an N-type semiconductor.

Now if I run a DC current through that simply by making it a part of a simple circuit with a powerful enough battery. Will peltier cooling happen?

I also have the inverse question. If I apply a temperature gradient to an single rod of N-type and make a simple circuit, will current flow?

Illustraded questions attached as file.

Offline Enthalpy

  • Chemist
  • Sr. Member
  • *
  • Posts: 4041
  • Mole Snacks: +304/-59
Re: Current through an N-type semiconductor
« Reply #1 on: October 13, 2015, 02:57:32 PM »
Hi Irresponsible, welcome here!

Down to the Peltier and Seebeck effects, this is how it should have worked, yes...

BUT.

The Peltier effect is so inefficient that only the best materials give you a net cooling. With random materials, like silicon an copper, the ohmic looses use to be so strong that you get heat everywhere - only a bit less heating at one end, a bit more at the other.

Peltier coolers use special semiconductors chosen for a high electric conductivity and a low heat conductivity. In metals the ratio of both is fixed because electrons conduct both, in chunks of q and 3kT/2, but in semiconductors the lattice conducts heat too, which spoils the effect because the obtained heat just flows to the intended cold side, and if you increase the current, ohmic losses put heat everywhere. Silicon is a huge heat conductor, as good as the best metals, among others because its crystal is regular and it consists of one element and almost one isotope. In constrast, working Peltier elements use II-VI or IV-VI semiconductors, to achieve a few % efficiency, sometimes >10%. Check Wiki, I believe they have a good article on this.

Getting electricity from a temperature difference is easier. Even if much heat flows directly by thermal conduction, if you achieve a temperature difference it still creates a voltage, especially if you exploit no current from it.

AND ALSO.

Flowing currents at both contacts supposes that you have ohmic contacts, not Schottky ones. Depending on process details, this can require additional steps, like heavy doping of the semiconductor near the metal.

Doping an existing crystal in depth isn't reasonable and isn't done. Imagine the dopant needs 1/4h to attain 1µm depth: because of the square law, 1mm depth takes 250,000h. Deep doping is made during crystal growth, in the melt.

Sponsored Links