Conversation

6/ But one of these experiments was a little _too_ good. The thermopower (the ratio of temperature difference applied to voltage difference created) rose by more than two orders of magnitude in a magnetic field. Even in conditions where our theory was not obviously applicable.
Image
1
5
7/ The published explanation was that this measurement was a confirmation of our theory. But I wasn't sure, so I set my graduate student Xiaozhou Feng to think critically about whether perhaps the theory I had written down was the wrong explanation.
1
7
8/ What a good decision that was! We ended up showing that there was a different mechanism at work in this material, which gave an even more dramatic field-enhancement of thermopower.
Image
2
13
9/ Even better, this mechanism turned out to be generic for all "compensated" systems, which are materials that carry electric current simultaneously through positive and negative charges. This opens the door to a much wider range of materials as efficient thermoelectrics.
Image
1
10
Replying to
I was just reading about the MMRTG used on Mars Rovers and they seem to have an awful power efficiency of ~5% (the plutonium generates ~2000W of heat, but the rovers get only 110W of electricity out of it). Do you think there's potential for higher efficiency using your results?
1
2
Replying to
I think there's definitely potential. It's still pretty far from a specific proposal for a working device, and at this stage I still wouldn't bet even odds on this working out. But there's a core idea that I like, and I hope will go somewhere.
1
1