+- Hicks and Dresselhaus 15 Years On

MIT issued a press release Nov. 20, 2007 briefly describing some of Dr. Mildred Dresselhaus’s thermoelectric work and drawing attention to her invited talk Monday Nov. 26, 2007 at the Fall MRS Meeting in Boston (Symposium U, Thermoelectric Power Generation, paper U2.4, 2:30pm).

She’s a great speaker and a world class scientist.  If can, go to her talk.  If you can’t make her talk, check out  the Dresselhaus Group homepage.

The original announcement from  The MIT Press  has been widely reproduced.  A Goolge search on the title of the MIT Press release yields:

And those are just from the first page of the Goolge search.  You must admire the MIT PR guys:  they do get the word out..

The institutions Dresselhaus and MIT are world class, among the best and I begrudge them nothing.  PRs are necessarily brief and energetically optimistic.   But given the attention in this case I thought I’d comment on a few claims from the PR that caught my attention:

  1. "Novel thermoelectric materials have already resulted in a new consumer product: a simple, efficient way of cooling car seats in hot climates."
  2. "Her work …  invigorated the field, and now there are real applications like seat coolers in cars."
  3. "People saw that paper and the field started"
  4. "Now there are conferences devoted to it."

The first two claims concern the Amerigon Climate Control Seats (CCS), which must be the most sucessful thermoelectric product to date.  There are two problems with claims 1 and 2:  first the CCS use old bismuth-telluride-based materials which haven’t been novel for 40 years and second the MIT work probably had nothing to do with the development of these products.  Certainly true that "Her work … invigorated the field" and also true "there are now real applications like seat coolers in cars" but as far as I am aware there is no causal connection between the two.

The third and fouth claims are also true in a way:  people did see the Hicks-Dresselhaus paper(s) and around that time the field really grew.  But it didn’t start then, it grew.  And today there are indeed conferences devoted to it, but there were then too.  Only smaller and fewer.  No question that Dresselhaus contributed mightily to the growth, but there were other equally important contributions as well.  Technically I’d cite off-hand the co-discovery of the Skutterudites independently by Slack and Caillat, Slack’s coining of the "Phonon-Glass, Electron-Crystal" concept, and of course Venkatasubramanian’s phonon-blocking superlattices.  Also crucial was the initial interest from  Dr. Rudy Buser, Director of the Army Night Vision Lab and then by Stuart Wolf (DARPA) and  John Pazik (Office of Naval Research) who, unlike Buser, actually provided R&D funds.

To be sure, it is hard to imagine what thermoelectrics would be like today without Dresselhaus, and I’d rather not try.  But the PR gives the definite impression that it is a one-trick pony which is far from true.

I’d also like to remark on what I view as the real significance of the early papers by Hicks and Dresselhaus (which I’ll call HD).  First the theoretical paper on quantum wells and then one on wires:

Hicks, L.D. and M.S. Dresselhaus, Effect of quantum-well structures on the thermoelectric figure of merit. Phys. Rev. B, 1993. 47(19): p. 12727-12731.

Hicks, L.D. and M.S. Dresselhaus, Thermoelectric Figure of Merit of a One-Dimensional Quantum Wire. Phys. Rev. B, 1993. 47(24): p. 16631-16634.

The quantum well paper concluded they had the "potential (empahsis in the original) to increase ZT by a factor of 13".  I can tell you this did indeed "invigorate" the field.   Probably the experimental paper referred to in the PR is this one

Hicks, L.D., et al., Experimental study of the effect of quantum-well structures on the thermoelectric figure of merit. Phys. Rev. B, 1996. 53(16): p. R10493-R10496.

which essentially confirms the Seebeck is enhanced more or less as described by the theory.  Nice work and everyone in the field should be familiar with these papers.

Today, you don’t hear as much talk about this sort of ‘quantum confinement’ and the focus is more on the phonons.  The Seebeck is indeed enhanced, but other ‘bad’ things happen too.  Lower mobility, for one.  Nothing close to the hoped for improvements has yet materialized due to this physics.  What we’ve seen for the HD ideas is perhaps tens of percent improvement in ZT, far less than the initial hopes.

It is possible, of course, that Seebeck enhancement is going on in some of today’s ZT~2+ ‘nanocomposites’, compensating for the loss of mobility and allowing lower thermal conductivity to play out in higher ZT values.  These are not simple matters and much R&D is still needed to sort out exactly what’s going on.  But it is far more complex than the initial HD ides.

Make no mistake:  the HD work did not lead to cooling car seats or any other commercialization that I’m aware.  Nor has it, by itself, resulted in really large ZT value execpt on paper..  Not yet, anyway.

For me, this does not diminish the importance of the HD contributions whatsoever.  Progress is rarely in a straight line.  The real significance of the HD work is that it allowed serious people to be optimistic about thermoelectrics after a long period of pessimism.  HD outlined a physically plausible, eminently testable approach with sufficient detail that anyone practiced in the art could pursue.  Also it  had a decent chance of really large improvements and came from credible people who were not previously in thermoelectrics, and didn’t need to be.  Taken as a whole there is no question this invigorated the field. 

That HD has, so far, come up short is a disappointment, of course.  Pauling said the key is "you have a lot of ideas and you throw away the bad ones."  In the early 1990’s I was looking for 6-10 ideas each of which had a 5% chance of working.  That way you’d get an even chance, or better, that one of  them would work.  If we could figure out winners in advance, we wouldn’t need so many great scientists like Millie Dresselhaus to work everything out.  But we can’t, so we do.

For the curious (or the glutton for punishment), I gave a paper at ECT2007 on  the last 15 years of thermoelectrics which discusses this history further.  You may download from this link:

Vining, C.B. ZT ~ 3.5:  Fifteen Years of Progress and Things to Come. in European Conference on Thermoelectrics, ECT2007. 2007. Odessa, Ukraine.