High Tc superconductors turn 20

[ZapperZ]

This past Sept. marked the 20th anniversary of the discovery of High-Tc superconductors. It was a discovery that turns physics, and especially condensed matter physics, upside down. A subject area that was thought to be 'dead' and fully matured, where we thought we knew everything that we were supposed to know, suddenly started revealing a whole new side that were never thought to be possible before. Certainly, there were no theoretical insights into what were to come next during the following years. Certainly, there probably would never again be (at least in my lifetime) the "Physics Woodstock" as the one that happened during the APS March Meeting in NY right after the discovery (although something similar did happen in 2001 in Seattle after the discovery of superconductivity in MgB2 - we called that Physics Woodstock West).

The revolution in the condensed matter that started out by this discovery affected ALL aspects of that field of study. Suddenly, strongly-correlated electron system, which permeates all of condensed matter, has a very prominent poster child in the form of high-Tc superconductors. The understanding that we got out of this material provided insights into a bunch of areas, some even beyond condensed matter (i.e. Laughlin's and company connection of the High-Tc phase diagram to the quark phase diagram).

This week's issue of Science (Science, 17 November, 2006) has a terrific recount of the history, difficulties, and future of High-Tc superconductors. Don't miss it if you have access to the journal.

Zz.

 

[Astronuc]

I think this work might be published in that issue.

http://www.bnl.gov/bnlweb/pubaf/pr/PR_display.asp?prID=06-122

UPTON, NY (November 16, 2006) - Research published online in the journal Science this week by Tonica Valla, a physicist at the U.S. Department of Energy’s Brookhaven National Laboratory, appears to resolve one mystery in the 20-year study of high-temperature (high T_c) superconductors — materials that lose their resistance to the flow of electricity at relatively high temperatures. The research shows that a “pseudogap” in the energy level of the material’s electronic spectrum is the result of the electrons being bound into pairs above the so-called transition temperature to the superconducting state, but unable to superconduct because the pairs move incoherently.

In conventional superconductors, which operate at much lower temperatures (near absolute zero), superconductivity occurs as soon as electron pairs are formed. But in the case of the high-T_c materials, the electrons, though paired, “do not ‘see’ each other,” Valla says, “so they cannot establish ‘phase coherence,’ with all the pairs behaving as a ‘collective.’”

The origin of this pseudogap, along with the mechanism for forming the pairs necessary for superconductivity, has been one of the biggest mysteries scientists have been trying to understand about high-T_c superconductors since their discovery some 20 years ago. Because of their higher operating temperatures (up to 134 kelvins at ambient pressure and up to 164 K under high pressure), high-T_c superconductors have much greater potential for real world applications, such as zero-loss power transmission lines, than do conventional superconductors.

 

[ZapperZ]

I think this work might be published in that issue.

http://www.bnl.gov/bnlweb/pubaf/pr/PR_display.asp?prID=06-122

Wow. Is that a familiar sight! :) I was a part of of the same group while I was at bookhaven, and Tony's picture is taken at the very same ARPES setup that I worked at. :)

It would help if the news ariticle cited the work (it isn't in that Science issue, but I'll double check). I follow the preprint coming out of my old group pretty closely, and none of the preprints they have online so far fit that description (I didn't know they were working on the LSCO system AND with Seamus Davis, who is at Cornell). I have a feeling they submitted this either to Nature or to Science, which is why they didn't put it online yet.

I will have to wait for the paper to see how convincing it is. If these preformed pairs really are the ones to eventually condense into the superfluid, I'd like to see how they explain why Tc drops even as the size of the gap increases in strength (signaling an increase in pair coupling) as one underdopes the material.

Zz.

 

[Gokul43201]

Sounds like this is what Kivelson and others had predicted in the Nature paper. I wonder what Randeria thinks - I recall that he too had proposed preformed pairs (but, I think, with short coherence lengths instead of large phase fluctuations).

Zz, when the paper comes out, don't forget to review it in the Noteworthy Papers thread.

 

[ZapperZ]

Sounds like this is what Kivelson and others had predicted in the Nature paper. I wonder what Randeria thinks - I recall that he too had proposed preformed pairs (but with short coherence lengths instead of large phase fluctuations).

Are you referring to Emery and Kivelson Nature paper that has that "X" on the phase diagram? Vic Emery had an office in the same hallway as I did at BNL before he passed away. We talked a little bit about it, and he was the one who gave me the idea that pre-formed pairs can occur without long-range coherence. Of course, I also didn't completely buy the stripe model that he was pushing, but the idea of pre-formed pairs did have some weight to it.

I think Kathy Levin's group at Chicago has the opposite scenario where the pre-formed pair competes against superconductivity, and that these do not eventually condenses. It would be interesting to see how this turns out.

Zz, when the paper comes out, don't forget to review it in the Noteworthy Papers thread.

I'll try!

Zz.

 

[Gokul43201]

Are you referring to Emery and Kivelson Nature paper that has that "X" on the phase diagram?

Yup. That's the one I think I recall.

 

[Lonewolf]

And we still have posters in our department celebrating the first five years...

 

[Gokul43201]

It would help if the news ariticle cited the work (it isn't in that Science issue, but I'll double check).

http://www.scienceonline.org/cgi/content/abstract/1134742v1

 

[ZapperZ]

Thanks, Gokul. As I suspected, they submitted this to one of the Big Two.

Unfortunately, since I'm not a member of AAAS, I'll have to wait for it to appear in the actual Science journal to get the full article.

Zz.

 

[ZapperZ]

Heads up!

The Stanford people have submitted, also to Science, a COMPETING paper.

http://arxiv.org/abs/cond-mat/0612048

In this one, they didn't study the LBCO system as the Brookhaven group did, but on the underdoped BSCO system. And they saw two distinct gaps - one along the anti-nodal direction that corresponds to the predominant gap seen in results so far and attributed as the pseudogap, but another along the nodal direction which does not get larger as one underdopes.

Their conclusion? The pairing that resulted in the pseudogap is NOT the precursor to superconductivity. These are not the pairs that will eventually form long range coherence below Tc.

I LOVE IT!

If Science practices the same thing as PRL, they will put thesse competing papers back-to-back in the same issue. This will heat up the debate and creates a lot of discussion and excitement for months or even years to come!

Zz.

 

[ZapperZ]

I have highlighted the Brookhaven paper, which appeared online in ArXiv over the holidays, in the https://www.physicsforums.com/showpost.php?p=1200941&postcount=33" thread. I will write my analysis of this paper on here, since that thread isn't meant for such a discussion.

First of all, a few background. When I was doing tunneling and photoemission spectroscopies, I was clearly leaning towards the same conclusion that the Valla et al. came to. So obviously, you should know a little bit of my bias in this area. However, now that I am technically out of that field, while I still think that conclusion is more convincing, I also see quite a few measurements that introduce some reason to also make the other scenario rather compelling (see the preprint from Stanford in this thread). So as an "outsider", this makes great drama! :)

Secondly, as I've mentioned, I used to do tunneling (my Ph.D research work) and photoemission (my postdoc work) spectroscopies. So the Valla et al. paper covers both of my expertise to a "T", since it is reporting both photoemission and tunneling results.

OK, on with the paper...

I find the "trend" reported in the paper very compelling, especially those shown in Fig. 2. These, to me, are the heart and soul of the whole paper. It clearly shows that, unless one believe in utter cosmic coincidence, the pseudogap pairs act like a duck, walk like a duck, and quack like a duck, where the "duck" here being the superconducting pairs. So if you buy these observations, you'll buy the conclusion of the paper.

However, and this is where we come to the fun part, there are issues.

1. There are no sharp peaks in the photoemission data, especially from the energy distribution curves (EDC). See Fig. 1. Now, in photoemission, a sharp peak in the EDC signifies the presence of well-defined Landau quasiparticles. It is well-known that in the pseudogap phase, photoemission spectrum shows no such peaks. In many cases, this has been attributed to the strong many-body effects that causes the Landau quasiparticles to no longer be a well-defined concept. Here, an "electron" or a "hole" with spin, charge, lifetimes, etc.. are no longer "good quantum numbers". If this is the case, then the language used in the paper isn't consistent. The "single-particle excitation gap" may no longer be a valid concept, because single-particle states are no longer well-defined. This is not a show-stopper, but it does give plenty of "wiggle room" for the other scenario.

2. If the pseudogap here is the same as the paring energy of the Cooper pairs but without the long range coherence, then the lack of any temperature dependence is troubling. See Fig. S2 in the supplimental section. This has been observed previously (I've measured this myself) and became one argument for the scenario that the pseudogap pairs are NOT the pre-formed pairs that condenses into a superfluid. So if the pseudogap pairs are really the pairing pairs, this is one aspect that do not mimick a duck.

[As a side note: as part of my doctoral dessertation, I studied the temperature evolution of the gap as measured in tunneling spectroscopy as one increases the temperature. Here, while the superconducting gap size reduces at T approaches Tc, I also introduced temperature effects in terms of an increase in scattering factor via the quasiparticle lifetimes. One can model it in such a way that even as the superconducting gap is trying to close as it approaches Tc, the increase in the scattering rate will prevent the tunneling gap from closing as quickly. So it will appear as if the observed gap isn't changing as much as you increase the temperature while still in the gapped state. So while it would be nice to see evidence that the pseudogap evolves with temperature, this isn't a showstopper for them either since I can easily show that an appropriate choice of scattering rates can mimick the same experimental evidence.]

3. Is the gap really a gap in the sense of states becoming unavailable due to some kind of phase transition, or is it just band structure (or some many body variation of band structure)? If it is a band structure related issue, that might explain the lack of temperature dependence. It would be a weird band structure (Fermi surface points) but such a many body system is supposed to have such a weird Fermi surface anyway.

OK.. my fingers are tired. That's it for now.

Zz.