The Role of Phonons in High Temperature Superconductivity

[staf9]

I'm trying to understand a bit more about high temperature superconductors. I'm familiar with BCS theory and a little bit about high temp superconductors, but I'm still confused about a bunch of stuff

Why exactly is copper oxide used in high temperature superconductors?

Why is this specific compound of copper used? What properties does it have that makes it so special?

If we took a step down the periodic table to silver would it all of a sudden set high temp superconductors back to only working at 30K instead of the 130K copper oxide affords us? I know the valence electrons in both molecules are the same, but there are other factors that come into play lower down the periodic table.

The only real reason I can think of right now is because it's cheap. If anyone could shed some light on this (and tell me where I'm wrong in any of the statements I've already made) it would be great

Edit: Had a couple questions answered already, my only real question now is could anyone tell me the relationship between the substance used and the temperature required for superconductivitiy?

 

[malawi_glenn]

hmm this is not nuclear- and particle physics. This is in atomic- and solid state forum.

 

[f95toli]

Since other people might be interested as well I will try to answer all your questions, not just the last one

Short introduction:

All known high-Tc superconductors have a crystal structure that include copper-oxide planes. These planes are the "superconducting" part of the structure; the other ions (ytterium and barium in YBCO) are belived to only have a "passive" role in that they only work as "spacers". There are also copper-oxide chains believed to work as charge reservoirs not direcly involved in electric transport.

Why exactly is copper oxide used in high temperature superconductors?
Why is this specific compound of copper used? What properties does it have that makes it so special?

Because the copper-oxide planes is what makes the material superconducting. There is nothing "special" about the copper or the oxygen expect that in this kind of crystal they form planes. Note that the high-Tc materials are not alloys; the elements are not randomly distributed throughout the material; each atom MUST be in the "right place" in order for the material to be superconducting. Simply mixing the right elements won't create a superconductor; the material it must have the right crystal structure (and the right oxygen doping but that is another story).

If we took a step down the periodic table to silver would it all of a sudden set high temp superconductors back to only working at 30K instead of the 130K copper oxide affords us? I know the valence electrons in both molecules are the same, but there are other factors that come into play lower down the periodic table.

If you replaced the copper by silver the material would not be superconducting. Even if copper and silver had the same electronic structure (valence electrons etc) it would still not be possible to create the right type of crystal structure using silver (the "size" of the ions matter).

The only real reason I can think of right now is because it's cheap. If anyone could shed some light on this (and tell me where I'm wrong in any of the statements I've already made) it would be great

Edit: Had a couple questions answered already, my only real question now is could anyone tell me the relationship between the substance used and the temperature required for superconductivitiy?

Cheap has nothing to do with it. Superconductivity is a "property", you can't just replace one element with another and expect to get an identical result (this is true in general in solid state physics). There are many different high-Tc superconductors (with different critical temperatures) but they ALL incorportate copper-oxide plane.

If someone could answer your last question he/she would get a Nobel prize. We still do not know WHY these materials become superconducting, this is one of the great unsolved mysteries in physics. We know that the copper-oxide planes and the layered structure are the two most important charachteristics, but so far no one has been able to figure out how it works; but we do know that you can't explain it using BCS theory.
Some of the BCS results are valid and the high-Tc materials obviously have a lot in common with conventional superconductors; but there are important differences as well and they can NOT be explained using any known theory.

These materials have been around for 20 years know and a HUGE amount of work has gone into understanding them. We do know a LOT about these materials but they are extremely complex due to their complicated structure (we do not even understand them fully in the normal state). They are also difficult to work with experimentally and it took many years to figure out how to make relatively "clean" experiements (many of the early results are quite frankly wrong, mostly due to bad sample preparation); most of the experiment that one day (hopefully) will give use the "key" to understanding how they work have only been done over the past few years (e.g STM studies of very good crystals, studies of infinite layers including superconductivity in a single atomic layer etc).

I should point out that this work HAS lead to a lot of "spin-off" effects: We have e.g learned how to grow thin films of very complicated compounds. Work that came out of the studies of the "macro-structure" (grain boundaries etc) has turned out to be important in many other "ordinary materials". The work on high-Tc has also given rise to a lot of the work being done on oxides that are not superconducting, e.g. many magnetic materials (e.g. CMR effects), ferroelectrics etc. So it has not been a waste of time.

 

[ZapperZ]

I'd like to add that there are other layered materials that also are superconducting, albeit at lower Tc. BKBO, the ruthenates, and even some cobaltates are also superconducting. So superconductivity in layered compounds isn't just restricted to just the cuprates. However, the cuprates are the only ones so far achieving those high Tc's.

Also note that even within the cuprates themselves, there could easily be a variety of "phenomena" that may make one cuprate compound different than another. There are already major differences between hole-doped cuprates (YBCO, BSCCO, LSCO, etc..) and the electron-doped cuprates (NCCO). So the "inert" doping of the rare-earth material may have some influence in its properties. Furthermore, within the hole-doped family itself, there can be dramatic differences. The phase diagram of LSCO, for example, has a "dip" at around 1/8th doping (I think, I'm pulling this off the top of my head) where Tc is completely suppressed. This doesn't occur with the other hole-doped cuprates.

So yes, as have been stated, the "recipe" and what is used in the cuprates are highly complex questions that are still be studied extensively. There is a huge body of experimental data that are out there, and they all have something different to say.

Zz.

 

[staf9]

Thanks very much for the explanations f95toli and Zz, looks like I still have a lot to learn about superconductivity.

If i have any questions in the future I'll post them in the right forum this time too!

 

[yuanyuan5220]

The 2-dimensional 1/2 spin Heisenberg model is the key problem in the copper oxide superconductor.
For the 1-D 1/2 spin antiferromagnetic Heisenberg model, it has been exactly solved by bethe ansatz. And the low-energy excitation can be analyzed be many methods, such as Bosonizaiton, RG...
In the 1 D model, the two-particle interaction is easy to handle, there are only four interaction forms:forward scattering, backward scattering, dispersion scattering, umklapp scattering. The topology of 2 D is totally different from 1 D model, and the two-particle interaction is extremely complicated. And in the strong-coupling condition, the mean field is always useless. This is why we still don't understand the 2-D 1/2 spin AF model. Is it has long range order, or just a spin liquid? can this model leads to the electron pairing?

All these problems needs answering

 

[ZapperZ]

The 2-dimensional 1/2 spin Heisenberg model is the key problem in the copper oxide superconductor.
For the 1-D 1/2 spin antiferromagnetic Heisenberg model, it has been exactly solved by bethe ansatz. And the low-energy excitation can be analyzed be many methods, such as Bosonizaiton, RG...
In the 1 D model, the two-particle interaction is easy to handle, there are only four interaction forms:forward scattering, backward scattering, dispersion scattering, umklapp scattering. The topology of 2 D is totally different from 1 D model, and the two-particle interaction is extremely complicated. And in the strong-coupling condition, the mean field is always useless. This is why we still don't understand the 2-D 1/2 spin AF model. Is it has long range order, or just a spin liquid? can this model leads to the electron pairing?

All these problems needs answering

Don't be so sure. The "phonon camp" would argue against you on that.

Zz.

 

[staf9]

Don't be so sure. The "phonon camp" would argue against you on that.

Zz.

Since phonons are bosons with integer spins because of their origin from quantizing vibration in a crystal lattice, and what yuanyuan went over included the 2-D 1/2 spin Heisenberg model, right? Or am I completely off? (just trying to get a better understanding)