Wednesday, February 20, 2013

Are elemental metals quantum critical?

I doubt it.

There is an interesting paper
Similarity of Scattering Rates in Metals Showing T-Linear Resistivity
J.A.N. Bruin, H. Sakai, R.S. Perry, A.P. Mackenzie

The central result is the figure below.

The graph shows the magnitude of the estimated scattering rate per Kelvin versus the Fermi velocity for a wide range of materials.
The line alpha=1 corresponds to a value of k_B/hbar, comparable to what one gets from a simple dimensional analysis or the "minimum viscosity limit" of "quantum hydrodynamic fluids" described by some theoretical models connected to string theory.

What worries me about this graph?
It is that elemental metals [copper, silver, aluminum, paladium, ...] lie on the same curve. As far as I am aware, they are not strongly correlated. They are nowhere near a quantum critical point. The resistivity is due to electron-phonon scattering. So given that they "accidentally" lie on this "universal" curve suggests to me that the significance of other materials lying close to it may not be of much significance.

On the other hand, the authors claim this "universality" arises because the electron-phonon scattering is "highly efficient" involving "high momentum scattering". They suggest similar scattering occurs in quantum critical metals.

A couple of earlier posts discussed my skepticism/confusion about similar claims about the significance of the magnitude of the linear resistivity.

A key piece of experimental evidence is needed to rigorously justify claims of quantum criticality: measurement of a correlation length which diverges at the quantum critical point.

But, perhaps I am missing something....

2 comments:

  1. “What worries me about this graph?
    It is that elemental metals [copper, silver, aluminum, paladium, ...] lie on the same curve. As far as I am aware, they are not strongly correlated”.

    Yes, you are right, these elemental metals are not strongly correlated. While the strongly correlated metals exhibit the properties of elemental ones, that is they demonstrate a quasi-classical behavior at low temperatures. Please, see: Quasi-classical physics and T-linear resistivity in both strongly correlated and ordinary metals, Phys. Rev. B 88, 115103 (2013),
    arXiv:1304.2068

    With warm regards,
    Vasily Shaginyan

    ReplyDelete
  2. More detailed consideration of this item can be found in "Theory of Heavy-Fermion Compounds; Theory of Strongly Correlated Fermi-Systems"
    http://www.springer.com/us/book/9783319108247

    ReplyDelete