Multi-soliton configurations in a doped antiferromagnetic Mott insulator

by Mona Berciu and Sajeev John
Department of Physics, University of Toronto


Notes:

  1. This poster summarizes a paper published in Phys. Rev. B 59 (23), pp. 15143-15159 (1999); you can find the .pdf file here .
  2. What makes a poster useful is having a person nearby, ready to explain it. If you have any questions, please contact me.
  3. Please click on each page to see it full size. To return, use the "Back" option.
  4. To return to my homepage, click here .


Summary of pages:

So let me emphasize once again some of the more attractive features of this model. We have shown that for intermediate values of U/t, at low and medium dopings, the stable configurations are liquids of meron-antimeron pairs of the spin-flux model. I remind you that the spin-flux itself is a consequence of long-range electron-electron interactions and the fermionic nature of the electrons.

Why is this very appealing?

  1. the existence of the meron-antimeron pairs, which may be the answer for the preformed pairs thought to exist in the cuprates. There is very strong bonding, of topological nature, between a vortex and an antivortex. Such a pair should be stable even in the presence of full 1/r Coulomb repulsion between the charged cores.
  2. the charged merons and antimerons are bosons . A liquid of such charged bosons must be a non-Fermi-liquid metal, since the charge carriers are not fermions. However, for higher dopings this liquid becomes unstable, and a transition to the conventional model describing a Fermi liquid takes place.
  3. the meron-antimeron pair creation upon doping provides a good scenario for understanding the destruction of the long range antiferromagnetic order at very low dopings, of about 0.02. From the picture on page 10 we can see that the spin of about 100 sites is indeed rotated out of the background orientation near a tightly bound meron-antimeron pair, containing two holes. However, short range antiferromagnetic correlations are preserved, since the rotated spins are locally antiferromagnetically ordered.
  4. this evolution of the magnetic state, from long range order to short range correlations as the doping is increased, reflects in the magnetic structure factor, which has the incommensurate splitting seen experimentally (see page 16 ).
  5. this picture may explain the unusual optical conduction of the cuprates. The broad mid-infrared peak that develops upon doping is associated with excitations on the gap levels trapped inside the vortex cores. This contribution is shown on page 15 . The missing part is the "Drude-like" tail, which is due to the overall translational motion of the charged vortices under the influence of the electric field (this motion is frozen in our static model). This picture explains why the broad mid-infrared peak is observed quite unchanged below the superconducting transition, while the Drude-tail collapses in a delta-function: superconductivity is related to how the charged vortices move, in other words how their flow becomes a superflow.


  6. Back to TOP     or     Back to my homepage