Friday, October 1, 2010

From chemical exotica to rich physics

I had a really nice meeting yesterday in the Clarendon lab with Amalia Coldea. She has been doing some magnetoresistance measurements on an interesting class of organic charge transfer salts, κ-β′′-(BEDT-TTF)2(PO-CONHC2H4SO3), where PO = 2,2,5,5 Tetramethyl-3-pyrrolin-1-oxyl Free Radical, described in a recent Chemistry of Materials paper.

Why should physicists care about such exotica?
A few things I think are particularly interesting about these new materials are the following.

  • The anion is a free radical (i.e., has a localised spin 1/2). These spins interact via an exchange interaction with the itinerant electrons in the BEDT-TTF layers. Thus, the system is something like a Kondo lattice model. In some senses this is the spin 1/2 analogue of the magnetic field induced superconductor lambda-(BETS)2FeCl4 which has spin-5/2 (see this PRB for a discussion of the relevant theory).
  • The crystal structure is such that there are alternating layers of BEDT-TTF molecules with two different stacking motifs (kappa and beta''). It is claimed that this leads to the two layers are doped away from half filling. The average filling is one half as for individual kappa and beta'' layers, but because they have different band structures the kappa (beta'') layers are at greater (less) than half filling. [This is a bit like what happens in TTF-TCNQ.] If correct this fulfills a long sought goal of doping organic charge transfer salts!
  • The alternation of beta'' and kappa layers also means the interlayer charge transport could be particularly interesting because the interlayer hopping integral could vary significantly with intralayer momentum, as it does in the cuprates. This could lead to unusual angle-dependent magnetoresistance, as described in a theory paper by Yagi and Iye.
What are the prospects of seeing Kondo lattice type physics in these materials? Not good, I fear. The scale of the Heisenberg exchange interaction J_eff between the free radical spins is estimated to be of the order of 1 K. Assuming this is due an RKKY interaction, J_eff ~ J^2 D(E_F) where J is the Kondo exchange interaction and D(E_F) is the metallic density of states, estimated to be about 6 states/eV from the Pauli susceptibility. 
The Kondo temperature ~ E_F exp( -1/JD(E_F)), using the above estimates I obtain JD(E_F) ~ 0.02 and so the Kondo temperature will be many orders of magnitude less than a mK and so not experimentally accessible. However, as for the reasons outline above there is a lot of other interesting strongly correlated electron physics to be explored in these materials!

No comments:

Post a Comment