Metallic Spin-Liquid Behavior of the Geometrically Frustrated Kondo Lattice Pr2Ir2O7

by S. Nakatsuji, Y. Machida, Y. Maeno, T. Tayama, T. Sakakibara, J. van Duijn, L. Balicas, J. N. Millican, R. T. Macaluso, and Julia Y. Chan

Journal club talk by Daniel Podolsky
Summary below by Jean-Sebastien Bernier



For the first CMP journal club of the fall semester, our discussion leader was Daniel Podolsky who presented a paper by Nakatsuji and coworkers on the possible observation of metallic spin liquid behavior in the geometrically frustrated Kondo lattice Pr$_2$Ir$_2$O$_7$.

Daniel began by presenting the material under study. He pointed out that the conduction electrons come from Iridium (Ir$^{4+}$) and that the local moments (Pr$^{3+}$) occupy the sites of a pyrochlore lattice (network of corner sharing tetrahedra). He noted that due to crystal field splitting the spins of the local moments are effectively Ising-like and point along an axis passing through the center of each tetrahedron. Daniel also explained that the Ising nature of the local moments makes this system unfrustrated if the RKKY interaction is antiferromagnetic. An AF RKKY interaction seems to be indicated by the large temperature susceptibility data. However, if the system is unfrustrated, it is difficult to understand why no long-range-order develops at the Curie-Weiss temperature.

Someone then asked how can we be sure that there is RKKY interaction in this system and not only superexchange. Daniel replied that the paper points out that $T^* = -20K$ is much higher than what should be expected from superexchange.

Daniel continued by presenting susceptibility data. He highlighted the unusual $\chi ~ \ln T$ behavior at low temperature and that no anomalies due to a magnetic transition were detected in the susceptibility except for freezing at $T_f = 120 mK$.

He then presented the resistivity data that shows the usual Kondo minimum and pointed out that $T_{Kondo} = 25 K$ is very large for such a poor metal.

Someone then asked if this material was also a poor metal at high temperatures. Daniel replied that he did not know.

Moving on, he presented that for $T_f = 120 mK < T < T_{CW} = 1.7 K$, the entropy and specific heat vary as $\sqrt{T}$ which leads to a very large entropy.

Someone then asked if $S(0)$ is known. Daniel answered that only $\Delta S$ was presented in this paper.

Finally, Daniel presented the global phase diagram put forward by the authors:

$T_f < T < T_{CW}$: metallic spin-liquid with $S \sim C \sim \sqrt{T}$
$T_{CW} < T < |T^*|$: Kondo screening, but incomplete since susceptibility diverges
$T > |T^*|$: decoupled Ir and Pr

... and he noted that the presence of a metallic spin-liquid did not sound too convincing to him.

Someone asked if any instability of the conduction electrons were observed. Daniel answered that there is no evidence in the data for such a phase transition.