Coexistence of spin-singlets and metallic behavior in simple cubic CsC60

TL;DR

This study reveals coexistence of localized spin-singlets and metallic behavior in simple cubic CsC60, highlighting inhomogeneous charge distribution and dynamic spin-singlet states affecting its electronic properties.

Contribution

It provides detailed NMR evidence for localized spin-singlets in CsC60 and their inhomogeneous distribution, advancing understanding of its metallic and insulating coexistence.

Findings

Presence of 12-15% anomalous C60 balls with a 15 meV gap

Spin-singlets are diluted within the lattice, not forming clusters

Spin-singlet lifetime decreases exponentially with temperature

Abstract

We present a detailed NMR study of simple cubic CsC60, the only metallic cubic fullerides known so far besides A3C60. 133Cs NMR signals the presence of about 12 to 15% of ``anomalous'' C60 balls, characterized by a 15 meV gap. We present different experimental observations supporting the idea that a spin-singlet (i.e. a C60^{2-}) is localized on these latter balls and stabilized by a JTD. A splitting of the 133Cs spectrum into three different lines, with distinct electronic environments, is observed which emerges naturally from such a situation. Quadrupole effects on 133Cs confirm an inhomogeneous distribution of the charge between C60 molecules. Analysis of the relative intensities of the 133Cs lines show that the spin-singlets do not form clusters at the local scale, but are diluted within the lattice. This is probably to avoid the occurrence of neighboring C60^{2-} and minimize electrostatic repulsion between these balls. Through spin-lattice relaxation measurements, we detect chemical exchange between the Cs sites above 100 K. From this, we deduce that the lifetime of a spin-singlet on a given ball decreases exponentially with increasing temperature (from 15 sec at 100 K to 3 ms at 130 K). The implications of the presence of such spin-singlets for the nature of the metallic state is discussed, in relation with a decrease of 1/T1T for 13C at low temperatures, which would be anomalous for a simple metal.