Unusual behavior on a triangular lattice
This project was stimulated by a set of intriguing experiments on NiGa2S4, a highly two-dimensional spin S=1 antiferromagnet with a triangular lattice structure. It came to focus initially through work of Nakatsuji and collaborators, who observed that it showed no sharp phase transitions, either ordering or freezing, down to the lowest measured temperatures, despite substantial antiferromagnetic interactions. Despite the lack of obvious order, it displays a quadratic T2 specific heat for T<10K, very consistent with the spin wave modes of an ordered magnet. Another puzzling feature was a two-peak specific heat, with the higher temperature peak located around the (absolute value of the) Curie-Weiss temperature, an unusually high temperature for entropy loss to occur in a frustrated magnet. More recent experiments with local probes (NMR, muSR, NQR) have revealed a gradual spin freezing phenomena which occurs at low temperature, with quasi-static but spatially disordered local moments.
A number of theories were proposed to describe the material. These included exotic spin nematic or quadrupolar (two words for the same thing) ground states, Kosterlitz-Thouless transitions, and Z2 vortex transitions.
^ TOPOur work:
We suggest that the experiments can be best understood by viewing the material as a system near a Quantum Critical Point (QCP) between a quadrupolar ground state and an antiferromagnetic one. As a consequence, there is an intermediate temperature range at which the system exhibits strong quadrupolar correlations in the paramagnetic state. The onset of these correlations gives rise to the higher temperature specific heat peak, and the onset of magnetic correlations to the lower temperature one.
Furthermore, we give a symmetry-based explanation for the unconventional spin freezing. This is based on the fact, known from experiment, that the dominant spin correlations occur at an incommensurate wavevector. We show that, as a consequence, even non-magnetic impurities act as random fields upon a component of the order parameter. This, combined with the two-dimensionality of the material, eventually leads to a description of the freezing in terms of the well-studied "gauge glass" problem.
Read the paper.
