Fig. 4: Transition to valence-bond-solid (VBS) state at ‘6K anomaly’.
From: Chasing the spin gap through the phase diagram of a frustrated Mott insulator
a, b The thermal expansion coefficient α yields similar in-plane anisotropy at the VBS transition of EtMe3P[Pd(dmit)2]2 (T⋆ = 24 K)7 and at the ‘6K anomaly’ of κ-(BEDT-TTF)2Cu2(CN)319. Note the different scales; data from refs. 7,19. c The rapid drop of NMR Knight shift K and ESR susceptibility χs (B∥a) at T⋆ = 6 K evidences the opening of a spin gap; data from refs. 16,44. d The phase diagram of κ-(BEDT-TTF)2Cu2(CN)3 assembled from R(T, εc) (Figs. 2c and 3) is underlaid with a false-color plot of \(d \, {{{{{\rm{ln}}}}}} \, R/d \, 1/T\) (red: insulating, blue: metallic). Inset: checkerboard area indicates range of our strain-dependent experiments. We find a clear back-bending of the insulator-metal boundary (cf. Fig. 1c) originating from the ‘6K anomaly’ at p = 0. Altogether, the ambient pressure data in b, c and dT⋆/dp < 0 in d provide solid evidence for a low-entropy spin-singlet phase with structural anisotropy -- the hallmarks of a VBS.
