Thirty-Year Anniversary of κ-(BEDT-TTF)2Cu2(CN)3: Reconciling the Spin Gap in a Spin-Liquid Candidate

Abstract

In 1991, the Argonne group led by Jack Williams [Inorg. Chem. 1991, 30, 2586–2588] reported the first synthesis of κ-(BEDT-TTF)2Cu2(CN)3. Although, originally, the focus was on the superconducting properties under pressure, this frustrated Mott insulator has been the most promising quantum-spin-liquid candidate for almost two decades, widely believed to host gapless spin excitations down to T→0. The recent observation of a spin gap by Miksch et al. [Science 2021, 372, 276 LP–279.] rules out a gapless spin liquid with itinerant spinons and puts severe constraints on the magnetic ground state. This review evaluates magnetic, thermal transport, and structural anomalies around T⋆=6 K. The opening of a spin gap yields a rapid drop of spin susceptibility, NMR Knight shift, spin-lattice relaxation rate, and μ-SR spin fluctuation rate, but is often concealed by impurity spins. The concomitant structural transition at T⋆ manifests in thermal expansion, THz phonons and 63Cu NQR relaxation. Based on the field dependence of T⋆, a critical field of 30–60 T is estimated for the underlying spin-singlet state. Overall, the physical properties are remarkably similar to those of spin-Peierls compounds. Thus, a strong case is made that the ‘6K anomaly’ in κ-(BEDT-TTF)2Cu2(CN)3 is the transition to a valence-bond-solid state and it is suggested that such a scenario is rather the rule than the exception in materials with strong magnetic frustration.

Chasing the spin gap through the phase diagram of a frustrated Mott insulator

“…1c) and, thus, we complete the phase diagram of κ-(BEDT--TTF) 2 Cu 2 (CN) 3 . Moreover, our transport data yield a gap size of 30-50 K for the spin-singlet state, which coincides with the unexpectedly large critical field of 60 T. 16 Among the QSL candidates, κ-(BEDT-TTF) 2 -Cu 2 (CN) 3 has been studied most intensely as it exhibits not only frustrated AFM exchange interactions (J = 250 K 26 ), but also a paradigmatic Mott transition. 21,24,27 This layered triangular-lattice compound with t /t close to unity shows no AFM order 26 and, at the same time, it is one of the best solid-state realizations of the single-band Hubbard model.…”

“…15 Crucially, the gap size observed here lines up with the critical field of order 60 T estimated in Ref. 16 , which explains why magnetic fields of order 10 T have little effect on the transition. 17,25 d ln R/d 1/T in the range T > 20 K reflects changes of the Mott-Hubbard gap.…”

“…[11][12][13][14] Such a VBS scenario is fully consistent with the report of a spin gap in the organic QSL candidate κ-(BEDT-TTF) 2 Cu 2 (CN) 3 that aroused great attention in the QSL community very recently. 15,16 While this finding rules out the widely assumed scenario of mobile spinons, 17 in agreement with thermal transport results, 18 it leaves open the possibility of gapped QSL phases. 3 The tunability of κ-(BEDT-TTF) 2 Cu 2 (CN) 3 -a pressure of 1.3 kbar triggers the transition from a Mott insulator to superconducting (SC) and metallic phases 5,6,[21][22][23][24] as illustrated in Fig.…”

“…4a,b, the crystal contracts along the caxis and expands along the other in-plane direction upon cooling. Also other probes 34,35 observed pronounced anomalies at 6 K due to modifications of the crystal structure and possible symmetry breaking, 16 which are typical fingerprints of a VBS transition. In addition, recent ESR studies revealed a rapid drop of spin susceptibility χ s due to the opening of a spin gap, 15 which agrees with NMR Knight shift data from Ref.…”

Controlling frustrated magnetism on the kagome lattice by uniaxial-strain tuning

“…Even after two decades of intense scrutiny, quantum spin liquids (QSL) remain an elusive state of matter [1][2][3][4]. Apart from growing evidence for the importance of disorder [5][6][7][8][9][10][11][12][13][14][15], geometrical frustration is considered decisive to suppress magnetic order in presence of strong antiferromagnetic (AFM) exchange interactions. The vast majority of QSL candidates are found in quasi twodimensional correlated electron systems with triangular [8,[16][17][18][19][20][21], honeycomb [22] or kagome [23,24] lattices.…”