The authors report ' C NMR spin-lattice relaxation rates 1/Tl and Knight shifts Ks in the quasitwo-dimensional organic superconductor «-(ET)2Cu[N(CN)i]Br ( T, = 11.6 K), for an aligned single crystal. The normal-state behavior is reminiscent of the high-T, cuprates, in which antiferromagnetic Auctuations and spin-gap behavior dominate. In the superconducting state, the data rule out the BCS electron-phonon mechanism as the source of the superconductivity, but support an unconventional pairing state with possible nodes in the gap function.The discovery of superconductivity in organic chargetransfer salts based on the BEDT-TTF ("ET") molecule has stimulated interest in understanding the electronic structure of their normal and superconducting states. ' The ET compound with the highest ambient pressure critical temperature is «-(ET)zCu[N(CN)2]Br (T, =11.6 K). This salt has a layered structure and quasi-twodimensional (2D) electronic conduction, similar to the cuprate superconductors. The reduced dimensionality, and low carrier density (-10 ' cm ), indicate that electron correlation effects may be important. The nature of the superconducting state in the ET salts (i.e. , BCS or unconventional) is unsettled at present. The low-temperature magnetic penetration depth has been studied, with some experiments supporting the existence of an isotropic BCS gap, and others an anisotropic gap.We have utilized ' C NMR to study the normal and superconducting states of tc-(ET)zCu [N(CN) 2]Br [' K-(ET)&Br"]. In agreement with the work of Mayafre et al. and Kawamoto et al. , we find that the normalstate behavior is not that of a simple metal, and that antiferromagnetic Auctuations, and spin-gap behavior may be present. Our data for «. -(ET)2Br below T, favor spinsinglet pairing, and a highly anisotropic energy gap, such as is present for d-wave, or anisotropic s-wave, orbital pairing. Thus, in contrast to the A3C6o (A =K,Rb, Cs) "Buckyball" superconductors, the superconductivity in tc-(ET)2Br arises from a mechanism other than the conventional BCS electron-phonon coupling.In a metal, the hyperfine interactions of the nuclei with the spins of the conduction electrons dominate the NMR properties. We find that in «. -(ET)2Br, both the isotropic Fermi contact and anisotropic dipolar (2p, ) interactions are present. For a simple metal, the Korringa law relates the Knight shift E, to the spin-lattice relaxation time T& . .1/T, T o= K, . Below T"a gap opens at the Fermi level, producing dramatic changes in the NMR properties.For BCS spin-singlet s-wave superconductors, K, decays to zero as T~O in a characteristic way. Just below T" 1!T,rises ("coherence peak"), but falls off exponentially at low T. These features are found in the NMR studies of A3C6o, clearly identifying it as a conventional BCS superconductor with an isotropic energy gap.' We note that the above results are drastically modified in the case of an anisotropic gap function.Previously, we have reported 'H NMR in the superconducting state of tc-(ET)@Br. " We showed that the NM...

Peaks in the magnetoresistivity of the layered superconductor κ-(BEDT-TTF)2Cu(NCS)2, measured in fields ≤ 45 T applied within the layers, show that the Fermi surface is extended in the interlayer direction and enable the interlayer transfer integral (t ⊥ ≈ 0.04 meV) to be deduced. However, the quasiparticle scattering rate τ −1 is such thath/τ ∼ 6t ⊥ , implying that κ-(BEDT-TTF)2Cu(NCS)2 meets the criterion used to identify interlayer incoherence. The applicability of this criterion to anisotropic materials is thus shown to be questionable.PACS numbers: 74.70. Kn, 78.20.Ls, 71.20.Rv Many correlated-electron systems which are of fundamental interest have very anisotropic electronic bandstructure. Examples include the "high-T c " cuprates [1,2], layered ruthenates [3], and crystalline organic metals [2,4]. Such systems may be described by a tightbinding Hamiltonian in which the ratio of the interlayer transfer integral t ⊥ to the average intralayer transfer integral t || is ≪ 1 [2,4,5]. The inequalityh/τ > t ⊥ [6] where τ −1 is the quasiparticle scattering rate [1, 2, 5], frequently applies to such systems, suggesting that the quasiparticles scatter more frequently than they tunnel between layers. The question has thus arisen as to whether the interlayer charge transfer is coherent or incoherent, i.e. whether or not the Fermi surface (FS) extends in the interlayer direction [2,4,5]. In this paper we have used magnetoresistance data to estimate the interlayer transfer integral in the highly anisotropic organic superconductor κ-(BEDT-TTF) 2 Cu(NCS) 2 . We find that the material obeys the inequalityh/τ > t ⊥ ; moreover, mean-free path in the interlayer direction is < ∼ 20% of the unit-cell height. Nevertheless, our data demonstrate a FS which is extended in the interlayer direction.κ-(BEDT-TTF) 2 Cu(NCS) 2 was selected for our experiments because it is perhaps the most thoroughly characterised quasi-two-dimensional (Q2D) conductor [4]. In contrast to the cuprates, the FS topology is well known from Shubnikov-de Haas (SdH) and de Haas-van Alphen (dHvA) studies [4] and from angle-dependent magnetoresistance oscillation (AMRO) [8] and millimetre-wave (MMW) experiments [9]; it consists of a pair of quasione-dimensional (Q1D) electron sheets plus a Q2D hole pocket (see Fig. 1a [10, 11]). The κ-phase BEDT-TTF salts are considered to be leading contenders for interlayer incoherence [5], and optical data may be interpreted as consistent with this suggestion [12]. Moreover, models for unconventional superconductivity in κ-phase BEDT-TTF salts invoke the nesting properties of the FS [11,13,14]; the degree of nesting might depend on whether the FS is a 2D or 3D entity (see [4], Section 3.5). Experimental tests for coherence in κ-(BEDT-TTF) 2 Cu(NCS) 2 are thus far deemed to be inconclusive [5]; e.g. semiclassical models can reproduce AMRO [8] and MMW data [9] equally well when the interlayer transport is coherent or "weakly coherent" [5].To examine how interlayer coherence might be detected, we use a tight-binding dispersio...

We present high precision measurements of the penetration depth of single crystals of κ−(ET)2Cu[N(CN)2]Br and κ−(ET)2Cu(NCS)2 at temperatures down to 0.4 K. We find that, at low temperatures, the in-plane penetration depth (λ ) varies as a fractional power law, λ ∼ T 3 2 . Whilst this may be taken as evidence for novel pair excitation processes, we show that the data are also consistent with a quasilinear variation of the superfluid density, as is expected for a d-wave superconductor with impurities or a small residual gap. Our data for the interplane penetration depth show similar features and give a direct measurement of the absolute value, λ ⊥ (0) = 100 ± 20 µm.PACS numbers: 74.70. Kn, 74.25.Nf Compounds of the family κ−(ET) 2 X have the highest transition temperatures of all organic superconductors known to date [1]. They have recently attracted considerable attention because of their similarity to the high T c cuprates and the possibility that they may also have a non-conventional paring state [2]. The two materials studied here, κ−(ET) 2 Cu[N(CN) 2 ]Br (T c ∼ 11.6 K) and κ−(ET) 2 Cu(NCS) 2 (T c ∼ 9.6 K), are highly anisotropic, layered, extreme type II superconductors. As in the cuprates, the superconducting phase in these materials is in close proximity to an antiferromagnetic phase. Both antiferromagnetic spin fluctuations and a pseudogap have been detected in NMR measurements in the normal state [3]. Neither the underlying pairing mechanism nor the symmetry of the order parameter has been conclusively established. Although NMR [4,5], specific heat [6] and thermal conductivity [7] measurements all suggest a non-conventional pairing state, results of penetration depth measurements have been inconsistent, with evidence for both conventional [8,9] and non-conventional [10][11][12] behavior. However, none of these penetration depth measurements have been performed over a temperature range (T /T c ) and a precision, comparable to those in the cuprates [13]. In this Letter, we present measurements of both the in-plane λ , and the interplane, λ ⊥ , penetration depths in κ−(ET) 2 Cu[N(CN) 2 ]Br and κ−(ET) 2 Cu(NCS) 2 at temperatures down to 0.4 K.Our measurements were performed on single crystals of κ−(ET) 2 Cu[N(CN) 2 ]Br and κ−(ET) 2 Cu(NCS) 2 which were grown at Argonne National Laboratory. Details of the growth procedures have been given elsewhere [14]. Penetration depth measurements were performed using a 13 MHz tunnel diode oscillator [15] mounted on a 3 He refrigerator. The low noise level [ ∆F F0 ≃ 10 −9 ], and low drift of the oscillator allows us to obtain high resolution data with a very small temperature spacing interval. The samples were attached, with a small amount of vacuum grease, to a sapphire rod which fitted inside the copper sense coil. The sense coil was calibrated using spheres of Aluminum. The sample temperature was measured with a calibrated Cernox thermometer attached to the other end of the sapphire rod. The samples were cooled slowly (0.1-1.0 K/min) to avoid introducing disorder [1...

049ChemInform Abstract The title compound is obtained by in situ electrocrystallization of bis(ethylenedithio)tetrathiafulvalene (ET) by use of a 1:1 mixture of CuBr and tetraphenylphosphoniumdicyanamide. An X-ray analysis (space group Pnma, Z=4) shows that the structure contains alternating layers of the donor molecule cations and polymeric mixed halo(pseudohalo)cuprate(I) anions. It is established that the title compound is the organic superconductor with the highest transition temp. found to date.