We have systematically investigated the global phase diagram for Li{x}M{y}HfNCl (M: molecule), demonstrating the independent controllability of carrier density x and interlayer spacing d. In LixHfNCl, the superconducting phase with almost constant T{c} of 20 K prevails for 0.15

The doping dependence of specific heat and magnetic susceptibility has been investigated for Li(x)ZrNCl superconductors derived from a band insulator. As the carrier concentration is decreased, the anisotropy of superconducting gap changes from highly anisotropic to almost isotropic. It was also found that, upon reducing carrier density, the superconducting coupling strength and the magnetic susceptibility are concomitantly enhanced in parallel with T(c), while the density of states at the Fermi level is kept almost constant. Theoretical calculations taking into account the on-site Coulomb interaction reproduced the experimental results, suggesting a possible pairing mediated by magnetic fluctuations, even in the doped band insulators.

Specific heat has been investigated in electron-doped Li x M y ZrNCl superconductors, in which interlayer spacing d can be controlled by the intercalation of organic molecules M. It has been revealed that, upon enlargement of d by molecule cointercalation, T c is enhanced concomitantly with the superconducting coupling strength whereas the density of states at the Fermi level is kept almost constant. Structural and spectroscopic analysis proved that the molecules are intercalated without affecting electronic nor vibrational states, simply increasing the interlayer spacing. These results lead us to conclude that the enhancement of T c is ascribed to the improvement of two dimensionality, which is consistent with a recent theory that discussed superconductivity in a band insulator on a honeycomb lattice with disconnected Fermi surfaces.

In this study, it is shown that in a layered nitride superconductor, i.e., LixZrNCl (0.07 ≤ x ≤ 0.21), the superconducting order parameter is highly anisotropic in a sample with x = 0.12, as inferred from both the temperature and the magnetic field dependences of the muon depolarization rate (σs, proportional to the superfluid density). Moreover, the tendency of strong anisotropy with an increase in x is indicated by the T dependence of σs. These observations are in good agreement with the recent theory that predicts the development of anisotropy in a d + id ′ gap upon carrier filling to the bands with disconnected Fermi surfaces on a honeycomb lattice. 74.20.Rp, 76.75.+i Layered nitrides such as β-M NCl (with M = Zr, Hf) are attracting considerable attention since they exhibit superconductivity upon intercalating alkaline metals [1,2]. While these nitrides have relatively high superconducting transition temperatures [T c ≃ 15 (25) K with M = Zr (Hf)], the density of states at the Fermi level [N (0)] is reported to be considerably lower than that of other superconductors having similar T c , as inferred from the results of the specific heat measurement of Li 0.12 ZrNCl [3] and magnetic susceptibility measurement of Li 0.48 (THF) y HfNCl (where THF refers to tetrahydrofuran) [4]. The small N (0) is in excellent agreement with the prediction of the theoretical investigation [5]. Such situation naturally leads to the question of whether or not the superconductivity in layered nitrides is fully understood on the basis of the conventional BCS theory with electron-phonon coupling. Moreover, it has been shown that one of these nitrides, i.e., Li x ZrNCl, exhibits further anomalies that are not expected to be present in the simplest situation presumed by the BCS theory.Li x ZrNCl has a lamellar structure consisting of alternating stacks of Zr-N double honeycomb layers and insulating Cl bilayers. Li atoms are intercalated into the van der Waals gap of the Cl bilayers to supply electrons to the conducting Zr-N layers. In addition to the above mentioned small N (0), it has been also revealed that the electron-phonon interaction is too weak to explain its high T c [3,5,6]. In general, in the case of two-dimensional electronic systems such as β−ZrNCl, N (0) may be only weakly dependent on band filling. Therefore, provided that superconductivity is explained by the conventional BCS theory, T c would not vary with the Li concentration (x). However, the fact is that while T c is independent of doping for x ≥ 0.15, it increases steeply below x ≃ 0.12, reaching a maximum (T c = 15.2 K at x = 0.06) and then suddenly transitioning into an insulating state for x ≤ 0.05 [7]. It might be worth noting that this tendency of T c to be high at a low carrier density x is opposite to that of underdoped cuprates.Another anomaly is reported in the magnetic field dependence of an electronic specific heat (Sommerfeld) coefficient γ in the mixed state of Li 0.12 ZrNCl. While γ is expected to be approximately proportional to the number of flux...

We present NMR measurements of the layered nitride superconductor Li x ZrNCl. The nuclear spin-lattice relaxation rate 1/T 1 shows that the coherence peak is strongly suppressed in Li x ZrNCl in contrast to conventional BCS superconductors. In the lightly doped region close to the insulating state, the system shows a gaplike behavior, i.e., pseudogap, that is characterized by a reduction in the magnitude of the Knight shift and 1/T 1 T . A higher superconducting (SC) transition temperature T c is achieved by coexisting with the pseudogap state. These unusual behaviors, which deviate from the ordinary BCS framework, are the key ingredients to understanding the SC mechanism of Li x ZrNCl.