Abstract
Based on a systematic analysis of the thermal evolution of the resistivities of Fe-based chalcogenides Fe1+δTe1-xXx (X = Se, S), it is inferred that their often observed nonmetallic resistivities are related to a presence of two resistive channels: one is a high- temperature thermally-activated process while the other is a low-temperature log-in-T process. On lowering temperature, there are often two metal-to-nonmetall crossover events: one from the high-T thermally-activated nonmetallic regime into a metal-like phase and the other from the log-in-T regime into a second metal-like phase. Based on these events, together with the magnetic and superconducting transitions, a phase diagram is constructed for each series. We discuss the origin of both processes as well as the associated crossover events. We also discuss how these resistive processes are being influenced by pressure, intercalation, disorder, doping, or sample condition and, in turn, how these modifications are shaping the associated phase diagrams.
Highlights
We discuss how these resistive processes are being influenced by pressure, intercalation, disorder, doping, or sample condition and, in turn, how these modifications are shaping the associated phase diagrams
An earlier electronic structure calculations predicted a low-carrier-density metallic character,[8] the isomorphous Fe1+δX compounds exhibit a variety of normal-state behavior: Fe1+δS is nonmetallic below TMNM ∼300 K [Ref.9] but TMNM can be strongly reduced by pressure.[10]
Based on the analysis of their resistivities and on the obtained phase diagrams, we identified two processes that are responsible for their bad metallic character as well as for shaping their phase diagrams: the first is a high-temperature (150 to 300 K) thermally-activated process[14] while the other is a low-temperature (< 100 K) log-in-T process.[15]
Summary
It is remarkable that the normal-state resistivities of Fe1+δX (X= Te, Se, S, or their solid solutions)[1,2,3,4] as well as those of intercalated AxFe2−ySe2 (A=K, Rb, Cs, Tl, ..)[5,6,7] are neither truly metallic nor truly insulating. In order to rationalize the variety of functional forms of ρ(T, P, x) and, in addition, so as to identify and evaluate the strength of the involved resistive channels, let us assume, based on earlier studies,[16,17,18,19] that the character of their normal-state is shaped by the combined influences of crystalline electric field interactions, electronic correlations, disorder, and band filling: based on the strength of these factors, the high-temperature (T >100 K) normal state could be either metallic, Mott insulator, or an intermediate orbital-selective Mott phase (OSMP) wherein some of the Fe 3d orbitals are localized while the others are itinerant.[14,16,17,18,19] The high-T phase of most of the studied intercalated AxFe2−ySe2 (as well as most of Fe1+δX1−xYx, see below) compounds is reported to be an OSMP.[14,16,17,18,19] We assumed that, within this OSMP, localized states are separated from itinerant ones by a mobility edge at Ec.[20] the thermal evolution of the resistivity depends on the relative strength of |Ec − EF | with respect to kBT : if Ec is not located in the 3d multiplet or that |Ec − EF | > kBT , transport is effected by a thermally-assisted hopping among the localized orbitals leading to a Mott variable range hopping resistivity (VRH): ρ(T. where d=2 (3) represents a 2- (3-) dimensionality and all other terms have their usual meaning. On cooling, two crossovers may be observed: one from the high-T activated regime into a metal-like phase (metal-I) at TXHT and another from the low-T log-in-T regime into a metallike phase (metal-II) at TXLT (see below); it is emphasized that, for Fe1+δ(Te1−xSex) 0.1 ≤ x ≤ 0.5, both TXLT and TXHT events (see Refs. 3, 26, and 27) are not accompanied by any visible symmetry-breaking process
Sign-in/Register to view highlights & summary 
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call