Intrinsic insulating ground state in transition metal dichalcogenide TiSe2

Abstract

The transition metal dichalcogenide TiSe2 has received significant research attention over the past four decades. Different studies have presented ways to suppress the 200 K charge density wave transition, vary low temperature resistivity by several orders of magnitude, and stabilize magnetism or superconductivity. Here we give the results of a new synthesis technique whereby samples were grown in a high pressure environment with up to 180 bar of argon gas. Above 100 K, properties are nearly unchanged from previous reports, but a hysteretic resistance region that begins around 80 K, accompanied by insulating low temperature behavior, is distinct from anything previously observed. An accompanying decrease in carrier concentration is seen in Hall effect measurements, and photoemission data show a removal of an electron pocket from the Fermi surface in an insulating sample. We conclude that high inert gas pressure synthesis accesses an underlying nonmetallic ground state in a material long speculated to be an excitonic insulator. arXiv:1809.09467v2 [cond-mat.str-el]

On the origin of the anomalous peak in the resistivity of TiSe2

“…[16] it was shown that the resistivity of TiTe 2 followed the Bloch-Grüneisen (B-G) formulation for scattering due to electron-phonon coupling. We therefore adopt this form of scattering rate for our model of TiSe 2 : the coefficient of the B-G scattering term is a free parameter which we set for good agreement on the absolute value of resistivity around ρ peak , and the Debye temperature of the B-G term is set to 250 K [2,14]. We also add a constant scattering term from elastic impurity scattering, which is set such that ρ exp T →0 = m * e /N e |e| 2 τ 0 .…”

“…extrinsic electron doping) reported by Di Salvo et al [1] [23]. Campbell et al [14] recently argued against any link between the height of the peak and the sample stoichiometry. However, our simulations are in line with the earlier intuition that the two are indeed related [1,3,5].…”

“…Still, the trend is consistent with experimental evidence. For example, the sample grown by sublimation in Di Salvo et al [1] with T peak ≈ 149 K, and the sample grown in Se flux by Campbell et al (T peak = 150 K [14]), which are likely to be the most stoichiometric samples (i.e. lowest extrinsic carrier densities), both display a higher ρ peak and a lower T peak compared with samples grown by iodine vapor transport, in which T peak is typically 165 K [1].…”

“…While our simulations support this idea in principle, in practice it may be difficult to precisely identify the midpoint of the step-like feature in the first derivative, given there is a substantially varying background. Taguchi et al [3], and more recently Campbell et al [14], argued that looking for a small but sharp peak in the second derivative is a better criterion for determining T CDW . In agreement with this, our analysis presented Fig.…”

“…On the other hand, some studies have interpreted the peak feature as a signature of the role of excitons in the CDW ordering [7][8][9][10][11]. Given the resurgence of interest in TiSe 2 in recent years, it is worthwhile to revisit the long-standing problem of its unusual transport properties [5,[12][13][14], aided by the availability of recent characterizations of its 3D electronic structure by angle-resolved photoemission spectroscopy (ARPES) [15].In this paper, we show that the overall temperaturedependence of the resistivity of TiSe 2 , including the anomalous peak, can be reasonably reproduced without accounting for the CDW at all. Our model is very simple, but it captures the essential ingredients necessary to qualitatively understand the transport properties of TiSe 2 .…”