- 日時 ： ６月27日（火） 午後１時３０分～
-
場所 ：理学部Ｆ棟５階 理論物質学セミナー室（Ｆ５２４）
-
講演者：三宅 和正 氏（阪大・強磁場）
-
タイトル：「Theory for Charge Kondo Effect and Valence Skipping Phenomenon:
Anomalous 125Te-NMR Relaxation in Pb0.99Tl0.01Te」
-
概要 ：
In ions with ns-electron configuration with large principal quantum
number n, (ns) 0 or (ns) 2 state appears but (ns) 1 state is missing.
For example, Tl3+ ((6s) 0 ) and Tl1+ ((6s) 2 ) exist but Tl2+ ((6s) 1 ) does not.
About a decade ago, a new type of Kondo effect and superconductivity have
been observed in Tldoped PbTe [1]. This phenomenon was shown to be
explained on the basis of the negative U (< 0) Anderson model which includes
a possibility of the degeneracy of d 0 and d 2 states for a wide range of doping
rate (d denotes the 6s state of Tl ion in the present case) [2], leading to the
charge Kondo effect and s-wave superconductivity [3]. While it had been
unknown what kind of microscopic mechanism exists behind the negative-U
ansatz and the charge Kondo effect in such materials, we proposed a new
microscopic model for the charge Kondo effect, and found on the basis of
numerical renormalization group calculations [4] that the charge Kondo
effect is caused by the pair hopping (PH) interaction Uph between the
conduction band and the localized ns electron with large n. The physical
reason why the Uph is relevant is attributed to the fact that ns wave function
has a large spatial extent, increasing a relative importance of the PH interaction
against the on-site Coulomb interaction. Quite recently, it was reported by
Mukuda at Osaka University [5] that the NMR relaxation rate 1/T1T of 125Te
adjacent to Te ion in Pb0.99Tl0.01Te exhibits sharp increase at temperature
T < 10K with decreasing T. It may be curious that valence fluctuations
between two magnetically inert valence states, (6s) 0 and (6s) 2 , causes
such an anomaly. However, by extending the theory developed in Ref. [4],
we found that the 1/T1T really exhibits the divergent T dependence around
the charge Kondo temperature TCK ∼ 10K [1] as an effect of the Kondo
renormalization on Uph just as in the case of magnetic Kondo effect. T
his is because the PH interaction is transformed to the sd-exchange
interaction through the canonical transformation `a la Shiba [6]
[1] Y. Matsushita et al., Phys. Rev. Lett. 94 (2005) 157002.
[2] A. Taraphder and P. Coleman: Phys. Rev. Lett. 66 (1991) 2814.
[3] M. Dzero and J. Schmalian: Phys. Rev. Lett. 94 (2005) 157003.
[4] H. Matsuura and K. Miyake, J. Phys. Soc. Jpn. 81 (2012) 113705.
[5] H. Mukuda, The JPS March Meeting 2017, 20pL21-10; private communications.
[6] H. Shiba, Prog. Theor. Phys. 48 (1972) 2171.