Distribution of critical temperature at Anderson localization

TL;DR

This paper investigates the distribution of critical temperatures at Anderson localization using a local mean-field theory that incorporates wave-function multifractality, revealing power-law behaviors and phase transition criteria relevant to quantum Griffiths phenomena.

Contribution

It introduces a local mean-field approach that includes wave-function multifractality to analyze critical temperature distributions at Anderson localization, highlighting phase transition behaviors.

Findings

Kondo temperature distribution exhibits a power-law tail as $T_K o 0$.

Ferromagnetic transition temperature distribution shows a power-law behavior below a critical interaction.

Typical transition temperatures indicate the presence of quantum Griffiths phenomena and smeared transitions.

Abstract

Based on a local mean-field theory approach at Anderson localization, we find a distribution function of critical temperature from that of disorder. An essential point of this local mean-field theory approach is that the information of the wave-function multifractality is introduced. The distribution function of the Kondo temperature ($T_{K}$) shows a power-law tail in the limit of $T_{K} \rightarrow 0$ regardless of the Kondo coupling constant. We also find that the distribution function of the ferromagnetic transition temperature ($T_{c}$) gives a power-law behavior in the limit of $T_{c} \rightarrow 0$ when an interaction parameter for ferromagnetic instability lies below a critical value. However, the $T_{c}$ distribution function stops the power-law increasing behavior in the $T_{c} \rightarrow 0$ limit and vanishes beyond the critical interaction parameter inside the ferromagnetic phase. These results imply that the typical Kondo temperature given by a geometric average always vanishes due to finite density of the distribution function in the $T_{K} \rightarrow 0$ limit while the typical ferromagnetic transition temperature shows a phase transition at the critical interaction parameter. We propose that the typical transition temperature serves a criterion for quantum Griffiths phenomena vs. smeared transitions: Quantum Griffiths phenomena occur above the typical value of the critical temperature while smeared phase transitions result at low temperatures below the typical transition temperature. We speculate that the ferromagnetic transition at Anderson localization shows the evolution from quantum Griffiths phenomena to smeared transitions around the critical interaction parameter at low temperatures.