Nuclear Magnetic Relaxation and Knight Shift Due to Orbital Interaction in Dirac Electron Systems

Hideaki Maebashi,Tomoki Hirosawa,Masao Ogata,Hidetoshi Fukuyama
DOI: https://doi.org/10.48550/arXiv.1709.02617
2017-09-08
Abstract:We study the nuclear magnetic relaxation rate and Knight shift in the presence of the orbital and quadrupole interactions for three-dimensional Dirac electron systems (e.g., bismuth-antimony alloys). By using recent results of the dynamic magnetic susceptibility and permittivity, we obtain rigorous results of the relaxation rates $(1/T_1)_{\rm orb}$ and $(1/T_1)_{\rm Q}$, which are due to the orbital and quadrupole interactions, respectively, and show that $(1/T_1)_{\rm Q}$ gives a negligible contribution compared with $(1/T_1)_{\rm orb}$. It is found that $(1/T_1)_{\rm orb}$ exhibits anomalous dependences on temperature $T$ and chemical potential $\mu$. When $\mu$ is inside the band gap, $(1/T_1)_{\rm orb} \sim T ^3 \log (2 T/\omega_0)$ for temperatures above the band gap, where $\omega_0$ is the nuclear Larmor frequency. When $\mu$ lies in the conduction or valence bands, $(1/T_1)_{\rm orb} \propto T k_{\rm F}^2 \log (2 |v_{\rm F}| k_{\rm F}/\omega_0)$ for low temperatures, where $k_{\rm F}$ and $v_{\rm F}$ are the Fermi momentum and Fermi velocity, respectively. The Knight shift $K_{\rm orb}$ due to the orbital interaction also shows anomalous dependences on $T$ and $\mu$. It is shown that $K_{\rm orb}$ is negative and its magnitude significantly increases with decreasing temperature when $\mu$ is located in the band gap. Because the anomalous dependences in $K_{\rm orb}$ is caused by the interband particle-hole excitations across the small band gap while $\left( 1/T_1 \right)_{\rm orb}$ is governed by the intraband excitations, the Korringa relation does not hold in the Dirac electron systems.
Mesoscale and Nanoscale Physics,Materials Science
What problem does this paper attempt to address?
The problem that this paper attempts to solve is to study the nuclear magnetic relaxation rate and Knight shift caused by orbital and quadrupole interactions in three - dimensional Dirac electron systems (for example, bismuth - antimony alloys). Specifically: 1. **Nuclear magnetic relaxation rate**: By using the latest results of dynamic susceptibility and dielectric constant, the paper derives a strict expression for the relaxation rate \(\frac{1}{T_1}\) caused by orbital and quadrupole interactions. In particular, the relaxation rate \(\left(\frac{1}{T_1}\right)_{\text{orb}}\) caused by orbital interactions shows an abnormal dependence on temperature \(T\) and chemical potential \(\mu\). When the chemical potential is within the band gap, the relaxation rate is proportional to the cube of the temperature with a logarithmic correction term; when the chemical potential is in the conduction band or valence band, the relaxation rate is proportional to the first power of the temperature with a logarithmic correction term. 2. **Knight shift**: The Knight shift \(K_{\text{orb}}\) caused by orbital interactions also shows an abnormal dependence on temperature \(T\) and chemical potential \(\mu\). In particular, at low temperatures and when the chemical potential is within the band gap, the Knight shift is negative and its absolute value increases significantly as the temperature decreases. 3. **Comparison and verification**: The paper also discusses the consistency of these results with previous research results on the relaxation rate of free - electron gas and Weyl fermion systems, and points out the reasons why the Korringa relationship no longer holds in Dirac electron systems, that is, the Knight shift is mainly caused by inter - band particle - hole excitation, while the relaxation rate is mainly determined by intra - band excitation. In summary, this paper aims to deeply understand the unique behavior of nuclear magnetic resonance characteristics (such as relaxation rate and Knight shift) in Dirac electron systems and provides theoretical explanations and predictions. The following are the key formulas: - When \(\mu\) is within the band gap: \[ \left(\frac{1}{T_1}\right)_{\text{orb}} \sim T^3\log\left(\frac{2T}{\omega_0}\right) \] - When \(\mu\) is in the conduction band or valence band: \[ \left(\frac{1}{T_1}\right)_{\text{orb}} \propto Tk_F^2\log\left(\frac{2|v_F|k_F}{\omega_0}\right) \] - Knight shift \(K_{\text{orb}}\): \[ K_{\text{orb}} = -\frac{2\alpha}{3\pi c^* c}\left[\log\left(\frac{E_\Lambda}{\Delta}\right)-\int_{\Delta}^{\infty}d\epsilon\left(1 - f(-\epsilon)+f(\epsilon)\right)\sqrt{\epsilon^2 - \Delta^2}\right] \] These results are helpful for further understanding orbital magnetism and related physical phenomena in Dirac electron systems.