Tight binding model on square lattice with stagger flux

Posted by Maggie on April 8, 2024

Tight binding model on square lattice with stagger flux

Landau guage

Dinosaur
Stagger flux on square lattice. Every neighbor plaquette with opposite flux. The gauge of every plaquette are fixed on red bond. (\#fig:1)

We adopt tranverse gauge to study this problem . The unit cell could be selected as Fig(\@ref(fig:1)) . Every cell has two freedom $A,B$ . The Bloch hamiltonian on the momentum space could be written as

\[\begin{align} H(\mathbf{k}) t=\sum\limits_{\mathbf{k}} \begin{pmatrix} \psi_{A}^{\dagger}(k)& \psi_{B}^{\dagger}(k) \end{pmatrix} \begin{pmatrix} 0 & f(k) \\ f^{*}(k) &0 \end{pmatrix} \begin{pmatrix} \psi_A(k) \\ \psi_B(k)\\ \end{pmatrix} \end{align}\]

where $f(k)$ is

\[\begin{align} f(k)=1+e^{\text{i} (k_x-k_y)} +e^{\text{i}(-k_x-k_y)} +e^{-\text{i}(2k_y+\phi)} \end{align}\]

The enenrgy spectrum is derived as

\[\begin{align} \varepsilon(k)=\pm \sqrt{ \cos^2k_x +\cos^2(k_y+\frac{\phi}{2}) +2\cos\frac{\phi}{2} \cos k_x \cos(k_y+\frac{\phi}{2}) } (\#eq:3) \end{align}\]

We can find the Dirac points of dispersion (\@ref(eq:3)) easily, namly

\[\begin{align} \left( \pm \frac{\pi}{2},\pm \frac{\pi}{2}-\frac{\phi}{2} \right) \quad \left( \mp \frac{\pi}{2},\pm \frac{\pi}{2}-\frac{\phi}{2} \right) \end{align}\]

We will give the linear dispersion on symmetric gauge cases.

Symmetric gauge

Dinosaur
Stagger flux on square lattice. Every neighbor plaquette with opposite flux. The black bond admits $\frac{\phi}{4}$ and red bonds admits $-\frac{\phi}{4}$. (\#fig:2)

We consider tight binding model on square lattice with stagger flux $\phi$. As shown on Fig(\@ref(fig:2)), the unit cell are selected as the plaquette with positive flux $\phi$. Every unit cell has two freedom $A,B$, the Bloch hamiltonian could be described with four component spinor $\displaystyle \mathbf{\psi}(k)= (\psi_{A}(k)\quad \psi_{B}(k) )^{\mathbf{T}}$ . The Bloch hamiltonian reads as

\[\begin{align} H(\mathbf{k}) t=\sum\limits_{\mathbf{k}} \begin{pmatrix} \psi_{A}^{\dagger}(k)& \psi_{B}^{\dagger}(k) \end{pmatrix} \begin{pmatrix} 0 & f(k) \\ f^{*}(k) &0 \end{pmatrix} \begin{pmatrix} \psi_A(k) \\ \psi_B(k)\\ \end{pmatrix} \end{align}\]

where $f(k)$ is

\[\begin{align} f(k)= e^{\text{i} \frac{\phi}{4}} +e^{\text{i}\left( \frac{\phi}{4} -2k_y\right) } +e^{\text{i}\left( k_x-k_y-\frac{\phi}{4} \right) } +e^{\text{i}\left( -k_x-k_y-\frac{\phi}{4} \right) } \end{align}\]

Hence, the enenrgy spectrum is just $\varepsilon(k)=\pm \mid f(k)\mid$

\[\begin{align} \varepsilon(k)=\pm 2\sqrt{\cos^2 k_x+\cos^2k_y+2\cos \frac{\phi}{2} \cos k_x\cos k_y} (\#eq:14) \end{align}\]

The dispersion relation (\@ref(eq:14)) shows the zero point at $(\pm \frac{\pi}{2},\pm \frac{\pi}{2}), (\mp \frac{\pi}{2},\pm \frac{\pi}{2})$ . We expand the linear dispersion relation at this points.

\[\begin{align} f(k) \mid _{(\frac{\pi}{2},\frac{\pi}{2})}&=t e^{\text{i} \frac{\phi}{4}} \left( 1- e^{-\text{i}(2k_y-\pi)} \right) +e^{-\text{i} \frac{\phi}{4}} \left( e^{\text{i}(k_x-k_y)} - e^{-\text{i}(k_x+k_y-\pi)} \right) \nonumber \\ & \approx 2t\text{i} e^{\text{i} \frac{\phi}{4}} (k_y-\frac{\pi}{2}) + 2\text{i}e^{-\text{i} \frac{\phi}{4}} (k_x-\frac{\pi}{2}) \nonumber \\ &=2t\left( (k_x-k_y)\sin \frac{\phi}{4} +\text{i}(k_x+k_y-\pi) \cos\frac{\phi}{4} \right) \end{align}\]

The low enenrgy physics could be described by

\[\begin{align} H(k)=2t \vec{\text{K}} \cdot \vec{\sigma} \qquad \vec{K}=\vec{k} -(\frac{\pi}{2},\frac{\pi}{2}) \end{align}\]