模型和能带, 学术

BBH高阶拓扑绝缘体模型(附Python代码)

BBH模型(Benalcazar-Bernevig-Hughes Model)的参考文献为:Wladimir A. Benalcazar, B. Andrei Bernevig, and Taylor L. Hughes, "Quantized electric multipole insulators", Science Volume 357:61-66 (2017).

模型:

h=[γ+λcos(kx)]Γ4+λsin(kx)Γ3+[γ+λcos(ky)]Γ2+λsin(ky)Γ1+δΓ0\begin{aligned} h=&[\gamma+\lambda cos(k_x)]\Gamma_4+\lambda sin(k_x)\Gamma_3\\ &+[\gamma+\lambda cos(k_y)]\Gamma_2+\lambda sin(k_y)\Gamma_1+\delta \Gamma_0 \end{aligned}

其中,

Γ0=τ3σ0=(1001−100−1)\Gamma_0=\tau_3 \sigma_0=\begin{pmatrix}1 & 0 & & \\0 & 1 & & \\& & -1 & 0 \\& & 0 & -1\end{pmatrix}
Γ1=−τ2σ1=−(0−i−i00i)\Gamma_1=-\tau_2\sigma_1=-\begin{pmatrix}& & 0 & -i \\& & -i & 0 \\0 & i & & \\i & 0 & &\end{pmatrix}
Γ2=−τ2σ2=−(0−11001)\Gamma_2=-\tau_2\sigma_2=-\begin{pmatrix}& & 0 & -1 \\& & 1 & 0 \\0 & 1 & & \\-1 & 0 & &\end{pmatrix}
Γ3=−τ2σ3=−(−i00ii0)\Gamma_3=-\tau_2\sigma_3=-\begin{pmatrix}& & -i & 0 \\& & 0 & i \\i & 0 & & \\0 & -i & &\end{pmatrix}
Γ4=τ1σ0=(100110)\Gamma_4=\tau_1\sigma_0=\begin{pmatrix}& & 1 & 0 \\& & 0 & 1 \\1 & 0 & & \\0 & 1 & &\end{pmatrix}

以上用到泡利矩阵的张量积:泡利矩阵以及泡利矩阵的张量积。

所以哈密顿量展开写为:

h=(δ0γ+λcos(kx)+iλsin(kx)γ+λcos(ky)+iλsin(ky)0δ−[γ+λcos(ky)]+iλsin(ky)γ+λcos(kx)−iλsin(kx)γ+λcos(kx)−iλsin(kx)−[γ+λcos(ky)]−iλsin(ky)−δ0γ+λcos(ky)−iλsin(ky)γ+λcos(kx)+iλsin(kx)0−δ)=(δ0γ+λexp(ikx)γ+λexp(iky)0δ−γ−λexp(−iky)γ+λexp(−ikx)γ+λexp(−ikx)−γ−λexp(iky)−δ0γ+λexp(−iky)γ+λexp(ikx)0−δ)\begin{aligned}h&=\begin{pmatrix}\delta & 0 & \gamma+\lambda cos(k_x)+i\lambda sin(k_x) & \gamma+\lambda cos(k_y)+i \lambda sin(k_y) \\0 & \delta & -[\gamma+\lambda cos(k_y)]+i \lambda sin(k_y) & \gamma+\lambda cos(k_x)-i \lambda sin(k_x) \\\gamma+\lambda cos(k_x)-i \lambda sin(k_x) & -[\gamma+\lambda cos(k_y)]-i \lambda sin(k_y) & -\delta & 0 \\\gamma+\lambda cos(k_y)-i \lambda sin(k_y) & \gamma+\lambda cos(k_x)+i \lambda sin(k_x) & 0 & -\delta\end{pmatrix}\\&=\begin{pmatrix}\delta & 0 & \gamma+\lambda exp(ik_x) & \gamma+\lambda exp(ik_y) \\0 & \delta & -\gamma-\lambda exp(-i k_y) & \gamma+\lambda exp(-i k_x) \\\gamma+\lambda exp(-i k_x) & -\gamma-\lambda exp(i k_y) & -\delta & 0 \\\gamma+\lambda exp(-ik_y) & \gamma+\lambda exp(i k_x) & 0 & -\delta\end{pmatrix}\end{aligned}

文献中给出的对应示意图:

元胞为(1,2,3,4)。把哈密顿量反傅里叶变换后,可以知道:

  • 元胞内部:3到1的跃迁项为γ\gamma;2到4的跃迁项为γ\gamma;4到1的跃迁项为γ\gamma;2到3的跃迁项为−γ-\gamma。
  • 元胞之间:1到3的跃迁项为λ\lambda;4到2的跃迁项为λ\lambda;1到4的跃迁项为λ\lambda;3到2的跃迁项为−λ-\lambda。

即:元胞内部跃迁项绝对值为γ\gamma,元胞间的跃迁项绝对值为λ\lambda。此外,2和3之间的跃迁要加个负号。

计算局域态密度(LDOS)的Python代码为:

"""
This code is supported by the website: https://www.guanjihuan.com
The newest version of this code is on the web page: https://www.guanjihuan.com/archives/8557
"""

import numpy as np
from math import *


def hamiltonian(Nx, Ny): 
    delta = 1e-3
    gamma = 1e-3
    lambda0 = 1
    h = np.zeros((4*Nx*Ny, 4*Nx*Ny))
    # 元胞内部跃迁
    for x in range(Nx):
        for y in range(Ny):
            h[x*Ny*4+y*4+0, x*Ny*4+y*4+0] = delta
            h[x*Ny*4+y*4+1, x*Ny*4+y*4+1] = delta
            h[x*Ny*4+y*4+2, x*Ny*4+y*4+2] = -delta
            h[x*Ny*4+y*4+3, x*Ny*4+y*4+3] = -delta

            h[x*Ny*4+y*4+0, x*Ny*4+y*4+2] = gamma
            h[x*Ny*4+y*4+0, x*Ny*4+y*4+3] = gamma
            h[x*Ny*4+y*4+1, x*Ny*4+y*4+2] = -gamma
            h[x*Ny*4+y*4+1, x*Ny*4+y*4+3] = gamma
            h[x*Ny*4+y*4+2, x*Ny*4+y*4+0] = gamma
            h[x*Ny*4+y*4+2, x*Ny*4+y*4+1] = -gamma
            h[x*Ny*4+y*4+3, x*Ny*4+y*4+0] = gamma
            h[x*Ny*4+y*4+3, x*Ny*4+y*4+1] = gamma

    # y方向上的元胞间跃迁
    for x in range(Nx):
        for y in range(Ny-1):
            h[x*Ny*4+y*4+0, x*Ny*4+(y+1)*4+3] = lambda0
            h[x*Ny*4+(y+1)*4+1, x*Ny*4+y*4+2] = -lambda0
            h[x*Ny*4+y*4+2, x*Ny*4+(y+1)*4+1] = -lambda0
            h[x*Ny*4+(y+1)*4+3, x*Ny*4+y*4+0] = lambda0

    # x方向上的元胞间跃迁
    for x in range(Nx-1):
        for y in range(Ny):
            h[x*Ny*4+y*4+0, (x+1)*Ny*4+y*4+2] = lambda0
            h[(x+1)*Ny*4+y*4+1, x*Ny*4+y*4+3] = lambda0
            h[(x+1)*Ny*4+y*4+2, x*Ny*4+y*4+0] = lambda0
            h[x*Ny*4+y*4+3, (x+1)*Ny*4+y*4+1] = lambda0
    return h


def main():
    Nx = 10
    Ny = 10
    fermi_energy = 0
    h = hamiltonian(Nx, Ny)
    green = np.linalg.inv((fermi_energy+1e-6j)*np.eye(h.shape[0])-h)

    x_array = []
    y_array = []
    DOS = []
    for x in range(Nx):
        for y in range(Ny):
            x_array.append(x*2+2)
            y_array.append(y*2+2)
            DOS.append(-np.imag(green[x*Ny*4+y*4+0, x*Ny*4+y*4+0])/pi)

            x_array.append(x*2+1)
            y_array.append(y*2+1)
            DOS.append(-np.imag(green[x*Ny*4+y*4+1, x*Ny*4+y*4+1])/pi)

            x_array.append(x*2+1)
            y_array.append(y*2+2)
            DOS.append(-np.imag(green[x*Ny*4+y*4+2, x*Ny*4+y*4+2])/pi)

            x_array.append(x*2+2)
            y_array.append(y*2+1)
            DOS.append(-np.imag(green[x*Ny*4+y*4+3, x*Ny*4+y*4+3])/pi)
    DOS = DOS/np.sum(DOS)
    Plot_2D_Scatter(x_array, y_array, DOS, xlabel='x', ylabel='y', title='BBH Model', filename='BBH Model')


def Plot_2D_Scatter(x, y, value, xlabel='x', ylabel='y', title='title', filename='a'): 
    import matplotlib.pyplot as plt
    fig = plt.figure()
    ax = fig.add_subplot(111)
    plt.subplots_adjust(bottom=0.2, right=0.8, left=0.2) 
    for i in range(np.array(x).shape[0]):
        ax.scatter(x[i], y[i], marker='o', s=1000*value[i], c=[(1,0,0)])
    ax.set_title(title, fontsize=20, fontfamily='Times New Roman')
    ax.set_xlabel(xlabel, fontsize=20, fontfamily='Times New Roman') 
    ax.set_ylabel(ylabel, fontsize=20, fontfamily='Times New Roman') 
    ax.tick_params(labelsize=15)  # 设置刻度值字体大小
    labels = ax.get_xticklabels() + ax.get_yticklabels() 
    [label.set_fontname('Times New Roman') for label in labels] # 设置刻度值字体
    # plt.savefig(filename+'.jpg', dpi=300) 
    plt.show()
    plt.close('all')


if __name__ == '__main__':
    main()

计算结果为:

换一种画图方式(Plot 3D Surface):

"""
This code is supported by the website: https://www.guanjihuan.com
The newest version of this code is on the web page: https://www.guanjihuan.com/archives/8557
"""

import numpy as np
from math import *


def hamiltonian(Nx, Ny): 
    delta = 1e-3
    gamma = 1e-3
    lambda0 = 1
    h = np.zeros((4*Nx*Ny, 4*Nx*Ny))
    # 元胞内部跃迁
    for x in range(Nx):
        for y in range(Ny):
            h[x*Ny*4+y*4+0, x*Ny*4+y*4+0] = delta
            h[x*Ny*4+y*4+1, x*Ny*4+y*4+1] = delta
            h[x*Ny*4+y*4+2, x*Ny*4+y*4+2] = -delta
            h[x*Ny*4+y*4+3, x*Ny*4+y*4+3] = -delta

            h[x*Ny*4+y*4+0, x*Ny*4+y*4+2] = gamma
            h[x*Ny*4+y*4+0, x*Ny*4+y*4+3] = gamma
            h[x*Ny*4+y*4+1, x*Ny*4+y*4+2] = -gamma
            h[x*Ny*4+y*4+1, x*Ny*4+y*4+3] = gamma
            h[x*Ny*4+y*4+2, x*Ny*4+y*4+0] = gamma
            h[x*Ny*4+y*4+2, x*Ny*4+y*4+1] = -gamma
            h[x*Ny*4+y*4+3, x*Ny*4+y*4+0] = gamma
            h[x*Ny*4+y*4+3, x*Ny*4+y*4+1] = gamma

    # y方向上的元胞间跃迁
    for x in range(Nx):
        for y in range(Ny-1):
            h[x*Ny*4+y*4+0, x*Ny*4+(y+1)*4+3] = lambda0
            h[x*Ny*4+(y+1)*4+1, x*Ny*4+y*4+2] = -lambda0
            h[x*Ny*4+y*4+2, x*Ny*4+(y+1)*4+1] = -lambda0
            h[x*Ny*4+(y+1)*4+3, x*Ny*4+y*4+0] = lambda0

    # x方向上的元胞间跃迁
    for x in range(Nx-1):
        for y in range(Ny):
            h[x*Ny*4+y*4+0, (x+1)*Ny*4+y*4+2] = lambda0
            h[(x+1)*Ny*4+y*4+1, x*Ny*4+y*4+3] = lambda0
            h[(x+1)*Ny*4+y*4+2, x*Ny*4+y*4+0] = lambda0
            h[x*Ny*4+y*4+3, (x+1)*Ny*4+y*4+1] = lambda0
    return h


def main():
    Nx = 10
    Ny = 10
    fermi_energy = 0
    h = hamiltonian(Nx, Ny)
    green = np.linalg.inv((fermi_energy+1e-6j)*np.eye(h.shape[0])-h)
    DOS = np.zeros((Ny*2, Nx*2))
    for x in range(Nx):
        for y in range(Ny):
            DOS[y*2+1, x*2+1] = -np.imag(green[x*Ny*4+y*4+0, x*Ny*4+y*4+0])/pi
            DOS[y*2+0, x*2+0] = -np.imag(green[x*Ny*4+y*4+1, x*Ny*4+y*4+1])/pi
            DOS[y*2+1, x*2+0] = -np.imag(green[x*Ny*4+y*4+2, x*Ny*4+y*4+2])/pi
            DOS[y*2+0, x*2+1] = -np.imag(green[x*Ny*4+y*4+3, x*Ny*4+y*4+3])/pi
    DOS = DOS/np.sum(DOS)
    Plot_3D_Surface(np.arange(1, 2*Nx+0.001), np.arange(1, 2*Ny+0.001), DOS, xlabel='x', ylabel='y', zlabel='DOS', title='BBH Model', filename='BBH Model')


def Plot_3D_Surface(x, y, matrix, xlabel='x', ylabel='y', zlabel='z', title='title', filename='a'): 
    import matplotlib.pyplot as plt
    from matplotlib import cm
    from matplotlib.ticker import LinearLocator
    fig, ax = plt.subplots(subplot_kw={"projection": "3d"})
    plt.subplots_adjust(bottom=0.1, right=0.65) 
    x, y = np.meshgrid(x, y)
    if len(matrix.shape) == 2:
        surf = ax.plot_surface(x, y, matrix, cmap=cm.coolwarm, linewidth=0, antialiased=False) 
    elif len(matrix.shape) == 3:
        for i0 in range(matrix.shape[2]):
            surf = ax.plot_surface(x, y, matrix[:,:,i0], cmap=cm.coolwarm, linewidth=0, antialiased=False) 
    ax.set_title(title, fontsize=20, fontfamily='Times New Roman')
    ax.set_xlabel(xlabel, fontsize=20, fontfamily='Times New Roman') 
    ax.set_ylabel(ylabel, fontsize=20, fontfamily='Times New Roman') 
    ax.set_zlabel(zlabel, fontsize=20, fontfamily='Times New Roman') 
    # ax.set_zlim(-1, 1)  # 设置z轴的范围
    ax.zaxis.set_major_locator(LinearLocator(2)) # 设置z轴主刻度的个数
    ax.zaxis.set_major_formatter('{x:.2f}')   # 设置z轴主刻度的格式
    ax.tick_params(labelsize=15)  # 设置刻度值字体大小
    labels = ax.get_xticklabels() + ax.get_yticklabels() + ax.get_zticklabels()
    [label.set_fontname('Times New Roman') for label in labels] # 设置刻度值字体
    cax = plt.axes([0.80, 0.15, 0.05, 0.75]) # color bar的位置 [左,下,宽度, 高度]
    cbar = fig.colorbar(surf, cax=cax)  # color bar
    cbar.ax.tick_params(labelsize=15) # 设置color bar刻度的字体大小
    for l in cbar.ax.yaxis.get_ticklabels():
        l.set_family('Times New Roman')
    # plt.savefig(filename+'.jpg', dpi=300) 
    plt.show()
    plt.close('all')


if __name__ == '__main__':
    main()

计算结果为:

评论功能已关闭,以下留言为历史存档,仅供查看。