稳态微扰理论 Théorie des perturbations stationnaires

稳态微扰理论用于解决含有微弱扰动的系统的问题。在这种方法中,系统的哈密顿量可以分解为两部分:一个是我们已经熟悉的,称为基态哈密顿量,而另一个是微扰哈密顿量,用来描述我们添加的微小扰动。这个方法的目标是通过微扰哈密顿量的小参数来展开能量和波函数,以便在扰动项较小的情况下解出系统的性质。

问题描述


对于无扰动的粒子:

\[ \widehat{H}_0\left|\psi_j^{(0)}\right\rangle=\varepsilon_j^{(0)}\left|\psi_j^{(0)}\right\rangle \]

其中,\(\widehat H_0\)被称为基态哈密顿量,\(\left|\psi_j^{(0)}\right\rangle\)被称为基态,\(\varepsilon_j^{(0)}\)被称为基态能量。

当粒子受到扰动:\(\widehat{W}=\eta \widehat{V}\),薛定谔方程转化为:

\[ \widehat{H}_0|\psi\rangle+\widehat{W}|\psi\rangle=\varepsilon|\psi\rangle \]

得到新的本征态和本征能量。

解决思路


使用受到不同扰动得态和能量表示最终的本征态和本征能量:

\[ |\psi\rangle=\sum_n \eta^n\left|\psi_i^{(n)}\right\rangle, \varepsilon=\sum_n \eta^n \varepsilon_i^{(n)} \]

得到新的薛定谔方程:

\[ \widehat{H}_0 \sum_{n=0} \eta^n|\psi_i^{(n)}\rangle+\eta \widehat{V} \sum_{n=0} \eta^n|\psi_i^{(n)}\rangle=\sum_{n=0} \eta^n \varepsilon_i^{(n)} \sum_{n^{\prime}=0} \eta^{n^{\prime}}|\psi_i^{\left(n^{\prime}\right)}\rangle \]

我们来看几个具体的例子。

  • 对于零阶的情况:

\[ \widehat{H}_0\left|\psi_i^{(0)}\right\rangle-\varepsilon_i^{(0)}\left|\psi_i^{(0)}\right\rangle=0 \]

  • 对于一阶的情况:

\[ \widehat{H}_0\left|\psi_i^{(1)}\right\rangle+\widehat{V}\left|\psi_i^{(0)}\right\rangle-\varepsilon_i^{(0)}\left|\psi_i^{(1)}\right\rangle-\varepsilon_i^{(1)}\left|\psi_i^{(0)}\right\rangle=0 \]

证明过程很简单,展开即可。

对于一阶的情况还可以考虑其总的概率密度为1:

\[ \|\left|\psi_i^{(0)}\right\rangle+\eta\left|\psi_i^{(1)}\right\rangle \|^2=1 \]

忽略\(\eta^2\)对应的项,有:

\[ \underbrace{\left\langle\psi_i^{(0)} \mid \psi_i^{(0)}\right\rangle}_1+2 \eta \underbrace{\operatorname{Re}\left(\left\langle\psi_i^{(1)} \mid \psi_i^{(0)}\right\rangle\right)}_0 \approx 1 \]

并近似认为第一项远大于第二项,从而有:

\[ \left\langle\psi_i^{(1)} \mid \psi_i^{(0)}\right\rangle=0 \]

非退化的情况


对于非退化的情况,有\(\left|\psi_j^{(0)}\right\rangle\ne\left|\psi_i^{(0)}\right\rangle \quad si \quad i\ne j\)。使用非退化的基表示本征态:

\[ \left|\psi_i^{(n)}\right\rangle=\sum_j c_{i, j}^{(n)}\left|\psi_j^{(0)}\right\rangle \]


先考虑一阶的情况。由于\(\left\langle\psi_i^{(1)} \mid \psi_i^{(0)}\right\rangle=0\),有:

\[ \left|\psi_i^{(1)}\right\rangle=\sum_{j \neq i} c_j^{(1)}\left|\psi_j^{(0)}\right\rangle \]

回想一阶的薛定谔方程:

\[ \widehat{H}_0\left|\psi_i^{(1)}\right\rangle+\widehat{V}\left|\psi_i^{(0)}\right\rangle-\varepsilon_i^{(0)}\left|\psi_i^{(1)}\right\rangle-\varepsilon_i^{(1)}\left|\psi_i^{(0)}\right\rangle=0 \]

\(\left\langle\psi_k^{(0)}\right|\)投影:

\[ \begin{gathered} \left.\left\langle\psi_k^{(0)}\left|\widehat{H}_0\right| \psi_i^{(1)}\right\rangle+\left\langle\psi_k^{(0)}|\widehat{V}| \psi_i^{(0)}\right\rangle\right. -\left\langle\psi_k^{(0)}\left|\varepsilon_i^{(0)}\right| \psi_i^{(1)}\right\rangle-\left\langle\psi_k^{(0)}\left|\varepsilon_i^{(1)}\right| \psi_i^{(0)}\right\rangle=0 \end{gathered} \]

  • 第一项和第三项得:\(\sum_{j \neq i} c_j^{(1)}\left(\varepsilon_j^{(0)}-\varepsilon_i^{(0)}\right) \delta_{j, k}\)
  • 第二项暂时不可求
  • 第四项得:\(\varepsilon_i^{(1)} \delta_{i, k}\)

从而得到:

\[ \sum_{j \neq i} c_j^{(1)}\left(\varepsilon_i^{(0)}-\varepsilon_j^{(0)}\right) \delta_{j, k}+\varepsilon_i^{(1)} \delta_{i, k}=\left\langle\psi_k^{(0)}|\widehat{V}| \psi_i^{(0)}\right\rangle \]

  • \(k = i\)时,\(\varepsilon_i^{(1)}=\left\langle\psi_i^{(0)}|\widehat{V}| \psi_i^{(0)}\right\rangle\)\(\varepsilon \approx \varepsilon_i^{(0)}+\eta\left\langle\psi_i^{(0)}|\widehat{V}| \psi_i^{(0)}\right\rangle\)
  • \(k \ne i\)时,\(c_k^{(1)}=\frac{\left\langle\psi_k^{(0)}|\widehat{V}| \psi_i^{(0)}\right\rangle}{\varepsilon_i^{(0)}-\varepsilon_k^{(0)}}\)\(|\psi\rangle \approx\left|\psi_i^{(0)}\right\rangle+\eta \sum_{k \neq i} \frac{\left\langle\psi_k^{(0)}|\widehat{V}| \psi_i^{(0)}\right\rangle}{\varepsilon_i^{(0)}-\varepsilon_k^{(0)}}\left|\psi_k^{(0)}\right\rangle\)

直接给出二阶结果:

\[ \begin{gathered}\varepsilon_i^{(2)}=\sum_{k \neq i} \frac{\left|\left\langle\psi_i^{(0)}|\widehat{V}| \psi_k^{(0)}\right\rangle\right|^2}{\varepsilon_i^{(0)}-\varepsilon_k^{(0)}} \\\varepsilon \approx \varepsilon_i^{(0)}+\eta\left\langle\psi_i^{(0)}|\widehat{V}| \psi_i^{(0)}\right\rangle+\eta^2 \sum_{k \neq i} \frac{\left|\left\langle\psi_i^{(0)}|\widehat{V}| \psi_k^{(0)}\right\rangle\right|^2}{\varepsilon_i^{(0)}-\varepsilon_k^{(0)}}\end{gathered} \]

兼并情况


兼并情况下,存在多个能量相同的状态,某一基态可以表示为:

\[ \left|\psi_i^{(0)}\right\rangle=\sum_g \alpha_{i, g}\left|\psi_{i, g}^{(0)}\right\rangle \]

在一阶,同样考虑薛定谔方程在\(\left\langle\psi_{i,g}^{(0)}\right|\)的投影:

\[ \begin{gathered} \left\langle\psi_{i, g}^{(0)}\left|\widehat{H}_0\right| \psi_i^{(1)}\right\rangle+\left\langle\psi_{i, g}^{(0)}|\widehat{V}| \psi_i^{(0)}\right\rangle -\left\langle\psi_{i, g}^{(0)}\left|\varepsilon_i^{(0)}\right| \psi_i^{(1)}\right\rangle-\left\langle\psi_{i, g}^{(0)}\left|\varepsilon_i^{(1)}\right| \psi_i^{(0)}\right\rangle=0 \end{gathered} \]

  • 第一项和第三项显然为0
  • 第二项和第四项:\(\sum_{g^{\prime}} \alpha_{i, g^{\prime}}\left\langle\psi_{i, g}^{(0)}|\widehat{V}| \psi_{i, g^{\prime}}^{(0)}\right\rangle-\alpha_{i, g^{\prime}}^{(1)} \delta_{g, g^{\prime}}\)

总的薛定谔方程转化为:

\[ \sum_{g^{\prime}} \alpha_{i, g^{\prime}}\left\langle\psi_{i, g}^{(0)}|\widehat{V}| \psi_{i, g^{\prime}}^{(0)}\right\rangle-\alpha_{i, g^{\prime}} \varepsilon_i^{(1)} \delta_{g, g^{\prime}}=0 \]

展开为:

\[ \begin{gathered} \sum_{g^{\prime}} \alpha_{g^{\prime}}\left\langle\psi_{i, 1}^{(0)}|\widehat{V}| \psi_{i, g^{\prime}}^{(0)}\right\rangle-\alpha_1 \varepsilon_i^{(1)}=0 \\ \sum_{g^{\prime}} \alpha_{g^{\prime}}\left\langle\psi_{i, 2}^{(0)}|\widehat{V}| \psi_{i, g^{\prime}}^{(0)}\right\rangle-\alpha_2 \varepsilon_i^{(1)}=0 \\ \vdots \\ \sum_{g^{\prime}} \alpha_{g^{\prime}}\left\langle\psi_{i, g}^{(0)}|\widehat{V}| \psi_{i, g^{\prime}}^{(0)}\right\rangle-\alpha_g \varepsilon_i^{(1)}=0 \end{gathered} \]

写作矩阵表示:

\[ \left(\begin{array}{cccc} V_{1,1} & V_{1,2} & \cdots & V_{1, g} \\ V_{2,1} & V_{2,2} & \cdots & V_{2, g} \\ \vdots & \vdots & \ddots & \vdots \\ V_{g, 1} & V_{g, 2} & \cdots & V_{g, g} \end{array}\right)\left(\begin{array}{c} \alpha_1 \\ \alpha_2 \\ \vdots \\ \alpha_g \end{array}\right)=\varepsilon_i^{(1)}\left(\begin{array}{c} \alpha_1 \\ \alpha_2 \\ \vdots \\ \alpha_g \end{array}\right) \]

即可得能量。

一个简单的示例


考虑一个二维势阱:

\[ \begin{aligned}&\varepsilon_{n_x, n_y}=\frac{\left(n_x^2+n_y^2\right) \pi^2 \hbar^2}{2 m a^2}, n_x, n_y \in \mathbb{N}^*\\&\psi_{n_x, n_y}(x, y)=\frac{2}{a} \sin \left(\frac{n_x \pi x}{a}\right) \sin \left(\frac{n_y \pi y}{a}\right)\end{aligned} \]

  • 对于第一能级,是一个非简并的状态,\(\varepsilon_1^{(0)}=\varepsilon_{1,1}=\frac{\pi^2 \hbar^2}{m a^2}\)
  • 对于第二能级,是一个兼并的状态,\(\varepsilon_2^{(0)}=\varepsilon_{1,2}=\varepsilon_{2,1}=\frac{5 \pi^2 \hbar^2}{2 m a^2}\)

考虑一个干扰:\(W(x, y)=\eta V(x,y) = \left(\frac{a}{2}\right)^2 \eta \delta\left(x-x_0\right) \delta\left(y-y_0\right)\)

考虑矩阵表示:

\[ \left(\begin{array}{cccc}V_{1,1} & V_{1,2} & \cdots & V_{1, g} \\V_{2,1} & V_{2,2} & \cdots & V_{2, g} \\\vdots & \vdots & \ddots & \vdots \\V_{g, 1} & V_{g, 2} & \cdots & V_{g, g}\end{array}\right)\left(\begin{array}{c}\alpha_1 \\\alpha_2 \\\vdots \\\alpha_g\end{array}\right)=\varepsilon_i^{(1)}\left(\begin{array}{c}\alpha_1 \\\alpha_2 \\\vdots \\\alpha_g\end{array}\right) \]

为了简单,考虑二阶情况

\[ \left(\begin{array}{ll}V_{1,1} & V_{1,2} \\V_{2,1} & V_{2,2}\end{array}\right)\binom{\alpha_1}{\alpha_2}=\varepsilon_i^{(1)}\binom{\alpha_1}{\alpha_2} \]

分别计算几个\(V_{i,j}\)

\[ \begin{aligned}V_{1,1} & =\langle 1,2|\widehat{V}| 1,2\rangle=\sin ^2\left(\frac{\pi x_0}{a}\right) \sin ^2\left(\frac{2 \pi y_0}{a}\right)=\frac{9}{16} \\V_{2,2} & =\langle 2,1|\widehat{V}| 2,1\rangle=\sin ^2\left(\frac{2 \pi x_0}{a}\right) \sin ^2\left(\frac{\pi y_0}{a}\right)=\frac{3}{16} \\V_{1,2} & =\langle 1,2|\widehat{V}| 2,1\rangle \\& =\sin \left(\frac{\pi x_0}{a}\right) \sin \left(\frac{2 \pi y_0}{a}\right) \sin \left(\frac{2 \pi x_0}{a}\right) \sin \left(\frac{\pi y_0}{a}\right)=\frac{3 \sqrt{3}}{16} \\V_{2,1} & =\langle 2,1|\widehat{V}| 1,2\rangle=\langle 1,2|\widehat{V}| 2,1\rangle=V_{1,2}=\frac{3 \sqrt{3}}{16}\end{aligned} \]

得:

\[ \left(\begin{array}{ll}V_{1,1} & V_{1,2} \\V_{2,1} & V_{2,2}\end{array}\right)=\frac{3}{16}\left(\begin{array}{cc}3 & \sqrt{3} \\\sqrt{3} & 1\end{array}\right) \]

计算本征值:

\[ \begin{aligned} \left|\begin{array}{cc} V_{1,1}-\varepsilon & V_{1,2} \\ V_{2,1} & V_{2,2}-\varepsilon \end{array}\right| & =\left(\frac{9}{16}-\varepsilon\right)\left(\frac{3}{16}-\varepsilon\right)-\left(\frac{3 \sqrt{3}}{16}\right)^2 \\ & =\varepsilon^2-\frac{3}{4} \varepsilon=\varepsilon\left(\varepsilon-\frac{3}{4}\right)=0 \end{aligned} \]

得到:

\[ \varepsilon_{2-}^{(1)}=0, \varepsilon_{2+}^{(1)}=\frac{3}{4} \]

计算本征向量:

\[ \begin{gathered}\varepsilon_{2-}^{(1)}=0 \Rightarrow\left(\begin{array}{cc}3 & \sqrt{3} \\\sqrt{3} & 1\end{array}\right)\binom{\alpha_1}{\alpha_2}=\binom{0}{0} \\|-\rangle=\frac{1}{2}(|1,2\rangle-\sqrt{3}|2,1\rangle) \\\varepsilon_{2+}^{(1)}=\frac{3}{4} \Rightarrow\left(\begin{array}{cc}-1 & \sqrt{3} \\\sqrt{3} & -3\end{array}\right)\binom{\alpha_1}{\alpha_2}=\binom{0}{0} \\|+\rangle=\frac{1}{2}(\sqrt{3}|1,2\rangle+|2,1\rangle)\end{gathered} \]