エーレンフェストの定理

まず、期待値の定義より

${\displaystyle {\begin{aligned}{\frac {\mathrm {d} ^{2}}{\mathrm {d} t^{2}}}\langle \mathbf {r} \rangle &={\frac {\mathrm {d} ^{2}}{\mathrm {d} t^{2}}}\int \psi ^{*}(\mathbf {r} ,t)\mathbf {r} \psi (\mathbf {r} ,t)\mathrm {d} \mathbf {r} \\&={\frac {\mathrm {d} }{\mathrm {d} t}}\int \left[{\frac {\partial \psi ^{*}}{\partial t}}\mathbf {r} \psi +\psi ^{*}\mathbf {r} {\frac {\partial \psi }{\partial t}}\right]\mathrm {d} \mathbf {r} \end{aligned}}}$

を得る。ここでシュレーディンガー方程式より

${\displaystyle {\begin{aligned}{\frac {\mathrm {d} }{\mathrm {d} t}}\int \left[{\frac {\partial \psi ^{*}}{\partial t}}\mathbf {r} \psi +\psi ^{*}\mathbf {r} {\frac {\partial \psi }{\partial t}}\right]\mathrm {d} \mathbf {r} &={\frac {\mathrm {d} }{\mathrm {d} t}}\int \left[-{\frac {1}{i\hbar }}({\hat {H}}\psi ^{*})\mathbf {r} \psi +\psi ^{*}\mathbf {r} {\frac {1}{i\hbar }}({\hat {H}}\psi )\right]\mathrm {d} \mathbf {r} \\&={\frac {1}{i\hbar }}{\frac {\mathrm {d} }{\mathrm {d} t}}\int \left[-\left\{-{\frac {\hbar ^{2}}{2m}}\nabla ^{2}+U(\mathbf {r} )\right\}\psi ^{*}\mathbf {r} \psi +\psi ^{*}\mathbf {r} \left\{{\frac {\hbar ^{2}}{2m}}\nabla ^{2}+U(\mathbf {r} )\right\}\psi \right]\mathrm {d} \mathbf {r} \\&={\frac {1}{i\hbar }}{\frac {\mathrm {d} }{\mathrm {d} t}}\int \left[{\frac {\hbar ^{2}}{2m}}\nabla ^{2}\psi ^{*}\mathbf {r} \psi -\psi ^{*}\mathbf {r} {\frac {\hbar ^{2}}{2m}}\nabla ^{2}\psi \right]\mathrm {d} \mathbf {r} \\&=-{\frac {i\hbar }{2m}}{\frac {\mathrm {d} }{\mathrm {d} t}}\int \left[\nabla ^{2}\psi ^{*}\mathbf {r} \psi -\psi ^{*}\mathbf {r} \nabla ^{2}\psi \right]\mathrm {d} \mathbf {r} \end{aligned}}}$

部分積分と、積分範囲が空間全体にわたること、及び波動関数は無限遠では0となるという仮定を用いると

${\displaystyle {\begin{aligned}\int \nabla ^{2}\psi ^{*}\mathbf {r} \psi \ \mathrm {d} \mathbf {r} &=[\nabla \psi ^{*}\mathbf {r} \psi ]_{-\infty }^{+\infty }-\int \nabla \psi ^{*}\nabla (\mathbf {r} \psi )\mathrm {d} \mathbf {r} \\&=-[\psi ^{*}\nabla (\mathbf {r} \psi )]_{-\infty }^{+\infty }+\int \psi ^{*}\nabla ^{2}(\mathbf {r} \psi )\mathrm {d} \mathbf {r} \\&=\int \psi ^{*}\nabla (\nabla \mathbf {r} \psi +\mathbf {r} \nabla \psi )\mathrm {d} \mathbf {r} \\&=\int \left[\psi ^{*}\nabla \psi +\psi ^{*}\nabla (\mathbf {r} \nabla \psi )\right]\mathrm {d} \mathbf {r} \\&=\int \left[2\psi ^{*}\nabla \psi +\psi ^{*}\mathbf {r} \nabla ^{2}\psi \right]\mathrm {d} \mathbf {r} \end{aligned}}}$

これらを用いると

${\displaystyle m{\frac {\mathrm {d} ^{2}}{\mathrm {d} t^{2}}}\langle \mathbf {r} \rangle =-i\hbar {\frac {\mathrm {d} }{\mathrm {d} t}}\int \psi ^{*}\nabla \psi \mathrm {d} \mathbf {r} =-i\hbar \int \left[{\frac {\partial \psi ^{*}}{\partial t}}\nabla \psi +\psi ^{*}\nabla {\frac {\partial \psi }{\partial t}}\right]\mathrm {d} \mathbf {r} }$

再度シュレーディンガー方程式を用いて

${\displaystyle {\begin{aligned}-i\hbar \int \left[{\frac {\partial \psi ^{*}}{\partial t}}\nabla \psi +\psi ^{*}\nabla {\frac {\partial \psi }{\partial t}}\right]\mathrm {d} \mathbf {r} &=-i\hbar \int \left[-{\frac {1}{i\hbar }}({\hat {H}}\psi ^{*})\nabla \psi +\psi ^{*}\nabla {\frac {1}{i\hbar }}({\hat {H}}\psi )\right]\mathrm {d} \mathbf {r} \\&=\int \left[\left\{-{\frac {\hbar ^{2}}{2m}}\nabla ^{2}+U(\mathbf {r} )\right\}\psi ^{*}\nabla \psi -\psi ^{*}\nabla \left\{-{\frac {\hbar ^{2}}{2m}}\nabla ^{2}+U(\mathbf {r} )\right\}\psi \right]\mathrm {d} \mathbf {r} \\&=-{\frac {\hbar ^{2}}{2m}}\int \left[\nabla ^{2}\psi ^{*}\nabla \psi -\psi ^{*}\nabla ^{3}\psi \right]\mathrm {d} \mathbf {r} +\int \left[U(\mathbf {r} )\psi ^{*}\nabla \psi -\psi ^{*}\nabla (U(\mathbf {r} )\psi )\right]\mathrm {d} \mathbf {r} \end{aligned}}}$

また部分積分を使うと、

${\displaystyle {\begin{aligned}\int \nabla ^{2}\psi ^{*}\nabla \psi \mathrm {d} \mathbf {r} &=[\nabla \psi ^{*}\nabla \psi ]_{-\infty }^{+\infty }-\int \nabla \psi ^{*}\nabla ^{2}\psi \mathrm {d} \mathbf {r} \\&=-[\psi ^{*}\nabla ^{2}\psi ]_{-\infty }^{+\infty }+\int \psi ^{*}\nabla ^{3}\psi \mathrm {d} \mathbf {r} \\&=\int \psi ^{*}\nabla ^{3}\psi \mathrm {d} \mathbf {r} \end{aligned}}}$

加えて

${\displaystyle {\begin{aligned}U\psi ^{*}\nabla \psi -\psi ^{*}\nabla (U\psi )&=U\psi ^{*}\nabla \psi -\psi ^{*}\nabla U\psi -U\psi ^{*}\nabla \psi \\&=-\psi ^{*}\nabla U\psi \end{aligned}}}$

を用いると、

${\displaystyle m{\frac {\mathrm {d} ^{2}}{\mathrm {d} t^{2}}}\langle \mathbf {r} \rangle =-\int \psi ^{*}\nabla U\psi \mathrm {d} \mathbf {r} }$

を得る。この右辺の積分は、期待値の導出法から${\displaystyle \nabla U}$の期待値であるから、

${\displaystyle m{\frac {\mathrm {d} ^{2}}{\mathrm {d} t^{2}}}\langle \mathbf {r} \rangle =-\langle \nabla U\rangle }$

となる。

この節の加筆が望まれています。

• 古典物理学