Problem 10.1
习题 10.1
来源: 第10章, PDF第181,182页
10.1 We saw that the Dirac equation predicted that there is an interaction between the electron spin and the magnetic field, , with strength . When the electron has angular momentum , such as in an atomic orbital, there is also a interaction and a spin-orbit coupling . The Dirac equation (along with symmetry arguments) predicts the strength of all three interactions, as well as other corrections. To see the effect of these terms on the hydrogen atom, we have to take the non-relativistic limit.
(a) For the Schrödinger equation, we need the Hamiltonian, not the Lagrangian. Find the Dirac Hamiltonian by writing the Dirac equation as . Write the Hamiltonian in terms of momenta rather than derivatives .
(b) Calculate in the Weyl representation for . Leave in terms of , and . Put back in the factors of and , keeping the charge dimensionless.
(c) Now take the square root of this result and expand in , subtracting off the zero-point energy , i.e. compute to order . Looking at the term, how big are the electron's electric and magnetic dipole moments?
(d) The size of the terms in this Hamiltonian are only meaningful because the spin and angular momentum operators have the same normalization. Check the normalization of the angular momentum operators and the spin operators by showing that they both satisfy the rotation algebra: .
(e) The gyromagnetic ratio, (sometimes called the -factor), is the relative size of the and interactions. Choose a constant magnetic field in the direction, then isolate the coupling in . Extract the electron gyromagnetic ratio by writing the entire coupling to the magnetic field in the Hamiltonian as , with . How could you experimentally measure (e.g. with spectroscopy of the hydrogen atom)?
(f) In spherical coordinates, the Schrödinger equation has an term. With spin, you might expect that this becomes , making the term proportional to the -factor. This is wrong. It misses an important relativistic effect, Thomas precession. It is very hard to calculate directly, but easy to calculate using symmetries. With no magnetic field, the atom, with spin included, is
still rotationally invariant. Which of or is conserved (i.e. commutes with )? Using this result, how does the spin-orbit coupling depend on ?
(g) There are additional relativistic effects coming from the Dirac equation. Expand the Dirac equation to next order in , producing a term that scales as .
(h) Now let us do some dimensional analysis - there is only one scale . Show that the electron's Compton wavelength, the classical electron radius, , the Bohr radius, , and the inverse-Rydberg constant, , are all times powers of . Are the splittings due to the term fine structure (), hyperfine structure () or something else? [Hint: write out a formula for the energy shift using time-independent perturbation theory, then see which of the above length scales appears.]
习题 10.1 - 解答
(a) 狄拉克方程在电磁场中的协变形式为 ,其中协变导数 。将其展开并分离出时间导数项: 左乘 并利用 ,得到: 将时间偏导数移到等式左边,并代入动量算符 (即 ): 定义狄拉克矩阵 和 ,即可得到狄拉克哈密顿量:
(b) 在 Weyl 表象中,狄拉克矩阵的形式为 ,。 因此 ,。 令 ,则 的矩阵形式为: 计算其平方: 利用泡利矩阵的恒等式 ,展开 : 由于 , 代入上式得到交叉项: 恢复光速 (动量项乘 ,质量项乘 ),得到:
(c) 对上式开平方并提取 因子: 在非相对论极限下( 展开),利用 : 减去静止能量 ,得到 阶的哈密顿量: 哈密顿量中与电磁场耦合的偶极矩项形式为 。 观察 项可知,电子的电偶极矩为 (无 项),磁偶极矩大小为 。
(d) 为了验证代数关系,设 。对于轨道角动量 : 利用 和 : 代回对易子并利用列维-奇维塔符号的缩并公式 : 而 。 故 。
对于自旋算符 : 故 。两者均满足标准旋转代数。
(e) 取恒定磁场 ,选择对称规范 。 展开动能项中的交叉项(库仑规范下 ,故 ): 自旋磁耦合项为 。 总磁耦合哈密顿量为: 定义玻尔磁子 (取电荷大小),对比形式 ,可提取出: 实验测量方法:可以通过氢原子的反常塞曼效应(Anomalous Zeeman effect)光谱学来测量。将氢原子置于外加磁场中,测量能级分裂的谱线间距,由于分裂正比于 ,从中即可拟合提取出 的值。
(f) 在没有外加磁场的情况下,孤立原子的空间旋转对称性未被破坏。 由于 是空间旋转的生成元,因此总角动量 是守恒的(与 对易),而 不守恒。 因为 守恒,自旋-轨道耦合项必须是旋转标量,即必须正比于 。由于磁矩相互作用给出的系数正比于 ,而 Thomas 进动(纯相对论运动学效应)会产生一个方向相反、大小相当于 的修正项,因此自旋-轨道耦合的强度依赖于 。
(g) 为了得到 项,我们在无电磁场()的情况下将狄拉克哈密顿量展开到 的下一阶。 利用泰勒展开 : 因此,缩放为 的相对论修正项为:
(h) 在自然单位制(或仅关注量纲与 的幂次)下,唯一的基本质量尺度是 。
- 康普顿波长 ,故 。
- 经典电子半径 ,故 。
- 玻尔半径 ,故 。
- 逆里德伯常数 ,故 。
能级移动分析: 利用不含时微扰论,能量移动为 。 在氢原子中,动量的特征尺度由玻尔半径决定:。 代入微扰项: 而未微扰的基态能量(里德伯能量)尺度为 。 因此,能量移动与原能级的关系为 。 这表明 项引起的能级分裂属于 。