Chapter 18 Key Takeaways: Degenerate Perturbation Theory and Fine Structure


Core Message

When energy levels of $\hat{H}_0$ are degenerate, standard perturbation theory breaks down because the "correct" zeroth-order states are not uniquely determined. The cure is to diagonalize the perturbation $\hat{H}'$ within each degenerate subspace, finding the basis in which perturbation theory can proceed. Applied to hydrogen, this technique reveals the fine structure — three corrections of order $\alpha^2 E_n$ that depend on $n$ and $j$ alone — the hyperfine structure and its astrophysically important 21 cm line, and the rich phenomenology of the Zeeman effect.


Key Concepts

1. Degenerate Perturbation Theory

Non-degenerate perturbation theory fails when $E_n^{(0)} = E_m^{(0)}$ for $n \neq m$ because the second-order energy correction contains terms with vanishing denominators. The resolution: within the $g$-fold degenerate subspace, construct the $g \times g$ matrix $W_{ij} = \langle i | \hat{H}' | j \rangle$ and diagonalize it. The eigenvalues are the first-order energy corrections, and the eigenvectors are the "correct" zeroth-order states from which standard perturbation theory may proceed.

2. Good Quantum Numbers

A quantum number is "good" if it labels eigenstates of operators that commute with both $\hat{H}_0$ and $\hat{H}'$. Finding good quantum numbers is equivalent to finding the correct basis without explicit diagonalization. For hydrogen fine structure: $n$, $l$, $j$, $m_j$ are good. For the strong-field Zeeman effect: $n$, $l$, $m_l$, $m_s$ are good.

3. Fine Structure of Hydrogen

Three corrections of order $\alpha^2 E_n$ collectively produce the fine structure:

  • Relativistic kinetic energy: $\hat{H}'_{\text{rel}} = -\hat{p}^4/(8m_e^3 c^2)$ — lowers all levels, largest for low $l$
  • Spin-orbit coupling: $\hat{H}'_{\text{SO}} \propto \hat{\mathbf{L}} \cdot \hat{\mathbf{S}} / r^3$ — splits levels by $j$ value, zero for $l = 0$
  • Darwin term: $\hat{H}'_{\text{Darwin}} \propto \delta^3(\mathbf{r})$ — affects only $l = 0$ states

4. The Fine Structure Constant

$\alpha = e^2/(4\pi\epsilon_0\hbar c) \approx 1/137.036$ — the dimensionless coupling constant of electromagnetism. It sets the scale of fine structure corrections relative to the gross structure: $E_{\text{FS}} \sim \alpha^2 E_n$.

5. Hyperfine Structure

The magnetic interaction between electron and nuclear spins splits energy levels by $\sim (m_e/m_p)\alpha^2 E_n$ — roughly 1000 times smaller than fine structure. The ground-state hydrogen hyperfine splitting produces the 21 cm line ($\nu = 1420$ MHz).

6. The Zeeman Effect

An external magnetic field $B$ adds $\hat{H}'_Z = \mu_B(\hat{L}_z + 2\hat{S}_z)B/\hbar$. The behavior depends on the ratio of $\mu_B B$ to the fine structure splitting: - Weak field: $j$, $m_j$ good; splitting $= g_j m_j \mu_B B$ - Strong field (Paschen-Back): $m_l$, $m_s$ good; splitting $= (m_l + 2m_s)\mu_B B$ - Intermediate field: must diagonalize full Hamiltonian


Key Equations

Equation Name When to Use
$W_{ij} = \langle i \| \hat{H}' \| j \rangle$ Perturbation matrix Degenerate perturbation: construct in degenerate subspace
$\det(\mathbf{W} - E^{(1)}\mathbf{I}) = 0$ Secular equation Find first-order energies by diagonalizing $\mathbf{W}$
$E_{\text{FS}}^{(1)} = -\frac{(E_n^{(0)})^2}{2m_ec^2}\left(\frac{4n}{j+1/2} - 3\right)$ Fine structure formula Total fine structure correction for hydrogen
$\langle \hat{\mathbf{L}}\cdot\hat{\mathbf{S}} \rangle = \frac{\hbar^2}{2}[j(j+1) - l(l+1) - s(s+1)]$ $\hat{\mathbf{L}}\cdot\hat{\mathbf{S}}$ expectation Spin-orbit coupling in coupled basis
$g_j = 1 + \frac{j(j+1) - l(l+1) + s(s+1)}{2j(j+1)}$ Lande $g$-factor Weak-field Zeeman splitting
$E_Z^{(1)} = g_j m_j \mu_B B$ Weak-field Zeeman energy When $\mu_B B \ll \Delta E_{\text{FS}}$
$\Delta E_{\text{HF}} = \frac{4}{3}g_p\alpha^2(m_e/m_p)\|E_1^{(0)}\|$ Ground-state hyperfine 21 cm line frequency

Key Expectation Values in Hydrogen

Quantity Value Used For
$\langle 1/r \rangle_{nl}$ $1/(n^2 a_0)$ Relativistic correction
$\langle 1/r^2 \rangle_{nl}$ $1/[n^3(l+1/2)a_0^2]$ Relativistic correction
$\langle 1/r^3 \rangle_{nl}$ $1/[n^3 l(l+1/2)(l+1)a_0^3]$ ($l \geq 1$) Spin-orbit coupling
$\|\psi_{n00}(0)\|^2$ $1/(\pi n^3 a_0^3)$ Darwin term, hyperfine

Decision Framework: Which Perturbation Theory to Use?

Is the unperturbed level degenerate?
├── NO → Use non-degenerate perturbation theory (Chapter 17)
└── YES → Is the perturbation diagonal in your chosen basis?
    ├── YES → Each diagonal element is E^(1); proceed as non-degenerate
    └── NO → Diagonalize H' within the degenerate subspace
        ├── All eigenvalues distinct → Degeneracy fully lifted; proceed
        └── Some eigenvalues equal → Residual degeneracy; go to 2nd order
            within remaining degenerate subspace

Common Misconceptions

Misconception Correction
"Degenerate perturbation theory is a fundamentally different formalism" It is the same perturbation theory, preceded by a basis change within the degenerate subspace.
"The Darwin term is a quantum gravity effect" It is a purely relativistic quantum effect (Zitterbewegung), unrelated to gravity.
"Spin-orbit coupling is always the dominant fine structure correction" For $s$-states ($l = 0$), spin-orbit vanishes and the relativistic + Darwin corrections dominate.
"The anomalous Zeeman effect violates theory" It is perfectly explained by spin and the Lande $g$-factor; "anomalous" is a historical misnomer.
"The 21 cm line is unobservable because the transition is too slow" The enormous number of hydrogen atoms in the universe compensates for the slow rate.
"Fine structure depends on $l$" The three corrections individually depend on $l$, but they combine to give a result depending only on $n$ and $j$.

Numerical Reference

Constant Symbol Value
Fine structure constant $\alpha$ $1/137.036$
Bohr magneton $\mu_B$ $5.788 \times 10^{-5}$ eV/T
Nuclear magneton $\mu_N$ $3.152 \times 10^{-8}$ eV/T
Proton $g$-factor $g_p$ $5.586$
Electron rest energy $m_ec^2$ $0.511$ MeV
Proton/electron mass ratio $m_p/m_e$ $1836.15$
Bohr radius $a_0$ $5.292 \times 10^{-11}$ m
21 cm frequency $\nu_{\text{HF}}$ $1420.405$ MHz

Energy Scale Hierarchy for Hydrogen

Level Scale Example ($n = 2$)
Gross structure (Bohr) $\alpha^2 m_e c^2 \sim 10$ eV $-3.4$ eV
Fine structure $\alpha^4 m_e c^2 \sim 10^{-4}$ eV $4.5 \times 10^{-5}$ eV
Lamb shift (QED) $\alpha^5 m_e c^2 \sim 10^{-6}$ eV $4.4 \times 10^{-6}$ eV
Hyperfine ($n = 1$) $\alpha^4 (m_e/m_p) m_e c^2 \sim 10^{-6}$ eV $5.9 \times 10^{-6}$ eV

Looking Ahead

  • Chapter 19 (Variational Principle): A complementary approximation method that does not require a solvable $\hat{H}_0$.
  • Chapter 21 (Time-Dependent Perturbation Theory): Transition rates between fine structure levels; selection rules for spectral lines.
  • Chapter 22 (Scattering): Cross-sections for scattering off hydrogen-like potentials.
  • Chapter 29 (Relativistic QM): The Dirac equation derives all three fine structure corrections at once.
  • Chapter 38 (Capstone): Full hydrogen simulation including fine structure, hyperfine, and Zeeman effects.