Skip to content

Spin-Orbit Coupling and Fine Structure

CSIR NET Physical Sciences Part C: spin-orbit coupling, Thomas factor, relativistic and Darwin terms, hydrogen fine structure, alkali doublets, Lamb shift.

  • 20 min read
  • Updated
On this page · 13 sections
  1. Prerequisites
  2. The Hook: The Yellow Line That Is Two Lines
  3. Where Spin-Orbit Coupling Comes From
  4. Hydrogen Fine Structure: Three Terms
  5. Spin-Orbit Coupling in Many-Electron Atoms
  6. Fine Structure in a Magnetic Field
  7. Memory Tricks
  8. Don’t Confuse
  9. Common Mistakes
  10. Practice
  11. Quick Revision Table
  12. Key Takeaways
  13. What’s Next

Prerequisites

The Hook: The Yellow Line That Is Two Lines

Where Spin-Orbit Coupling Comes From

The magnetic field in the electron’s frame

In the rest frame of the nucleus the electron moves with velocity \( \mathbf v \) through an electric field \( \mathbf E \). In the electron’s own frame, to first order in \( v/c \), there is a magnetic field

$$\mathbf B = -\frac{1}{c^2}\,\mathbf v\times\mathbf E$$

For a central potential energy \( V(r) \) of the electron, the field is \( \mathbf E = \frac{1}{e}\frac{dV}{dr}\hat{\mathbf r} \) (the electron charge is \( -e \)). Using \( \mathbf r\times\mathbf v = \mathbf L/m \):

$$\mathbf B = \frac{1}{emc^2}\frac{1}{r}\frac{dV}{dr}\,\mathbf L$$

The electron carries a spin magnetic moment \( \boldsymbol\mu = -\frac{g e}{2m}\mathbf S \approx -\frac{e}{m}\mathbf S \) (\( g \approx 2 \)). Its energy \( -\boldsymbol\mu\cdot\mathbf B \) is then \( \frac{1}{m^2c^2}\frac{1}{r}\frac{dV}{dr}\mathbf L\cdot\mathbf S \).

The Thomas factor of one half

The electron’s rest frame is not inertial: it accelerates around the nucleus. Transforming correctly between successive rest frames adds a kinematic rotation of the spin, Thomas precession, which cancels exactly half of the naive result. The correct Hamiltonian is

$$\boxed{\hat H_{SO} = \frac{1}{2m^2c^2}\,\frac{1}{r}\frac{dV}{dr}\,\hat{\mathbf L}\cdot\hat{\mathbf S}}$$

For hydrogen-like ions, \( V = -\frac{Ze^2}{4\pi\epsilon_0 r} \), so

$$\hat H_{SO} = \frac{Ze^2}{8\pi\epsilon_0 m^2c^2}\,\frac{1}{r^3}\,\hat{\mathbf L}\cdot\hat{\mathbf S}$$

The Dirac equation produces this term, factor of \( \frac{1}{2} \) included, with no extra argument. That is one of its early successes, discussed in Klein-Gordon and Dirac Equations.

Why the good quantum numbers change

\( \hat{\mathbf L}\cdot\hat{\mathbf S} \) does not commute with \( \hat L_z \) or \( \hat S_z \) separately, but it commutes with \( \hat L^2 \), \( \hat S^2 \), \( \hat J^2 \) and \( \hat J_z \). So once spin-orbit coupling is included, \( m_l \) and \( m_s \) stop being good quantum numbers and the states are labelled \( \lvert n, l, s, j, m_j\rangle \). In that coupled basis the operator is diagonal:

$$\langle\hat{\mathbf L}\cdot\hat{\mathbf S}\rangle = \frac{\hbar^2}{2}\left[j(j+1) - l(l+1) - \tfrac{3}{4}\right] = \begin{cases} \dfrac{l\hbar^2}{2} & j = l + \tfrac{1}{2} \\[2mm] -\dfrac{(l+1)\hbar^2}{2} & j = l - \tfrac{1}{2} \end{cases}$$

For \( l = 0 \) only \( j = \frac{1}{2} \) exists and \( \langle\mathbf L\cdot\mathbf S\rangle = 0 \).

The weighted sum \( (2j_>+1)\frac{l}{2} - (2j_<+1)\frac{l+1}{2} = (2l+2)\frac{l}{2} - 2l\,\frac{l+1}{2} = 0 \). Spin-orbit coupling splits a level without moving its centre of gravity. This trace check is worth doing in every Part C question.

Hydrogen Fine Structure: Three Terms

The fine-structure Hamiltonian for a hydrogen-like atom has three pieces, all of order \( \alpha^2 \) relative to the Bohr energy \( E_n = -\frac{13.6\ \text{eV}\,Z^2}{n^2} \).

TermOperatorOriginActs on
Relativistic kinetic\( -\dfrac{\hat p^4}{8m^3c^2} \)expansion of \( \sqrt{p^2c^2 + m^2c^4} \)all \( l \)
Spin-orbit\( \dfrac{1}{2m^2c^2}\dfrac{1}{r}\dfrac{dV}{dr}\hat{\mathbf L}\cdot\hat{\mathbf S} \)spin moment in the orbital field, with Thomas factor\( l \geq 1 \)
Darwin\( \dfrac{\pi\hbar^2}{2m^2c^2}\dfrac{Ze^2}{4\pi\epsilon_0}\delta^3(\mathbf r) \)smearing of the electron over a Compton wavelength\( l = 0 \) only

The relativistic kinetic correction

\( \sqrt{p^2c^2 + m^2c^4} = mc^2 + \frac{p^2}{2m} - \frac{p^4}{8m^3c^2} + \cdots \). Writing \( p^2 = 2m(E_n - V) \) and using the hydrogen expectation values of \( 1/r \) and \( 1/r^2 \):

$$E_{\text{rel}}^{(1)} = -\frac{E_n^2}{2mc^2}\left[\frac{4n}{l + \frac{1}{2}} - 3\right]$$

It is always negative and largest for small \( l \), since low-\( l \) orbits come close to the nucleus where the electron is fastest.

The spin-orbit shift

Using \( \left\langle\frac{1}{r^3}\right\rangle = \frac{Z^3}{n^3a_0^3\,l\left(l + \frac{1}{2}\right)(l + 1)} \) for \( l \geq 1 \):

$$E_{SO}^{(1)} = \frac{E_n^2}{mc^2}\,\frac{n\left[j(j+1) - l(l+1) - \frac{3}{4}\right]}{l\left(l + \frac{1}{2}\right)(l + 1)}$$

The Darwin term

The Darwin term is a contact interaction, so only \( s \) states, which have \( \lvert\psi(0)\rvert^2 = \frac{Z^3}{\pi n^3a_0^3} \neq 0 \), feel it:

$$E_D^{(1)} = \frac{E_n^2}{mc^2}\,2n \qquad (l = 0)$$

The combined result

Adding the three, for both \( j = l \pm \frac{1}{2} \) and including \( l = 0 \), the \( l \) dependence cancels:

$$\boxed{E_{fs}^{(1)} = \frac{E_n^2}{2mc^2}\left[3 - \frac{4n}{j + \frac{1}{2}}\right]}$$

or, equivalently,

$$\boxed{E_{nj} = -\frac{13.6\ \text{eV}\,Z^2}{n^2}\left[1 + \frac{(Z\alpha)^2}{n^2}\left(\frac{n}{j + \frac{1}{2}} - \frac{3}{4}\right)\right]}$$

Read off the consequences:

  1. The energy depends on \( n \) and \( j \) only. \( 2s_{1/2} \) and \( 2p_{1/2} \) stay degenerate; so do \( 3p_{3/2} \) and \( 3d_{3/2} \).
  2. Every level is pushed down, since \( j + \frac{1}{2} \leq n \) makes \( \frac{4n}{j + 1/2} \geq 4 > 3 \), so the bracket is negative. The shift is largest in magnitude for the smallest \( j \).
  3. Scaling: \( E_{fs} \propto Z^4/n^3 \) at fixed \( j \). The fine-structure splitting in He\( ^+ \) is \( 2^4 = 16 \) times that in hydrogen for the same \( n \).
  4. Size: \( \frac{E_n^2}{2mc^2} = \frac{\alpha^2\lvert E_1\rvert}{4n^4} \) for hydrogen. With \( \alpha^2 \approx 5.3\times10^{-5} \), fine-structure shifts are of order \( 10^{-5} \)–\( 10^{-4} \) eV for \( n = 1, 2 \).

Worked numbers for hydrogen

LevelFine-structure shift (eV)
\( 1s_{1/2} \)\( -1.81\times10^{-4} \)
\( 2s_{1/2},\ 2p_{1/2} \)\( -5.66\times10^{-5} \)
\( 2p_{3/2} \)\( -1.13\times10^{-5} \)
\( 3s_{1/2},\ 3p_{1/2} \)\( -2.01\times10^{-5} \)
\( 3p_{3/2},\ 3d_{3/2} \)\( -6.71\times10^{-6} \)
\( 3d_{5/2} \)\( -2.24\times10^{-6} \)

The \( n = 2 \) splitting, \( 2p_{3/2} - 2p_{1/2} \), is \( \frac{2E_2^2}{mc^2} = 4.53\times10^{-5} \) eV, which is about 10.9 GHz or 0.365 cm\( ^{-1} \).

For the \( 2p \) states the pieces are worth knowing separately: the relativistic term is \( -2.64\times10^{-5} \) eV for both, and spin-orbit adds \( +1.51\times10^{-5} \) eV (\( j = \frac{3}{2} \)) or \( -3.02\times10^{-5} \) eV (\( j = \frac{1}{2} \)). For \( 2s_{1/2} \) the relativistic term is \( -1.47\times10^{-4} \) eV and the Darwin term \( +9.06\times10^{-5} \) eV. Both routes land on the same \( -5.66\times10^{-5} \) eV for \( j = \frac{1}{2} \).

graph TD
    A["Bohr level n = 2 (8 states with spin)"] --> B["2p 3/2 : 4 states, shift -1.13e-5 eV"]
    A --> C["2s 1/2 and 2p 1/2 : 4 states, shift -5.66e-5 eV"]
    C --> D["Lamb shift (QED): 2s 1/2 lies above 2p 1/2"]
    B --> E["Hyperfine: coupling to proton spin, much smaller"]
    D --> E

Beyond fine structure

  • The Lamb shift. The Dirac equation keeps \( 2s_{1/2} \) and \( 2p_{1/2} \) exactly degenerate. Lamb and Retherford measured in 1947 that \( 2s_{1/2} \) lies higher by about 1058 MHz (about \( 4.4\ \mu \)eV), roughly a tenth of the fine-structure splitting. It is explained by quantum electrodynamics: the coupling of the electron to the vacuum fluctuations of the quantised field.
  • Hyperfine structure. The electron’s magnetic moment also couples to the proton’s. The effect is smaller than fine structure by roughly \( m_e/m_p \). In the hydrogen ground state it splits \( F = 1 \) from \( F = 0 \); the transition between them is the 21 cm (about 1420 MHz) line of radio astronomy.

The hierarchy, Bohr \( \gg \) fine \( \gg \) Lamb \( \gtrsim \) hyperfine, is a favourite statement-checking question.

Spin-Orbit Coupling in Many-Electron Atoms

Alkali doublets

In an alkali atom the single valence electron moves in a screened potential that is not \( 1/r \). Penetration of the core lifts the \( l \) degeneracy (the \( 3s \) and \( 3p \) levels of sodium differ by about 2.1 eV), and spin-orbit coupling then splits every \( l \geq 1 \) level into \( j = l \pm \frac{1}{2} \). Write the coupling as

$$\hat H_{SO} = \frac{A_{nl}}{\hbar^2}\hat{\mathbf L}\cdot\hat{\mathbf S}$$

with \( A_{nl} > 0 \) for a single electron. For \( l = 1 \): the \( j = \frac{3}{2} \) level is at \( +\frac{A}{2} \), the \( j = \frac{1}{2} \) level at \( -A \), and the splitting is \( \frac{3}{2}A \). In general the doublet splitting is \( A\left(l + \frac{1}{2}\right) \).

For sodium, the \( D_2 \) line (\( 3p_{3/2} \to 3s_{1/2} \)) is at 589.0 nm and the \( D_1 \) line (\( 3p_{1/2} \to 3s_{1/2} \)) at 589.6 nm. The splitting is

$$\Delta E = hc\left(\frac{1}{589.0\ \text{nm}} - \frac{1}{589.6\ \text{nm}}\right) \approx 2.1\ \text{meV} \approx 17\ \text{cm}^{-1}$$

about 50 times larger than hydrogen’s \( n = 2 \) fine structure, because the \( 3p \) electron penetrates close to an unscreened nucleus where \( dV/dr \) is large. The same mechanism makes the doublet grow quickly down the alkali group.

LS coupling and the Landé interval rule

In light atoms the electrostatic interaction between electrons dominates, so spins couple to a total \( S \) and orbital momenta to a total \( L \) first, giving a term \( ^{2S+1}L \). Spin-orbit coupling, \( A\,\mathbf L\cdot\mathbf S \) with the totals, then splits each term into levels \( ^{2S+1}L_J \), with \( J = \lvert L - S\rvert, \ldots, L + S \). Since

$$E(J) - E(J-1) = \frac{A}{2}\left[J(J+1) - (J-1)J\right] = AJ$$

the gap between adjacent levels is proportional to the larger \( J \). This is the Landé interval rule. For a \( ^3P \) term the \( J = 0, 1, 2 \) intervals are in ratio \( 1:2 \); for \( ^3D \) (\( J = 1, 2, 3 \)) they are \( 2:3 \).

The sign of \( A \) is set by Hund’s third rule: for a sub-shell less than half full the lowest \( J \) lies lowest (a normal multiplet, \( A > 0 \)); more than half full, the highest \( J \) lies lowest (inverted, \( A < 0 \)). In very heavy atoms spin-orbit coupling becomes comparable to the electrostatic terms and \( jj \) coupling takes over; the full treatment belongs to atomic physics.

Fine Structure in a Magnetic Field

Once the levels carry \( j \), a weak magnetic field \( B\hat z \) (Zeeman energy much smaller than the fine-structure splitting) shifts each \( m_j \) state by

$$\Delta E = g_J\,\mu_B B\,m_j, \qquad \boxed{g_J = 1 + \frac{j(j+1) + s(s+1) - l(l+1)}{2j(j+1)}}$$

with \( \mu_B = e\hbar/2m \) and the electron spin \( g \) taken as 2. The projection theorem is what makes this work: within a level of fixed \( j \), the magnetic moment \( -\frac{\mu_B}{\hbar}(\mathbf L + 2\mathbf S) \) acts like its projection on \( \mathbf J \).

Level\( g_J \)Number of Zeeman components
\( ^2S_{1/2} \)22
\( ^2P_{1/2} \)2/32
\( ^2P_{3/2} \)4/34
\( ^2D_{3/2} \)4/54
\( ^2D_{5/2} \)6/56

Because \( g_J \) differs between the upper and lower levels, the sodium D lines split into more than three components: the anomalous Zeeman effect. In a strong field, where \( \mu_BB \) exceeds the fine-structure splitting, spin and orbit decouple, \( m_l \) and \( m_s \) become good again and \( \Delta E = \mu_BB(m_l + 2m_s) \): the Paschen–Back regime. Which regime applies is decided by comparing \( \mu_BB \) with the splitting, so for hydrogen \( n = 2 \) (splitting \( 4.5\times10^{-5} \) eV, while \( \mu_B \approx 5.79\times10^{-5} \) eV/T) the crossover is already near 1 T.

Memory Tricks

Don’t Confuse

ThisvsThat
Fine structure: depends on \( n, j \); order \( \alpha^2 \) of BohrHyperfine: involves nuclear spin, \( F \); smaller by roughly \( m_e/m_p \)
Dirac/fine structure: \( 2s_{1/2} \) and \( 2p_{1/2} \) degenerateLamb shift (QED): \( 2s_{1/2} \) slightly above \( 2p_{1/2} \)
Hydrogen: \( 2s \)–\( 2p \) differ only by tiny QED effectsSodium: \( 3s \)–\( 3p \) differ by about 2 eV (core penetration), not spin-orbit
Term \( ^{2S+1}L \) (electrostatic)Level \( ^{2S+1}L_J \) (after spin-orbit); state adds \( M_J \)
Weak field: \( g_J\mu_BBm_j \), anomalous ZeemanStrong field: \( \mu_BB(m_l + 2m_s) \), Paschen–Back
Normal multiplet (less than half full): lowest \( J \) lowestInverted (more than half full): highest \( J \) lowest

Common Mistakes

Practice

All questions below are original practice questions written for this page. They are not taken from any CSIR NET paper. Use \( E_1 = -13.6 \) eV, \( mc^2 = 0.511 \) MeV and \( hc = 1240 \) eV nm where needed.

Level 1: Foundation

  1. Which hydrogen \( n = 2 \) levels remain degenerate when fine structure (but not the Lamb shift) is included?
  2. What is \( \langle\mathbf L\cdot\mathbf S\rangle \) for a \( d \) electron in a \( j = \frac{5}{2} \) level?
  3. Between which two levels of a \( ^3P \) term is the interval largest, according to the Landé rule?
Answers
  1. \( 2s_{1/2} \) and \( 2p_{1/2} \). The fine-structure energy depends only on \( n \) and \( j \).
  2. \( \hbar^2 \). \( \frac{\hbar^2}{2}\left[\frac{35}{4} - 6 - \frac{3}{4}\right] = \frac{\hbar^2}{2}\times2 \). Equivalently \( \frac{l}{2}\hbar^2 \) with \( l = 2 \).
  3. Between \( J = 2 \) and \( J = 1 \). Intervals are \( AJ \) with \( J \) the upper value: \( A \) for \( 1 \)–\( 0 \), \( 2A \) for \( 2 \)–\( 1 \).

Level 2: Part B Style

Practice Q1. The splitting between the \( 2p_{3/2} \) and \( 2p_{1/2} \) levels of hydrogen is closest to:

(a) \( 4.5\times10^{-5} \) eV (b) \( 4.5\times10^{-3} \) eV (c) \( 1.8\times10^{-4} \) eV (d) \( 4.4\times10^{-6} \) eV

Answer: (a)

\( \Delta E = \frac{E_2^2}{2mc^2}\left[\left(3 - \frac{8}{2}\right) - \left(3 - \frac{8}{1}\right)\right] = \frac{4E_2^2}{2mc^2} = \frac{2(3.4)^2}{0.511\times10^6}\ \text{eV} = 4.5\times10^{-5}\ \text{eV} \).

Option (c) is the \( 1s \) shift, and (d) is the size of the Lamb shift.

Practice Q2. The fine-structure splitting between \( 2p_{3/2} \) and \( 2p_{1/2} \) in He\( ^+ \), relative to the same splitting in hydrogen, is:

(a) 2 (b) 4 (c) 8 (d) 16

Answer: (d) 16

The shift is \( \propto E_n^2/mc^2 \) and \( E_n \propto Z^2 \), so the splitting scales as \( Z^4 = 16 \). Option (b) is the Bohr-energy scaling \( Z^2 \).

Practice Q3. In the weak-field Zeeman effect, the Landé factor of the \( ^2P_{3/2} \) level is:

(a) 2/3 (b) 1 (c) 4/3 (d) 2

Answer: (c) 4/3

\( g_J = 1 + \frac{\frac{15}{4} + \frac{3}{4} - 2}{2\cdot\frac{15}{4}} = 1 + \frac{5/2}{15/2} = \frac{4}{3} \). Option (a) is the value for \( ^2P_{1/2} \).

Practice Q4. A \( ^3D \) term shows three levels. With a spin-orbit constant \( A > 0 \), the energies of \( J = 1, 2, 3 \) relative to the term centre (before spin-orbit) are:

(a) \( -3A, -A, +2A \) (b) \( -A, 0, +A \) (c) \( +2A, -A, -3A \) (d) \( -2A, -A, +3A \)

Answer: (a)

\( \langle\mathbf L\cdot\mathbf S\rangle/\hbar^2 = \frac{1}{2}[J(J+1) - 6 - 2] \): \( J = 1 \to -3 \), \( J = 2 \to -1 \), \( J = 3 \to +2 \). Check the trace: \( 3(-3) + 5(-1) + 7(2) = 0 \). Intervals \( 2A \) and \( 3A \), ratio \( 2:3 \), as the interval rule requires. Option (c) lists the energies in the order of an inverted multiplet (\( J = 3 \) lowest), which would need \( A < 0 \); its magnitudes are not those of any \( ^3D \) term.

Level 3: Part C Style

Practice Q5. For the \( n = 3 \) level of hydrogen, including fine structure but not the Lamb shift or hyperfine effects, consider:

  1. There are exactly three distinct energies.
  2. The \( 3d_{3/2} \) and \( 3p_{3/2} \) levels have the same energy.
  3. The highest of the three lies about \( 1.8\times10^{-5} \) eV above the lowest.
  4. All three lie below the Bohr energy \( -13.6/9 \) eV.

Which are correct?

(a) 1 and 2 only (b) 1, 2 and 4 only (c) 1, 2, 3 and 4 (d) 2 and 3 only

Answer: (c) 1, 2, 3 and 4

Energies depend on \( j \) only: \( j = \frac{1}{2}, \frac{3}{2}, \frac{5}{2} \), so three levels (1 true), and \( 3p_{3/2} \), \( 3d_{3/2} \) share \( j = \frac{3}{2} \) (2 true).

With \( \frac{E_3^2}{2mc^2} = \frac{(1.512)^2}{1.022\times10^6} = 2.24\times10^{-6} \) eV, the brackets \( 3 - \frac{12}{j + 1/2} \) are \( -9, -3, -1 \), giving shifts \( -2.01\times10^{-5} \), \( -6.7\times10^{-6} \) and \( -2.2\times10^{-6} \) eV. The spread is \( 8\times2.24\times10^{-6} = 1.8\times10^{-5} \) eV (3 true), and every shift is negative (4 true).

Practice Q6. The sodium D lines are measured at 589.0 nm and 589.6 nm. Model the \( 3p \) level as split by \( \frac{A}{\hbar^2}\mathbf L\cdot\mathbf S \). Consider:

  1. The longer-wavelength line comes from \( 3p_{1/2} \).
  2. The splitting of the \( 3p \) level is about 2.1 meV.
  3. The spin-orbit constant is \( A \approx 1.4 \) meV.
  4. Relative to the unsplit \( 3p \) energy, \( 3p_{3/2} \) moves up by about 0.7 meV.

Which are correct?

(a) 1 and 2 only (b) 2 and 3 only (c) 1, 2 and 3 only (d) 1, 2, 3 and 4

Answer: (d) 1, 2, 3 and 4

With \( A > 0 \), \( j = \frac{1}{2} \) lies lower, so its transition to \( 3s \) has less energy and longer wavelength: 589.6 nm is \( D_1 \) from \( 3p_{1/2} \) (1 true).

\( \Delta E = 1240\left(\frac{1}{589.0} - \frac{1}{589.6}\right) \) eV \( = 2.14\times10^{-3} \) eV (2 true; the precise wavelengths 588.995 and 589.592 nm give 2.13 meV).

For \( l = 1 \) the splitting is \( \frac{3}{2}A \), so \( A = 1.43 \) meV (3 true). The \( j = \frac{3}{2} \) level moves by \( +\frac{A}{2} = 0.71 \) meV and \( j = \frac{1}{2} \) by \( -A \) (4 true). Check: \( 4(0.71) + 2(-1.43) \approx 0 \).

Practice Q7. An electron in a hydrogen-like ion of charge \( Z \) is in a \( 2p \) state. Which statements about the fine-structure Hamiltonian are correct?

  1. Without the Thomas factor, the predicted \( 2p_{3/2} - 2p_{1/2} \) splitting would double.
  2. The Darwin term contributes to \( 2p_{1/2} \).
  3. The relativistic kinetic correction shifts \( 2p_{1/2} \) and \( 2p_{3/2} \) equally.
  4. In the \( \lvert m_l, m_s\rangle \) basis, the spin-orbit operator is diagonal.

(a) 1 and 3 only (b) 1, 2 and 3 only (c) 3 and 4 only (d) 1 and 4 only

Answer: (a) 1 and 3 only
  1. The \( 2p \) splitting comes entirely from spin-orbit coupling (the other two terms do not depend on \( j \) at fixed \( l \)), and the Thomas factor halves it. True.
  2. The Darwin term needs \( \psi(0) \neq 0 \), so it acts only on \( s \) states. False.
  3. \( -\frac{E_n^2}{2mc^2}\left[\frac{4n}{l + 1/2} - 3\right] \) depends on \( l \), not \( j \). True.
  4. \( \mathbf L\cdot\mathbf S = L_zS_z + \frac{1}{2}(L_+S_- + L_-S_+) \) mixes \( \lvert m_l, m_s\rangle \) states with the same \( m_l + m_s \). False.

Quick Revision Table

IdeaFormula or fact
Spin-orbit Hamiltonian\( \frac{1}{2m^2c^2}\frac{1}{r}\frac{dV}{dr}\mathbf L\cdot\mathbf S \) (Thomas factor included)
\( \langle\mathbf L\cdot\mathbf S\rangle \)\( \frac{l}{2}\hbar^2 \) for \( j = l + \frac{1}{2} \); \( -\frac{l+1}{2}\hbar^2 \) for \( j = l - \frac{1}{2} \)
Good quantum numbers\( n, l, s, j, m_j \) (not \( m_l, m_s \))
Relativistic term\( -\frac{E_n^2}{2mc^2}\left[\frac{4n}{l+1/2} - 3\right] \)
Darwin term\( +\frac{E_n^2}{mc^2}2n \), \( s \) states only
Total fine structure\( \frac{E_n^2}{2mc^2}\left[3 - \frac{4n}{j+1/2}\right] \); depends on \( n, j \)
Scaling\( \propto Z^4\alpha^2/n^3 \)
Hydrogen \( 2p \) splitting\( 4.5\times10^{-5} \) eV, about 10.9 GHz
Lamb shift\( 2s_{1/2} \) above \( 2p_{1/2} \) by about 1058 MHz (QED)
Doublet splitting\( A(l + \frac{1}{2}) \); \( l = 1 \): \( \frac{3}{2}A \)
Interval rule\( E(J) - E(J-1) = AJ \)
Landé factor\( g_J = 1 + \frac{j(j+1) + s(s+1) - l(l+1)}{2j(j+1)} \)

Key Takeaways

What’s Next