Limitations of Bohr — fails for multi-electron atoms, fine structure: Visual walkthrough

Step 1 — Bohr's world: one energy per orbit

WHY we start here: to see a splitting, we must first see the thing that is supposed to be single. If energy is set by $n$ alone, there is literally no room for one line to become two — so the splitting must come from physics Bohr left out. We are hunting for the missing knob.

PICTURE: Look at the figure — three concentric chalk circles labelled $n = 1, 2, 3$ . Each circle carries one horizontal energy bar on the right. One track, one energy. That is the whole of Bohr.

Figure — Limitations of Bohr — fails for multi-electron atoms, fine structure

For the full derivation of this line see Bohr Model Derivation.

Step 2 — The electron carries a hidden compass (spin)

WHY this matters: a magnet has two things it can do — point one way or the opposite way. That is a two-choice knob. If the electron's energy can depend on which way this compass points, then one Bohr line could become two lines. This is our candidate for the missing variable.

PICTURE: The figure shows the electron drawn as a small sphere with a curved arrow around it (the spin), and beside it the equivalent bar magnet with its north pole marked. The spin points either up or down — the two pale-yellow arrows.

Step 3 — From the electron's seat, the nucleus orbits it (a current loop)

WHY this step: a loop of current makes a magnetic field through its centre — and you, the electron, sit at that centre. So the electron sits inside a magnetic field that its own orbital motion creates. This is the field the spin-compass will feel. We needed a field for the compass to react to; the orbit itself supplies it.

PICTURE: Two panels. Left = "lab view": electron circles a fixed nucleus. Right = "electron's view": nucleus circles the electron, forming a current loop, with a blue field arrow $B$ poking straight up through the electron's location.

Step 4 — Compass meets field: two energies (spin–orbit coupling)

WHY this is the whole trick: Bohr's single energy $E_{n}$ now splits into two — one slightly below, one slightly above — depending on whether spin sits with or against the orbital field. We have found how one line becomes two.

PICTURE: The single energy bar from Step 1 splits into two chalk-pink bars: "spin with field" pushed down, "spin against field" pushed up. The gap between them is the fine-structure splitting.

Step 5 — One label to name each split level: total angular momentum $j$

WHY we need $j$ : the split energies depend on whether spin and orbit reinforce or fight, and $j$ captures exactly that in a single number. "Spin with orbit" gives the larger $j$ ; "spin against orbit" gives the smaller $j$ . So $j$ is the missing variable Bohr never had — energy will now depend on $n$ and $j$ .

PICTURE: Two vector-addition triangles. Top: $L$ and $S$ aligned → long $J$ , $j = l + \frac{1}{2}$ . Bottom: $L$ and $S$ opposed → short $J$ , $j = l - \frac{1}{2}$ . Two arrows, two $j$ values, two energies.

Step 6 — The degenerate case: when there is nothing to split ( $l = 0$ )

WHY we must show this: the contract says cover every case. With $l = 0$ , only one value of $j$ is possible ( $j = \frac{1}{2}$ , since $l - \frac{1}{2}$ would be negative and is forbidden), so spin–orbit splitting vanishes. The line stays single. Fine structure needs $l \geq 1$ to show its two-way split.

PICTURE: Left panel $l = 0$ : $L = 0$ , a single un-split bar (no splitting). Right panel $l = 1$ : $L \neq = 0$ , the bar splits into two. Side by side so the difference is unmistakable.

Step 7 — How big is the split? The number $α^{2}$

WHY this size and not bigger: $α$ measures how fast the electron moves compared to light. Since $α \approx 1/137$ is small, $α^{2} \approx 5 \times 1 0^{- 5}$ is tiny. That is precisely why you need a high-resolution spectrometer to see the split — it is 20,000 times smaller than the main line spacing.

PICTURE: A zoom-in: the main H-α line on a coarse scale, then a magnifying-glass blow-up revealing two lines separated by a gap labelled $\sim α^{2} E_{n}$ .

The full Sommerfeld–Dirac level, showing energy depends on $j$ : $E_{n, j} \approx - 13.6 \frac{Z ^{2}}{n ^{2}} [1 + \frac{Z ^{2} α ^{2}}{n} (\frac{1}{j + \frac{1}{2}} - \frac{3}{4 n})]$ The bracket is the fine-structure correction; because it contains $j$ , Bohr's $n$ -only formula could never produce it.

The one-picture summary

Recall Feynman retelling — the walkthrough in plain words

Start with Bohr's picture: the electron runs on a race-track, and each track has just one energy — nothing else can change it. But the real electron spins like a top, and a spinning charge is a tiny magnet. Now here's the clever bit: if you sit on the electron, you see the nucleus whizzing around you in a circle — and a charge going in a circle makes a magnetic field right where you're sitting. So the electron's little magnet is sitting inside a field made by its own orbit. A magnet in a field has two energies — pointing with the field (low) or against it (high). Bang: the one energy becomes two. We give each a name using $j$ , which is just "spin arrow plus orbit arrow, added up." If the electron isn't orbiting at all ( $l = 0$ ), there's no field, so no split — one line stays one line. And when there is a split, it's teeny — about $1/137$ squared, or five parts in a hundred thousand — so you need a really sharp spectrometer to catch it. That tiny doubling is fine structure, and Bohr missed it because his map had no knob for spin.

Recall Quick self-check

Why can Bohr's $E_{n}$ never show fine structure? ::: It depends on $n$ alone; splitting needs a second variable ( $j$ ), which Bohr's model lacks. What two ingredients create the splitting? ::: The electron's spin (a magnetic compass) and the orbital motion (which creates the field it feels) — together, spin–orbit coupling. Why does an $s$ orbital ( $l = 0$ ) show no spin–orbit split? ::: $L = 0$ means no orbital magnetic field, so only $j = \frac{1}{2}$ is allowed — a single level. Roughly how big is the split relative to the main energy? ::: About $α^{2} \approx (1/137)^{2} \approx 5 \times 1 0^{- 5}$ .