Quantum numbers — n (principal), l (azimuthal), mₗ (magnetic), mₛ (spin): Visual walkthrough

Before line one, four plain words we will lean on:

Definition Four words we need first

Wave — a pattern that has a value (height) at every point in space and can wiggle up and down. A guitar string held at both ends is a wave.
Angle $ϕ$ (phi) — how far around a circle you have walked, measured from a starting mark. One full lap is $2 π$ (that's just $36 0^{\circ}$ written in radians — a way of measuring angle where a full circle is $2 π$ instead of $360$ ).
Angular momentum — the amount of "going-around motion" a thing has. A spinning top and a planet orbiting the Sun both have it. More going-around = more angular momentum. We give it the symbol $L$ (the little arrow means it points in a direction, like which way a spinning wheel's axle points).
The imaginary unit $i$ — a special number defined by one rule: $i^{2} = - 1$ . There is no ordinary number that squares to a negative, so mathematicians invented one and named it $i$ . Its magic use for us: multiplying by $i$ rotates things by a quarter-turn. So while an ordinary wiggle like $cos$ moves up and down on a line, a wave built with $i$ moves around a circle — which is exactly what we need for a wave that loops around the nucleus.

Step 1 — An electron is a standing wave, so it must "fit"

WHAT. Picture the electron not as a ball but as a wave wrapped in a loop around the nucleus, like a ripple running around a circular pond.

WHY. A wavefunction must be single-valued: at any one point in space it has exactly one height. If the wave did not close up smoothly after one lap, the same point would have two different heights — impossible. So only waves that fit a whole number of humps into the loop survive.

PICTURE. In the figure, the green wave completes exactly $3$ full humps around the ring and joins itself perfectly — it survives. The coral wave has $2.5$ humps: after one lap it comes back at the wrong height (the two red dots don't meet). That wave is forbidden.

Step 2 — Counting the humps gives us $m_{l}$

WHAT. Give the number of humps a name. Call it $m_{l}$ — the magnetic quantum number. The wave that wraps around the loop is written

$Φ (ϕ) = e^{i m_{l} ϕ}$

Term by term, right where each symbol sits:

$Φ (ϕ)$ — the height of the wave at angle $ϕ$ (the loop is measured by $ϕ$ ).
$ϕ$ — how far around the loop we've walked.
$m_{l}$ — how many full humps the wave packs into one lap.
$i$ — the imaginary unit from the definitions box. It is what turns a plain up-down wiggle into a wave that travels around the circular loop.
$e^{i (\dots)}$ — this is the compact way of writing a wave that goes around. You may read $e^{i θ}$ as "the point on a unit circle at angle $θ$ "; because of the $i$ , growing $θ$ walks the point around the circle, and each time $θ$ grows by $2 π$ the point returns home.

WHY this exact form? Because it is the only shape whose "wiggle rate" stays constant all the way around the loop — which is exactly what a free go-around motion demands. Any other shape would speed up and slow down for no reason.

PICTURE. The figure shows the height of $Φ$ as we walk once around: for $m_{l} = 1$ it makes one hump, for $m_{l} = 2$ two humps, for $m_{l} = 3$ three. Count the peaks = read off $m_{l}$ .

Now impose "fit after one lap." Walking one extra lap ( $ϕ \to ϕ + 2 π$ ) must return the same height:

$e^{i m_{l} (ϕ + 2 π)} = e^{i m_{l} ϕ} ⟹ e^{i 2 π m_{l}} = 1$

The right side, $e^{i 2 π m_{l}} = 1$ , is true only when $m_{l}$ is a whole number (the point returns exactly home only after a whole number of laps around the circle). Fractions land somewhere else on the circle. So:

Step 3 — What "asking the wave for its going-around" means: $L_{z} = m_{l} ℏ$

WHAT. In quantum mechanics, every measurable quantity is found by applying a matching operator — a little instruction — to the wavefunction. The operator that reads off going-around motion about the up-axis (the $z$ -axis — we simply pick "straight up" as our reference direction) is

$\hat{L}_{z} = - i ℏ \frac{\partial}{\partial ϕ}$

Term by term:

$\hat{L}_{z}$ — the "measure the up-axis going-around" instruction (the hat means operator).
$\frac{\partial}{\partial ϕ}$ — "see how fast the wave changes as you step around the loop $ϕ$ " (a rate-of-change along the loop). This is the natural thing to look at, because going-around motion is about how the wave twists as $ϕ$ advances.
$- i ℏ$ — the fixed quantum conversion factor that turns "twist-rate around the loop" into genuine angular momentum (the $ℏ$ sets the units; the $- i$ undoes the $i$ that lives inside the wave).

WHY apply it to our wave? Because our loop-wave $Φ (ϕ) = e^{i m_{l} ϕ}$ from Step 2 has a beautifully simple twist-rate. Take the rate of change:

$\frac{\partial}{\partial ϕ} e^{i m_{l} ϕ} = i m_{l} e^{i m_{l} ϕ}$

so feeding it into the operator gives

$\hat{L}_{z} Φ = - i ℏ (i m_{l}) e^{i m_{l} ϕ} = (m_{l} ℏ) same wave back e^{i m_{l} ϕ}$

The wave comes back unchanged except for a number in front, and that number is the measured value:

$L_{z} = m_{l} ℏ$

(The two $i$ 's multiply to $i^{2} = - 1$ , which cancels the minus sign — that is why the operator carries a $- i$ .) A wave that returns itself times a number is called an eigen-wave; the number in front is the value you actually measure. So $m_{l}$ counted humps in Step 2, and here that exact same $m_{l}$ pops out as the up-axis angular momentum in units of $ℏ$ .

PICTURE. The vector $L$ is drawn as a tilted arrow; its vertical shadow is $L_{z}$ . Different $m_{l}$ give arrows tilted at different fixed angles, each casting a shadow that is a whole number of $ℏ$ : $+ 2ℏ, + ℏ, 0, - ℏ, - 2ℏ$ .

Step 4 — The polar-angle wave forces integer $l$ , and $l$ caps $m_{l}$

WHAT. Winding around the loop ( $ϕ$ ) was only part of the motion. The wave also rises and dips as you go from the north pole, down to the equator, to the south pole — this is the polar angle $θ$ (theta): $θ = 0$ is straight up, $θ = π$ is straight down. The total going-around motion (all directions, not just the $z$ -shadow) is read by a second operator, $\hat{L}^{2}$ . Solving the $θ$ -part of the Schrödinger equation for hydrogen atom gives its eigenvalue:

$\hat{L}^{2} Y = l (l + 1) ℏ^{2} Y ⟹ ∣ L ∣ = l (l + 1) ℏ$

Term by term:

$\hat{L}^{2}$ — the "measure the total going-around, squared" instruction.
$Y (θ, ϕ)$ — the combined angular wave (loop part $\times$ pole-to-pole part).
$l$ — a whole number $0, 1, 2, \dots$ labelling how much total going-around; bigger $l$ = longer arrow.
$l (l + 1)$ — comes straight out of the polar equation, not from a triangle. Here is why it's this and not just $l^{2}$ :

WHERE $l (l + 1)$ and integer $l$ come from. The $θ$ -part of the equation is solved by special humpy functions called Legendre functions $P_{l}^{m_{l}} (cos θ)$ . Just like the loop wave had to close up, the polar wave has to stay finite at both poles ( $θ = 0$ and $θ = π$ ). That "no blow-up at the poles" demand only succeeds when:

$l$ is a non-negative whole number (else the function shoots to infinity at a pole), and
the loop-count satisfies $∣ m_{l} ∣ \leq l$ (else the same Legendre function is identically zero — the wave vanishes everywhere, which is no wave at all).

So the polar angle is what actually forces $l$ to be an integer and pins the range of $m_{l}$ :

$m_{l} = - l, - l + 1, \dots, 0, \dots, + l ⟹ (2 l + 1) values$

WHY the "shadow" picture still works. The result $∣ m_{l} ∣ \leq l$ looks like a shadow never beating its stick: $L_{z} = m_{l} ℏ$ (a shadow) can never exceed $∣ L ∣ = l (l + 1) ℏ$ (the stick). Both stories agree — the Legendre rule is the exact reason, the shadow is the picture that makes it memorable.

PICTURE. A fixed-length arrow of length $l (l + 1) ℏ$ (here $l = 2$ ) is drawn in several allowed tilts. Its shadow reaches at most $l ℏ = 2ℏ$ but never the full length — because the polar wave has no finite solution beyond $∣ m_{l} ∣ = l$ . Count the allowed tilts: $5 = 2 l + 1$ .

Step 5 — The radial wave must be finite at the centre and die out far away: $n$ appears and caps $l$

WHAT. So far we handled the two angular directions ( $ϕ$ around, $θ$ pole-to-pole). The electron-wave also spreads outward from the nucleus — the radial direction $r$ (distance from the centre). Its shape is the radial function $R (r)$ , and solving the radial Schrödinger equation gives it two survival demands:

Finite at the centre: $R (r)$ must not blow up at $r = 0$ (an infinite wave at the nucleus is unphysical).
Dies out far away: $R (r) \to 0$ as $r \to \infty$ (the electron is bound, so it can't have appreciable wave infinitely far out).

WHY this forces integer $n$ and $l \leq n - 1$ . Meeting both demands at once only works for special discrete energies. The surviving radial functions turn out to have a whole number of radial nodes (places where $R (r)$ crosses zero), and that node count is exactly

$radial nodes = n - l - 1.$

Term by term:

$n$ — the principal quantum number, the whole number that labels which discrete energy survived; it counts total nodes $+ 1$ .
$l$ — the angular label from Step 4.
$n - l - 1$ — the number of times the radial wave crosses zero. A count of things can never be negative, so we need $n - l - 1 \geq 0$ , i.e.

$l \leq n - 1 equivalently n \geq l + 1.$

If you tried $l = n$ , the formula asks for $- 1$ radial nodes — impossible — so no finite, decaying radial wave exists. That is the honest reason a shell $n$ can only host shapes up to $l = n - 1$ (and why the "budget" felt right: more angular humps leaves fewer allowed radial nodes).

PICTURE. For $n = 3$ : the allowed subshells $l = 0, 1, 2$ each shown with their radial wave $R (r)$ — finite at $r = 0$ , decaying as $r$ grows — and their node counts $n - l - 1 = 2, 1, 0$ . The forbidden $l = 3$ would need $- 1$ nodes and is crossed out.

Step 6 — The hidden fourth number: spin $m_{s}$

WHAT. The three numbers $n, l, m_{l}$ came straight from a wave in 3D space. Experiments (a single spectral line splitting into two) forced a fourth, purely quantum label: spin, $m_{s}$ , which takes only two values.

$m_{s} = + \frac{1}{2} or - \frac{1}{2}$

WHY only two. Spin is an intrinsic two-way property — "up" or "down" — with no classical picture (the electron is not literally spinning; a point spinning fast enough would break the light-speed limit). It doubles every spatial address into two.

PICTURE. A single beam of atoms passes through a magnet and splits into exactly two spots — up-deflected and down-deflected. Two spots ⇒ two spin states, no more.

Step 7 — Add up every address: the $2 n^{2}$ result

WHAT. Now multiply the whole nested tree for a shell $n$ :

$states = spin m_{s} 2 \times orbitals in shell l = 0 \sum n - 1 (2 l + 1)$

Term by term:

$\sum_{l = 0}^{n - 1}$ — add up over every allowed subshell (Step 5).
$(2 l + 1)$ — orientations in each subshell (Step 4).
$\times 2$ — the two spins (Step 6).

WHY it collapses so neatly. The sum $1 + 3 + 5 + \dots + (2 n - 1)$ is the sum of the first $n$ odd numbers, which is always $n^{2}$ . (Look at the figure: odd-sized L-shaped layers stack into a perfect $n \times n$ square.) So:

$\sum_{l = 0}^{n - 1} (2 l + 1) = n^{2} ⟹ states = 2 n^{2}$

PICTURE. Odd numbers $1, 3, 5$ drawn as nested L-shaped "gnomons" filling a $3 \times 3$ grid ( $= 9 = n^{2}$ for $n = 3$ ), then doubled by two spin colours to give $18 = 2 n^{2}$ .

The one-picture summary

The whole derivation as one nested tree. Each level's count is decided by the level above it — read top to bottom like Russian dolls. This single figure holds every rule we built.

Recall Feynman retelling — the walkthrough in plain words

Picture the electron as a wave wrapped around the nucleus. For it to make sense, it has to close up on itself after one lap — so it can only fit a whole number of humps. Count those humps: that's $m_{l}$ , and because you can't fit half a hump, $m_{l}$ must be a whole number, positive or negative depending on which way the wave winds.

To get the going-around motion out of this wave, we apply the operator $- i ℏ \partial / \partial ϕ$ ; the loop-wave comes right back multiplied by $m_{l} ℏ$ , so the up-axis angular momentum is $L_{z} = m_{l} ℏ$ .

But the wave also rises and dips from pole to pole (angle $θ$ ). Keeping that part finite at both poles only works when $l$ is a whole number and $∣ m_{l} ∣ \leq l$ — the true reason behind the "shadow can't beat the arrow" picture, with total length $l (l + 1) ℏ$ .

The wave also spreads outward (distance $r$ ). Staying finite at the nucleus and dying away far out forces discrete energies labelled by $n$ , whose radial wave has $n - l - 1$ zero-crossings. A count can't be negative, so $l \leq n - 1$ .

Finally, experiments split one beam of atoms into exactly two — so there's a hidden two-way tag, spin $m_{s} = \pm \frac{1}{2}$ , that has no picture of an actual spinning ball.

Multiply the whole nested tree — $2$ spins, times $2 l + 1$ tilts, summed over $l = 0$ to $n - 1$ — and the odd numbers $1, 3, 5, \dots$ stack into a perfect square $n^{2}$ , giving the famous $2 n^{2}$ electrons per shell. Every rule was just "the wave has to fit."

Recall Quick self-check

Why must $m_{l}$ be an integer? ::: The wave $e^{i m_{l} ϕ}$ must return to the same height after one lap ( $ϕ \to ϕ + 2 π$ ), forcing $e^{i 2 π m_{l}} = 1$ , true only for integer $m_{l}$ . Which operator gives $L_{z}$ , and what does it return on $e^{i m_{l} ϕ}$ ? ::: $\hat{L}_{z} = - i ℏ \partial / \partial ϕ$ ; it returns $m_{l} ℏ$ times the same wave, so $L_{z} = m_{l} ℏ$ . Why is $l$ an integer and $∣ m_{l} ∣ \leq l$ ? ::: The polar ( $θ$ ) Legendre functions stay finite at both poles only for integer $l$ with $∣ m_{l} ∣ \leq l$ . Why is $l \leq n - 1$ ? ::: Radial nodes $= n - l - 1$ must be $\geq 0$ for a finite, decaying radial wave. Where does $2 n^{2}$ come from? ::: $2 \times \sum_{l = 0}^{n - 1} (2 l + 1) = 2 n^{2}$ , since the first $n$ odd numbers sum to $n^{2}$ .

Connections

Schrödinger equation for hydrogen atom — the wave equation whose fitting conditions produced $n, l, m_{l}$
Shapes of atomic orbitals (s, p, d) — what different $l$ values look like in space
Zeeman effect — switching on a field that finally makes $m_{l}$ matter for energy
Pauli exclusion principle — why each address holds at most one electron
Aufbau principle and electron configuration — filling those addresses in order
Hund's rule — how spin $m_{s}$ decides the filling within a subshell