Quantum Numbers and Orbital Angular Momentum

The algebra of angular momentum is settled before this article begins. The operators $\hat L^2$ and $\hat L_z$ , their commutator $[\hat L_x,\hat L_y]=i\hbar\hat L_z$ , and the reading of $\hat{\mathbf L}$ as the generator of rotations are built and owned by aq-qm-orbital-angular-momentum-operators; the eigenfunctions $Y_\ell^m$ , the eigenvalue equations $\hat L^2 Y_\ell^m=\ell(\ell+1)\hbar^2 Y_\ell^m$ and $\hat L_z Y_\ell^m=m\hbar Y_\ell^m$ , and the magnitude $|\mathbf L|=\sqrt{\ell(\ell+1)}\,\hbar$ are built and owned by aq-qm-spherical-harmonics. Both are recalled here by name, not re-derived. What this article owns is the piece that turns that machinery into spectroscopy: the labeling system $(n,\ell,m_\ell,m_s)$ — the four numbers that name every atomic state — together with the vector model, space quantization, the $s,p,d,f$ notation, and the counting that fixes the shape of the periodic table. Two facts drive everything and are worth stating up front, because both are routinely mis-stated. First, the length of the angular-momentum vector is $\sqrt{\ell(\ell+1)}\,\hbar$ , never $\ell\hbar$ — and because $\sqrt{\ell(\ell+1)}>\ell$ , the vector can never lie flat along its own quantization axis. Second, its direction is quantized too: the projection $L_z=m_\ell\hbar$ is restricted to $2\ell+1$ discrete values, so the vector is pinned to a set of allowed cones. This directional quantization — space quantization — is the genuinely non-classical content, and Stern and Gerlach made it visible in 1922. From these two facts and the Pauli principle the entire architecture of atomic shells follows.

Key takeaways

Every bound electron state is labeled by four quantum numbers: principal $n$ (energy, size), orbital $\ell$ (angular-momentum magnitude), magnetic $m_\ell$ (its projection), and spin $m_s = \pm\tfrac12$ .
The ranges interlock: $n = 1,2,3,\dots$ ; $\ell = 0,1,\dots,n-1$ ; $m_\ell = -\ell,\dots,+\ell$ ( $2\ell+1$ values); $m_s = \pm\tfrac12$ .
Magnitude: $|\mathbf L| = \sqrt{\ell(\ell+1)}\,\hbar$ , strictly larger than the maximum projection $\ell\hbar$ — so $\mathbf L$ can never point along $z$ .
Space quantization: $L_z = m_\ell\hbar$ confines $\mathbf L$ to $2\ell+1$ cones with half-angles $\cos\theta = m_\ell/\sqrt{\ell(\ell+1)}$ ; for $\ell=2$ , $m=2$ that angle is $35.26^\circ$ .
The letters $s,p,d,f$ are fossils of pre-quantum spectroscopy (sharp, principal, diffuse, fundamental) mapped onto $\ell = 0,1,2,3$ .
Counting states: $\sum_{\ell=0}^{n-1}(2\ell+1) = n^2$ orbital states, $2n^2$ with spin — the shell capacities $2, 8, 18, 32$ that set the periodic table's row lengths.
Key idea: the four quantum numbers, the magnitude $\sqrt{\ell(\ell+1)}\,\hbar$ , and directional (space) quantization together form the label system on which every atomic spectrum, magnetic splitting, and shell structure is built.

The core insight

Angular momentum is quantized in both magnitude and direction: $|\mathbf L| = \sqrt{\ell(\ell+1)}\,\hbar$ and $L_z = m_\ell\hbar$ with $-\ell \le m_\ell \le \ell$ . Because the magnitude always exceeds the maximum projection $\ell\hbar$ , the vector never lies along the axis — it is confined to $2\ell+1$ discrete cones (space quantization). The labels $(n,\ell,m_\ell,m_s)$ built from this, plus the Pauli principle, name every state and count every shell.

Where this lives today

The four quantum numbers are the working coordinates of every precision atomic technology. Laser cooling and magneto-optical traps address individual $m_\ell$ and $m_s$ sublevels, using the Zeeman shift of equation (6) to tune atoms into resonance and push them toward rest; optical atomic clocks are stabilized on transitions between specific $(n,\ell,m)$ states chosen for their insensitivity to stray fields. The Stern–Gerlach idea — reading a quantum label off a spatial deflection — is the conceptual ancestor of qubit state readout, where a two-valued spin projection is mapped to a measurable signal. At the fundamental end, the same labels organize the search for new physics: measurements of the electron's magnetic moment (the $\mu_B$ of equation (6), refined to parts in a trillion) test quantum electrodynamics, and comparisons of $m_\ell$ sublevel structure across isotopes probe nuclear size. The scaffolding Sommerfeld guessed at in 1916 is now the readout basis of the most accurate measurements humans make.

The physical picture

Picture the orbital angular momentum as an arrow of fixed length, $|\mathbf L| = \sqrt{\ell(\ell+1)}\,\hbar$ , that is not free to point anywhere. Its shadow on the $z$ -axis is quantized — $L_z = m_\ell\hbar$ can take only integer steps from $-\ell\hbar$ to $+\ell\hbar$ — so the arrow is confined to a discrete family of cones, one per $m_\ell$ . A classical vector could tilt at any angle; the quantum one is allowed only these. And notice the arrow is always longer than its tallest allowed shadow ( $\sqrt{\ell(\ell+1)} > \ell$ ), so even the steepest cone leaves it tilted away from the axis. The vector can approach $z$ but never lie along it.

Figure

The vector model for ℓ = 2: L has fixed length √6 ℏ ≈ 2.449 ℏ but only five allowed tilts L_z = m_ℓℏ; even the steepest (m = +2) sits at θ = 35.26° from z, never on the axis

The five cones are the whole of $\ell = 2$ : five values of $m_\ell$ , five allowed orientations, all sharing the one length $\sqrt6\,\hbar \approx 2.449\,\hbar$ . The topmost, $m = +2$ , tilts at $35.26^\circ$ — not zero — because $2\hbar$ is the tallest shadow a $\sqrt6\,\hbar$ arrow can cast. That residual tilt is not a technicality: it is the geometric face of the uncertainty principle, since a vector lying exactly along $z$ would have $L_x = L_y = 0$ sharply, which the non-commuting components forbid. Now assemble the states. Each shell $n$ opens $\ell = 0,\dots,n-1$ ; each $\ell$ opens $2\ell+1$ values of $m_\ell$ ; each of those is doubled by spin. The running total is $2n^2$ — and those numbers are the lengths of the periodic table's rows.

Figure

The quantum-number tree: n opens ℓ = 0…n−1, each ℓ opens 2ℓ+1 values of m_ℓ, each doubled by spin — summing to the shell capacities 2n² = 2, 8, 18, 32

Read the totals down the right edge: $2, 8, 18, 32$ . They are $2n^2$ , and they are exactly the number of elements in the first four periods (or their theoretical maxima). The abstract angular-momentum spectrum, counted with the Pauli principle, is the reason the periodic table has the shape it does.

The core equations

The labeling system rests on the recalled eigenvalue structure applied to a single electron. The four quantum numbers, with their interlocking ranges, are the complete address of a hydrogenic state:

n = 1,2,3,\dots; \quad \ell = 0,1,\dots,n-1; \quad m_\ell = -\ell,\dots,+\ell; \quad m_s = \pm\tfrac12 \qquad (1)

where: $n$ — principal quantum number (energy $E_n = -13.6\,\mathrm{eV}/n^2$ , and size); $\ell$ — orbital quantum number (subshell $s,p,d,f$ ); $m_\ell$ — magnetic quantum number ( $2\ell+1$ orientations); $m_s$ — spin projection. The bound $\ell \le n-1$ comes from radial-series termination (owned upstream); the range of $m_\ell$ comes from the ladder that generated $Y_\ell^m$ . The magnitude and projection of the vector, recalled from aq-qm-spherical-harmonics, are the numbers $\ell$ and $m_\ell$ actually stand for:

|\mathbf L| = \sqrt{\ell(\ell+1)}\,\hbar, \qquad L_z = m_\ell\hbar \qquad (2)

where: $|\mathbf L|$ — length of the orbital angular-momentum vector; $L_z$ — its projection on the quantization axis. The gap between them is the whole story: $\sqrt{\ell(\ell+1)} > \ell$ for every $\ell > 0$ , so the maximum shadow $\ell\hbar$ falls short of the length, and the vector is always tilted. Writing $|\mathbf L| = \ell\hbar$ is the single most common error in the subject. The allowed tilts follow directly by taking the ratio of projection to length — the cone half-angles of the vector model:

\cos\theta_{m} = \frac{m_\ell}{\sqrt{\ell(\ell+1)}} \qquad (3)

where: $\theta_m$ — angle between $\mathbf L$ and $+z$ for the state $m_\ell = m$ . The steepest cone, $m = \ell$ , gives the minimum angle $\theta_{\min} = \arccos\!\big(\ell/\sqrt{\ell(\ell+1)}\big)$ , which is $45^\circ, 35.26^\circ, 30^\circ$ for $\ell = 1,2,3$ — shrinking toward $0$ as $\ell$ grows, the classical limit. Counting the states in a shell is a finite sum that collapses to a perfect square (Derivation B):

g_n = \sum_{\ell=0}^{n-1}(2\ell+1) = n^2 \quad(\text{orbital}), \qquad 2n^2 \ (\text{with spin}) \qquad (4)

where: $g_n$ — orbital degeneracy of shell $n$ ; the factor $2$ is the two spin states. The values $2n^2 = 2, 8, 18, 32$ are the shell capacities. The maximum orbital angular momentum a shell can hold is bounded, and — a sharp break from Bohr — it falls short of $n\hbar$ :

|\mathbf L|_{\max} = \sqrt{(n-1)n}\,\hbar < n\hbar \qquad (5)

where: the largest $\ell$ in shell $n$ is $\ell = n-1$ , giving $|\mathbf L|_{\max} = \sqrt{(n-1)n}\,\hbar$ . Bohr's model assigned the ground state $L = n\hbar = \hbar$ ; the correct answer is that the $n=1$ ground state has $\ell = 0$ and zero orbital angular momentum. The ratio $\sqrt{(n-1)n}/n \to 1$ only as $n \to \infty$ — the correspondence principle recovering the classical value. Finally, the labels acquire measurable energy once a magnetic field breaks the isotropy, through the orbital magnetic moment:

\mu_{L,z} = -m_\ell\,\mu_B, \qquad \Delta E = \mu_B B, \qquad \mu_B = \frac{e\hbar}{2m_e} = 9.274\times10^{-24}\ \mathrm{J\,T^{-1}} \qquad (6)

where: $\mu_B$ — the Bohr magneton; $\Delta E$ — the splitting between adjacent $m_\ell$ sublevels in a field $B$ , equal to $57.9\ \mu\mathrm{eV}$ at $1$ T. Without a field the $m_\ell$ states are degenerate and the label is invisible; a field makes space quantization observable as the Zeeman effect.

Derivation AThe allowed cone angles of the ℓ = 2 vector

The goal is to compute the five orientations the $\ell = 2$ angular-momentum vector may take, turning the abstract eigenvalues $|\mathbf L| = \sqrt6\,\hbar$ and $L_z = m_\ell\hbar$ into a concrete table of tilt angles — the vector model made quantitative. The strategy is elementary trigonometry on the recalled eigenvalues, and the payoff is a direct demonstration that the vector never lies along $z$ : the smallest angle is not zero.

Step 1. Fix the length and list the projections. For $\ell = 2$ , $|\mathbf L| = \sqrt{2(2+1)}\,\hbar = \sqrt6\,\hbar \approx 2.449\,\hbar$ , and $L_z = m_\ell\hbar$ with $m_\ell = -2,-1,0,+1,+2$ . Why: the magnitude is fixed by $\ell$ alone, so all five states are arrows of the same length differing only in tilt.

Step 2. Form the angle from projection over magnitude: $\cos\theta_m = L_z/|\mathbf L| = m_\ell/\sqrt6$ . Why: $\theta$ is the polar angle of the vector, and its cosine is the ratio of the $z$ -component to the length — the definition of a direction cosine.

Step 3. Evaluate. $m = \pm2$ : $\cos\theta = \pm2/\sqrt6 = \pm0.8165 \Rightarrow \theta = 35.26^\circ,\,144.74^\circ$ . $m = \pm1$ : $\cos\theta = \pm1/\sqrt6 \Rightarrow 65.91^\circ,\,114.09^\circ$ . $m = 0$ : $\theta = 90^\circ$ . Why: these five angles are the cones drawn in the vector-model figure — the complete set of orientations $\ell = 2$ permits.

Step 4. Read the punchline. The smallest angle is $35.26^\circ$ , not $0^\circ$ , because $|\mathbf L| = \sqrt6\,\hbar$ exceeds the maximum projection $2\hbar$ . Why: the vector cannot align with $z$ — a sharp $L_z = |\mathbf L|$ would force $L_x = L_y = 0$ , which the non-commuting components forbid; the residual tilt is the uncertainty principle in geometric form. *Verified: $|\mathbf L|/\hbar = \sqrt2,\sqrt6,\sqrt{12}$ for $\ell = 1,2,3$ and the $\ell = 2$ cone angles (35.26° at $m=2$ ) in verify_aq-optics-quantum-numbers-angular-momentum.py.*

Go deeper · Derivation B · 1 minWhy the shell count is exactly n²

Derivation BWhy the shell count is exactly n²

The goal is to prove $\sum_{\ell=0}^{n-1}(2\ell+1) = n^2$ — that the orbital states of shell $n$ number exactly $n^2$ , doubling to $2n^2$ with spin, and to see why the periodic table's rows are $2, 8, 18, 32$ . The strategy is a direct evaluation of the arithmetic series; the payoff is the bridge from angular-momentum labels to chemical shell structure.

Step 1. Count the states in each subshell. A given $\ell$ has $2\ell+1$ values of $m_\ell$ , hence $2\ell+1$ orbital states; the shell $n$ contains the subshells $\ell = 0, 1, \dots, n-1$ . Why: $m_\ell$ ranges over $2\ell+1$ integers, and the allowed $\ell$ stop at $n-1$ — the two range rules of equation (1).

Step 2. Sum the odd numbers: $\sum_{\ell=0}^{n-1}(2\ell+1) = 1 + 3 + 5 + \dots + (2n-1)$ . Why: substituting $\ell = 0,1,\dots,n-1$ into $2\ell+1$ produces exactly the first $n$ odd integers.

Step 3. Evaluate. The sum of the first $n$ odd integers is $n^2$ (each pair of consecutive squares differs by an odd number: $n^2 - (n-1)^2 = 2n-1$ ). So $g_n = n^2$ . Why: this is the classic telescoping identity — the running total of odd numbers builds the squares, giving $1, 4, 9, 16$ for $n = 1,2,3,4$ .

Step 4. Double for spin. Each orbital state holds two electrons ( $m_s = \pm\tfrac12$ ), so the shell capacity is $2n^2 = 2, 8, 18, 32$ . Why: the Pauli principle forbids two electrons from sharing all four numbers, so each $(n,\ell,m_\ell)$ orbital admits exactly the two spin states — and these capacities are the lengths of the periodic table's rows. *Verified: $\sum_{\ell=0}^{n-1}(2\ell+1) = n^2$ symbolically, and $2n^2 = 2,8,18,32$ for $n \le 4$ , in verify_aq-optics-quantum-numbers-angular-momentum.py.*

Limiting cases

Case	Result	Meaning
$\ell = 0$ ( $s$ -state)	$\|\mathbf L\| = 0$ , no cones	no orientation to quantize — spherically symmetric, nothing for a field to grip
Large $\ell$	$\sqrt{\ell(\ell+1)}\,\hbar \to \ell\hbar$ , $\theta_{\min} \to 0$	cones crowd toward a continuum; the classical vector returns (correspondence)
Circular state $\ell = n-1$	$\|\mathbf L\|_{\max} = \sqrt{(n-1)n}\,\hbar < n\hbar$	the most Bohr-like state — yet still short of Bohr's $n\hbar$
Ground state $n=1$	$\ell = 0$ , $L = 0$	Bohr said $L = \hbar$ ; quantum mechanics says zero — the model's sharpest failure
Strong field $B$	$E \supset \mu_B B\,(m_\ell + 2m_s)$	$m_\ell, m_s$ become good labels — the Zeeman and Paschen–Back regimes
Non-Coulomb core	$E_{n\ell}$ depends on $n$ and $\ell$	screening lifts the accidental degeneracy; the quantum defect of alkali spectra

Common trap|L| = ℓℏ, and the vector points along z when m = ℓ

Both halves are wrong, and they are the same error twice. The magnitude of orbital angular momentum is $|\mathbf L| = \sqrt{\ell(\ell+1)}\,\hbar$ , not $\ell\hbar$ : for $\ell = 1$ it is $\sqrt2\,\hbar \approx 1.414\,\hbar$ , not $\hbar$ ; for $\ell = 2$ it is $\sqrt6\,\hbar \approx 2.449\,\hbar$ , not $2\hbar$ . And precisely because the length exceeds the largest projection $\ell\hbar$ , the vector can never lie along the quantization axis — even the stretched state $m_\ell = \ell$ tilts at $\arccos(\ell/\sqrt{\ell(\ell+1)}) > 0$ . The reason is not a convention but the uncertainty principle: $\hat L_x$ and $\hat L_y$ do not commute with $\hat L_z$ , so they cannot all be sharp, and a vector pinned to $z$ would demand exactly that. Space quantization is a quantization of tilt, not a vector nailed to the axis.

Worked examples

1 · The n = 3 shell, enumerated (confidence builder)

List every subshell and orbital of the $n = 3$ shell, count the states with and without spin, and give the five $\ell = 2$ cone angles.

Step 1. Subshells: $\ell = 0, 1, 2$ , written $3s, 3p, 3d$ . Orbital counts $2\ell+1$ : $3s$ has $1$ , $3p$ has $3$ , $3d$ has $5$ — $9$ orbitals, matching $n^2 = 9$ .

Step 2. With spin, each orbital holds two electrons: $2 \times 9 = \mathbf{18}$ states, matching $2n^2 = 18$ . This is the theoretical capacity of the third shell.

Step 3. The $3d$ vector ( $\ell = 2$ ) has $|\mathbf L| = \sqrt6\,\hbar$ and cones at $\cos\theta = m/\sqrt6$ : $\theta = 35.26^\circ, 65.91^\circ, 90^\circ, 114.09^\circ, 144.74^\circ$ for $m = 2, 1, 0, -1, -2$ . Five orientations, none along the axis.

Step 4. Sanity check on energy: in pure hydrogen all nine orbitals share $E_3 = -13.6/9 = -1.51$ eV (accidental degeneracy); in sodium, penetration splits $3s < 3p < 3d$ . $n = 3$ : 9 orbitals, 18 states, $3d$ cones from $35.26^\circ$ to $144.74^\circ$ .

2 · Minimum tilt and the Bohr shortfall (exam level)

Find the minimum angle between $\mathbf L$ and $z$ for $\ell = 1, 2, 3$ , and the maximum $|\mathbf L|$ a shell $n$ can hold. Compare to Bohr's $n\hbar$ .

Step 1. Minimum angle is the steepest cone, $m = \ell$ : $\theta_{\min} = \arccos(\ell/\sqrt{\ell(\ell+1)})$ . For $\ell = 1$ : $\arccos(1/\sqrt2) = \mathbf{45^\circ}$ . For $\ell = 2$ : $\arccos(2/\sqrt6) = \mathbf{35.26^\circ}$ . For $\ell = 3$ : $\arccos(3/\sqrt{12}) = \arccos(\sqrt3/2) = \mathbf{30^\circ}$ .

Step 2. The trend is monotone: $45^\circ \to 35.26^\circ \to 30^\circ$ , shrinking toward $0$ as $\ell$ grows. A high- $\ell$ state aligns ever more closely with $z$ — the quantum cones approaching the classical continuum of orientations.

Step 3. Maximum magnitude in shell $n$ is $\ell = n-1$ : $|\mathbf L|_{\max} = \sqrt{(n-1)n}\,\hbar$ . For $n = 2$ : $\sqrt2\,\hbar$ ; $n = 3$ : $\sqrt6\,\hbar$ ; $n = 4$ : $\sqrt{12}\,\hbar = 2\sqrt3\,\hbar$ .

Step 4. Compare to Bohr. Bohr's circular orbit had $L = n\hbar$ ; the correct maximum is $\sqrt{(n-1)n}\,\hbar$ , always smaller ( $\sqrt2 < 2$ , $\sqrt6 < 3$ , $\sqrt{12} < 4$ ), and the ground state has $L = 0$ , not $\hbar$ . The ratio $\sqrt{(n-1)n}/n \to 1$ only as $n \to \infty$ . $\theta_{\min} = 45^\circ, 35.26^\circ, 30^\circ$ ; $|\mathbf L|_{\max} = \sqrt{(n-1)n}\,\hbar < n\hbar$ .

Physics of the portrait

The concept at the center of this article is not a person but an idea that outlived the theory that named it: space quantization, the claim that a direction in space can be quantized.

Origins · 1 minspace quantization and how this was discovered

space quantization

the quantization of angular-momentum direction, restricting L_z to m_ℓℏ so the vector tilts only at discrete cone angles · atomic physics

Space quantization — Richtungsquantelung, the quantization of direction — was introduced by Arnold Sommerfeld in 1916 as he extended Bohr's model to elliptical orbits in a magnetic field, positing that an orbit's plane could take only discrete orientations. It was a startling claim: that space itself imposes a discrete set of allowed tilts. In 1922 Otto Stern and Walther Gerlach set out to test it, sending a beam of silver atoms through an inhomogeneous magnetic field; a classical magnet would smear into a continuous band, but the beam split cleanly into two — discrete deflection, space quantization made visible. The irony, unrecognized for years, is that silver's valence electron is in an $\ell = 0$ state with no orbital angular momentum at all: what Stern and Gerlach actually quantized was electron spin (its two-valued projection, detailed in aq-optics-electron-spin-magnetic-moment), a phenomenon not yet discovered. The old quantum theory even had the magnitude wrong — Sommerfeld's $|\mathbf L| = \ell\hbar$ was replaced in 1926 by Schrödinger's $\sqrt{\ell(\ell+1)}\,\hbar$ . Yet the core idea survived every correction: angular momentum is quantized in direction as well as magnitude, and the Stern–Gerlach experiment remains the textbook demonstration of it.

History

1896

— Zeeman observes the magnetic splitting of spectral lines

1907

— Bergmann's fundamental series completes the sharp, principal, diffuse, fundamental naming behind s, p, d, f

1916

— Sommerfeld introduces space quantization (Richtungsquantelung) in the old quantum theory

1922

— Stern and Gerlach split a silver beam, demonstrating spatial quantization

1925

— Pauli's exclusion principle adds a two-valued fourth quantum number

1925

— Uhlenbeck and Goudsmit identify that fourth number as electron spin

1926

— Schrödinger's wave mechanics gives the magnitude √(ℓ(ℓ+1)) ℏ, correcting the old-quantum ℓℏ

Going further

Texts. Foot, Atomic Physics, Ch. 2–3 · Griffiths and Schroeter, Introduction to Quantum Mechanics, 3rd ed., Ch. 4 · Demtröder, Atoms, Molecules and Photons, Ch. 5 · Bransden and Joachain, Physics of Atoms and Molecules, Ch. 3 · Saleh and Teich, Fundamentals of Photonics, Ch. 13.

Provenance. Ranges of $n, \ell, m_\ell, m_s$ , node count $n-\ell-1$ , degeneracies $n^2$ and $2n^2$ , and eigenvalues $\ell(\ell+1)\hbar^2$ , $m_\ell\hbar$ are standard (Griffiths Ch. 4; Foot Ch. 2–3; Bransden and Joachain Ch. 3). Energy $-13.6\,\mathrm{eV}/n^2$ and $\mu_B = 9.274\times10^{-24}\,\mathrm{J\,T^{-1}}$ are CODATA/textbook values. A verifier should reproduce the eigenvalue magnitudes $\sqrt{\ell(\ell+1)}\,\hbar$ , the cone angles $\cos\theta = m_\ell/\sqrt{\ell(\ell+1)}$ , the degeneracy sum $\sum(2\ell+1) = n^2$ , and the maximum- $|\mathbf L|$ bound $\sqrt{(n-1)n}\,\hbar < n\hbar$ . All are reproduced in verify_aq-optics-quantum-numbers-angular-momentum.py.

Prerequisites. Solution of the Schrödinger equation for hydrogen · The Bohr model of hydrogen · Hydrogenic wavefunctions and radial distributions. Leads to. Electron spin and the spin magnetic moment · Fine structure and spin-orbit coupling · The Zeeman effect · Multi-electron atoms and the central-field approximation · Hyperfine structure.

Why is |L| = √(ℓ(ℓ+1)) ℏ and not ℓℏ, which would be so much tidier?

Because $\ell\hbar$ is only the maximum projection, not the length. The ladder that builds the $Y_\ell^m$ (owned by aq-qm-spherical-harmonics) fixes $\hat L^2 = \hat L_-\hat L_+ + \hat L_z^2 + \hbar\hat L_z$ , and evaluating it on the top rung $m = \ell$ gives $\ell(\ell+1)\hbar^2$ , not $\ell^2\hbar^2$ . The extra $\ell\hbar^2$ is the contribution of the transverse components $L_x, L_y$ , which cannot be zero even when $L_z$ is maximal — the uncertainty principle again. Sommerfeld's old quantum theory did use $\ell\hbar$ ; it was wrong, and the discrepancy shows up in the correct multiplicities and the vector-model angles.

Is the atom physically tilted at these cone angles, wobbling around z?

The vector model is a bookkeeping picture, not a literal wobbling arrow. What is true and sharp: $|\mathbf L|$ and $L_z$ have definite values simultaneously. What is not sharp: $L_x$ and $L_y$ , which are completely uncertain (their averages vanish but their spreads do not). The cone represents that uncertainty — the vector is somewhere on the cone with uniform azimuth, never at a definite point on it. So the atom is not a tilted top precessing at a fixed angle; it is a state in which the projection is known and the transverse direction is maximally unknown. The cone is the geometry of that ignorance.

Where do the letters s, p, d, f actually come from?

From nineteenth-century spectroscopy, long before anyone knew what $\ell$ was. Analysts of alkali spectra sorted the emission lines into families by appearance: the sharp series (narrow lines), the principal series (strongest, seen in absorption), the diffuse series (broad), and — added by Bergmann in 1907 — the fundamental series (in the infrared). When quantum mechanics later assigned each family to transitions from a definite orbital type, the initials $s, p, d, f$ were simply kept, and the sequence continues alphabetically ( $g, h, \dots$ , skipping $j$ ). The labels are pre-quantum fossils that happen to have survived into the quantum era — a small piece of history embedded in every orbital diagram.

Why does m_ℓ not matter unless there is a magnetic field?

Because in an isolated atom space is isotropic — there is no preferred axis, so all $2\ell+1$ orientations of a given $\ell$ have identical energy. The label $m_\ell$ exists, but it is a degeneracy, invisible in the spectrum. A magnetic field breaks the isotropy: it defines a $z$ -axis and gives each orientation a different energy through $\Delta E = \mu_B B\,m_\ell$ (equation 6), splitting one spectral line into several. That is the Zeeman effect, and it is precisely the experimental appearance of space quantization — the field converts a hidden label into a measurable energy ladder. Remove the field and the sublevels collapse back into one.

Did Stern and Gerlach really demonstrate orbital space quantization?

They demonstrated space quantization, but not of the kind they thought. Silver's valence electron is a $5s$ state with $\ell = 0$ , so its orbital angular momentum is zero and cannot be what deflected the beam. The two-way split came entirely from the electron's spin magnetic moment, whose projection takes two values — a phenomenon not discovered until Uhlenbeck and Goudsmit in 1925, three years later. So the experiment is a valid, even decisive, demonstration that a magnetic moment's orientation is quantized (space quantization as a principle), but the specific angular momentum it revealed was spin, not orbital motion. It is one of physics' great productive accidents.

How does the fourth quantum number and the 2n² count come together?

Spin adds $m_s = \pm\tfrac12$ as an internal degree of freedom the Schrödinger equation does not contain — grafted on after Uhlenbeck and Goudsmit, and demanded by Pauli's exclusion principle, which forbids two electrons from sharing all four numbers $(n, \ell, m_\ell, m_s)$ . Each spatial orbital $(n, \ell, m_\ell)$ therefore holds exactly two electrons, one of each spin. Since a shell has $n^2$ orbitals (Derivation B), it holds $2n^2$ electrons: $2, 8, 18, 32$ . This is why the periodic table's rows have those lengths and why the noble gases fall where they do — the entire architecture is the four quantum numbers plus one exclusion rule.