Where Classical Physics Breaks

Seven experiments that classical physics got provably, catastrophically wrong — and the single quantum idea each one forced into existence. In every panel the classical prediction is computed and drawn as the failing foil, right beside the honest quantum result. This is the crime scene; the deep First-Principles entries are the theory it demanded.

Planck 1900 · Einstein 1905 · Rutherford–Bohr 1911–13 · de Broglie 1924 · Stern–Gerlach 1922 · Pauli 1925 · Davisson–Germer 1927
HONESTY TIER  A · DIRECT  every curve is a live computation; each classical foil is labelled as the wrong prediction it is
7
classical crises
1900–27
the quantum revolution
1
idea forced by each
live
every panel computed
 ›  Quantum Mechanics  ›  Where Classical Physics Breaks

By 1900 classical physics — Newton's mechanics, Maxwell's electromagnetism, and statistical thermodynamics — looked complete. Then a handful of ordinary measurements returned answers the theory could not merely refine, but flatly could not produce: infinite energy in a warm oven, atoms that should live a hundredth of a nanosecond, particles that interfere with themselves, and matter that has no business taking up space. Each crisis below is shown twice: what a rigorous classical calculation predicts, and what actually happens. The gap between them is not a rounding error. It is the width of quantum mechanics.

On honesty: this series never lets a picture claim more than the mathematics allows. Here the quantum results are direct computations — the Planck law, the Born-rule distribution, the exact barrier transmission. The classical curves are equally real calculations (Rayleigh–Jeans, the Larmor radiated-power spiral, a two-band particle histogram), drawn deliberately as the failing foil so you can see exactly where and how classical physics diverges from the world. Nothing here is a hand-wave in either direction.

I · The Ultraviolet Catastropheoven → ∞ energy · quantization II · The Photoelectric Effectcolour, not brightness · photons III · The Collapsing Atom16-picosecond death · stationary states IV · One Particle, Two Slitsself-interference · superposition V · Through the Wallforbidden crossing · tunnelling VI · The Beam That Split in TwoStern–Gerlach · quantized spin VII · Why Matter Takes Up Spaceno collapse · Pauli exclusion
CRISIS I

The Ultraviolet Catastrophe

TIER A · DIRECT
Classical predicts
Equipartition gives every wave-mode of the oven the same energy \(kT\). Counting modes, the Rayleigh–Jeans law says radiated power \(\propto 1/\lambda^4\) — soaring to infinite energy at short wavelengths.
What actually happens
Real ovens glow a finite, bell-shaped spectrum that peaks and then falls to zero in the ultraviolet. A hot body does not blast out infinite blue light.
Quantum idea forced
Energy comes in lumps \(E=h\nu\). High-frequency modes cost too much to excite and freeze out — Planck's law. The birth of the quantum, 1900.
Peak λ (Wien): 500 nm
Rayleigh–Jeans at 100 nm: —× Planck
Planck: \(\;u(\lambda,T)=\dfrac{8\pi hc}{\lambda^5}\dfrac{1}{e^{hc/\lambda kT}-1}\;\) vs Rayleigh–Jeans: \(\;u_{\rm RJ}=\dfrac{8\pi kT}{\lambda^4}\to\infty.\) The two agree at long λ and diverge without limit toward the blue — the "catastrophe."
CRISIS II

The Photoelectric Effect

TIER A · DIRECT
Classical predicts
Light is a wave, so its energy is set by intensity. A bright enough beam of any colour should, given time, shake electrons loose — and brighter light should eject faster electrons.
What actually happens
Below a threshold frequency, no electrons come out at any brightness. Above it, they emerge instantly; raising brightness makes more electrons, never faster ones.
Quantum idea forced
Light arrives as photons of energy \(h\nu\). One photon, one electron: \(K_{\max}=h\nu-\phi\). Einstein 1905 — the reality of the quantum of light.
Metal: sodium, φ = 2.28 eV
Kmax = 0.72 eV
\(K_{\max}=h\nu-\phi\). Threshold \(\nu_0=\phi/h\) → for sodium, \(\lambda_0=544\) nm (green). Redder than that: nothing, no matter how intense. This linear \(K\)-vs-\(\nu\) slope is Planck's constant, measured.
CRISIS III

The Collapsing Atom

TIER A · DIRECT
Classical predicts
An orbiting electron is an accelerating charge, so by Maxwell it must radiate, lose energy, and spiral into the nucleus in ~1.6 × 10⁻¹¹ s. Every atom in the universe should have died instantly.
What actually happens
Atoms are stable for the age of the universe, and emit only sharp, discrete spectral lines — not the continuous smear a death-spiral would radiate.
Quantum idea forced
The electron occupies stationary states — standing waves of definite energy \(\hat H\psi=E\psi\) — with a lowest rung it cannot fall below. Bohr 1913 → Schrödinger 1926.
classical clock: 0.0 ps
quantum 1s: stable ∀ t
Larmor: radiated power \(P=\dfrac{e^2 a^2}{6\pi\varepsilon_0 c^3}\) drains the orbit; integrating gives a fall time \(\tau=\dfrac{a_0^3}{4 r_e^2 c}\approx 1.6\times10^{-11}\) s. Quantum mechanics replaces the orbit with \(|\psi_{1s}|^2\), a cloud with no lower state to radiate into.
→ deep dive: The Hydrogen Atom (exact |ψₙₗₘ|², the Bohr orbit as labelled foil)
CRISIS IV

One Particle, Two Slits

TIER A · DIRECT
Classical predicts
A particle goes through one slit or the other. Fire them one at a time and the screen should show two bands — the simple sum of two single-slit piles. No stripes.
What actually happens
Electrons sent one at a time — never two in the apparatus together — still pile up into an interference pattern of many fringes. Each particle interferes with itself.
Quantum idea forced
The electron travels as a superposition of both paths; the amplitudes add, and \(|\psi_1+\psi_2|^2\) carries a cross term. Landing spots are Born-rule random. Tonomura 1989.
electrons landed: 0
quantum: fringes forming
Screen intensity \(I(x)\propto \cos^2\!\big(\tfrac{\pi d\,x}{\lambda L}\big)\,\mathrm{sinc}^2\!\big(\tfrac{\pi a\,x}{\lambda L}\big)\). Turn on a which-path marker and the cross term dies: the fringes collapse to the classical two-band envelope. Complementarity, made of dots.
CRISIS V

Through the Wall

TIER A · DIRECT
Classical predicts
A ball with less energy than a hill is high can never reach the other side. It always rolls back. Transmission is exactly zero for \(E<V_0\).
What actually happens
Quantum particles routinely appear on the far side of barriers they cannot classically surmount — alpha decay, the Sun's fusion, the scanning tunnelling microscope, flash memory.
Quantum idea forced
The wavefunction doesn't stop at the wall; it decays as \(e^{-\kappa x}\) inside and re-emerges with reduced amplitude. A finite transmission \(T\). Gamow 1928.
Transmission T = 0.6%
Exact rectangular barrier: \(T=\Big[1+\dfrac{V_0^2\sinh^2(\kappa L)}{4E(V_0-E)}\Big]^{-1}\), \(\kappa=\sqrt{2m(V_0-E)}/\hbar\). The transmitted packet's area is \(T\); shrink \(L\) and \(T\) climbs exponentially — the physics that makes an STM image single atoms.
CRISIS VI

The Beam That Split in Two

TIER A → C
Classical predicts
Atomic magnets enter the field pointing every which way, so a magnetic-field gradient should deflect them by a continuous range of angles — a single smeared vertical band on the screen.
What actually happens
The beam splits into exactly two sharp spots — up and down — with nothing in between. Angular momentum along the measured axis is quantized.
Quantum idea forced
The electron carries intrinsic spin-½ with only two projections, \(m_s=\pm\tfrac12\). Measurement yields an eigenvalue, never the classical continuum. Stern–Gerlach 1922; spin, Uhlenbeck–Goudsmit 1925.
up: 0 · down: 0
quantum: two spots
Deflection \(\propto \mu_z\,\partial B/\partial z\). Classically \(\mu_z\) ranges continuously over \([-\mu,\mu]\); quantum-mechanically \(\mu_z=-g\mu_B m_s\) takes only two values. The Bloch-sphere state (Tier C) lives in Hilbert space, not on the screen — the screen only ever shows the two eigen-outcomes.
CRISIS VII

Why Matter Takes Up Space

TIER A → B
Classical predicts
Electrons should all settle into the single lowest-energy state. Every atom would be a tiny inert blob of the same size, there'd be no periodic table, no chemistry, and matter could be squeezed to nothing.
What actually happens
Electrons stack into shells (2, 8, 8, 18…), giving every element its distinct chemistry, giving atoms their size, and giving solids and white-dwarf stars a pressure that resists crushing.
Quantum idea forced
No two identical fermions share a state: the antisymmetric wavefunction vanishes if they do. Electrons are forced up the ladder. Pauli 1925; spin–statistics, Pauli 1940.
neon · 1s² 2s² 2p⁶
Antisymmetry: \(\Psi=\tfrac{1}{\sqrt2}[\psi_a(1)\psi_b(2)-\psi_b(1)\psi_a(2)]\to 0\) if \(a=b\). Each orbital holds two electrons (spin ↑↓); filling 1s²2s²2p⁶… is the periodic table. Switch exclusion off and every electron collapses to the ground level — a universe with no structure.

The pattern behind all seven

Read together, the crises tell one story. Classical physics fails wherever action approaches the scale of \(h\) — the quantum of action, \(6.6\times10^{-34}\) J·s. When the relevant energies, sizes, and times are large compared to \(h\), the quantum corrections are invisible and Newton and Maxwell are superb. When they shrink to atomic scale, the same equations predict infinities, instant death, and impossible zeros. Every fix in this page is the same move seen from a different angle: quantities that classical physics treats as continuous — energy, angular momentum, occupation — are actually discrete, and objects that it treats as either wave or particle are actually both. That is the whole of what changed in 1925, and everything downstream — the laser, the transistor, the chemical bond, the stability of the star you're orbiting — is a consequence.

Boundary of validity

Blackbody

Planck and Rayleigh–Jeans curves are the exact closed forms, plotted directly. Wien peak and the RJ/Planck ratio are computed live from the constants.

✓ exact formulas
Photoelectric

\(K_{\max}=h\nu-\phi\) with sodium's real work function; ejection speed \(\propto\sqrt{K}\). Photon arrival is illustrative animation, not a QED amplitude.

✓ Einstein relation
Collapsing atom

Classical fall time is the Larmor result (1.56×10⁻¹¹ s). The quantum cloud is the true \(|\psi_{1s}|^2\) radial density. Timescale mapped to seconds for viewing.

✓ Larmor + |ψ|²
Double slit

Dots are sampled from the exact two-slit intensity (Born rule). It is a screen-distribution model, not a 2-D TDSE — that full solve lives in the Duality entry.

△ distribution model
Tunnelling

Transmission \(T(E)\) is the exact rectangular-barrier formula. The moving packet is an envelope whose reflected/transmitted areas are set to R and T — not a live split-step (see Quantum Wells for that).

△ exact T, envelope packet
Stern–Gerlach

Two outcomes with the correct ±deflection; classical foil is the continuous \(\mu_z\) band. Spot scatter is illustrative. The Bloch state is Tier C, deliberately not drawn on the screen.

✓ eigen-outcomes
Pauli

Real subshell capacities (2, 2, 6, 2, 6…) and build-up order; configurations match the periodic table for Z ≤ 18. Level energies are schematic (ordering exact, spacing not to scale).

✓ configurations exact

References

R.A.V.E.N. Research Labs · First Principles · "Where Classical Physics Breaks" — the seven experiments that forced the quantum, each classical foil computed and drawn as the failing prediction. Companion to the Astrophysics & Equations Quantum Mechanics Roadmap.  ·  teaching-grade for the classroom, research-grade under the hood.