The Quantum Leap as a Two-Level Rotation E = hf as the Beat of Two Clocks
Bohr's quantum leap has been a scandal since 1913: the electron changes level and emits a photon, but the formalism says nothing about what happens during the change. This note gives the leap a continuous, geometric picture — the picture the series always wanted — and gets it right by placing the rotation on the correct object. A single energy eigenstate does not, by itself, leap: its time evolution e^{−iEt/
Abstract
1The Scandal, and the Right Geometric Answer
In 1913 Bohr placed electrons on discrete levels and had them jump between, emitting light of frequency f = ΔE/h. The prediction was triumphant and the mechanism outrageous: where is the electron mid-jump? Schrödinger hated it — 'if all this damned quantum jumping were really here to stay, I should be sorry I ever got involved.' The honest geometric answer is that there is no instant jump; there is a continuous rotation. The only question is what rotates. The mistake to avoid is rotating a single level. The right object is the two-level state.
2Two Clocks
Give each level its quaternion label — energy on the real axis, angular momenta on the imaginary — and read its time evolution. An eigenstate of energy E evolves as
|ψ(t)⟩ = e^{−iEt/
a steady rotation of phase at rate ω = E/
3The Leap Is a Two-Level Rotation
A transition is a rotation in the two-dimensional subspace spanned by the levels it connects — the Bloch sphere of a two-state system — driven by the electromagnetic field (the Rabi rotation). During it the atom is a genuine superposition,
|Ψ(t)⟩ =
with the state point travelling from the north pole (pure 1s) toward the south (pure 2p) as α and β trade weight. The midpoint |α| = |β| is the Bloch equator — the 'transfigured' instant, neither level, a coherent both-at-once that no energy measurement can catch mid-flight. This is a real rotation of a real state, and it is where the leap actually lives.
4The Light Is the Beat
What radiates is the atom's electric dipole, and in the superposition it carries the relative phase of the two clocks: ⟨d⟩(t) ∝ Re[ α*β · ⟨1s| d |2p⟩ · e^{−i(E_2p − E_1s)t/
5The Duration Is Physical
Because the leap is a driven rotation of a real state, its timescale is set by dynamics, not geometry. Driven, the transition takes a Rabi period 2π/Ω_R, with the Rabi frequency Ω_R proportional to the dipole matrix element ⟨1s|d|2p⟩ times the field amplitude — a fast leap in a strong field, a slow one in a weak field. Free, the excited level decays with its spontaneous-emission lifetime — about 1.6 ns for 2p — fixed by the same matrix element through the Einstein A-coefficient. There is no parameter-free 'rotation time': the leap is as quick as the coupling makes it, and that is a measurable, standard quantity.
6What Is Geometric, What Is Standard
Geometric (claimed). The leap is a rotation — of the two-level state, on its Bloch sphere. E = hf is a geometric identity — the beat of two Compton clocks. The photon carries energy on the real axis and angular momentum on the imaginary, and mediates the transition by exchanging both. These are the series' own readings, and they are correct.
7The Sentence
References
N. Bohr, Phil. Mag. 26, 1 (1913); M. Planck (1900); the Rabi two-level model, the Einstein A/B coefficients, Fermi's golden rule and the Wigner–Eckart theorem (standard quantum optics and quantum mechanics); and the notes of this series (the Tick Count — mechanics as clock-counting; Schrödinger from the Geometry — the shutter/Compton clock; the State Quaternion label). Verification script: quantum_leap_check.py (stdlib). This edition supersedes the earlier draft that rotated a single state quaternion from +1 to −1. (Citations from memory; the literature-verification pass applies.) Acknowledgment: drafting and numerical checks by machine (Claude, Anthropic), in conversation; the instinct that the quantum leap is a continuous rotation and that E = hf is geometric is the author's, and both are here placed on their correct footing — the rotation on the two-level Bloch sphere, and E = hf as the beat of two clocks.
8Verification
The companion scripts, with their recorded output. Each script's docstring states what it establishes and what it does not; the Source tab shows the file itself, unedited.
quantum_leap_check.py — quantum_leap_check
(1) THE BUG: a single state's rotation is a global phase (same ray)
e^{-iEt/hbar} multiplies |psi> by a phase -> the ray |psi><psi| is unchanged
-> unobservable, not a transition. Q -> -Q (180 deg) is also the same ray.
So 'the quaternion rotates +1 -> -1' cannot model 1s -> 2p.
(2) THE FIX: the photon frequency is the BEAT of the two eigen-phases
dE = E2p - E1s = 10.204 eV (a DIFFERENCE, not a single eigenvalue)
nu = dE/h = 2.4674e+15 Hz
lambda = c/nu = 121.5 nm (Lyman-alpha, measured 121.567 nm)
Mechanism: the superposition a|1s> + b|2p> has a dipole
<d>(t) ~ Re[ a* b <1s|d|2p> e^{-i(E2p-E1s)t/hbar} ], oscillating at (E2-E1)/hbar;
that oscillating dipole radiates at the photon frequency. E = h f, as a beat.
(3) THE DURATION is physical, not geometric
- driven: the Rabi period 2pi/Omega_R, Omega_R proportional to the dipole
matrix element times the field amplitude;
- free: the 2p spontaneous-emission lifetime ~ 1.6 ns.
Neither is a parameter-free 'rotation time' read from omega = E/hbar.
SURVIVES: the transition IS a rotation -- on the Bloch sphere of {|1s>,|2p>} --
and E=hf IS geometric -- the beat of the two Compton-clock phase rotations of
the Tick-Count / Schroedinger notes. What fails is putting it on ONE quaternion
and reading a duration from a single-state rotation.
# -*- coding: utf-8 -*-
"""The Quantum Leap, diagnosed and repaired. Pure standard library.
THE BUG (Quaternion Quantum Leap paper): a transition is modelled as a rotation
of one state quaternion 'from one real eigenvalue to another', with angular
velocity omega = E/hbar identified as the photon frequency. But the time
evolution of a single eigenstate, e^{-iEt/hbar}, is a GLOBAL PHASE: it leaves the
physical ray invariant, so it is unobservable and is not a transition. A 180-deg
rotation of one quaternion is Q -> -Q, the SAME ray. So the '+1 -> -1' leap cannot
be a transition between two energy eigenstates, and omega = E/hbar of a single
state is not the photon frequency.
THE FIX (standard QM): a transition is a rotation in the 2-STATE subspace
{|1s>,|2p>} -- a Bloch-sphere / Rabi rotation driven by the field. The OBSERVABLE
(the radiating dipole) oscillates at the BEAT frequency (E2 - E1)/hbar, which IS
the photon frequency: E = h f, geometrically, as the beat of two phase-clocks.
The duration is the Rabi period (driven) or the spontaneous lifetime (free), set
by the dipole matrix element -- not a free geometric rotation time.
"""
h = 4.135667696e-15 # eV s
c = 2.99792458e8
E1s = -13.606
E2p = -13.606/4.0
print("(1) THE BUG: a single state's rotation is a global phase (same ray)")
print(" e^{-iEt/hbar} multiplies |psi> by a phase -> the ray |psi><psi| is unchanged")
print(" -> unobservable, not a transition. Q -> -Q (180 deg) is also the same ray.")
print(" So 'the quaternion rotates +1 -> -1' cannot model 1s -> 2p.\n")
print("(2) THE FIX: the photon frequency is the BEAT of the two eigen-phases")
dE = E2p - E1s
nu = dE/h
lam = c/nu*1e9
print(f" dE = E2p - E1s = {dE:.3f} eV (a DIFFERENCE, not a single eigenvalue)")
print(f" nu = dE/h = {nu:.4e} Hz")
print(f" lambda = c/nu = {lam:.1f} nm (Lyman-alpha, measured 121.567 nm)")
print(" Mechanism: the superposition a|1s> + b|2p> has a dipole")
print(" <d>(t) ~ Re[ a* b <1s|d|2p> e^{-i(E2p-E1s)t/hbar} ], oscillating at (E2-E1)/hbar;")
print(" that oscillating dipole radiates at the photon frequency. E = h f, as a beat.\n")
print("(3) THE DURATION is physical, not geometric")
print(" - driven: the Rabi period 2pi/Omega_R, Omega_R proportional to the dipole")
print(" matrix element times the field amplitude;")
print(" - free: the 2p spontaneous-emission lifetime ~ 1.6 ns.")
print(" Neither is a parameter-free 'rotation time' read from omega = E/hbar.\n")
print("SURVIVES: the transition IS a rotation -- on the Bloch sphere of {|1s>,|2p>} --")
print("and E=hf IS geometric -- the beat of the two Compton-clock phase rotations of")
print("the Tick-Count / Schroedinger notes. What fails is putting it on ONE quaternion")
print("and reading a duration from a single-state rotation.")