Skip to content

12.4

Antiparticles and the negative-energy sea

With no floor to the spectrum, what keeps atoms stable? Dirac's answer was to fill the vacuum: the negative-energy sea. A hole in the sea is a positron — written on paper in 1931, found in a cloud chamber in 1932.

Recommended first

After this section you should be able to

  • Explain why the negative-energy solutions can neither be deleted nor left alone: a bottomless spectrum would collapse every atom within nanoseconds
  • Lay out the Dirac-sea picture: how Pauli exclusion blocks the downward cascade, and how a hole behaves as a positively charged particle
  • Recount the prediction of the positron and Anderson's cloud-chamber discovery, and give a modern application of annihilation (PET)
  • Point out the three hard flaws of the Dirac sea and the modern field-theory (Feynman-Stückelberg) reading

At the close of last section, the Dirac equation had won the probability and lost the spectrum: every momentum carries a negative-energy solution E=p2c2+m2c4E=-\sqrt{p^2c^2+m^2c^4}, with levels running from mc2-mc^2 all the way down to negative infinity. This section tells the story of the boldest patch in the history of physics — the patch itself was later dismantled, but what it predicted stayed, and collected a Nobel Prize.

How concrete the disaster is

First compute the consequences of a “bottomless spectrum”, to appreciate how lethal the problem is.

In a quantum system, excited states decay to lower levels and emit photons — hydrogen’s 2p1s2p\to1s takes only about 1.6 ns. Now the Dirac equation says: there are levels below 1s1s, below mc2-mc^2, and no lowest rung. Then nothing stops the cascade:

e(1s)    e(E<mc2)+γ    e(Emc2)+γ+γ    (12.4.1)e^-(1s)\;\to\;e^-(E<-mc^2)+\gamma\;\to\;e^-(E\ll-mc^2)+\gamma+\gamma\;\to\;\cdots\tag{12.4.1}

Each rung down releases another photon; the energy falls ever more negative, the photons multiply without end. Estimates put the rate of this “downward radiation” at nanosecond scale — no stable atom should exist anywhere in the universe, and you and I should at this moment be a puff of gamma rays. This is worse than KG’s negative probability: negative probability merely defied interpretation, while a bottomless spectrum collides head-on with the observed fact that things exist.

Can we just delete the negative-energy solutions? No — same reason as last section: incompleteness. Worse, in an external field the transition matrix elements between positive- and negative-energy states are non-zero — the dynamics itself will ferry particles into the states you deleted.

Dirac turns the tables: fill the vacuum

In 1930 Dirac proposed a scheme that still reads as breathtakingly audacious.

From paper to cloud chamber

The birth of this prediction was not dignified. Dirac at first did not dare invent a new particle and proposed that the hole “is the proton” — after all, the proton was then the only known positive particle. The idea was swiftly executed by mathematics: Weyl proved the hole’s mass must exactly equal the electron’s, and the proton is 1836 times heavier; besides, “the electron in a hydrogen atom falling into its own proton-hole” would mean hydrogen atoms annihilate spontaneously. In 1931 Dirac accepted his fate and wrote it down in black and white: there should exist a new particle “as yet unobserved by experimental physics, with the same mass as the electron and opposite charge”.

In 1932, Carl Anderson at Caltech saw a track in a cosmic-ray cloud-chamber photograph: its curvature in the magnetic field showed positive charge, but the ionisation density and radius of curvature showed a mass far below the proton’s — consistent with an electron. To pin down the particle’s direction of flight he placed a 6 mm lead plate across the middle of the chamber: after crossing the plate the particle lost energy and curved more sharply, fixing the direction — and with it the sign of the charge, beyond dispute. The positron had arrived — the first time humanity found a particle in an equation before finding it in nature. Dirac received the Nobel Prize in 1933, Anderson in 1936.

Dismantling the scaffolding: the modern view

The Dirac sea did great service, but it is scaffolding, not the building. Three hard flaws:

  1. An infinite background. The filled sea carries infinite negative energy and infinite charge density, waved away with “a uniform background cannot be observed”; gravity does not accept the excuse — energy gravitates however uniform it is.
  2. What about bosons? The KG equation has negative-energy solutions too, but spin-0 particles do not obey Pauli exclusion, so “fill the sea” fails in principle. Yet pions have antiparticles all the same — clearly the explanation missed the root.
  3. Asymmetric bookkeeping. The electron is a “particle” and the positron a “hole”, yet in experiment the two are perfectly symmetric in status.

What comes next

Spin emerged from the algebra (12.3), antiparticles from the negative-energy solutions (this section) — and the Dirac equation is not done giving gifts. The last one is also the most precisely verified by experiment: the electron’s magnetic moment. Put the Dirac equation in a magnetic field, take the non-relativistic limit, and g=2g=2 drops out automatically; expand one order further and the fine-structure trio hand-stitched in section 6.6 (relativistic correction, spin-orbit coupling, the Darwin term) appears in one stroke, all by itself.

Section 88 of 106 · use to turn the page