← quantum · contents

§ Phase ix · Pauli & The Periodic Table · Chapter 01

Pauli exclusion, revisited

In the winter of 1924, a 24-year-old Austrian physicist with a sharp tongue and a thicker accent stared at the spectrum of helium and concluded that nature was running a peculiar accounting rule. Two electrons cannot share the same complete address. He wrote it down without a derivation, almost apologetically. Sixteen years later he supplied the theorem behind it, and in the meantime that single accounting rule turned out to hold up the periodic table, the rigidity of solids, and the corpses of dead stars.

In the autumn of 1924, Wolfgang Pauli was twenty-four years old, the youngest privatdozent at the University of Hamburg, and already famous for two things: a Handbuch article on relativity that Einstein had called “no work done by an adult could have surpassed it,” and a habit of telling distinguished professors that their lectures were rubbish. He smoked. He kept the hours of a cabaret performer. He arrived at colloquia, listened for two minutes, announced “this is not even wrong,” and went home. The Bohr model of the atom was eleven years old and crumbling. The spectroscopists of Tübingen and Zürich had built the most precise instruments ever turned on a glowing gas, and the lines they were measuring refused to fit the rules Bohr and Sommerfeld had handed them. Pauli was reading their tables every night.

The puzzle, in plain language, was this. Heat hydrogen and it glows with a set of sharp lines whose colors Niels Bohr had derived in 1913 by a beautifully unscientific argument involving electrons that moved only on certain circular orbits. The rule worked for hydrogen. For everything heavier, the simple Bohr picture failed in messy ways. Alkali metals, lithium and sodium and the rest, showed pairs of lines where one was expected. Magnetic fields split lines in patterns the old quantum theory could not explain. Helium, the second-simplest atom in the universe, refused to be calculated at all. Bohr had been working on a “building-up” principle, an Aufbauprinzip in German, that imagined the atom assembling itself one electron at a time as the nuclear charge climbed. The picture fit the broad pattern of the periodic table well enough that nobody could dismiss it, but the specific assignments of electrons to shells were guesswork. Why exactly two electrons in the innermost shell of every atom? Why exactly eight in the next? Bohr called the rule “an a priori postulate” and changed the subject.

Pauli refused to change the subject. He had grown up in Vienna in the circle of Ernst Mach, the philosopher-physicist who insisted that a scientific concept that could not be measured had no business in physics. The “shells” of the Bohr-Sommerfeld atom were a piece of bookkeeping, not a measurement. If two electrons fit into a shell and not three, there had to be a reason in the formalism, not in the storytelling. Through the winter of 1924 he stared at the term diagrams of the alkali metals, the patterns of fine and hyperfine splittings, the “anomalous” Zeeman effect that nobody could compute, and slowly convinced himself that an electron in an atom needed not three but four quantum numbers to specify its state. The fourth was new and two-valued. He did not yet know what it physically was. He called it a “classically non-describable two-valuedness” and refused, in print, to give it a friendlier name. Within a year, two graduate students in Leiden, George Uhlenbeck and Samuel Goudsmit, would identify Pauli’s mysterious fourth label as the intrinsic angular momentum of the electron, what we now call spin.

With the four quantum numbers in hand, Pauli formulated the rule. It is one sentence. No two electrons in an atom can have all four quantum numbers the same. Each electron occupies one of the (n, ℓ, m_ℓ, m_s) addresses; once an address is filled, the next electron must take a different one. Count the addresses in the innermost shell: n equals 1 forces ℓ equals 0, which forces m_ℓ equals 0, leaving two spin states. Two addresses. Two electrons. That is why helium completes a shell. Count the addresses in the next shell: n equals 2 allows ℓ equals 0 or 1, which adds one s-orbital and three p-orbitals, four spatial states, eight addresses with spin. That is why neon completes a shell. The row lengths Mendeleev had drawn on a deck of cards in 1869, the noble gases that ended each period like a row of full stops, all of it had been waiting for an Austrian to count.

Pauli published in the Zeitschrift für Physik in March 1925. The paper is one of the shortest pieces of foundational physics in the literature. It contains no derivation, only a postulate. He wrote, in his own German prose: “there can never be two or more equivalent electrons in the atom for which, in strong fields, the values of all quantum numbers … coincide. If an electron in the atom is characterized by the quantum numbers … then this state is ‘occupied’.” He called it the Ausschliessungsprinzip, the principle of exclusion. He offered no mechanism. Bohr, reading the paper in Copenhagen, wrote back that the rule was “an extraordinarily important step forward,” and that it was “unbelievable that nature should be so peculiar.” Nature was so peculiar. The rule fit every measured term diagram. It worked.

It was a clean rule, but it lacked a reason. Pauli himself was uncomfortable with this. A postulate that fits the data is a useful rule of thumb, but a postulate without a derivation is the sort of thing nature can revoke without warning. Why should two electrons refuse to share an address? The answer arrived, in installments, from three different directions, between 1926 and 1940. The first installment was an observation by Heisenberg and Dirac, working independently, in the summer of 1926. They wrote down what the two-electron wavefunction ought to look like. Call the two electrons “particle 1” and “particle 2,” and let a and b label two single-particle states the electrons might occupy. A naive guess would write ψ(r₁, r₂) = ψ_a(r₁) ψ_b(r₂): particle 1 is in state a, particle 2 is in state b. But this is wrong, because particles 1 and 2 are identical. There is no measurement, in principle or in practice, that distinguishes them. The wavefunction must not depend on which label we put on which particle.

That leaves two possibilities. Either the wavefunction is symmetric under exchange, ψ(r₁, r₂) equals plus ψ(r₂, r₁), in which case particles 1 and 2 are interchangeable in a friendly way that lets them pile into the same state. Or the wavefunction is antisymmetric, ψ(r₁, r₂) equals minus ψ(r₂, r₁), which costs nothing as long as the two particles are in different states but produces a magical zero the moment they try to occupy the same one. Try it: set r₁ equals r₂ in an antisymmetric wavefunction. You get ψ(r, r) equals minus ψ(r, r), which is only consistent with ψ(r, r) equals zero. The probability of finding two identical particles in the same place, in the same state, vanishes. The Pauli principle is not an extra postulate. It is what antisymmetry means for a two-particle wavefunction.

So which particles are antisymmetric and which are symmetric? Through the late 1920s the answer accumulated empirically. Electrons, protons, neutrons: antisymmetric. Photons, helium-4 nuclei, the still-undiscovered pions: symmetric. The two families came to be called fermions after Enrico Fermi, who had used the antisymmetric statistics in 1926 to model an electron gas, and bosons after Satyendra Nath Bose, the Indian physicist whose 1924 letter to Einstein had introduced the symmetric statistics for photons. Fermi and Dirac shared the antisymmetric credit (their Fermi-Dirac statistics is the name in the textbooks); Bose and Einstein shared the symmetric (Bose-Einstein statistics). The empirical pattern was striking: every particle with half-integer spin (one half, three halves) turned out to be a fermion. Every particle with integer spin (zero, one, two) turned out to be a boson. Spin and statistics were linked. No exception had ever been observed. But “no exception observed” is not a theorem, and Pauli, ever the Machian, refused to leave a connection that important to coincidence.

In 1940, at the Institute for Advanced Study in Princeton, Pauli supplied the theorem. The argument is, even now, one of the deepest results in quantum field theory; the modern presentation runs to a dozen pages of relativistic algebra. The substance, in plain English, is this. Any field theory consistent with special relativity and with the requirement that probabilities be positive forces a precise tie between the spin of a particle and the symmetry of its multi-particle wavefunction. Half-integer spin forces antisymmetry. Integer spin forces symmetry. The two options of 1926 were not arbitrary; they were the only two consistent choices, and which one applies to a given particle is determined by its spin. The result is the spin-statistics theorem, and it is what makes the exclusion principle a law instead of an observation. Pauli’s 1940 paper closes with characteristic dryness: “we want to mention that an interpretation of our result which attempts to claim the principle as something to be admired but not understood is, in our opinion, untenable.” Translation: stop being mystical about it. The proof is on the previous page.

The first time most students meet Pauli exclusion, it sounds like a piece of bookkeeping about atoms. It is much more than that. Strip away the chemistry and what remains is a rule about how many fermions can be packed into a region of space at a given energy. In a single atom that rule produces the periodic table. In a block of metal that same rule produces a Fermi sea, a sloshing collection of conduction electrons stacked one to a state up to an energy ceiling that depends only on density. In the cooling core of a sunlike star the rule produces electron degeneracy pressure, a stiffness in the gas that has nothing to do with thermal motion and everything to do with the fact that the bottom of the energy ladder is full and the next electron has nowhere cheap to go. Chandrasekhar in 1931 used exactly this argument to show that a white dwarf above 1.4 solar masses must collapse, because at some point even degeneracy pressure runs out. Below that mass the dwarf sits there, glittering against gravity, held up by Pauli. A neutron star, a kilometer-wide ball of nuclear matter, is held up by the same principle applied to neutrons: nucleon degeneracy pressure, a force whose only origin is the antisymmetry of fermion wavefunctions.

ψ(r₁, r₂) = ψ_a(r₁) · ψ_b(r₂)

ψ_F(r₁, r₂) = (1/√2) [ ψ_a(r₁) ψ_b(r₂) − ψ_a(r₂) ψ_b(r₁) ]

ψ_F → (1/√2) [ ψ_a(r₁) ψ_a(r₂) − ψ_a(r₂) ψ_a(r₁) ] = 0

It is worth pausing on the sheer reach of one short sentence. Without Pauli exclusion, every electron in every atom would slump into its 1s ground state, a tiny puddle of charge a fraction of an angstrom across, and atoms would all look chemically identical, like featureless dense crumbs. Chemistry would not exist as a science of distinctions. With Pauli exclusion, electrons stack into a tower of orbitals whose top floors stick out in different shapes depending on how full the lower floors are, and the chemical personality of an atom (its color, its reactivity, its bonding habits) is exactly the shape of that top floor. The periodic table is the cross-section of the tower. Without Pauli exclusion the white dwarf at the heart of the Sun’s eventual remnant would collapse to a point; with Pauli exclusion it sits there forever, glowing dimly. Without Pauli exclusion the neutrons in a neutron star would have nothing to push back against gravity, and pulsars would be black holes; with Pauli exclusion they spin for a billion years.

We will spend the next few chapters cashing this in. The Hartree-Fock method takes the Slater determinant of this chapter and turns it into a working calculation, the workhorse of computational chemistry from 1930 to the present. The full machinery of multi-electron atoms (which orbitals fill in which order, why the transition metals have the chemistry they do, why iron is magnetic and copper is not) all rests on the antisymmetry we have just derived. Pauli is the lock; spin is the key it turns; the periodic table is the door it opens. The same Wolfgang Pauli will show up again in this book wearing different hats: as the inventor of the Pauli matrices, the two-by-two algebra of spin we met in Phase iv; as the man who proposed the neutrino in 1930 to save energy conservation in beta decay; as the proposer of the spin-statistics theorem itself, the deep reason that integer spin pairs with bosons and half-integer with fermions. One physicist, four immortal contributions, all of them following from his refusal, as a twenty-four-year-old in Hamburg, to accept that nature could be peculiar without being calculable.