← quantum · contents

§ Phase xii · Quarks & Hadrons · Chapter 01

Color confinement

In 1964 a brash, polylingual prodigy at Caltech wrote down a model of protons and neutrons that worked too well to be true, then spent decades insisting the parts he had invented were not real. The parts were quarks. They are real. And the universe enforces a strange rule about them: you are never, ever allowed to hold one alone.

By the early 1960s, particle physics had a politely catastrophic problem. The accelerator labs at Berkeley, Brookhaven, and CERN had spent a decade smashing protons into other protons and counting the wreckage. The wreckage was a zoo. Pions, kaons, lambdas, sigmas, xis, omegas, rhos, etas, an alphabet soup of short-lived particles that decayed in microseconds or nanoseconds and left a tangle of bubble-chamber tracks behind them. By 1962 the count was over a hundred. Robert Oppenheimer joked that the Nobel Prize in physics should go to the physicist who did NOT discover a new particle that year. None of these things looked elementary. They could not all be on the same level as the electron. Yet no one could see what they were made of.

Into this mess walked a Caltech professor named Murray Gell-Mann, who had already, at the age of 32, given the field two of its most useful organizing tricks: strangeness (a new conserved quantum number that explained why kaons hung around for nanoseconds instead of decaying in 10⁻²³ seconds) and the Eightfold Way (a grouping of the particle zoo into octets and decuplets based on an SU(3) symmetry he had borrowed from mathematics). The Eightfold Way had even predicted a missing particle, the omega-minus, before it was found in a bubble chamber at Brookhaven in 1964. Gell-Mann was, by then, the brightest theorist of his generation, a man who read fourteen languages, corrected his colleagues’ Greek etymologies, and once, when introduced to the queen of Belgium, complimented her on the regional dialect she happened to be speaking. He was insufferable. He was also usually right.

What Gell-Mann did in early 1964 was the natural next step. If the hundred-odd hadrons fell into neat SU(3) patterns, it was because they were built out of three more fundamental things. He worked out the rules. The building blocks would have to carry electric charge of +2/3 or -1/3, fractions never before seen in nature. Three of them, glued together, would make a proton or a neutron or any of the other baryons. One of them paired with an antiparticle of one of them would make a meson. The model accounted for every hadron in the zoo, predicted the missing ones, got the masses approximately right, and unified the whole spectrum under one symmetry. Gell-Mann wrote it up in a two-page paper and, casting around for a name, lifted a nonsense word from a line of James Joyce’s Finnegans Wake: “Three quarks for Muster Mark.” Quarks.

Gell-Mann did not, in 1964, believe his own model. He told colleagues that quarks were a “mathematical convenience,” a useful bookkeeping device to keep the SU(3) algebra straight, and that they probably did not exist as physical particles. There were good reasons for the skepticism. Free quarks should have been easy to find. A fractional electric charge is unmistakable in any detector built for integer-charged particles; a +2/3 quark in a Millikan oil-drop experiment would announce itself by sitting between the steps. Cosmic-ray searches turned up nothing. Accelerator searches turned up nothing. Oyster-shell searches (yes, this happened) turned up nothing. By the late 1960s, no one had ever caught a fractional-charge particle anywhere, and the assumption that quarks were real things you could in principle separate had begun to look untenable.

Then came the experiment that changed everything. In 1968 a Stanford team led by Jerome Friedman, Henry Kendall, and Richard Taylor fired very energetic electrons (about 20 GeV) at protons at the SLAC accelerator and measured how they scattered. The experiment was the high-energy analog of Rutherford’s gold foil. If the proton were a smooth blob of charge, the electrons would deflect gently. They did not. They bounced off something hard. The angular distribution of the scattered electrons indicated that the proton contained pointlike charged constituents, three of them, of fractional charge, behaving as if they were nearly free. James Bjorken and Richard Feynman analyzed the data and announced, in Feynman’s deliberately neutral language, that the proton was built out of “partons.” The partons looked exactly like Gell-Mann’s quarks. The Nobel Committee gave Gell-Mann the 1969 Prize in physics for the model and gave Friedman, Kendall, and Taylor the 1990 Prize for confirming it.

There was, though, a glaring contradiction. Inside the proton, the quarks behaved as if they barely felt each other at all. Outside the proton, they had never been seen. A force that lets particles roam free at short range but somehow forbids their separation at long range is unlike any other force in nature. Gravity, electromagnetism, and the weak force all weaken with distance. This new force did the opposite. It got stronger the harder you tried to separate the quarks, and the energy cost of pulling them apart was, as far as anyone could tell, infinite. The principle had a name, color confinement, before there was any theory to derive it from. It was a property of the universe that had been measured but not understood.

The theory that explained both halves of the puzzle was built between 1972 and 1973 by a small, intense group of theorists. Gell-Mann himself, in collaboration with Harald Fritzsch and Heinrich Leutwyler, proposed quantum chromodynamics as the gauge theory of three colors. The idea was that each quark carries one of three color charges (red, green, or blue), each antiquark carries an anti-color, and the force between them is mediated by eight massless particles called gluons, which themselves carry color and so can interact with each other. That self-interaction is the key. A photon does not carry electric charge and so does not couple to other photons; QED is, in a deep sense, polite. A gluon carries color charge and couples to other gluons; QCD is rude. The gluon-gluon interactions tangle the field, pull the lines of force into a narrow tube between any two color charges, and refuse to let the tube spread out the way an electric field does.

In the same months that Gell-Mann’s group was sketching the theory, two young Princeton theorists named David Gross and Frank Wilczek, working with the graduate student method of “lots of paper and not much sleep,” computed what QCD does at short distances. So did Hugh David Politzer, a graduate student at Harvard. They found, independently and almost simultaneously in the spring of 1973, a result that made everyone sit up. The QCD coupling constant, the number that sets the strength of the strong force, runs in the opposite direction from QED’s. In QED, a charge looks slightly stronger as you probe closer to it. In QCD, a color charge looks weaker. At very short distances and high energies, quarks behave almost as if they were free particles. The phenomenon was named asymptotic freedom, and it explained, at a stroke, why the SLAC electrons had seen pointlike quarks inside the proton. The Stanford team had probed deep enough that the strong force had effectively turned itself off. Gross, Wilczek, and Politzer shared the 2004 Nobel Prize for the discovery.

So here, by mid-1973, was the picture. Quarks carry color. Color comes in three values. The only allowed combinations are colorless ones: three quarks together with one of each color (a baryon), or a quark with an antiquark of matching anti-color (a meson). At short distances the color force fades and quarks roam almost free inside the hadron. At long distances the force does not fade. It grows. And this is where the genuinely strange behavior begins, the behavior that gives this chapter its title.

Imagine you grab a quark inside a proton and pull. The color force between it and its partners pulls back, and the pull does not weaken. The lines of color field that connect the quark to its mates do not fan out into space the way electric field lines do; the gluon self-interactions hold them together into a narrow tube, sometimes called a flux tube or a QCD string. Pull the quark, and you stretch the string. The energy stored in the string grows linearly with the stretching distance, at roughly one GeV of energy per femtometer of separation. That is a colossal force at the subatomic scale, about fourteen tons of force in human units, acting on something the size of a proton. Stretch the string far enough, and you have stored enough energy in it that the vacuum can do something dramatic. From the quantum vacuum, a new quark-antiquark pair pops into existence. The new quark grabs the original’s color partner. The new antiquark grabs the original. The string snaps in two. Where there was one over-stretched hadron, you now have two normal-length hadrons, each color-neutral, each placid. You will never get a free quark this way. You will only ever get more hadrons.

V(r) = -4/3 · α_s / r  +  σ · r

The picture that emerges is profoundly strange and profoundly clean. The vacuum of QCD, the supposedly empty space inside any hadron, is in fact a turbulent soup of virtual quark-antiquark pairs and gluon fluctuations. The “empty” interior of a proton contains, on any given snapshot, far more than three quarks; it contains a sea of momentary pairs popping in and out, plus a coordinating fog of gluons. The three quarks that we use to label the proton (the two ups and one down of the static picture) are properly called valence quarks, and they account for only a small fraction of the proton’s mass. About 99% of the proton’s mass is QCD binding energy: the quarks themselves contribute only the last 1%. Most of you is glue. Most of any object you have ever weighed, when you weigh atoms and nuclei and add it all up, is the energy of the QCD vacuum holding quarks together. E = mc² has rarely had a more thoroughly literal application.

And what about ordinary nuclear physics, the protons and neutrons sticking together inside an atom, the binding energy that Hans Bethe used to power his sun? That is not the strong force in its primary form. Protons and neutrons are color-neutral hadrons; from the outside, they have no color charge to attract one another with. What they have is a faint residue, in the same way that two electrically neutral atoms still feel a faint van der Waals attraction because their internal charges fluctuate. Nuclear physics is the van der Waals tail of QCD: a small, range-limited residual attraction between objects that are, in the deep accounting, color-neutral. The cleanest demonstration of this is the pion exchange model of nuclear binding, originally proposed by Hideki Yukawa in the 1930s long before quarks existed. Yukawa was, retrospectively, modeling the residue of a force he had no way to know about. The full theory took another forty years to write down.

It is worth pausing on what confinement, as a fact about the universe, actually means. The electron is a particle in the deep sense: you can isolate it, accelerate it, hold it in a Penning trap, measure its magnetic moment to fourteen decimal places. The quark is something else. It is a real degree of freedom in the QCD Lagrangian. It carries momentum and energy. It scatters off other quarks at short distances exactly as theory predicts. But it is never, anywhere in the universe, available as a free particle. Try to make one alone and the vacuum reaches in and gives you a hadron instead. The strong force does not let go. We have been doing high-energy experiments for sixty years, including the highest-energy collisions ever attempted by humans at the Large Hadron Collider, and not one isolated quark has ever been registered in a detector. The rule is exact. It is enforced by the vacuum itself.

The chapter that follows will catalog the hadrons that confinement does allow: the mesons (quark-antiquark pairs) and the baryons (three-quark bundles), the protons and neutrons that build all ordinary matter, and the more exotic states that flicker briefly through accelerator data. But the central fact will not change. Quarks come in colored triples, the colors must always cancel, and the universe will go to any length, including manufacturing fresh pairs of quarks out of nothing, to make sure you never catch one alone.