Loops (Quantum Field Theory)

When two electrons repel each other, they exchange particles of light in the process — so-called photons, the messenger particles of the electromagnetic force. The simplest picture shows just a single photon flying from one electron to the other. In reality, though, the process is richer: on its way, the photon can briefly split into a particle-antiparticle pair that immediately vanishes again. These short-lived detours form closed curves in the theory’s graphical representations — the so-called Feynman diagrams — which are precisely the loops. The more loops one includes in the calculation, the more accurately the result matches what is measured in experiment.

Loops as a Benchmark for Precision

Quantum field theory makes predictions by calculating a quantity as a sum of ever-smaller corrections. The first term of this sum is called the tree diagram or “tree level” — it contains no loops. Each further term adds an additional loop and refines the result. Physicists then speak of one-loop, two-loop, or three-loop order.

That sounds abstract, but it has very concrete consequences. The electron’s intrinsic magnetic spin — the so-called anomalous magnetic moment — is among the most precisely measured quantities in all of physics. The theoretical prediction agrees with experiment to more than ten decimal places. This agreement is only possible because physicists have calculated corrections up to five-loop order — a computational feat that required decades of human and machine effort.

Loops are thus not a flaw in the theory but its precision tool. Anyone who leaves them out gets a rough approximation. Anyone who includes them gets one of the most precise agreements between theory and experiment known to science.

Virtual Particles and the Divergence Problem

The particles inside a loop are virtual. This means: they exist only for an extremely short time and can in the process take on properties that would be forbidden for real particles — such as any arbitrary energy and momentum. Since one must sum over all possible energies, some loop integrals mathematically run off to infinity. For a long time, this was a serious problem.

The solution is called renormalization. The basic idea: one rewrites the infinite contributions as corrections to quantities that one doesn’t know directly from the theory anyway — namely, the mass and charge of the particles. These quantities are then taken from experiment. After this rewriting, the infinities disappear, and the remaining calculation yields finite, meaningful results. Richard Feynman, Julian Schwinger, and Sin-Itiro Tomonaga received the Nobel Prize in Physics in 1965 for this idea.

A common misconception is that renormalization is a trick to hide errors. In fact, it is mathematically rigorously justified and is now regarded as one of the deepest concepts in theoretical physics.

Loops in Research and Technology

Loops show up everywhere quantum field theory meets high precision. At CERN, the particle accelerator near Geneva, collisions of protons are analyzed. To extract anything meaningful from the measurement data, one needs theoretical predictions that include loop corrections — otherwise one cannot decide whether a deviation points to new physics or is simply a known loop correction that was overlooked.

Related concepts also play a role in the development of quantum computers. Error-correction methods in quantum information science borrow techniques from quantum field theory, including approaches derived directly from the treatment of loops.

In headlines, loops usually appear only indirectly — for instance, when there is talk of a “precision measurement at the LHC” or a “test of the Standard Model.” Behind this there is almost always an elaborate loop calculation that first makes the comparison between theory and experiment possible.