A crowd pours out of a narrow passage. Some run in a tight group in one direction, like jetty events in nuclear collisions. Others scatter evenly in all directions, like isotropic events. The shape of particle debris, which physicists call transverse sphericity, turned out to be a handy tool for hunting a rare phenomenon.
In collisions of lead nuclei at enormous energies, the chiral magnetic effect can arise: an ultrastrong magnetic field momentarily separates positive and negative charges along its axis. This effect is predicted by theory, but it's hard to pick out from background processes — the usual collective flow of particles and decays of short-lived resonances (particles that quickly decay into others). Previously, events were classified by the flow vector, but that method suffers from circularity: the classifier itself is contaminated by the same background it's supposed to suppress.
The authors simulated lead-nucleus collisions at an energy of 5.02 TeV using the AMPT transport model, including a realistic implementation of the chiral magnetic effect. They split events into jetty and isotropic based on transverse sphericity and compared the signal in the two classes. It turned out that in isotropic events the background contribution from elliptic flow and resonance decays is significantly weaker, and the signal-to-background ratio is higher. This means that selecting isotropic events is a cleaner way to search for the chiral magnetic effect.
This approach complements existing methods and could help confirm or refute the existence of the chiral magnetic effect in experiments. If the effect is reliably isolated, it would be an important step toward understanding the properties of the super-hot matter born in relativistic nuclear collisions.
The magnetic field in lead-nucleus collisions at the Large Hadron Collider can reach $10^{18}$ gauss, a billion times stronger than the most powerful magnets on Earth, but it lasts only a tiny fraction of a second.