polarization oddity

If you apply an electric field to a chunk of metal, electrons will tend to migrate to one side, leaving it more negative and the opposite side positive--polarized. A proton is made of charge quarks (and neutral gluons and various virtual particles), which under an electric field should also, on the average, migrate to one side or the other. I say "on the average" because they don't stay put anywhere, but move quite rapidly.

A Science News article on "Protons might be stretchier than they should be." points to a Nature research paper about an anomaly in the polarizability they calculate for the proton.

They look at the reaction: electron hits a proton to produce an electron plus a proton plus a gamma ray. In short, e+p→e+p+γ, where the γ goes somewhere undetected. Because they measure the momentum of the electron and proton they can reconstruct the missing mass and select the events with missing mass near 0 to pick those events with a photon. There's a much larger set of events set where the third particle is something else (e.g. a pion). So far so good. They can predict the photon's momentum, and that quantity squared is handy for calculations--that's what you see in the plots in the second link.

A photon has (is) an electric and magnetic field and that field interacts with the proton. They claim that the interaction is higher than expected for a photon energy of about .59 GeV, or a q^2 of .35GeV^2. Everybody expects a smooth curve--they see a bump (as others did before). It's not as big a bump as the older experiments, but their measurements are a bit more accurate.

What does this mean? It could be that the proton has some unexpected structure, and at certain electric fields it reacts more strongly than at higher or lower fields. That sounds like some kind of resonant effect, and by an opportune pun there are particles called "resonances" that might be relevant.

For example, if you bounce a pion off a proton you'll get an interaction rate that varies smoothly with energy, until you get close to a center of mass energy of 1440 MeV/c^2. Then the interaction rate jumps, and the jump is due to creating a new short-lived particle.

You might expect something similar here, like p+e → N + e → p + γ + e. The unknown "N" would have a mass somewhere around the proton mass plus the photon momentum--and there's a resonance near that. However the cross section for the 1520 resonance decaying to p + γ is quite small (110MeV width * .004 branching fraction = 0.4MeV), which means the production rate is going to be quite small also, so it probably doesn't contribute much, unless I shouldn't have ignored the phase space factors.

Very odd. I wonder if the different groups use related monte-carlo programs.

UPDATE: If they looked at the reaction rate while requiring the missing mass to be that of a pi-zero, they might see a bump in the rate at that q-squared.