jump to navigation

Radiation zero or anomalous couplings ? March 21, 2007

Posted by dorigo in news, physics, science.

Continuing my crusade against particle physics lingo, this time I come to talk about “radiation zero”. Radiation zero is a nickname for a dip in the angular distribution of photons produced together with W bosons in the center-of-mass of proton-antiproton collisions.

…Ok, ok, give me another chance to explain it.

W boson production is by itself an interesting process in the collisions we study at the Tevatron with the CDF and D0 detectors – but by now, we have collected more than a million events of that kind. So, while we still use those large datasets to measure as precisely as we can the W mass and its angular distribution in the detector (the latter is actually quite interesting since it is a tale-telling hint at the detailed inner structure of the proton) we also move to study fancier processes, which – being more complex – have interesting features we can only see with the large datasets we now have.

One such process is the production of a W in association with an energetic photon. You smash a proton with an antiproton, and what do you get ? A W boson together with a single quantum of light; but not ordinary light! One photon which is a thousand billion times more energetic than those emitted from a light bulb.  

W-gamma production (gamma is a high-energy photon in the language of nuclear and subnuclear physics) occurs through one of the Feynman diagrams shown below. Think of it as either of the two following mechanisms:

  1. one quark emits a energetic photon, and then annihilates with a colleague, producing a W boson (see two graphs on top of figure);
  2. two quarks annihilate producing a W boson, and the latter emits a energetic photon (see graph on bottom of figure).


Ok, you may say, that indeed is a fancy process – but what is radiation zero, anyway ? Well, radiation zero is a region at a particular angle with respect to the incoming projectiles where W bosons are not found radiating photons. In other words, the process of creation of a W-gamma pair is suppressed for a particular direction of flight of the W.

The process responsible for turning off the light around the W is called destructive interference: the local merging of two ondulatory phenomena whose amplitudes meet with opposite signs, zeroing the total intensity.

Quantum interference is cool, but there are several places more suited to observe it than in high-energy hadron collisions. There, it is usually hard to put it in evidence, because of the complexity of signal extraction. So when we can do it, we are justly proud about it. 

Now, how do we actually do it ? Right, because the flight direction of a W is not exactly something we can measure! W bosons can be seen only when they decay to a lepton-neutrino pair. And the neutrino escapes our detectors unseen, so we only have the lepton to measure – but one particle alone does not tell us where the originating body (the W) was going.

We content ourselves to measure the angle between the lepton and the photon. The distribution of that angle retains some memory of the initial depletion of W emission in the “radiation zero” direction, and it also shows a dip. One like that found by D0 last fall:

In the plot, you see the data (points with error bars) plotted as a function of the angle between lepton and photon. It does show a dip, despite the background contamination (estimated in grey)! The W-gamma content predicted is in white, and the blue zones highlight the uncertainty in the total event rate in each bin.

That is cool. Even more cool, in my opinion, is to know that while the Standard Model predicts the dip at zero (actually, -0.3) in the plot, a anomaly in the strength with which W and photon interact (the “coupling”) would soften the dip. So we can look at the plot, and figure out whether such “anomalous couplings” exist or not!

That is what D0 does in this other plot shown below: after subtracting the expected background content bin by bin, the data is shown compared to the SM (in blue) and to the SM with just enough anomalous couplings added to make the rate best agree with observation. 

The red histogram is marginally in better agreement with the data than the Standard Model alone. Is that an indication that anomalous couplings are the cause ? No. The SM alone is a good model of the data. Moreover, one cannot really talk of a better agreement with SM+A.C., because the normalization of the latter has been tuned to best fit the data. The chosen model for anomalous couplings would instead predict a much larger overall rate of W-gamma events observed in the amount of collisions analyzed by D0.

All in all, a neat piece of analysis.

%d bloggers like this: