First measurement of the forward-backward charge asymmetry in top quark pair production

Introduction

Of the six quarks in the standard model, the top quark was the last to be discovered. It was first observed at Fermilab in 1995 by the CDF and D0 (DZero) collaborations. The top quark¹ is the most massive fundamental particle that we know. It is so massive that Fermilab's Tevatron  accelerator, which collides protons with anti protons, remains the only place on earth where it can be created and studied. Top quarks are usually produced in pairs: a quark (from the proton) and an anti quark (from the anti proton) annihilate and produce a top quark and an antitop quark through the strong interaction. The direction of the incident proton (anti proton) is called the forward (backward) direction. The forward-backward charge asymmetry is then simply a numerical answer to the question: is it the top quark or the antitop quark that is produced in a direction closer (in angle) to the forward direction, or in short, which one is “more forward”? The sources of the asymmetry in the standard model are subtle, yet there is no fundamental reason for this production to be charge-symmetric. And while it is difficult to calculate the size of the asymmetry as predicted by the standard model, all existing calculations predict that it is small, below 10%. Where an asymmetry of 0% means that the production is charge symmetric, and 100% means that it’s always the top quark that is more forward. The standard model also firmly predicts that the asymmetry should be smaller when additional particles are created.

¹The term “top quark” usually refers both to the top (positively charged) and to the antitop (negatively charged) quarks, with the obvious explicit exceptions when defining the charge asymmetry

Reconstructing top events

Since top quarks have very short lifetimes, they cannot be directly detected in the experimental apparatus. Rather, their particle remnants (products of top quark decays) leave signatures in the detector. In the standard model, a top quark decays almost exclusively to a b-quark and a W boson . The W is an unstable particle whose decay gives rise to several possible combinations of other daughter particles. In this analysis we focus on events in which one of the W bosons decays into two light quarks, and the other into two leptons. Specifically, we look for an electron (e) or a muon (μ) and its associated neutrino. These leptons leave clear signatures in our detectors, which are useful in singling out the very rare top production (only 1 in 10¹⁰ collisions yields top quarks). The tracking detectors are used to reconstruct the electric charge of the leptons, which tells us which of the b-quark and W boson combinations are from the top, and which is from the anti-top. The two b and two light quarks produce “jets” (remnants of quarks) which we use together with the charged lepton to fully reconstruct the decay. Having fully reconstructed the event, we can then determine which of its two top quarks was more forward.

Results

The figures below show the distribution in variable that discriminates between top-pair production and background processes. The left figure shows events where the top quark is more forward, and the right figure shows events where the antitop quark is more forward. The figures show the sample decomposition, with the top-pair signal shown in red.

Figure 1 - Events where the top quark is more forward than the antitop quark

Figure 2 - Events where the antitop quark is more forward than the top quark

We measure an asymmetry of (12±8)% in our data, which is consistent with the standard model prediction of (1±1)%. We also observe the trend predicted by the standard model, as we measure an asymmetry of (19±9)% when we do not identify any additional jets and an asymmetry of (-16±16)% when we do.

The hard part

Clearly, contributions to the asymmetry are not measured for those top events that fail our selection criteria. Furthermore, the selected events are not reconstructed perfectly, and so the classification of events into the two above plots will be wrong some of the time. As an extreme case consider a top quark and an antitop quark produced at the same angle relative to the incoming proton. It is then essentially arbitrary which one will be reconstructed as having the smaller angle (more forward). This analysis was designed to limit these difficulties, and includes a detailed characterization of the remaining effects.

Implications for new physics

Since the standard model predicts a small asymmetry, any deviation from that implies the presence of new physics. To demonstrate the procedure, we can set limits on the fraction of top pairs produced in the decay of a possible new Z’ resonance, instead of through the strong interaction as predicted by the standard model. Z’ resonances appear in many possible extensions of the standard model, and their decays may be analogous to the asymmetric decays of the well studied Z boson, which yields asymmetries at the 30% level. In this kind of situation, even a wide resonance that can not be identified as a narrow resonance in the top-antitop mass spectrum may be revealed through a contribution to top-pair charge asymmetry.

       Outlook

As the Tevatron experiments collect more data, the statistical uncertainties on the asymmetry will be greatly reduced. In fact, the statistics already in hand can yield a result twice as precise as this first measurement. Should a deviation from the standard model persist, it could become statistically significant. Should the deviation decrease, we will be able to place more stringent limits on top-pair production mechanisms beyond the standard model.

If you would like to know more about the measurement, you can read the Physical Review Letter or our detailed webpage. You can also contact the primary author (A. Harel) directly.

April 30, 2008