Measurement of Single-top Quark Production with ATLAS Data

Measurement of Single-top Quark Production with ATLAS Data


Single-top production processes have been studied using 0.7 of data from 7 TeV center-of-mass energy proton-proton collisions collected with the ATLAS detector at the LHC. Single-top is electroweak top production and the standard model includes three production modes. Each contains a Wtb vertex, allowing the possibility of a direct measurement of the CKM matrix element . Single-top could also be sensitive to new physics, such as flavor changing neutral currents or heavy bosons. Using cut-based selections, a limit of pb is set for dilepton production and pb for -channel production. For the -channel measurement, both cut-based and neural network analyses are performed and the cross-section is measured to be pb, where pb is expected according to standard model.

I Introduction

Single-top is electroweak top-quark production through three different standard model production modes: -channel, (associated production) and -channel. The -channel has the largest expected cross-section for 7 TeV proton-proton collisions ( pb) (1) followed by ( pb) (2), and -channel ( pb) (3). The -channel and cross-sections in particular are at least 30 times larger at the LHC than at the Tevatron, while the -channel cross-section is about 4 times as large (4). The larger cross-sections will allow precision measurements of the cross-sections, tests of the Wtb coupling, direct measurements of the CKM matrix element  (5); (6), and searches for new physics such as flavor changing neutral currents and heavy bosons (7).

Single-top production was first observed at the Tevatron in 2009 by the D0 (8) and CDF (9) collaboration, and the -channel production mode of single-top alone was first observed by D0 (10). Here, results from ATLAS (11), a detector located at the LHC, are discussed. Of the data collected, 0.7 is used to measure the -channel (12) cross-section, set a  (13) limit, and set the first ATLAS -channel (14) limit.

Ii t-channel Analysis

The -channel analysis is performed using two different techniques, a cut-based selection and a neural network approach. In both cases the analyses begin by reducing the large multijet and +jets backgrounds via a common selection which forms a sample enriched with events containing a -channel final state. Events with at least 2-jets with transverse momentum of and are selected, where exactly 1 jet must be b-tagged using a secondary vertex b-tagging algorithm with a 50% b-tagging efficiency and a rejection of about 270 for light-quarks. The analyses are performed using 2-jet or 3-jet multiplicity bins. Events are selected which have exactly 1 isolated, central () muon or electron with which is associated with a corresponding lepton trigger. The events are also required to have missing transverse energy, , to account for the undetected neutrino from the decay. Finally, is required to further reduce multijet contributions, where is the reconstructed transverse mass of the determined from the lepton and . After all of these selections, the signal yield divided by the background yield (S/B) is approximately 0.1.

The background normalization uses theoretical cross-sections for most processes (, , -channel, diboson, and +jets backgrounds) except the large +jets and multijet backgrounds. In the case of multijets, the kinematic distribution shapes are determined using a “jet-electron” method, where the typical electron trigger is replaced by a jet trigger, and the normalization is determined via a binned maximum likelihood fit to the  distribution in data (15). This method is also used for the muon event selection. The +jets heavy flavor fractions and normalization are determined using three control regions in the cut-based analysis and via a fit of the neural network output distribution in that analysis.

The cut-based analysis applies four selections in addition to the initial selection: , , and . Here, ljet refers to the untagged jet with the largest  value and bjet refers to the b-tagged jet. Additionally, is sum of the transverse momentum of the leading two jets in , the lepton , and  while is the reconstructed top quark mass. Four channels are used in this analysis, containing 2 or 3-jets and positively or negatively charged leptons. This channel division makes use of the expected lepton charge asymmetry in p-p collisions due to the excess of valence u-quarks versus d-quarks. Such a division would be ineffective at a p- collider, where no lepton asymmetry is expected. After the cut-based analysis selections, the best channel (2-jets with positive leptons) has a S/B value of 1.2. The distribution of the reconstructed top mass for 2-jets (combined lepton charge channels) is shown in Figure 1 and the signal excess can be clearly seen.

Figure 1: Reconstructed top mass distribution for 2-jet events after all -channel cut-based analysis selections except the reconstructed top mass selection (12). The -channel signal contribution is normalized to the measured combined cut-based analysis -channel cross section.

A neural network analysis is also performed using the NeuroBayes (17); (16) program with 13 variables, including the variables used in the cut-based analysis and lepton charge, used in its channel division. The other 8 variables include the invariant mass of the two jets, , the lepton and , , the  of the untagged jet, the mass of the b-tagged jet, and between the b-tagged jet and reconstructed boson. The variables are combined, including correlations, into one discriminant. Unlike the cut-based selection, no additional selections are applied after initial selection and all events contain exactly 2-jets.

The -channel cross-sections are extracted using a profile likelihood technique for the cut-based analysis (CB) and a maximum likelihood fit for the neural network analysis (NN). The inclusive cross-sections obtained from the various methods are pb (CB 2-jets), pb (NN), and pb (CB 3-jets), all of which are consistent with each other within at least two standard deviations.

The final result is constructed as the combination of the two and three jet channels from the cut-based analysis, pb, compared to pb expected. The measured cross-section is consistent with the predicted standard model -channel cross-section within about 1.1 standard deviations. The uncertainty of this measurement is dominated by systematic uncertainties, with b-tagging, jet energy scale, and initial/final state radiation being particularly large contributors to the cross-section uncertainty.

Iii Wt Analysis

The analysis considers the dilepton final state only, using three different channels: electron-electron, electron-muon, and muon-muon. For this analysis, at least one jet is required with , as well as exactly two isolated, triggered, central muons or electrons with and corresponding . Additionally, the sum of the angular differences in between the leptons and  direction is used to reduce the  contamination, , and the reconstructed dilepton invariant mass is used to remove events coming from boson decays, , for events where both leptons are electrons or both are muons. Here, refers to the reconstructed mass of the two leptons and refers to the boson mass.

The background cross-section normalization is determined via data-based techniques with the exception of the diboson background, for which the theoretical cross-section is used. A matrix method is used to estimate the fake lepton (multijet and +jets) background. The  is estimated using a  dominated region orthogonal to the signal region, , while orthogonal data samples formed using two uncorrelated variables,  and , are used to determine the Drell-Yan background. The largest background, , is determined using a  dominated region orthogonal to the signal region defined as events with at least 2-jets (see Figure 2). The non- events in this region are subtracted from the data and the result is compared to the expected value to determine a scale factor. This result is then propagated into the signal region. Correlated effects of systematic uncertainties affecting both  and have been taken into account.

Figure 2: Number of jets after initial event selection (13). The 1-jet bin is used for the final selection, while the 2 or more jets region is used for the  background normalization.

The final selection requires exactly 1 jet (see Figure 2). The cross-section and limit are extracted using a profile likelihood technique. The resulting cross-section is pb. The 95% confidence level (CL) upper limit on production is pb observed, pb expected. The total uncertainty is dominated by data statistics, jet energy scale, jet energy resolution, and jet reconstruction efficiency uncertainties.

Iv s-channel Analysis

The -channel analysis event selection is similar to the -channel cut-based analysis selection (see Section II) with two differences: 2 central () jets are selected and at least 1 b-tagged jet is required. The backgrounds are also estimated using the same techniques, but selecting events with at least 1 b-tagged jet. The cut-based selection includes a requirement of exactly 2 b-tagged jets, as well as , , , , , and . Here, is the reconstructed top quark mass using the leading jet in pt (bjet2 is the other jet). After these selections, S/ is 0.98, an improvement from 0.26 using the initial selection only.

The -channel cross-section and limit are extracted using a profile likelihood technique. The 95% CL upper limit on the production of -channel events is pb observed, where pb is expected. This is about 5 times the standard model cross-section. More data statistics will improve this result as data statistics is the dominant uncertainty.

V Conclusion

ATLAS single-top production results are estimated using 0.7 of data from 7 TeV proton-proton collisions. The -channel cross-section is measured to be pb. The analysis determines a cross-section of pb, corresponding to a 95% CL limit of pb. Finally, a 95% CL limit of pb is set for the single-top -channel production cross-section.


  1. N. Kidonakis, Next-to-next-to-leading-order collinear and soft gluon corrections for t-channel single top quark production, Phys. Rev. D 83 (2011) 091503.
  2. N. Kidonakis, Two-loop soft anomalous dimensions for single top quark associated production with a W- or H-, Phys. Rev. D 82 (2010) 054018.
  3. N. Kidonakis, NNLL resummation for s-channel single top quark production, Phys. Rev. D 81 (2010) 054028.
  4. N. Kidonakis, Single top quark production at the Fermilab Tevatron: Threshold resummation and finite-order soft gluon corrections, Phys. Rev. D 74 (2006) 114012.
  5. N. Cabibbo, Unitary Symmetry and Leptonic Decays, Phys. Rev. Lett. 10 (1963).
  6. M. Kobayashi and T. Maskawa, CP-Violation in the Renormalizable Theory of Weak Interaction, Prog. Theor. Phys. 49 (1973).
  7. T. M. P. Tait and C. P. Yuan, Single top quark production as a window to physics beyond the standard model, Phys. Rev. D 63 (2000) 014018.
  8. D0 Collaboration, Observation of Single Top-Quark Production, Phys. Rev. Lett. 103( 2009) 092001.
  9. CDF Collaboration, First Observation of Electroweak Single Top Quark Production, Phys. Rev. Lett. 103 (2009) 092002.
  10. D0 Collaboration, Model-independent measurement of -channel single top quark production in collisions at TeV, arXiv:1105.2788[hep-ex], submitted to Phys. Lett. B.
  11. ATLAS Collaboration, The ATLAS Experiment at the CERN Large Hadron Collider, J. Inst. 3 (2008) S08003.
  12. ATLAS Collaboration, Measurement of the -channel Single Top-Quark Production Cross Section in 0.70 of pp Collisions at  TeV with the ATLAS detector, ATLAS-CONF-2011-101 (2011),
  13. ATLAS Collaboration, Search for Wt associated production in dilepton final states with 0.70 of 7 TeV pp collision data in ATLAS, ATLAS-CONF-2011-104 (2011),
  14. ATLAS Collaboration, Search for s-channel Single Top-Quark Production in pp Collisions at  TeV, ATLAS-CONF-2011-118 (2011),
  15. ATLAS Collaboration, Measurement of the top quark-pair production cross section with ATLAS in pp collisions at  TeV, Eur. Phys. J. C71 (2011) 1577.
  16. M. Feindt and U. Kerzel, The NeuroBayes Neural Network Package, Nucl. Instrum. Meth. A559 (2006) 190-194.
  17. M. Feindt, A Neural Bayesian Estimator for Conditional Probability Densities, arXiv:0402093v1 [].
This is a comment super asjknd jkasnjk adsnkj
The feedback must be of minumum 40 characters
The feedback must be of minumum 40 characters
Comments 0
Request comment
The feedback must be of minumum 40 characters
Add comment
Loading ...

You are asking your first question!
How to quickly get a good answer:
  • Keep your question short and to the point
  • Check for grammar or spelling errors.
  • Phrase it like a question
Test description