Forward Physics – Appendix for Annual Report 2014

Forward Physics – Appendix for Annual Report 2014

(in PDF)


Appendix: CDF – Highlights of the Year 2014:

  1. Single Top Quark Cross Section Measurement

For  the  single-top  quark  analysis  at  CDF,  events  with  the  following  characteristics  were  selected:  a  lepton  (either  an  electron  or  a  muon)  with  high  transverse  energy,  a  large imbalance of transverse momentum, which indicates the presence of a neutrino, and two or three jets, at least one of which must originate from a bottom quark.1

By using artificial neural networks in the analysis,  a single-top quark cross section of 3.04 +0.57/-0.53 picobarns was measured (Figure 1).


Figure 1: Results of the two-dimensional fit for σ(s) and σ(t+Wt). The black circle shows the best-fit value, and the 68.3%, 95.5% and 99.7% confidence regions are shown as shaded areas. The Standard Model (SM) predictions are also included with their theoretical uncertainties.


  1. Searching for boosted tops

When the jets are Lorentz-boosted into a jet-jet center-of-mass system, the ones representing  primary  light  quarks  are  expected  to  have  a  lower  mass,  and  can  be  used  for  added  background discrimination. In Figure 2 below, the measured effective mass of a jet is plotted against the mass of another one belonging to the selected pair of jets. A pair of jets for which each individual jet has a mass between 40 and 60 GeV/c2 dominate. The measured masses are consistent with the predictions of Quantum Chromodynamics (QCD).2

In case the high mass jet-jet pairs originate from a decay of a top quark (with the measured mass of 173.34 ± 0.76 GeV/c2) a cluster of events in which both jets have masses between 140 and 200 GeV/c2 is expected (see Figure 2). Although there are roughly 30 such events in the  data,  it  is  only  slightly  more  than  is expected  from  an  unlikely  production  of  two  very  massive jets by the light quarks.

The data was used to set an upper limit for the rate of top quarks being produced at these very high energies of 40 femtobarns, i.e., they were produced in less than in a single collision in every trillion events. This is to be compared to the Standard Model expectation of top quark production exhibiting two jets of 5 femtobarns.


Figure 2: Observed jet masses (mjet1 vs mjet2) plotted against each other. The measurement shows clustering at masses between 40 to 60 GeV/c2 coinciding with the QCD expectations.


  1. Lepton asymmetry observed in top quark decay

In 2014, the lepton asymmetry measurement was completed by using the full Tevatron Run II data sample with two charged leptons from top pair decays. The resulting asymmetry in the two-lepton mode (Figure 3) was measured to be 7.2 ± 6.0%.3

After combining this with the previous measurement, based on the data with only one charged lepton from top pair decays, the final CDF measurement of this asymmetry was 9.0  +2.8/-2.6% (see lower figure). While there are several competing theoretical predictions for the asymmetry, the most current theoretical prediction is 4%.

The new result has a high impact for the top quark studies and places further constraints on the  QCD  predictions.  The   CDF  analyses  will  continue,  particularly  in  studies  of  the asymmetry of the top quark pairs in the two-lepton mode. Measurements of the asymmetry of the bottom quark pairs that probe the same physics question are also on the way.


Figure 3. A comparison of forward-backward asymmetry measured in this experiment (DIL, dilepton mode), in an earlier CDF measurement in one-lepton mode (L+J, lepton + jets), and in their combination.


  1. Is the top hiding a charged Higgs?

According to the Standard Model, the heaviest of the six quarks – the top quark – decays into a charged lepton, a neutrino and a bottom quark (Figure 4). In most analyses at the Tevatron and  in  the  LHC  experiments,  this  decay  involves  an  intermediate W boson.  The  W boson decays  into  either  an  electron  or  a  muon  (accompanied  with  their  unseen  neutrinos)  are selected in these analyses.

The W boson can also decay into the tau lepton, which is heavier than an electron or a muon. (see figure below). The tau lepton decay channel is less explored than the other two because it is more challenging to identify. After the tau lepton decays in the detector, the CDF level 1 track trigger — instrumentation that helps with the rapid selection of important events among  the  hundreds  of  trillions  that  occur  inside  the  detector  —  selects  the  tau  leptons.  Finally, offline software reconstructs the tau from its fragments.

The W boson is not the only particle that could decay into a tau. Certain new particles, like additional  Higgs  bosons,  could  provide  further  decay  channels  with  a  preference  for  heavy particles, including taus.

A top quark, then — through either a W boson or a new particle such as a Higgs — decays into a tau lepton, a tau neutrino, and a bottom quark. Analysing this decay channel will help scientists assess the effects of new physics.

CDF measures the number of taus produced from top-antitop pairs. The number is the product of the production of top-antitop pairs, called the cross section, and the fraction of the pairs that decay into taus, called the branching fraction. New physics could modify either rate, so when evidence of new physics is seen, it is not obvious which of the two is affected. With two unknowns, a single equation will not help to  solve the problem, a second equation using  a second set of data is needed.

This CDF analysis separates – for the first time – single-tau from two-tau events to effectively get the second equation for the two unknowns. We can then measure the branching fraction of the top quark into a tau lepton independently of the cross section.4

The result is that the branching fraction for a top quark into a tau lepton, a tau neutrino and a bottom quark is 9.6 ± 2.8%, which is in agreement with the Standard Model.

This measurement, which depends on our understanding of the upper figure, can be used to calculate the branching fraction shown in the lower figure.

A branching fraction of a top quark into a charged Higgs boson is excluded (in the mass range from 80 to 140 GeV/c2) and a bottom quark at the 5.9 percent level. The exclusion is at the 95% confidence level and is comparable with recent measurements at the LHC.


Figure 4. The Feynman diagram depicting production and decay of a top-antitop pair. In the upper part, the top quark decays into a W+ and an anti-b quark. The W+ then decays into a tau lepton and a neutrino.


  1. The top, the bottom and everything in between

In  collisions  of  high-energy  beams,  new,  heavy  particles  are  usually  produced  in  pairs:  a matching  antiparticle  for  every  particle.  This  is  the  case  for  the  bottom  quark  and  the  top quark, both discovered at Fermilab. But every now and then, the weak force is responsible for the collision, and so particles can be transmuted from one kind into another (see Figure 5). The weak interaction can also produce a pair of dissimilar particles, a bit like giving birth to fraternal  twins  rather  than  identical  ones.  Studying  the  rate  of  such  production  tells  us volumes about the secrets of the weak force and possible new interactions that may mimic it, especially if they favor heavy particles such as the top quark.

This measurement describes a search for such a process: a top quark and an associated anti-bottom quark are produced via s-channel W boson exchange (see figure below). It is similar to an earlier reported study on evidence for the s-channel process used in a search in what is known as the lepton-plus-jets mode.5

The  search  described  in  this  column  is  performed  in  the  so-called  missing-energy-plus-jets mode.  The  production  mechanism  is  the  same  as  the  earlier  CDF  search  —  the  top  quark decays to a W boson and a bottom quark, and the W boson subsequently decays to a lepton and a neutrino. The two bottom quarks, circled in green in the above figure, produce jets that have  long-lived,  heavy  B hadrons  in  them.  If  the  lepton  is  not  identified,  we  use  it  in  the missing-energy-plus-jets analysis; otherwise the earlier analysis makes use of it.

One challenge of analysing data with jets and missing energy is that they can be mimicked by events with only jets in them. These jets-only events can be mismeasured, resulting in large amounts  of  fake  missing  energy.  Some  of  these  mismeasured  events  then  contaminate  the sample  of  events  used  to  search  for  the  single-top  signal.  Scientists  use  sophisticated algorithms to reduce the amount of contamination from these events, and then use the rejected data  to  estimate  the  amount  that  remains.  Other  algorithms  reduce  the  contamination  from other sources.

The measured  cross section in this analysis is 1.12 +0.61/-0.57 picobarns. When combined with  the  earlier  lepton-plus-jets  result,  the  cross  section  is  1.36  +0.37/-0.32  picobarns.  The addition of the missing-energy analysis increases the sensitivity of the combination by more than 10 percent compared with the lepton-plus-jets result alone. This analysis forms the CDF contribution to the Tevatron combined observation of the s-channel single top quark process.


Figure 5. The diagram depicting the s-channel single-top quark production mechanism searched for in the analysis; the process selects signal events in which the lepton is not identified.


  1. CDF documentation for the W boson mass measurement

In Tevatron Run I (1990), 1 722 W-boson events were used to obtain the W boson mass of 79 910 ± 390 MeV/c2. Between Run I and Run II, all of the tracking chambers were rebuilt in order to improve accuracy. A new measurement was based on more than 60 times the number of W boson events in 1990, and brought the mass uncertainty down to 390 MeV/c2. The W mass was now measured to be 80 413 ± 48 MeV/c2 based on 115 092 events. This was the world’s single most precise measurement of the W mass. For the 2012 measurement, the goal was to reduce the 48 MeV/c2 uncertainty below the previous world average of 23 MeV/c2.

The  largest  single  systematic  uncertainty  in  this  measurement,  ±  23  MeV/c2,  was  due  the uncertainty  in  the  calibration  of  the  momentum  of  the  decay  electron  and  muon.  The collaboration improved the calibration of the electron energy scale using, in decays of both the W and  the Z boson,  the  ratio  of  its  energy  to  its  momentum.  This  calibration  technique was validated by applying it to the measurement of the mass of the Z boson decaying into two electrons;  the  momentum  calibration  was  verified  by  using  the J/ψ-  and Υ-to-muon  pair decays  and  cross-checked  by  using  the Z-to-muon  pair  mass  measurement.  In  the  current measurement,  the  uncertainty  of  the  electron  and  muon  energy  scale  is  reduced  to  ±  7 MeV/c2.

The 2012 measurement of the W boson mass by the CDF experiment, 80 387 ± 19 MeV/c2, was based on 470 126 candidates in which a W decays into an electron and a neutrino, and on 624 708 candidates in which a W decays into a muon and a neutrino. The combined world average now yielded a W boson mass of 80 385 ± 15 MeV/c2 (Table 1).

What has changed since February 2012? The Higgs boson has been discovered, and its mass has been determined to a high accuracy, allowing a prediction of the W boson mass of 80 359 ±  11  MeV/c2.  The  comparison  of  this  prediction  with  the  combined  world  average  places bounds on non-Standard Model physics.6


Table 1. Uncertainties in units of MeV/c2 in the latest result for the mass of the W boson, which was determined to be 80 387 ± 19 MeV/c2. The total uncertainty amounts to 0.02%.


  1. An improved search for a dijet resonance

The   CDF   collaboration   recently   reported   another   study   of   collisions   giving   rise   to a W or Z boson and two quarks (jets).7 The data were purposefully selected to be completely independent of the data that gave rise to the excess in the earlier study. The original data set required  a  lepton  with  a  high  transverse  momentum;  the  current  analysis  vetoes  any  event with a high-transverse-momentum lepton. The idea was to cross-check earlier results and, at the same time, to probe the scenario in which the hypothetical new particle would truly exist and  would  also  appear  together  with  a Z boson,  as  suggested  by  several  scientists.  In the 2014 analysis, additional corrections to reconstructed jets in simulated events were applied. These  corrections  more  accurately  model  particle  showers  that  are  initiated  by  both quarks and gluons.

The  original  selection  resulted  in  more  than  2  million  events.  The  principal  background  is multi-jet events, which are produced by the strong interaction. After using up-to-date analysis tools,  the  number  of  multi-jets  was  reduced  to  6  280  ±  1  190  (see  above  figure).  The experiment  found  2  900  ±  183  diboson  events  (WW, WZ, ZZ).  This  number  of  diboson events translates into a measured cross section of 13.8 +3.0/-2.7 picobarns. This number is in agreement with the Standard Model value of 16.8 ± 1.0.

The most important result of this analysis is that no anomalies (no second peaks) are observed in the dijet mass spectrum (Figure 6). This story summarises beautifully many of the salient features  of  the  scientific  journey:  the  appearance  of  an  experimental  anomaly  in  a  well-established framework, leading to great excitement; the process of independently checking the validity of the result; and finally the improved understanding of nature that inherently follows either its confirmation or disproval. Here, what appeared to be a potential game changer to particle physics ended up producing a sounder understanding of important physics processes.


Figure 6. Dijet invariant mass distribution with fit results overlaid for events passing all selection criteria (top) and the same background-subtracted distribution with the fitted diboson contribution overlaid (bottom).


  1. Trilepton events using a blind analysis

This   search   looked   for   the   production   of   three   leptons   plus   missing   energy   and   is characterized by three well-understood backgrounds. The lepton class of particles comprises electrons, muons and taus.8

CDF analysts studied extensively all backgrounds, both in counting and kinematics, in a large number of dilepton and trilepton control regions before they were convinced they understood the  Standard  Model  as  it  manifests  itself  in  multilepton  final  states.  They  did  not  allow themselves to look at the signal region before this process was fully concluded, making this a “blind” search for new physics.

The  final  step  was  to  uncover  the  signal  region  and  study  the  observed  events.  The  two highest-energy leptons must be an electron or a muon, while the lowest-energy lepton can be an electron, a muon or a tau. An excess of trilepton events was observed at low dilepton mass. Scientists  observed  34  electron  pairs  with  a  third  lepton  when  only  20  ±  4  were  expected. They also observed 19 muon pairs with a third lepton when 13 ± 2 were expected. The results are displayed in Figure 7.

The probability that such an excess over the same energy range (between 10 and 85 GeV/c2) was produced by a statistical fluctuation from background events anywhere in the spectrum is 3.2%  (1.85  sigma  effect),  which  is  not  threatening  to  the  Standard  Model.  Although  new physics  was  not  observed,  the  understanding  of  Standard  Model  multileptonic  processes  at CDF  reached  an  unprecedented  level,  which  could  stimulate  new  searches  for  even  more unexpected signals.


Figure 7. The dilepton mass distribution of electron- or muon-pair + lepton events for the Standard Model background, CDF data and a SUSY benchmark (stacked on top of the Standard Model background) in the signal region.


1 Measurement of the Single Top Quark Production Cross Section and |Vtb| in Events with One Charged Lepton, Large Missing Transverse Energy, and Jets at CDF by the CDF Collaboration (Timo Antero Aaltonen (Helsinki U. & Helsinki Inst. of Phys.) et al.). Jul 15, 2014. 8 pp. FERMILAB-PUB-14-229-E e-Print: arXiv:1407.4031 [hep-ex]

2 Studies of high-transverse momentum jet substructure and top quarks produced in 1.96 TeV proton-antiproton collisions , by the CDF Collaboration (Timo Antero Aaltonen (Helsinki U. & Helsinki Inst. of Phys.) et al.). Jul 13, 2014. 51 pp. FERMILAB-PUB-14-228-E  e-Print: arXiv:1407.3484 [hep-ex] | PDF

3 Measurement of the inclusive leptonic asymmetry in top-quark pairs that decay to two charged leptons at CDF , by the CDF Collaboration (Timo Antero Aaltonen (Helsinki U. & Helsinki Inst. of Phys.) et al.). Apr 14, 2014. 8 pp. Published in Phys.Rev.Lett. 113 (2014) 042001  FERMILAB-PUB-14-096-E  DOI: 10.1103/PhysRevLett.113.042001   e-Print: arXiv:1404.3698 [hep-ex] | PDF

4 Study of Top-Quark Production and Decays involving a Tau Lepton at CDF and Limits on a Charged-Higgs Boson Contribution, by the CDF Collaboration (Timo Antero Aaltonen (Helsinki U. & Helsinki Inst. of Phys.) et al.). Feb 26, 2014. 9 pp. Published in Phys.Rev. D89 (2014) 091101 FERMILAB-PUB-14-036 DOI: 10.1103/PhysRevD.89.091101 e-Print: arXiv:1402.6728 [hep-ex] |

5 Search for s-Channel Single-Top-Quark Production in Events with Missing Energy Plus Jets in pp¯ Collisions at s√=1.96  TeV  Timo Antero Aaltonen (Helsinki U. & Helsinki Inst. of Phys.), Silvia Amerio (INFN, Padua & Padua U.), Dante E Amidei (Michigan U.), Anton Iankov Anastassov (Fermilab & Northwestern U.),Alberto Annovi (Frascati), Jaroslav Antos (Comenius U. & Kosice, IEF), Giorgio Apollinari, Jeffrey A Appel (Fermilab), Tetsuo Arisawa (Waseda U.), Akram Muzafarovich Artikov (Dubna, JINR) et al., Feb 16, 2014. 8 pp. Published in Phys.Rev.Lett. 112 (2014) 23, 231805  FERMILAB-PUB-14-027-E  DOI: 10.1103/PhysRevLett.112.231805  e-Print: arXiv:1402.3756 [hep-ex] | PDF

6 Precise measurement of the W -boson mass with the Collider Detector at Fermilab  CDF Collaboration (Timo Antero Aaltonen (Helsinki U. & Helsinki Inst. of Phys.) et al.). Nov 4, 2013. 40 pp.  Published in Phys.Rev. D89 (2014) 7, 072003  FERMILAB-PUB-13-515-E  DOI: 10.1103/PhysRevD.89.072003 e-Print: arXiv:1311.0894 [hep-ex] | PDF

7 Invariant-mass distribution of jet pairs produced in association with a W boson in pp¯ collisions at s√=1.96 TeV using the full CDF Run II data set , by the CDF Collaboration (T. Aaltonen (Helsinki Inst. of Phys.) et al.). Feb 27, 2014. 14 pp. Published in Phys.Rev. D89 (2014) 9, 092001  FERMILAB-PUB-14-037-E DOI: 10.1103/PhysRevD.89.092001 e-Print: arXiv:1402.7044 [hep-ex]

8 Search for new physics in trilepton events and limits on the associated chargino-neutralino production at CDF , by the CDF Collaboration (Timo Antero Aaltonen (Helsinki Inst. of Phys. & Helsinki U.) et al.). Sep 28, 2013. 9 pp. Published in Phys.Rev. D90 (2014) 012011 FERMILAB-PUB-13-446 DOI: 10.1103/PhysRevD.90.012011 e-Print: arXiv:1309.7509 [hep-ex] | PDF