CDF Logo Measurement of the ratio of the cross section of ttbar+Photon to ttbar and a Search for New Physics in the Lepton Photon Missing Et and B-tagged Jet channel.
Public Page 		Exotics Logo
This search for physics beyond the Standard Model, uses a "signature-based" analysis. This strategy does not focus on a specific model for new physics, but allows for a larger chance of observing new physics by not directly probing one distinct example. Here we present results for anomalous production of lepton + photon + missing transverse energy + one jet which is b-tagged using 6.0 fb-1 of integrated luminosity from ppbar collisions at sqrt(s) 1.96 TeV, collected using the CDF II detector. The events for the searches include production of two gauge bosons, W and photon, and two third-generation quarks, top and bottom. This search is an extension of a previous search done at CDF with 1.9 fb-1 of data collected at the CDF II detector,
Phys. Rev. D 80, 011102(R) (2009)

A search for the production of top pairs with an additional photon, ttbar+Photon, is a natural extension of this signature-based search, in that ttbar+Photon is characterized by the signature of a high-PT lepton, photon, b-tagged jet, and MET. When one in addition requires large total transverse energy HT (the sum of the transverse energies of the lepton, photon, jets and MET) and 3 or more jets, radiative top-pair events dominate the SM predictions.

The search was further refined by require photons with transverse momentum below 25 GeV to have a chi-squared value less than 6. We looked at a control region of leptons, photons and missing transverse energy, to find true photons, and fake photons, and plotted the efficiency versus the purity of the cut. A cut on the chi-squared value of a photon with transverse energy less than 25 GeV was justified.

We find 85 Lepton + Photon + MET + b events versus a standard model expectation of 99.1 +/- 9.3 events. We observe 30 ttbar +photon (with chi-squared cut) candidate events versus a standard model expectation of 26.9 +/- 3.4 events . Assuming the difference between the observed number of predicted non-top SM total is due to top production, we measure the ttbar+photon cross-section to be 0.18 +/- 0.07 pb . This compares well to the 1.9 fb-1 analysis measurement of 0.15 +/- 0.08 pb , and the theoretical cross- section measurement of 0.17 +/- 0.03 pb . Finally we find the ratio of production cross section of ttbar+photon, and ttbar to be 0.024 +/- 0.009. The probability for non-top backgrounds alone to produce this 30 or more events is 0.0015 or 3.0 standard deviations.


We have estimated the background due to events with jets misidentified as lepton+photon+MET+b signature by using the anti-electron method. We identify samples of photon+b+anti-electrons, and photon+b+lepton. We compare assume the difference between data and MC simulation in the MET sideband is due to the QCD background. The photon+b+anti-electrons distribution is scaled to fit the difference between data and MC prediction in the sideband region. Our estimate on the amount of QCD is the sum of scaled photon+MET+b+anti-electron events in the signal region of MET.

The number of "jet faking photon" events expected in the lepton+photon+MET+b signature is determined by looking at objects which pass many of the photon cuts except for isolation, and to plot the isolation shape of all such candidates. The shape is well-modeled with the isolation shape of electrons from Z>ee decays (for lower isolation values) and a straight line (for higher isolation values). The straight line represents the fake photons and is propagated into our cut range for isolation, and the area under this line is the amount of jets faking photons in our signal.

The number of "electron faking photon" events expected in the lepton+photon +MET+b signature is determined by measuring the photon ET spectrum in lepton+electron+MET+b samples, and then multiplying by the probability of an electron being misidentified as a photon, which is measured in Z>ee events in which one of the electrons radiates a high-ET photon.

To estimate the size of the mistag background, each jet in the lepton+photon+MET+pretagged jet sample is weighted by its mistag rate. The mistag rate per jet is measured using a large inclusive-jet data sample. For consistency with the top pair production analysis we use the Method 2 mistag background.

In order to make the background estimate as precise as possible it is neccessary to compensate for the fact that some data-driven backgrounds have overlapping samples. For example, the sample of jets faking photons, and the mistagging of light-flavor jets as b-tagged jets, have a large overlap.

The sample for jets faking photons in the lepton+photon+MET+b signature requires significant MET, a lepton, a jet to fake a photon, and a b-tagged jet. While the sample for a mistagged light-flavor jet is MET,a lepton, a jet to fake a b-jet, and a photon. An event that has MET, a lepton, a jet able to fake a photon, and a jet able to fake a b-jet, and a b-jet, will then be counted once in the jets faking photons sample, and once in the mistag category.

The double counting overlap is removed from background predictions in both the lepton+photon+MET+b

Tau faking photon background is estimated from ttbar pythia sample by selecting tau > hadrons > photon events using MC information and then applying same analysis cuts as for data.

The production of semileptonic ttbar+photon with one of the W bosons decaying leptonically (e, mu and tau channels) and another hadronically is estimated from the MadGraph MC. The production of dileptonic ttbar+photon with both W bosons decaying leptonically (e, mu and tau channels) is also estimated from the MadGraph MC. The SM background from a production of W boson and a photon, accompanied by QCD production of heavy flavor quarks in the processes Wbbar+photon, Wccbar+photon, and Wc+photon, is estimated from MadGraph MC.

Backgrounds from ZZ, single top + photon are estimated to be negligible.

Final Plots and Tables:

Idea for how small this measured cross section is!
Plot showing the size of the ttg cross section relative to "standard candles" EPS or GIF
Full summary of Signal and Backgrounds for ttbar Analysis
Table summary of signals and backgrounds of ttbar Analysis EPS or GIF
Plots From ttbar Analysis:
Missing Transverse Energy Spectrum for Electrons and Muons EPS or GIF
Transverse Momentum of Leptons from ttbar/ttbar decays EPS or GIF
Transverse Mass of Leptons and Missing Et EPS or GIF
Transverse Energy of Leading Jet EPS or GIF
Transverse Energy of Leading Jet EPS or GIF
Full summary of Signal and Backgrounds for ttbar + photon Analysis
Table summary of signals and backgrounds of ttbar + photon Analysis EPS or GIF

Plots from the ttbar + photon side of the analysis:
Transverse Energy of the Observed Lepton, EPS or GIF
Plot of Transverse Energy of Photon from ttbar+photon selection, EPS or GIF
Missing Transverse Energy in the ttbar+photon selection, EPS or GIF
Number of Jets Distribution from the ttbar+photon selection, EPS or GIF, that belong under the same header.
Full summary of Signal and Backgrounds for lepton + photon + met + b-quark Analysis
Table summary of signals and backgrounds of ttbar + photon Analysis EPS or GIF

Plots from the lepton + photon + Missing Transverse Energy + b-jet side of the analysis:
Transverse Energy of the Observed Lepton, EPS or GIF
Plot of Transverse Energy of Photon from lepton, photon, missing transverse energy, bjet selection, EPS or GIF
Missing Transverse Energy in the lepton, photon, missing transverse energy, bjet selection, EPS or GIF
Number of Jets Distribution from the lepton, photon, missing transverse energy, bjet selection, EPS or GIF, that belong under the same header.
Summary of results slide EPS or GIF, that belong under the same header.

Page page: http://www-cdf.fnal.gov/~bauerbac/index.html was last updated February 9th, 2011 by Ben Auerbach
Contact, Legal Notices, and Fermilab Policy on Computing