Search for a SM Higgs boson in the diphoton final state at CDF

1. Search for a SM Higgs boson in the diphoton final state at CDF Karen Bland On behalf of the CDF Collaboration • Higgs production and decay • Photon ID Selection • Advantages of Diphoton Search • Analysis Description • Results • Summary √s = 1.96 TeV Pheno Symposium May 10, 2010 Madison, WI

2. hγγfinal state Higgs production Gluon Fusion:σ ≈ 1000 fb* • Low mass search: Focus on 100 – 150 GeV/c2 • Overall σ: ~1300 fb: larger overall cross section than other channels • Br(hγγ) < 0.0025: smaller branching ratio than other channels • Signal Expectation: ≈ 16 events produced with 5.4 fb-1 of data ≈ 2 events after acceptances and efficiencies included Associated Production:σ ≈ 225 fb* Vector Boson Fusion: σ ≈ 70 fb* *σ for √s = 1.96 TeV p-pbar collisions and Mh = 120 GeV/c2

3. The Central Electromagnetic (EM) Calorimeter Hadronic Calorimeter Electromagnetic Calorimeter • CEM (Central EM calorimeter) • Made of alternating sheets of scintillator and lead • Energy resolution: ~13.5%/√E + 2% • |η| < 1.1 • 24 wedges distributed in ϕ (cracks between wedges) • CES (central EM shower maximum detector) • Six radiation lengths into CEM • Refines position measurement of the EM cluster so can match to a track

4. Identifying a Photon Detector profile consistent with a prompt photon: • Compact and isolated EM cluster • No track (no electric charge) • Not in a jet (no color charge)

5. Centralγ Selection • FiducialEM cluster in well instrumented region of CES • Had/EMMajority of calorimeter energy in CEM (not hadronic) • Cal IsolationPhoton EM cluster is isolated • Lateral shower shape in CESComparison to test beam consistent with prompt photon (rather than those from neutral pions, for example) • Track Isolation: No high pT tracks leading to CES cluster and pT of tracks near cluster restricted • 2nd CES clusterEnergy of any 2nd CES cluster restricted to reject neutral mesons decaying to two photons

6. Advantages of Using PhotonsMass resolution and signal acceptance Large signal acceptance • ≈ 13% overall signal acceptance • Will double with inclusion of forward photons • Great mass resolution • σ/Mγγabout 4 x smaller than best jet algorithms • Can then just search for narrow resonance! • Sideband fits can be used to estimate background

7. Advantages of using PhotonsCalibration Process: Z e+e– • Study photon ID efficiencies • Compare data and MC to derive simulation correction factor • Check electromagnetic energy scale Use of Ze+e– process ensures small uncertainties on ID efficiencies, data-MC scale factor, and energy scale Diphoton event Dielectron event Can calibrate EM calorimeter response to photons using electrons…

8. Analysis Overview • Event Selection: • Use standard CDF photon ID • Select central 2 photons w/ Mγγ> 30 GeV/c2 • Data: • Diphoton triggers • ~5.4 fb-1 from between Feb. 2004 – July 2009 • Signal MC: • Generated using PYTHIA 6.2, CTEQ5 PDFs and the standard CDF UE tune-A • 100 – 150 GeV/c2 in 10 GeV/c2 intervals • Scale factors derived from Ze+e– studies • Small signal mass resolution reduces search to a bump hunt – we search for a narrow peak over a smooth background • Data-Driven Background Model: • Use fits to sideband of data to predict background shape in signal region • Search for Resonance • Search for resonance over background • If no sign of resonance, then set 95% CL limits on σ x Br

9. Data-Driven Background Model • Composition • Real SM photons via QCD interactions • 1 or 2 jets faking a photon • Background Model • Fit to sideband region of the Mγγ distribution • Excluding 12 GeV/c2 window around signal test mass • Interpolate fit into signal region • Fit in signal region used as background model Higgs test mass window of 120 GeV/c2 shown. Process repeated for each test mass.

10. Visually Search for Resonance • Use background model obtained from sideband fit • Interpolated fit used for data-fit residuals • Inspect residuals for signs of resonance • Repeated for each Higgs test mass • Results: No resonance observed, so then set limits hγγ production…

11. Systematic Uncertainty • Signal • Acceptance and efficiency (in table) • Cross section: • σggH (12%) • σVH ( 5%) • σVBF (10%) • Luminosity: 6% • Background • 4% rate uncertainty • Obtained from studies allowing normalization of fit to vary in the signal region

12. Limits on hγγ • 12 GeV/c2 signal region for each test mass used to set upper limits set on σ  Br relative to SM prediction • Expected and observed limits in good agreement • Most sensitive for range 110 – 130 GeV/c2 Will contribute sensitivity to SM Higgs at Tevatron, particularly around 120 GeV/c2 • For Comparison… • Limits similar or better for current results from hττ (2.0 fb-1) • Comparable to channels like ZHllbb at 140 and 150 GeV/c2

13. Limits and Mγγ

14. Summary and Comments • Summary • Searched for SM hγγ using 5.4 fb-1 of data and data-driven background model • Expected 95% C.L. limits on (σ  Br) / SM are about 20 x SM at 120 GeV/c2 • Plans • Forward photons to soon be added (will double acceptance) • Use results to contribute to Higgs combination, particularly in 110–130 GeV/c2 region • Potential Improvements in Future • Add photons that convert to e+e– pairs • Improve photon ID efficiency

15. Backup Slides

16. The Collider Detector at Fermilab (CDF) A multipurpose detector that observes: • electrons • photons • jets • muons • neutrinos

17. Method used to set limits The likelihood as a function of cross section: • Nid, Nib, and Nis are the number of data, bkg, and sig events in the ith bin • A is acceptance • is efficiency • is luminosity • Ntots is the total number of signal events passing selection requirements The 95% confidence limit was obtained by finding the value of σ95 for which:

18. Limit results

19. Limit results

20. Tevatron Combined Limits

21. Channels Used to Obtain CDF Combined Limits