1. Sinéad Farrington University of Glasgow for the CDF Collaboration European Physical Society Aachen, 17 th -23 rd July 2003 B Lifetimes and Flavour Tagging at CDF Run II

2. Sinead Farrington EPS Aachen 17th-23rd July 2003 2 Outline Overview of B Physics at CDF B lifetimes: –Exclusively reconstructed B’s decaying via J/  –Semileptonically decaying B’s Flavour Tagging

3. Sinead Farrington EPS Aachen 17th-23rd July 2003 3 B Physics at CDF Run II pp collisions produce a wide spectrum of B hadrons in a challenging environment Tevatron upgrade to 1.96 TeV CoM energy  larger b production cross section -

4. Sinead Farrington EPS Aachen 17th-23rd July 2003 4 B Triggers at CDF Run II Di-Muon (J/  ) p T (  )>1.5 GeV/c J/  modes at low (conventional) p T ( J/  )  0 GeV/c Measure x-section (See talk by Tara Shears) J/  Yield=2x Run I l + displaced track p T (e/  )>4 GeV/c Semileptonic modes p T (trk)>2 GeV/c Lifetimes,flavour tagging 120  m<d 0 (trk)<1mm B Yields 3x Run I Two displaced tracks p T (trk)>2 GeV/c Hadronic modes 120  m<d 0 (trk)<1mm Charm Physics, B 0 s mixing NEW!  (bb) <<  (pp)  B events are selected with specialised triggers: Indivdual data samples of about 140pb -1 accumulated - -

5. Sinead Farrington EPS Aachen 17th-23rd July 2003 5 B Hadron Lifetimes According to the spectator model all B mesons would have the same lifetime… but that is not the whole story! The HQE predicts by how much the lifetimes differ. Objective 1: Measure lifetimes accurately to form input to theories Objective 2: Prove the detector and triggers are working Objective 3: Measure  s /  s B 0 s  l D s is an admixture of CP even and CP odd states Pair it with fully CP even or CP odd state (or use polarisation analysis) and measure  s /  s PDG Values

6. Sinead Farrington EPS Aachen 17th-23rd July 2003 6 B Hadron Lifetimes EXCLUSIVE J/  trigger clean fully reconstructed lifetime unbiased low statistics J/  K+K+ B+B+   SEMILEPTONIC lepton+displaced track trigger clean partially reconstructed - need MC to unfold ct lifetime biased good statistics Reconstruct decay length by vertexing Measure p T of decay products B 0  J/  K 0* B +  J/  K +  b  J/   B 0 s  J/   B 0  l D ( * )-  b  l  c B 0 s  l D  s

7. Sinead Farrington EPS Aachen 17th-23rd July 2003 7 Fit Methodology: Simultaneous fit of M(B)  signal fraction, define sidebands c  (B)  lifetime Background Parameterisation: Signal Contribution: Exclusive B  J/  X Lifetimes

8. Sinead Farrington EPS Aachen 17th-23rd July 2003 8  b  J/   Lifetime and Crosscheck CONTROL SAMPLE

9. Sinead Farrington EPS Aachen 17th-23rd July 2003 9 Exclusive B  J/  X Lifetimes: B 0 s Meson CDF Run II Preliminary B+B+ 1.63  0.05  0.04 ps B0dB0d 1.51  0.06  0.02 ps B0sB0s 1.33  0.14  0.02 ps bb 1.25  0.26  0.10 ps B 0 s  J/   is mostly CP even  measures mostly  L component Access  s /  s via angular analysis – coming soon World’s largest sample of fully reconstructed B s mesons

10. Sinead Farrington EPS Aachen 17th-23rd July 2003 10 Extraction of Semileptonic Lifetimes Lepton+ displaced track trigger implemented successfully for the first time!  samples rich in B hadron semileptonic decays Follow the standard methodology: reconstruct the D decay near to lepton B decay not fully reconstructed  extract the  factor from Monte Carlo: extract lifetime from decay length: but with one crucial complication: lifetime bias from the displaced vertex trigger: 4 GeV lepton + 2 GeV track with 120  m<d 0 <1mm  Emulate trigger with Monte Carlo and model the lifetime bias which is factored into the fit function.

11. Sinead Farrington EPS Aachen 17th-23rd July 2003 11 K Factor and Bias from MC K Factor Distribution P(K) Reconstruction efficiency including trigger bias, as a function of proper lifetime Normalised to unit area Arbitrary vertical scale

12. Sinead Farrington EPS Aachen 17th-23rd July 2003 12 B Semileptonic Channels M(pK  ) (GeV/c 2 ) apply PID CDF Run II Preliminary 70pb -1  b  l  c l DsDs DD

13. Sinead Farrington EPS Aachen 17th-23rd July 2003 13 Semileptonic Decay Fit Model Unbinned maximum likelihood fit to c  (B) –Background is parameterised by delta function and positive exp convoluted with Gaussian resolution: Free parameters:  D  E + f +  G –Signal: exp convoluted with Gaussian resolution, K factor distribution, P(K), and bias function,  –Maximum likelihood function:

14. Sinead Farrington EPS Aachen 17th-23rd July 2003 14 Semileptonic Lifetime Fits  b   c Lifetime statistical error projections B+B+  0.04 ps B0dB0d  0.06 ps B0sB0s  0.07 ps bb  0.09 ps Highest statistical accuracy lifetime fits in B 0 s and  b : Control Samples Fit results show discrepancy with world average Many crosschecks Currently working to understand all effects of trigger bias etc. CDF Run II Preliminary 70pb -1 B 0 s   D s ct(B)(mm)

15. Sinead Farrington EPS Aachen 17th-23rd July 2003 15 Flavour Tagging Development OPPOSITE SIDE Soft Lepton Tag* Opposite side K Tag Jet charge tag OPPOSITE SAME SAME SIDE Same side K/  Tag* Determine B flavour at production and decay to measure –B 0 mixing, B 0 s mixing, CP violation Figure of merit:  D 2 –  : tag efficiency –D: Right-Wrong/Right+Wrong CDF has a battery of tags which are being developed further

16. Sinead Farrington EPS Aachen 17th-23rd July 2003 16 Same Side Tag Asymmetry Identify fragmentation pion Measure raw asymmetry: Test tagger on B +  should see NO oscillations! B  l D 0 B +  J/  K + For B 0 s mixing apply same side K tag: very powerful

17. Sinead Farrington EPS Aachen 17th-23rd July 2003 17 Tagger Studies in Semileptonic Sample lepton + displaced track trigger provides high statistics sample of ~150000 B (and D) semileptonic decays Strategy: trigger lepton  estimate of B flavour at production count numbers of opposite side tags can subdivide sample by muon system and p T rel parameterise relative dilution for use in likelihood fit cross check consistency with partially reconstructed lepton+D +,0 consistent results  believe dilution in lepton+D s This number is UNBIASED since we’re using an independent (and high statistics) control sample   track p(e/  +D) P.V. Detailed sample composition studies: –Mass cut removes D decays: 2<M(l+track)<4GeV/c 2 –Background subtraction variable separates B’s from background: signed IP of displaced track

18. Sinead Farrington EPS Aachen 17th-23rd July 2003 18 Soft Muon Tag  D 2 Soft Muon Tagger  D 2 =0.660  0.093 % CDF Run I:  D 2 =0.584  0.082 % Identify  charge on opposite side Sufficient statistics to examine parameterisation of dilution as function of p T rel Method will be used to evaluate, optimise and compare all taggers Unbiased way of optimising taggers Next step: measure  D 2 for a reconstructed channel Background subtraction has been used to obtain measurement of  D 2

19. Sinead Farrington EPS Aachen 17th-23rd July 2003 19 Summary Lifetime measurements made in clean B  J/  channels: Lepton + displaced vertex trigger has been implemented at CDF for the first time –lifetime measurements ongoing: will give best statistical accuracy in B 0 s and  b –Flavour Tagging procedure has been established and two taggers have been optimised and evaluated in a non-biased way Measurements of  s /  s coming soon and first steps towards B 0 s mixing are happening! CDF Run II Preliminary B+B+ 1.63  0.05  0.04 ps B0dB0d 1.51  0.06  0.02 ps B0sB0s 1.33  0.14  0.02 ps bb 1.25  0.26  0.10 ps

20. Sinead Farrington EPS Aachen 17th-23rd July 2003 20 K Factor and Bias from MC ELECTRONS:

21. Sinead Farrington EPS Aachen 17th-23rd July 2003 21 Major upgrades at CDF provide Increased geometric coverage Extended muon system new 3D silicon tracker faster response drift chamber Triggers: large bandwidth, background reduction, displaced vertex triggering

22. Sinead Farrington EPS Aachen 17th-23rd July 2003 22 Determine B flavour at production and decay to measure –B 0 mixing, B 0 s mixing, CP violation Figure of merit:  D 2 –  : tag efficiency –D: Right-Wrong/Right+Wrong CDF has a battery of tags which are being developed further Run II’s lepton+displaced vertex trigger  large sample of semileptonic decays allows us to measure , D and  D 2 in data and optimise the taggers can then apply them in any sample without bias

23. Sinead Farrington EPS Aachen 17th-23rd July 2003 23