Recent results from CLEO & the CESR/CLEO Charm factory

1. Recent results from CLEO & the CESR/CLEO Charm factory yDoDo, DoK-p+ K+ K- p+ p- Ian Shipsey, Purdue University CLEO Collaboration

2. I am completely deaf • please write down your questions. • Pass them up to me • I will read out your question before answering it.

3. The CKM Matrix • The parameters of the Standard Electroweak Model are: • The 4 quark mixing parameters reside in CKM matrix Weak eigenstates & fermion masses and mixings Mass eigenstates • * In SM , A, ,  are fundamental parameters • Does the CKM fully explain quark mixing? CP Violation? • To detect new physics in flavor changing sector must know CKM well • Must overdetermine the magnitude and phase of each element

4. CKM Matrix Status Vud/Vud 0.1% Vus/Vus =1% Vub/Vub 25% l 1 l e l eig B n n n K n p p  Free/bound W cs Vcd/Vcd 7% Vcs/Vcs =11% Vcb/Vcb 5% l l B D n n l D D n K p Vtd/Vtd =36% Vtb/Vtb 29% Vts/Vts 39% 1 eib Bd Bd Bs Bs Vud, Vus and Vcb are the best determined due to flavor symmetries: I, SU(3), HQS. Charm (Vcd & Vcs) and rest of the beauty sector (Vub, Vtd, Vts) are poorly determined. Theoretical errors on hadronic matrix elements dominate.

5. B decays & the unitarity triangle Goals for the decade: precision measurements of Vub, Vcb, Vts, Vtd, Vcs, Vcd, , , . Test SM description of CP violation and search for new physics.  Unitarity : SM: side & angles measured in many processes are self consistent Otherwise: new physics B    B   bul (B  //l ) B/BK*  Vtd Vub* B   /K  & other charmless decays Vtd/Vts   BK BJ/Ks  BDK Vcb* Rates: sides of triangle CP asymmetries & rates : angles bcl (B D*/Dl ) Base Vcb well known Other sides poorly known angles   unknown.  soon / 25%

6. CLEO at CESR   OFF ON • CLEO at CESR e+e- storage ring s=10.58 GeV • Operation at Y(4S) just above BB threshold • 107 BB, cc,  CLEO II/II.V (1990-99) • 7 x 106 BB. CLEO III (2000-01) Turned Off June 25 1st results at LP01 • B’s produced nearly at rest • No Bs or b-baryons • 25% of hadronic cross section is BB • CLEO 4pi solenoidal detector • si + drift chamber in 1.5T field • CsI calorimeter • muon identification s

7. B  / :  via penguin,tree interference CKM phases are measured via interference tree From CKM counting expect: * B  is mostly penguin * B  is mostly tree  Two approaches: measure rates or Acp both contain information on the product of  and the (unknown) strong phase difference  between contributing amplitudes penguin   

8. B Reconstruction • Suppress background: • and energy flow • E,p constraints do not fix B direction: exploit (4S) transverse polarization • maximise differences between sig and bkgd Fisher discriminant • multidimensional unbinned maximum likelihood fit to use all available information • dE/dx K/ separation 2 kinematic constraints: E,p • Each B has energy equal to the beam • For signal events E peaks at 0 (E = 25-100 MeV • Constrain E=0 To charm final states

9. B / CLEO II/II.V 20 ± 7 80 ± 12  N( N(

10. B / Summary    * General agreement theory * Large Br(B K)/Br(B)  severe penguin pollution complicating extraction of  at BaBar/Belle * B  0 0 submitted to PRL Good agreement between BABAR/Belle/CLEO

11. Measuring  via penguin,tree interference • Comparison of rates between modes related by I or SU(3) allow a low statistics determination of . • Example: (Fleischer –Mannel) • Recent improved theoretical treatment of hadronic B decays: QCD factorization (Beneke, Buchalla, Neubert Sachrajda hep/ph0104110) provides amplitudes & strong phases if R <1 bound on  Theory Theory uncertainty All input parameters Vub/Vcb Earlier model  R* (Babar/Belle/CLEO Average)  (deg) Good global fit to all K/ data. Data not yet precise enough to determine  Near future: R*/R*=10% =110 need ~ 75M BB Future: measures  need ~175M BB

12. Search for Direct CP Violation in B  CLEO 4 significance * Summary: no evidence for direct CP violation in five modes * Statistics limited >0.12 * Systematic error 0.02 mostly dE/dx * Very large Acp are excluded * BABAR/Belle/CLEO results in good agreement * Precision Acp & will require very large data sets!  One of the motivations for a next generation B Factory dd Acp dd NBB CLEO II (Data error bars are 1  & 90% CL )

13. Inclusive EM penguins: bs g s b q B (H-) • No tree level FCNC in SM • Sensitive to new physics in loop H-… • Calculated to NLO in SM ( 3.3 - 3.7 ± 0.33) ´ 10-4 • Measure: inclusive  spectrum • Past: Branching ratio & Acp. • Now: (+ shape of gspectrum) • not sensitive to new physics  N E  • Mean : <Eg > ~ mb /2 • Width: non-perturbative interactions • between b quark and light degrees of • freedom in hadron (Fermi motion) • Both mean & width needed for • extraction of Vcb & Vub from B Xl quark level hadron level

14. b s: Measuring the  spectrum Continuum Signal vs. or Expected raw contributions: • Signal: isolated g 2.0 < E g < 2.7 GeV • In principle: Measure g spectrum for ON and OFF resonance and subtract • But:b sg isn’t only source of g • I B  Xg: p 0gg gg previous analysis photon cut at 2.2 GeV, now model and subtract, significantly reduces model dependence • II huge continuum background: reduce by: • I shape cuts • II leptons (suppression and tagging) • III Identify strangeness containing hadronic system recoiling againstg  N E 

15. hep-ex/0108032 PRL on-line pub. Dec 3 2001 b s b  s g Preliminary • Full Cleo II + II.V datasetBFmeasured for ~90% of full spectrum (2.0 GeV cutoff) Theory: ( 3.3 - 3.7 ± 0.33) ´ 10-4 (Chetyrkin, Misiak,& Münz/ Kagan &Neubert,Gambino&Misiak) • Expt. & theory agree • Expt. error close to theoretical uncertainty • not much room for new physics • Belle (BCP4) measures: (2.25 GeV cutoff) ON-OFF data: b sg B(b  s g) BB bkgd MC E  Subtract BB bkgd: 1st moment: (CLEO) ‹Eg › = 2.346± 0.032 ± 0.011 GeV E 

16. Determination of Vcb & Vub • Semileptonic decays are used to determine the quark couplings as they simpler than hadronic decays i.e. the strong interaction is confined to the lower vertex • Vcb2 for final states with charm (D /D* etc.) • Vub2 for final states without charm ( //…) • Since of necessity we must work with hadrons rather than quarks, theory is needed to relate the underlying quark decay to hadronic reality: Semileptonic: Vcb or Vub or u Hadronic: hadronic: or // Hadronic: or u Theory converts rate to quark coupling Experiment measures rate or // • Two approaches inclusive B X l  or exclusive B  D* l , Bl

17. Determination of Vcb from B  D*  + • Heavy Quark Effective Theory • For the form factor (strong interaction physics) which measures the probability that the c quark forms a D* is unity. • Corrections for finite are 2nd order for BD*l and calculable • Since this is a 0 - 1- S,P,D wave decay large rate near • Measure and extrapolate to b c B meson Light d.o.f. b b quark is nearly at rest b spin has little effect on energy At c quark nearly at rest, light degrees of freedom unaware of flavor change B D*+ + Osaka (2000) , now also B D*0 + Rome (2001) (systematics limited ~1/3 of the CLEO data set)

18. Vcb from B  D*  + Form factor Phase space CLEO Consistency at the 7% CL theory Dominant sys errors: slow, form factors Single most precise excl. Vcb & B(DK)

19. A Road-map for inclusive |Vcb| first moment of g spectrum measure b sg spectrum first moment of hadronic mass spectrum measure B Xclnspectrum measure semileptonic width Spectator model (free quark decay) made rigorous by HQET+OPE a controlled expansion in s and 1/MB. Schematically: ~hyperfine interaction M(B*)-m(B) relates m(b) to m(B) ~average kinetic energy b quark in B meson 2 = 0.12 GeV2 1 Expt: Theory: predicts b sg first moment (with parameter ) predicts B Xclnfirst moment (with parameters and ) predicts semileptonic width in terms of , and (Falk, Ligeti, Luke, Wise, Savage, Manohar, Bauer, Bigi)

20. Want B Xclnhadronic mass distribution Identify lepton (P>1.5 GeV) Hermiticity: p Calculate hadronic recoil mass from  Drop because PBis small Fit spectrum with B Dl B D*l  B  XHl  (XH = D**, D*n… various models for XH) Find moments of true MX2spectrum B XclnHadronic Mass Moments Observed recoil mass: ON-OFF CLEO 2001 Fit D* D XH  Hep-ex/0108033 PRL on-line pub 3 Dec 2001

21. L andl1 bs 1st moment f(  ) Preliminary CLEO 2001 moments . BXl 1st moment f(l1 L) PRL on-line pub3 Dec 2001

22. Combine Extraction of |Vcb| CLEO 2001 hep-ex/0108033 Measured sl with L and l1: B(B Xcln ) = (10.390.46)% [CLEO] Preliminary (3.2% error !) (smallest error yet. PRL online 12/3/01) Inclusive assumes quark : hadron duality. Moments (hadr. now lep. soon) can validate inclusive method. Inclusive & exclusive methods both needed. Agreement will give confidence in Vcb determination

23. Determination of Vub • the quark process bul is simple • Theoretically difficult to calculate strong interaction effects when a heavy B meson becomes a light / (no Heavy Quark symmetry) • theoretical uncertainties enter twice, 1st the shape of the form factors determines the acceptance and hence Br • 2nd , the absolute normalization is needed for Vub • Severe background: bcl~ x100 bul lead to measurements in small regions of phase spacelarge extrapolation to obtain Vub • Two approaches: inclusive and exclusive The dangers of extrapolation: Inclusive methods: To distinguish bu from bc theoretically: better better q2 spectrum > mhad spectrum > Elepton spectrum But experimental difficulty is in opposite order Endpoint Rest of phase space

24. Determination of Vub • the quark process bul is simple • Theoretically difficult to calculate strong interaction effects when a heavy B meson becomes a light / (no Heavy Quark symmetry) • theoretical uncertainties enter twice, 1st the shape of the form factors determines the acceptance and hence Br • 2nd , the absolute normalization is needed for Vub • Severe background: bcl~ x100 bul lead to measurements in small regions of phase spacelarge extrapolation to obtain Vub • Two approaches: inclusive and exclusive The dangers of extrapolation: Inclusive methods: To distinguish bu from bc theoretically: better better q2 spectrum > mhad spectrum > Elepton spectrum But experimental difficulty is in opposite order Endpoint Rest of phase space

25. New Lepton endpointInclusive Determination of Vub • 1% of lepton spectrum is bul • Go beyond kinematic limit for b cl • Experiment measures Bub(end) : B(BX) in endpoint region Challenges: Limited understanding of decay spectrum/form factors Large extrapolation to get Vub (5-20% bu in endpoint) = fu(end) The endpoint is most influenced by the Fermi motion of the b quark in the B meson Uncertainty can be reduced by using bs photon spectrum shape parameters (mb and pFermi) to calculate fu(end) bs a laboratory for b  ul fu(end) = 0.1380.034 Neubert; Leibovich, Low, Rosthstein, DeFazio-Neubert

26. New Lepton endpointInclusive Determination of Vub After continuum, & background suppression: • endpoint :(2.2 pl 2.6 GeV/c), • Nub(end) = 1874 ± 123 ± 326 • Bub(end) = (2.35 ± 0.15 ± 0.45)  10–4. ON-OFF bcl MC Apply f u(end) Excess is b ul & QED radiative correction: 5% Use: Momentum (GeV/c) Hoang,Ligeti,Manohar, hep-ph/9811239 |Vub| = (4.09 ± 0.14 ± 0.66)  10–3. |Vub| / |Vub| =16% Dominant error: theory fu(end) The most precise inclusive determination to date

27. sin2 & Vub/Vcb From CLEO data Vub/Vcb is determined to 17% What are the implications ? • BABAR & Belle immediate objective: (sin2)mixing • Mixing :box diagrams new physics may enter: (sin2+ )mixing • The goal compare (sin2)mixing to sin2CKM i.e  from Vub/Vcb •  depends strongly on |Vub/Vcb| but weakly on  for 450 <  < 1100 • Take 450 <  < 1100 B0 Phase of Vtd =  |Vub/Vcb| 0.101  0.017 sin2CKM 0.95 < @ 90% CL sin2CKM = 0.74 0.09  0.08 (1st error Vub, 2nd error  range)

28. sin2 & Vub/Vcb |Vub/Vcb| = 0.101  0.017 sin2CKM = 0.74 0.09  0.08 (Assumes 450 <  < 1100) This agrees well with sin2 from BABAR and Belle: (sin2)mixing= 0.79  0.10 A significant consistency check of the CKM mechanism of CPV. Checks of this type, with increasing precision, will be the hallmark of heavy flavor physics in this decade. (Average computed by Belle) (sin2)mixing

29. B /l Vub Exclusive reconstruction • Since heavy  light HQET does not help • Theoretically difficult: evolution of form factor over large q2 range. • Hermiticity:  reconstruction: • * Model dependence dominates B l  CLEO (Bl) • Sort between models:probe q2 distribution no discriminating power at high lepton energy. New results soon B /l  El> 1.0 GeV

30. Future of Vub • future of Vub: • inclusive: reduce extrapolation error fit or large region error depends on size of accessible region 5% may be possible • ultimate exclusive method (may be more precise): ~100% phase space ~20% phase space u   b HQS I d   c Is related to at the same E(corrections O(1/M)) Assume unitarity. Measure in Dl, : calibrate lattice to 1% Lattice error on ~ 3% expected unquenched (Cornell/FNAL) Extract Vub at BaBar/Belle using B l & calibrated lattice calc. of But: need absolute Br(D l) and high quality neither exist

31. Importance of measuring fD & fDs: Vtd & Vts 1.8% ~15% (lattice) ~5% (lattice) Lattice predicts fB/fD & fBs/fDs with small errors if precision measurements of fD & fDs existed (they do not) could substitute in above ratios to obtain precision estimates of fB & fBs and hence precision determinations of Vtd and Vts Similarly fD/fDs checks fB/fBs

32. Summary of Recent CLEO Results • >60 rare B decays observed by CLEO. Branching ratios in good agreement with theory. No CPV observed. In almost all cases BABAR/Belle confirm CLEO results, & in some cases extend them. This trend will accelerate • New Vcb from BD*l (to <5%) • New Vcb from moments analysis of bs & BXl (to <5%) • New Vub from endpoint of lepton spectrum, where faction of rate in endpoint given by analysis of bs spectrum. • Provides a useful constraint on sin 2. This is the beginning of the era of precision cross checks of the b sector of the CKM matrix. To make this cross check much more precise theory needs measurements of absolute charm semileptonic branching ratios and form factors • Vtd & Vts extraction: lattice needs precision measurements of charm meson decay constants Most Precise Single Determi nations |Vub| = (4.09 ± 0.14 ± 0.66)  10–3.

33. CLEO-c : context SuperBaBar L =10 36 1010 BB/year CLEO Major contributions to B/c/ physics But, no longer competitive. Last Y(4S) run ended June 25 ‘01 BB /year year (*) CLEO No longer Operating at Y(4S) Best numbers as of Dec 3 ‘01 ON OFF CESR/CLEO(*) 1.3 16.0 6.7 34 KEKB/Belle 5.4 29.1 3.7 79 PEPII/BABAR 4.3 52.5 6.5 114

34. CLEO-c : the context • Flavor Physics: is in “the sin2 era” akin to precision • Z. Over constrain CKM matrix with precision • measurements. Limiting factor: non-pert. QCD. This Decade LHC may uncover strongly coupled sectors in the physics that lies beyond the Standard Model The LC will study them. Strongly-coupled field theories are an outstanding challenge to theoretical physics. Critical need for reliable theoretical techniques & detailed data to calibrate them. The Future Example: The Lattice Complete definition of pert & non. Pert.QCD. Matured over last decade, can calculate to 1-5% B,D,,… Charm at threshold can provide the data to calibrate QCD techniques  convert CESR/CLEO to a charm/QCD factory “CLEO-c/CESR-C”

35. CLEO-c Physics Program • flavor physics: overcome the non pert. QCD roadblock • strong coupling inPhysics beyond the Standard Model • Physics beyond the Standard Model in unexpected places: • CLEO-c: precision charm abs. branching ratio measurements Semileptonic decays: Vcs Vcd unitarity & form factors Leptonic decays : decay constants Abs D hadronic Br’s normalize B physics Tests QCD techniques in c sector, apply to b sector • Improved Vub, Vcb, Vtd & Vts • CLEO-c: precise measurements of quarkonia spectroscopy & • decay provide essential data to calibrate theory. • CLEO-c: D-mixing, charm CPV, rare decays of charm and tau. CLEO-c will help build the tools to enable this decade’s flavor physics and the next decade’s new physics.

36. CESR-C • Modify for low energy operation: • add wigglers for transverse cooling (cost $4M) Expected machine performance: • Ebeam ~ 1.2 MeV at J/

37. CLEO-c Proposed Run Plan 2002: Prologue: Upsilons ~1-2 fb-1 each at Y(1S),Y(2S),Y(3S),… Spectroscopy, matrix element, Gee, B hb 10-20 times the existing world’s data (started Nov 2001) 2003: y(3770) – 3 fb-1 30 million DD events, 6 million tagged D decays (310 times MARK III) C L E O - c 2004: MeV – 3 fb-1 1.5 million DsDs events, 0.3 million tagged Ds decays (480 times MARK III, 130 times BES) 2005: y(3100), 1 fb-1 & y(3686) –1 Billion J/y decays (170 times MARK III, 20 times BES II) A 3 year program

38. 1.5 T now,... 1.0T later 83% of 4p 87% Kaon ID with 0.2% p fake @0.9GeV 93% of 4p sp/p = 0.35% @1GeV dE/dx: 5.7% p @minI 93% of 4p sE/E = 2% @1GeV = 4% @100MeV CLEO III Detector  CLEO-c Detector Trigger: Tracks & Showers Pipelined Latency = 2.5ms Data Acquisition: Event size = 25kB Thruput < 6MB/s 85% of 4p For p>1 GeV

39. How well does CLEO III work? 1st results from CLEO III data at LP01 Preliminary result using ~1/2 of the CLEO III data Clean K/  separation at ~2.5 GeV using RICH Rest of reconstruction technique similar to previous CLEO analyses B(B-D0K-) = ( 3.8 ±1.3) x10-4 CLEOIII (2.9 ± 0.8) x10-4 CLEO II Good agreement: between CLEO III & CLEO II

40. 1st results from CLEO III data at LP01 Yield BR(BK)(x10-6) BK CLEOIII CLEO(1999) (Preliminary) Good agreement: between CLEOIII & CLEO II & with BaBar/Belle CLEO III works well

41. A typical Y(4S) event: y(3770) events: simpler than Y(4S) events A typical y(3770) event: CLEO III detector has: excellent tracking resolution excellent photon resolution maximum hermeticity Excellent particle ID Flexible triggering High throughput DAQ • The demands of doing physics in the 3-5 GeV range are easily met by the existing detector. • The CLEO Collaboration (150 physicists/ 18 universities & growing) has a history of diverse interests spread over b, c, tau, resonance and QCD studies & great enthusiasm for CLEO-c • BUT: B Factories : 400 fb-1  500M cc by 2005, what is the advantage of running at threshold?

42. Advantages of Running on Threshold Resonances • Charm events produced at threshold are extremely clean • Large , low multiplicity • Pure initial state: no fragmentation • Signal/Background is optimum at threshold • Double tag events are pristine • These events are key to making absolute branching fraction measurements • Neutrino reconstruction is clean • Quantum coherence aids D mixing and CP violation studies

43. Tagging Technique, Signal Purity • (3770) DD s ~4140  DsDs • Charm mesons have many large branching ratios (~1-15%) • High reconstruction eff •  high net tagging efficiency ~20% ! Anticipate 6M D tags 300K Ds tags: D Kp tag. S/B ~5000/1 ! Ds  (KK) tag. S/B ~100/1 MC MC Log scale! Log scale! Beam constrained mass

44. Absolute Branching Ratios ~ Zero background in hadronic tag modes *Measure absolute Br (D X) with double tags Br = # of X/# of D tags # of D's is well determined Double tags are pristine MC Decay s L Double PDGCLEOc fb-1 tags (dB/B %) (dB/B%) D0K-p+ 3770 3 53,000 2.40.6 D+  K- p+p+ 3770 3 60,000 7.20.7 Ds fp 4140 3 6,000 251.9 dB/B Includes Stat, sys & bkgd errors CLEO-c sets absolute scale for all heavy quark measurements

45. fDs from Absolute Br(Ds m+n) |VCKM|2 |fD|2 l n • Measure absolute Br (Ds mn) • Fully reconstruct one D (tag) • Require one additional charged track and no additional photons MC • Compute MM2 • Peaks at zero for Ds+  m+n • decay. • Expect resolution of ~Mpo Ds tag Ds mn Vcs, (Vcd) known from unitarity to 0.1% (1.1%)

46. Semileptonic Decays |VCKM|2 |f(q2)|2 D0 pln Lattice D0 pln Tagged Events Low Bkg! D0Kln D0 pln CLEO-C U = Emiss - Pmiss Assume 3 generation unitarity: for the first time measure complete set of charm PS  PS & PS  V absolute form factor magnitudes and slopes to a few% with ~zero bkgd in one experiment. Stringent test of theory! If theory O.K. determine Vcs and Vcd

47. CLEO-C Impact semileptonic dB/B, Vcd, & Vcs PDG CLEO-c Absolute rates : Vcs /Vcs = 1.6% (now: 11%) Vcd /Vcd = 1.7% (now: 7%) Use CLEO-c validated lattice + B factory Br/p/h/lv for ultra precise Vub

48. Unitarity Constraints |Vud|2 + |Vus|2 + |Vub|2 = 1 ?? With current values this test fails at 2.5 |Vcd|2 + |Vcs|2 + |Vcb|2 = 1 ?? With CLEO –c can test to ~3% Make the second row sensitivity comparable to the 1st row cu triangle |VubVcb*| |VudVcd*| Compare ratio of long sides to 1.3% Also major contributions to Vub Vcb Vtd Vts |VusVcs*|

49. Compare to B factories f Ds FDs at a B factory Scale from CLEO analysis • Search for Ds* -> Dsg, Ds -> mn • Directly detect g, m, Use hermeticity of detector to reconstruct n • Backgrounds are LARGE! • Precision limited by systematics of background determination • FDs Error ~23% now (CLEO) • 400 fb-1~6-9% CLEO-c CLEO signal 4.8fb-1 Excess of m over e fakes Background measured with electrons Ds mn DM=Mmng -Mmn Mass Difference M GeV/c

50. Comparison between B factories & CLEO-C BaBar 400 fb-1 CLEO-c 3 fb-1 Statistics limited abcdefghi Systematics & Background limited Current