1 Evidence for a Near-Threshold Structure in the J/  from B +  J/  K + Decay at CDF Kai Yi University of Iowa (for CDF Collaboration) Fermilab, March 17, 2009

2 History—from strange to bottom discovery  - discovery J/  (cc) discovery  (bb) discovery BNL FNAL Heavy flavor spectroscopy helped turn quarks into a reality! B +  J/  K + decay includes b,c,s,u quarks SLACBNL

3 Quark Model Six quarks exist in Standard Model charm/bottom/top called heavy no free quarks, they have to be bound top so heavy, it decays before forming bound states Baryons are bound states of three quarks Mesons are bound states of quark and anti-quark quarkonia: quark and its own antiquark Quark model works pretty well Challenged by newly discovered charmonium-like states c c b u d

4 X(3872) X(3872)  J/  +  - M=3871.8±0.7±0.4 MeV  < % CL PRL 91, ?? mass ~70 MeV > predictions (2003)  +  - peak at high value like a ρ (2003) J PC = 1 ++ or 2 -+ CC

5 Y(3940)  J/ , Y(4260)  J  +  Above DD && DD* threshold, tiny Branching Fraction expected New mass and width from BaBar: M   2.0,   5 MeV at the J/  threshold ? Well above DD && DD* threshold, tiny Branching Fraction expected J PC =1 --, plus Y(4350), Y(4660) too many 1 -- ? PRL 94, Y(3940)  J/  ω M ≈ 3940±11 MeV  ≈92±24 MeV PRL 95, Y(4260)  J/  +  - M ≈ 4259±8 MeV  ≈ MeV

6 Z(4430) +  (2S)  PRL 100, Z(4430) +  (2S)  + M ≈ 4433±4 MeV  ≈ 44 ±17 MeV Needs confirmation The first charged charmonium-like new state, if confirmed Many more new states… They do not (easily) fit into charmonium Beyond (qq) mesons: exotic mesons?

7 Exotic Mesons—QCD predictions c u u c Multi-quark mesons molecule, diquark-antidiquark Hybrid mesons quark-antiquark-gluon Glueball gluonic color singlet states Explore more channels to understand How about J/ψ ϕ ? GeV, VV, C=+) (cc) with a mass above GeV, expect tiny branching fraction g g c g c

8 Charmonium hybrid  J/  ϕ ? PRD 57, 5653 (1998) F. Close et al accessible J/ψ ϕ is also accessible to 0 -+, 2 -+ (0 -+,1 -+,2 -+ ) masses are expected near Y(4260), E. Eichten

9 Multi-quark states  J/  ϕ ? arXiv: (new) N. V. Drenska et al J/ψ ϕ is well motivated! How to search? Inclusive? Challenge! Through B decays!

10 Search structures  J/  through B decays Experimentally easy to search through clean B  J/  K channel -- taking advantage of B lifetime and narrow B mass window -- B  J/  ϕ K is OZI suppressed, so low physics background vacuum polarization PRL 84,1393 gluon coupling PRL 91,

11 The current status statistically limited, no structures reported The current status through B  J/  K : BaBar 2003, PRL 91, B + 13 B 0 J/  and ee CLEO, PRL 84,1393 J/  and ee, 10 B + and B 0

12 B hadrons at Tevatron are: copiously produced boosted --vertex separation --boost low p T daughters CDF has: excellent mass resolution excellent vertex resolution reasonable hadron PID CDF detector Why search at CDF? mass

13 Muon: μ ID ToF: TOF COT: track p dEdx Silicon: track p vertex CDF detector

14 CDF Di-muon trigger Scale to ~2.7 fb -1, ~ 20M J/  with silicon hits used for this analysis p T > 1.5 GeV both muons

15 CDF hadron PID 1.5σ dE/dx parameterized using D* +(-) → D 0 π +(-) sample dE/dx efficiency ~100% dEdx residual CDF Time-of-flight: Tevatron store /23/2001 Excellent resolution Time-of-Flight acceptance+efficiency ~60% Make use of both dEdx and ToF for hadron PID summarizing dEdx and ToF into a log-likelihood ratio Main background: prompt pions, need PID to suppress

16 CDF hadron PID 1.5σ 3.0σ K-π separation Typical B decay daughter momentum ~GeV

17 Analysis strategy I) Reconstruct B + as: B +  J/   + J/   +  -  +  - II) Search for structure in J/  mass spectrum inside B + mass window ++ -- ++ -- ++ J/   B+B+ L xy primary vertex secondary vertex Search? Vertex separationParticle Identification

18 I) Reconstruct B +  J/  K +

19 The challenge Start with typical requirements for B hadron at CDF: --p(  2) for B + vertex fit>1% --p T (track)>0.4 GeV, -- >=4 r-  silicon hits --p T (B + )>4 GeV --mass window: J/  (  50 MeV) and  (  7 MeV) constrain  +  - to J/  PDG mass value NOT applied yet: L xy and kaon PID B + mass Typical hadron collider environment

20 Applying L xy B + mass +L xy >500 um Maximize S/√(S+B) for B +  J/  K + signal, has nothing to do with J/  Maximized cuts: L xy >500  m, kaon LLR>0.2 L xy Reduce background by a factor of ~200

21 Applying L xy and kaon PID 75  10 define  3  as B + signal region (17.7 MeV obtained from MC) Purity ~80% in B + region Gaussian function mean fixed to PDG rms fixed to resolution (5.9 MeV) Kaon PID reduce background by a factor of ~100 clear B +  J/  K + signal

22 Control channels We also reconstruct two control channels with similar cuts: ~3 000 B s  J/ , ~ B +  J/  K + before L xy and kaon LLR cuts Clean control signals after L xy and kaon LLR cuts cross check and efficiency evaluation B s  J/  ~ 800 events B +  J/  K + ~21k events

23 Background reduction factor Reduce background by a factor of by using L xy and kaon PID cuts while keeping about 20% of signal as estimated by control channels. Any background under  ? Before L xy >500 um, kaon LLR>0.2After L xy >500 um, kaon LLR>0.2

24 Verify B +  J/  K + Investigate components of B + peak --relax K+K- mass window to: [1.0,1.04] MeV --do B + sideband subtraction for K + K - --fit to sideband subtracted K + K - mass A P-wave relativistic BW only fit to data with  2 probability 28%, no f 0  K + K - or K + K - phase space components with our ϕ mass window Conclusion pure B +  J/  K + for B+ peak negligible B +  J/  f 0 K +, J/  K + K - K + components

25 II) Search for structures in J/ψϕ spectrum from B

26 Monte Carlo

27 Investigate J/  mass spectrum in MC MC simulated phase space, full detector simulation MC events smoothly distributed in Dalitz plot No artifacts in the J/  mass spectrum How about reflections from other B hadrons?  M=m(  +  - K + K - )-m(  +  - )

28 Investigate J/  mass spectrum in MC We simulate generic B hadron decays with a J/  in the final state and we identified a contamination channel: B s  (2S) ,  (2S)  J/  +  - Bs contamination at  M>1.56 GeV, cut it off for simplification (B s  (2S) ,  (2S)  J/  +  - )  B +  J/  K + due to kaon mis-identification 20 times Luminosity of data

29 Data

30 Search for structures in J/  mass--Data Three-body Phase Space Background shape is different from data An near threshold enhancement is observed  M=m(  +  - K + K - )-m(  +  - )

31 Robustness test Extensive cross checks by varying L xy, kaon PID LLR, B + mass window, vertex probability, # of silicon hits,… Robust against variations More signal but with more background L xy >100  m p(vertex-  2 )>10 -5 SVX hits  3 Default cuts  M=m(  +  - K + K - )-m(  +  - )

32 Search for structures in J/  mass--Data We model the Signal (S) and Background (B) as: S: S-wave relativistic Breit-Wigner B: Three-body decay Phase Space  M=m(  +  - K + K - )-m(  +  - ) Yield =14  5  m =  2.9 (stat) MeV/c 2 Width = (stat) MeV √(-2log(L max /L 0 ))=5.3, need Toy MC to determine significance for low statistics Convoluted with resolution (1.7 MeV)

33 Significance study We determine significance from simulation (Toy MC): --Using Three-body decay Phase Space only to generate the  m spectrum --Find the most significant fluctuation for each trial anywhere in  m between 1.02 and 1.56 GeV width between 1.7 MeV (resolution) and 120 MeV (10X observed width) --Count it if -2log(L max /L 0 ) (-2Δln) ≥ -2Δln value in data P-value: 9.3X10 -6 (3.1 million total trials) corresponding to 4.3  -2Δln

34 Significance study Cross check: --fit -2Δln distribution with  2 Probability Density Function to get n (d.o.f): --calculate p value using  2 PDF as cross check -2Δln Most conservative: Phase Space and flat background for non-B background, 3.8  P-value by integrating  2 PDF: 6.5X10 -6, corresponding to 4.3 , consistent with counting Best estimation of significance More conservative

Including systematics: Yield =14  5  m =  2.9 (stat)  1.2 (syst) MeV/c 2 Mass =  2.9 (stat)  1.2 (syst) MeV/c 2 (adding J/ψ mass) Width = (stat)  3.7(syst) MeV Significance: at least 3.8σ for most unphysical conservative background Width indicates a strong decay tentatively name it as Y(4140) 35 Results

36 What is it? Y(4140) Well above charm pair threshold Expect tiny BF to J/  Does not fit into charmonium Close J/  threshold like Y(3940) arXiv: [hep-ph] molecular?

37 Opportunities Determine J PC ( c=+ )? Need statistics --increase efficiency, reduce background --add more data,  5σ --investigate efficiencies against angles? … More channels for this structure? --open charm pair? Note: Search for potential more structures? B +  (2S)  K +, B +  K +, B s  J/   (nS) , …

38 Summary Mass =  2.9 (stat)  1.2 (syst) MeV/c 2 Width = (stat)  3.7(syst) MeV J PC = ? ?+ tentatively name it as Y(4140) Stay tuned! CDF

39 Backup 1 ~ 400 fb -1 m ee  J/  ee 220 fb -1 m  J/  ee is difficult but not impossible Trigger is gone 

40 Backup 2 Not close from the PDF comparison although they both have C=+ X(4160)  D*D*

41 Backup 3  M=m(  +  - K + K - )-m(  +  - )

Tevatron

43 Exotic J PC For qq meson system, let L to be the orbital angular momentum. The meson spin J is given by |L-S|<J<|L+S|, where S=0 (antiparallel quark spin) or 1 (parallel quark spin) The parity P and charge parity C of the meson system can be expressed as: P=(-1) L+1 C=(-1) L+S In the configuration of P=(-1) J, S=1, CP=+1,  Exotic J PC (not allowed for qq meson): 0 --, 0 +-,1 -+,2 +-,… But exotic mesons can have these JPC due to additional degree of freedom. Identify exotic J PC is helpful to identify exotic mesons