www.lnf.infn.itesperimentiflatevpapersferrara2 by yurtgc548

VIEWS: 0 PAGES: 32

									                               Stefano Bianco
                       Laboratori Nazionali di Frascati




CPConf 2000, Ferrara    Results on Charm Mixing and CPV - S.B.   1
                                      Outline

•   Basic Mixing Phenomenology
•   FOCUS Detector
                                                                     J.Wiss, Fermilab Seminar Apr 2000
                                                                        FOCUS Coll. (J.M. Link et al.)
                                                                         Phys.Lett.B485:62-70,2000


•   Part I: Charm Mixing and yCP                                                  FOCUS Coll. (J.M. Link et al.)
                                                                                       hep-ex/0005037


•   Part II: Search for CP violation
•   Part III: Measurement of D*+ p+ (K+p-)
•   Conclusions and Outlook           Preliminary
                                                                              J.M. Link, DPF2000 Columbus (USA)




     CPConf 2000, Ferrara   Results on Charm Mixing and CPV - S.B.                                   2
                                                              _
                                              Formalism of P0-P0 mixing
The time evolution of flavour eigenstates P 0 , P 0
is given by the Schrödinger equation
                                                                                x 1            y  1
          P 0   H 11           H12  P 0 
       i  0                            0                        mixing can be described by the ratio     r
        t  P   H 21
                                H 22  P 
                                           
                         m jk  i jk / 2                                       ( P 0  P 0  f ) x 2  y 2
                 H jk                                                         r                   
                                                                                    ( P 0  f )        2
Processes which allow P  P mixing appear
                                  0       0

in H12 and H21.
If H12  0 , H 21  0       P 0 , P 0 are not
mass eigenstates. Mass eigenstates are
                                                                                                   |x|>0    mixing is
(assuming CP conservation)                                            |y|>0        mixing
                                                                                                           caused by
                                                                      is caused by the
                              
                 P,2  1 / 2 P0  P 0
                  1                                                  short-lived
                                                                      component                      P0  P 0
                                                                                                             genuine

and mixing amplitude is                                               disappearing                 transitions(f=2)
                         1t                   
                  2                  t                              rapidly, leaving
                                t
  P 0 P 0 (t )         e 1  e  2e 2 cos(mt )                    behind the long-
                                                                    lived component,
                                                                      made of P 0 and P 0
Mass and width differences are parametrized                                 y splits                   x splits
by                                                                                                     masses
             m                                                          lifetimes
          x                      y
                                    2
          CPConf 2000, Ferrara                Results on Charm Mixing and CPV - S.B.                        3
                    K0, D0, B0 _ Where’s The Difference ?


                                                                0           KK,pp
                                                             D                      D0
                                                             K0             p0,h    K0
           SHORT DISTANCE                                            LONG DISTANCE

 m 2
          md
            2
                
                2
    q
                            Also: recent OPE-based limits
     2
  m ms2,c ,b
    W
                            I.Bigi, N.G.Uraltsev hep-ph/0005089


                        q   SD/LD            x=m/             y=/2)

           K 0 ( ds )   c      SD~LD          0.48                   100%

           D 0 ( cu )   s     SD<<LD            ?           THIS MEASUREMENT
            0           t     SD>>LD          0.75                     ?
           Bd ( db )
           Bs0 ( sb )   t     SD>>LD           >22                     ?




CPConf 2000, Ferrara        Results on Charm Mixing and CPV - S.B.                       4
                       Theoretical “guidance”
                                                     From compilation of H.N.Nelson
                                                            hep-ex/9908021

                                                        Triangles are SM x
                                                         Squares are SM y
                                                         Circles are NSM x
                                                                            3
                                                           x, y SM  10
                                                     _
                                              “D0D0 mixingis a
                                              window on New
                                                 Physics”
CPConf 2000, Ferrara     Results on Charm Mixing and CPV - S.B.                  5
                                                             _
           TECHNIQUES FOR STUDYING D0D0 MIXING (I)



Measure yCP, the y parameter                                       Measure r
    for CP eigenstates                                        ( D 0  D 0  f ) x 2  y 2
                                                           r                   
                                                                 ( D  f )
                                                                      0
                                                                                     2
                                     2
CP K  K    K  K 
CP K S   K S          These Results
                              Also E791, BELLE

CP K p   mixed                                          HADRONIC           SEMILEPTONIC
                                                            DECAYS               DECAYS
                           (1 +2)/2
                                                         f  K p           f  K 
                                                                                     
                                                                                        

                                                         E791, CLEO, ALEPH          E791




   CPConf 2000, Ferrara   Results on Charm Mixing and CPV - S.B.                      6
                                                                                   _
                 TECHNIQUES FOR STUDYING D0D0 MIXING (II)
                        ( D 0  D 0  f ) x 2  y 2
Measure r            r                                                    via    HADRONIC
                                                                                    DECAYS
                                                                                                     f  K p 
                           ( D 0  f )        2
                                                   Cabibbo-Favoured
                               mixing                   Decay

                                                                                     Strong phase d
                               D0  D 0  K p                                   between CFD and DCSD
                                   c    s                 u                          mixes x and y
                           D   0
                                   u    s   c             s   K
                                                              p                        y  y cosd  x sin d
D* tags D0
 flavour
                                             Vcs          d
                                                    Vud                                
                                                                                        x  x cosd  y sin d
                                                          u
D p D
*              0
                                                              u
                                                                  K
                                                    Vus
                                   c         Vcd              s
                                                              d
                           D   0
                                   u                          u   p
                                   D 0  K p 
                        Double Cabibbo-Suppressed Decay


                                                                          ( D 0  f )    t         t2           
                                                                  rWS                  e rDCS  r   y rDCS t 
                                                                          ( D 0  f )               2            

      CPConf 2000, Ferrara              Results on Charm Mixing and CPV - S.B.                                   7
                                                                          _
              TECHNIQUES FOR STUDYING D0D0 MIXING (III)
                     ( D 0  D 0  f ) x 2  y 2                   via   SEMILEPTONIC
Measure r         r                                                        DECAYS      f  K    
                        ( D 0  f )        2
                                                 Cabibbo-Favoured
                              mixing                  Decay




                              D 0  D 0  K    
                                  c    s               u
                         D    0
                                  u    s   c           s   K
 D* tags D0
  flavour
                                           Vcs             
                                                           
D*  p  D0




                                                                                  ( D 0  f )
                                                                          rWS                 r
                                                                                  ( D  f )
                                                                                       0



       CPConf 2000, Ferrara            Results on Charm Mixing and CPV - S.B.                     8
                                                       _
     TECHNIQUES FOR STUDYING D0D0 MIXING (Concl.)


• Direct measurement of yCP requires good
lifetime resolution
• Measuring rws in hadronic decays requires
     to disuntangle rDCS from r;
     and the knowledge of d
• Measuring rws in semileptonic decays
provides the r parameter, but it is
experimentally more difficult


 CPConf 2000, Ferrara   Results on Charm Mixing and CPV - S.B.   9
                         Vertexing
                                                           Spectrometer at
                                                   FNAL Wide Band Photon
                                                   Beam
                                     Cerenkov




                                                                      Over 1 million
                                                                      reconstructed!
                             PWC

Successor to E687. Designed to study charm particles
produced by ~200 GeV photons using a fixed target
spectrometer with updated Vertexing, Cerenkov, EM
Calorimeters, and Muon id capabilities. Member groups
from USA, Italy, Brazil, Mexico, Korea.
  CPConf 2000, Ferrara       Results on Charm Mixing and CPV - S.B.             10
         The FOCUS Collaboration

         Univ. of California-Davis, CBPF-Rio de Janeiro, CINVESTAV-Mexico City,
           Univ. Colorado-Boulder, FERMILAB, Laboratori Nazionali di Frascati,
            Univ. of Illinois-Urbana-Champaign, Indiana Univ.-Bloomington,
                         Korea Univ.-Seoul, INFN and Univ.-Milano,
Univ. of North Carolina-Asheville, INFN and Univ.-Pavia, Univ. of Puerto Rico-Mayaguez,
             Univ. of South Carolina-Columbia, Univ. of Tennessee-Knoxville,
       Vanderbilt Univ.-Nashville, Univ. of Wisconsin-Madison, Yonsei Univ.-Seoul




  CPConf 2000, Ferrara     Results on Charm Mixing and CPV - S.B.                  11
   Detector Performance related to lifetime study
• Segmented target:                             • Vertex algorithm is driven by D
– 62% of charm decay in air.                    candidates:           Detachment cut
– small absorption correction.                  – The Fit variable ( reduced proper
• Extremely good proper time                                         
                                                time t     N  / c )
                                                acceptance is flat.
resolution: ()/(D0) = 8 % (30fs)
                                                • Excellent ( and flexible)
– No resolution convolution                     Cerenkov identification:
systematics/error inflation 
                                                –Minimized systematics on particle
Binned likelihood method.
                                                ID misidentification.

                              primary vtx            BeO
                        BeO   secondary vtx
                                                                  tarsil
                                       tarsil




                                                                          Z (cm)
 CPConf 2000, Ferrara    Results on Charm Mixing and CPV - S.B.                    12
                   Selection of D  Kp, KK
                                              • Common base cuts: detachment (l /
• Important: Get a sample with                > 5), Kaon ID (Wp-WK > 4)
    a flat efficiency over t’                 • Tagged sample:
                                                | M ( D*)  M ( D)  145.4 |  3 MeV
     Kp                                       • Or inclusive sample:
       Base is
       l/ >7 &                               More Cerenkov cuts : pion ID (W*-Wp
       kaon>1                                 > -2), Kaon ID 2 ( WP-WK > -2 )
                                                                             p  p2
                                              Mom sym :   0.7 where   1
                                                                             p1  p2
                                              Primary vertex in target (PIT)
                                                                               
                                              Resolution:  60 fs where   l
                                                                               c
                                                 D  Kp: 119738 selected
                                                 D  KK: 10331 selected
  CPConf 2000, Ferrara   Results on Charm Mixing and CPV - S.B.                 13
          DKK signal for several cleanups

                       l / > 5                                    l / > 5
                                   Wp                              Wp-WK>4
                       -WK>1
                       Y=16532                                     Y=10331
                       S/N=2.3                                     S/N=2.7
                                  Kp
                                  reflection



                        l / > 9
                                                                   l / > 9
                        Wp-WK>1
                                                                   Wp-WK>4
                        Y=11528                                    Y=7151
                        S/N=4.3                                    S/N=5.7




CPConf 2000, Ferrara      Results on Charm Mixing and CPV - S.B.              14
  Fit technique: non-parametric background
  • Binned likelihood: 20  200 fs                             1.15
                                                                1.1          Kp
  • Acceptance/absorption f(t’)                                              KK
                                                               1.05
  correction by MC ~ nearly 1
                                                                 1
  • i: signal in each t’ bin                                  0.95

  • bi: background from sidebands                               0.9          f(t’) vrs t’ (fs)
                                                               0.85
  • Fit  Kp, B Kp (but see next                                  0          500    1000     1500       2000        2500       3000
  page)                                                                    • option: w or w/o B-tie term.

                   f ( t 'i ) exp(
                                       t 'i                                      ini e i             (2 B)( N1  N2 ) e (2 B ) 
                                                )
                                                B bi           w  2   i ln                  2 ln                             
 i  ( N s  B)                                                        
                                                                                  ni !         
                                                                                                  
                                                                                                               ( N1  N 2 )! 
                     i f ( t 'i )e  t 'i    
                                                       i bi            
         25000

         20000                                           Ns                                           DKp
         15000

         10000                     N1                                             N2
           5000                                                  B
               0
                   1.74             1.79             1.84                 1.89         1.94        1.99            2.04

     CPConf 2000, Ferrara                     Results on Charm Mixing and CPV - S.B.                                        15
                               Fitting technique
• The KK sample has some Kp
reflection at its side; B2 = Kp +                       • Background under KK signal 
nominal  one more player in fitting.                   B1 + (B2 - Kp reflection)
• Subtract Kp reflection by a mass fit.                 • Simultaneous time evolution
– The reflection mass shape from MC.                    fit of both Kp and KK histos.

– The subtraction level by the mass fit.                • 4 variables in the time fit:

– Time evolution of the reflection from                  Kp, yCP, B Kp, B KK
t (Kp)



  4500
  4000
  3500                                                                      DKK
  3000
  2500             Region B1                    Region B2
  2000
  1500                                                                Kp reflection
  1000
   500
     0
     1.74        1.79
   CPConf 2000, Ferrara    1.84
                            Results       1.89         1.94
                                      on Charm Mixing and CPV - S.B.1.99       2.04      2.09
                                                                                           16
                       Fitted time evolutions
                                               100000

                                                10000                                   Kp
        Kp              l/ > 5                 1000

                        Wp-WK>4                  100

                                                  10
                                                            back+signal
                                                   1
                                                            background
        KK                                              0       1000      2000   3000        4000


                                               10000

                                                1000                                    KK
                                                 100

                                                  10
   • Background subtracted                         1
   and f(t’) corrected time                             0       1000      2000   3000         4000

   evolution of Kp and KK
   events in the final fit.

CPConf 2000, Ferrara     Results on Charm Mixing and CPV - S.B.                          17
                     Ycp: Results and Systematics
100%
                CL
50%                                                                 3 K ID  3 l/  B-tie or not
                                                                     15/20 bin = 36 fit variants
 0%
                                                                    shown.
8%
            YCP          B- tie                    B- tie          yCP = 3.421.390.74 %
6%

4%                                                                  Sample standard deviation
2%
                                                                    of fit variants is 0.63.

0%
               15 bin                        20 bin

                                                                  (Kp) = 409.21.3?? fs
412


410
                                                                   Sample standard deviation of
408
                                                                   fit variants 0.3.
            Kp                                                    Absolute lifetime systematics
406
                                                                   not ready until we analyze
                                                                   K3p, etc.
       CPConf 2000, Ferrara       Results on Charm Mixing and CPV - S.B.                   18
                    Additional systematic checks
8%
                                          1/2 sidebands                         Could curvature in
6%                                                                              reflection line-
4%                                                                              shapes cause the
                                                                                sideband
2%
                                                                                background to
0%
                                                                                fail? Let’s reduce
                                                                                the sideband size
     Results are nearly identical to standard fits
                                                                                by half and see...
8%
                             f(t')=1
6%                                                                              How important is
                                                                                the f(t’)
4%
                                                                                correction? Set
2%                                                                              f(t’)=1 and rerun
                 15 bins                          20 bins
0%                                                                              all the fits.
      CPConf 2000, Ferrara             Results on Charm Mixing and CPV - S.B.                19
            Consistency of D* and full sample
                         D* only path
10%
                         B- tie
8%
                                                         B- tie       Further suppress
6%                                                                    non-charm and
4%                                                                    non-Do reflection
2%                                                                    by looking at D*-
0%                                                                    only sample.
-2%             15 bins                       20 bins

                          Inclusive                                      D* tag
                           +D* tag                                        only
                              l/ > 5
                              Wp-WK>4




  CPConf 2000, Ferrara       Results on Charm Mixing and CPV - S.B.               20
    Comparisons to CLEO, E791 and BELLE
                                            The comparison to CLEO is valid only if
                                            one assumes a small strong phase
      FOCUS D0  K  K  / K p 
      FOCUS                                 difference d.
      CLEOII.V         D 0  K p 
                                            About the same sensitivity to the CLEO CP
                       D  K  
                               
      E791                                  constrained fit, but the opposite sign!
                        0




                                                                        FOCUS

                                                     yCP = 3.42  1.39  0.74 %

                                                            Recent Measurements
         E791     D0  K  K  / K p             E791: yCP = 0.8  2.9  1 %
         BELLE     D0  K  K  / K p            CLEO: -5.8 % < y’ < 1% ( 95% CL)
95% CL                                             BELLE prelim.: yCP =


CPConf 2000, Ferrara           Results on Charm Mixing and CPV - S.B.                 21
              Search for CP asymmetry in charm decay
 CP asymmetries may show via the interference of two distinct amplitudes with phases
 d1 and d2 which contain both a weak (CKM) and a strong (FSI) contribution.
                                             A  aeid1  beid 2
The weak contribution in the phases changes sign under CP, while the strong one does
not. The most accessible decays which contain two weak phases are the SCSD (tree and
penguin diagrams). The CP asymmetry will then be


                                                          
                                                   2 ab* sin d 2  d 1 
                                     2
                             A A
                              2

                    CP                 
                             A A
                              2      2
                                                                
                                             a  b  2 ab* cosd 2  d 1 
                                               2     2




   Buccella et al predict state specific asymmetries in the range of 0.002  0.14 %
   The Do asymmetry is complicated by a direct as well mixed contribution but
   expectation are that the D° might be smaller than D+



      CPConf 2000, Ferrara        Results on Charm Mixing and CPV - S.B.        22
Search for CP asymmetry in charm decay (DKK)



                                                        Kp MC




CPConf 2000, Ferrara   Results on Charm Mixing and CPV - S.B.   23
Search for CP asymmetry in charm decay (DKKp)


• Cabibbo
                                  D+                             D–
suppressed
mode.
D+  KK+p+                                    D S+                   D S

D  KK+p



• Cabibbo
favored mode.
D+  K p +p+
D  K+p p


 CPConf 2000, Ferrara   Results on Charm Mixing and CPV - S.B.               24
                          CP asymmetry results

     Decay Mode        FOCUS            E791 (previous best published)
    D+K K+ p+ +0.006  0.011  0.005 0.014  0.029
    D0K K+    0.001  0.022  0.015 0.010  0.049  0.012
    D0p p+              +0.048  0.039  0.025 0.049  0.078  0.025


• No evidence for CP violation.
• Our limits on       K  K p  are much better than K  K 
    – Need to use tagged Do sample which cuts our sample by 80%.
• 2~3 times better limit than the previous published measurements.
    …also, New Unpublished CLEO limits on KK, pp (CIPANP,
Quebec City 2000)



   CPConf 2000, Ferrara     Results on Charm Mixing and CPV - S.B.        25
   Measurement of D*+ p+ (K+p-)
                           Event Selection
• Very loose Cerenkov based particle id cuts on K and p.
• The D0 candidate is used as a seed to find the production vertex.
• The production vertex has at least 2 tracks in addition to the D0.
• The production vertex is required to be within 1 of target material.
• Production and decay vertices are required to be well formed (CL>1%).
• D0 daughter tracks inconsistent with coming from the production vertex.
• The vertex separation L/L>5.
• Cut Kp pairs with high momentum asymmetry and low D0 momentum.



CPConf 2000, Ferrara   Results on Charm Mixing and CPV - S.B.             26
   Measurement of D*+ p+ (K+p-)
                First You Have to Tag the D0 Flavor
 • The decay D+*gD0p+ is used to identify the D0 flavor.
 • So we study the D*D0 mass difference.




                              D0                                   D*D0
                                                                 (CF-like Tags)
    Kp+p0

             K+K
                       p+p




CPConf 2000, Ferrara    Results on Charm Mixing and CPV - S.B.                    27
   Measurement of D*+ p+ (K+p-)
The Worst BG is CF Kp Double Mis-id
                                              So we use a tight Cerenkov based
                                                                   m
                                              mis-id cut in a 4 window about
                                              the D0 with Kp reconstructed as pK.




                                                                 Standard cuts.
 The double mis-id m is                                         Double mis-id cut.
 indistinguishable from the
 correctly identified signal.
CPConf 2000, Ferrara   Results on Charm Mixing and CPV - S.B.                          28
                          Measurement of D*+ p+ (K+p-)
                                                    A New Method
                       • Divide the data into 1 MeV wide bins in m, and fit the D0 in each bin.
                       • Fit the KK and pp reflections with Monte Carlo events.

                                                                                  • Fit BG to a polynomial.
D*D mass difference




                                                                                  • Fit D0 to a gaussian.
                                                                              .
                                                                                                  DCS-like tags
                                                                                                 146<m<147 MeV




                                                             .   .

                                                       Kp mass

                       CPConf 2000, Ferrara   Results on Charm Mixing and CPV - S.B.
                                                                                       A total of 80 fits!
                                                                                                         29
       Measurement of D*+ p+ (K+p-)
                              Fit the m Distributions
           • Fitted D0 yields are plotted in the appropriate m bins.
           • Background is fit to: f (m)  a (m  mp )  b(m  mp ) .
                                                      12           32



           • DCS signal is fit directly to the CF histogram signal region.

                           Yield  35901 196                    Yield  172.9  33.7




CLEO:
(0.332
0.064
0.040)%           rWS= (0.482  0.093)% Preliminary! Stat err only
    CPConf 2000, Ferrara       Results on Charm Mixing and CPV - S.B.                   30
                                         Conclusions
                                            •Measurement of yCP, asymmetry between
     FOCUS
     FOCUS D
                  0
                       K  K  / K p     the KK(CP=1) and Kp(CP=mixed) lifetimes
     CLEOII.V         D 0  K p 

     E791             D 0  K    
                                                               yCP = 3.42  1.39  0.74 %
                                                               (Kp) = 409.4  1.34 ?? fs

                                            •New limits on CPV asymmetries
                                              Decay Mode        FOCUS            E791 (previous best published)
                                             D+K K+ p+ +0.006  0.011  0.005 0.014  0.029
                                             D0K K+    0.001  0.022  0.015 0.010  0.049  0.012
                                             D0p p+         +0.048  0.039  0.025 0.049  0.078  0.025
       E791     D0  K  K  / K p 

       BELLE D0  K  K  / K p 

95% CL                                       •   Preliminary       Measurement of D*+ p+ (K+p-
                                                                 Assuming No Mixing
                                                 rDCS = (0.482+-0.093)% (stat only)
 CPConf 2000, Ferrara                    Results on Charm Mixing and CPV - S.B.                          31
                                             Outlook
                                                •FOCUS yCP result becomes more compatible
                                                to CLEO r’ measurement if a very large strong
                  0
                       K  K  / K p 
     FOCUS
     FOCUS D                                    phase d is allowed
     CLEOII.V         D 0  K p 

                      D 0  K    
                                                •Fairly intense theory production to explain
     E791
                                                the CLEO and FOCUS mixing results
                                                     1.    I.I.Bigi, ICHEP2K Osaka Plenary talk hep-ph/0009021
                                                     2.    A.Golutvin, ICHEP2k Osaka Plenary talk
                                                     3.    A.Petrov, hep-ph/0009160
                                                     4.    Yosef Nir, hep-ph/0008226
                                                     5.    D.Atwood et al., hep-ph/0008090
                                                     6.    J. P. Silva, hep-ph/0007214
                                                     7.    J. L. Rosner, hep-ph/0007194
                                                     8.    J.L. Rosner, hep-ph/0005258
                                                     9.    S.Bergmann et al., Phys.Lett.B486:418-425,2000
                                                     10.   I.I. Bigi, hep-ph/0005089

       E791     D0  K  K  / K p            •Expect new FOCUS results on r’ soon
       BELLE D0  K  K  / K p               •Expect new CLEO and FOCUS results on r
95% CL                                          from semilep analysis soon
                                                •The FOCUS 2.2 evidence for charm mixing
                                                should be verified soon by B-factories

 CPConf 2000, Ferrara                    Results on Charm Mixing and CPV - S.B.                             32

								
To top