A. Drutskoy, University of Cincinnati Results and prospects of Y(5S) running at Belle. March 14, 2008, Lausanne, Switzerland. LPHE seminar Results and prospects of Y(5S) at Belle, March 14, 2008, Lausanne A. Drutskoy LPHE seminar

Outline Recent Belle measurements at Y(5S). Prospects of B s meson (and other) studies at Y(5S). Introduction. Conclusion. My thoughts (speculations) about Y(5S) -> Y(6S) ->…. 2

CLEO PRL 54, 381 (1985)  ( 5S )  (4S)  (6S) e+e+ e-e- B B _ e + e - hadronic cross section Resonance to continuum hadron production ratios are Y(4S)/Cont ~ 1./3.5 and Y(5S)/Cont ~ 1./10. bb (cc,ss,uu,dd)

CLEO PRL 54, 381 (1985)  (5S) Running at Y(4S) and Y(5S) e + e - -> Y(5S) -> BB, B * B, B * B *, BB , BB , B s B s, B s * B s, B s * B s * where B * -> B  and B s * -> B s  Asymmetric energy e + e - colliders (B Factories) running at Y(4S) : Belle and BaBar _ ______ 1985: CESR (CLEO,CUSB) ~ 0.1 pb -1 at Y(5S) 2003: CESR (CLEO III) ~ 0.42 fb -1 at Y(5S) __ B s rate is ~10-20% => high lumi e + e - collider at Y(5S) -> B s factory. 2005: Belle, KEKB ~ 1.86 fb -1 at Y(5S) CLEO PRL 54, 381 (1985)  ( 5S )  (4S) 2006, June 9-31: Belle, KEKB ~21.7 fb -1 at Y(5S) e + e - ->Y(4S) -> BB, where B is B + or B 0 meson M(Y(5S)) =  8 MeV/c 2 (PDG)  (Y(5S)) = 110  13 MeV/c 2 (PDG) 4

Belle First Y(5S) runs at the KEKB e + e - collider 3.5 GeV e + 8 GeV e - Electron and positron beam energies were increased by 2.7% (same Lorentz boost  = 0.425) to move from Y(4S) to Y(5S). No modifications are required for Belle detector, trigger system or software to move from Y(4S) to Y(5S). Integrated luminosity of ~1.86 fb -1 at 2005 and ~21.6 fb -1 at 2006 was taken by Belle detector at Y(5S). The same luminosity per day was taken at Y(5S) as it is at Y(4S). Very smooth running 5

Off-resonance run 5S run Exceeded 1.2fb -1 /day for the first time at Y(4S). ~22 fb -1 Correction factor(1.056) is necessary for 5S run due to smaller Bhabha Cross section. New 2006 runs at Y(5S) at Belle Daily luminosity Belle collected data at Y(5S): June 9-June 31, 2006 => 21.7 fb -1 6

Belle + BaBar > 1 ab -1 Integrated luminosity Belle (10 Mar 08) All: ~ 778 fb -1, Cont: ~ 68 fb -1, Y(5S): ~24 fb -1 7

hadronic events at  (5S) u,d,s,c continuum bb events B s * B s B s B s * B s * channel B s events b continuum  5S) events Hadronic event classification CLEO PRL 54, 381 (1985)  (5S) B 0, B + events N(bb events) = N(hadr, 5S) - N(udsc, 5S) f s = N(B s ( * ) B s ( * ) ) / N(bb) 8

Number of bb events, number of B s events How to determine f s = N(B s ( * ) B s ( * ) ) / N(bb)? Bf (Y(5S) -> D s X) / 2 = f s Bf (B s -> D s X) + (1- f s ) Bf (B -> D s X) xx 2. Bf (B -> D s X) is well measured at the Y(4S). 1. Bf (B s -> D s X) can be predicted theoretically, tree diagrams, large. bb at Y(5S) B B s _ N cont (5S) = N cont (E=10.519)* L (5S) / L (cont) * (E cont /E 5S ) 2 (  5S /  cont ) Y(5S) :Lumi = ± (stat) fb -1 N bb (5S) = 561,000 ± 3,000 ± 29,000 events Cont (below 4S) : ± (stat) fb -1 N bb (5S) / fb -1 = 302,000 ±15,000 CLEO: N bb (5S)/ fb -1 = 310,000 ± 52,000 Continuum event yield (uu,dd,ss,cc) is estimated using data taken below the Y(4S): _ => 5% uncertainty from luminosity ratio

Inclusive analyses: Y(5S)->D s X, Y(5S)->D 0 X Y(5S) D s ->   ± 100 ev After continuum subtraction and efficiency correction: Bf (Y(5S) -> D s X) 2 = (23.6 ± 1.2 ± 3.6) % => points:5S hist: cont D 0 -> K -  + D s ->   + points:5S hist: cont P/P max <0.5 f s = N(B s ( * ) B s ( * ) ) / N(bb) Bf (Y(5S) -> D 0 X) 2 = (53.8 ± 2.0 ± 3.4) % = (18.0 ± 1.3 ± 3.2 )% N bb (5S) = 561,000 ± 3,000 ± 29,000 events L = 1.86 fb -1  (Y(5S)->bb) = (0.302 ± 0.015) nb at E=10869MeV 10

Signature of fully reconstructed exclusive B s decays B s B s, B s *B s, B s *B s * Figures (MC simulation) are shown for the decay mode B s -> D s -  + with D s - ->   -. M bc = E* beam 2 – P* B 2, The signals for B s B s, B s *B s and B s * B s * can be separated well. e + e - -> Y(5S) -> B s B s, B s *B s, B s *B s *, where B s * -> B s  M bc vs  E  E = E* B – E* beam Reconstruction : B s energy and momentum, photon from B s * is not reconstructed. Two variables calculated: MC Zoom M bc vs  E 11

Exclusive B s -> D s ( * )+  - /  - and B s -> J/  /  decays B s -> D s +  - B s -> D s * +  - B s -> D s ( * )+  - B s -> J/  7 evnts in B s * B s *3 evnts in B s * B s *9 evnts in B s * B s *4 evnts in B s * B s * N(B s *B s *) / N(B s ( * ) B s ( * ) ) = (93± 7 9 ± 1)% Potential models predict B s * B s * dominance over B s *B s and B s B s channels, but not so strong <M BC < 5.429GeV/c 2 B s * N ev =20.3 ± 4.8 Conclusions: 1. Belle can take ~30 fb -1 per month. 2. Number of produced B s at Y(5S) is ~10 5 /fb B s *B s * channel dominates over all B s ( * ) B s ( * ). 4. Backgrnds in exclusive modes are not large. Data at Y(5S), 1.86 fb -1 12

Number of B s in dataset hadronic events at Y(5S) bb events B s events bb continuum included N ev = 561,000 ± 3,000 ± 29,000 N ev = 101,000 ± 7,000 ± 19,000 f(B s *B s *) = (93 ± 7 9 ± 1 )% N ev = 94,000 ± 7,000 ± 20,000 B s * B s * channel Lumi = fb -1 ~10 5 B s mesons per 1 fb -1 at Y(5S) f s = (18.0 ± 1.3 ± 3.2 )% Biggest uncertainty comes from f s systematics. How to improve it (3 times)? 13

How to measure f s with 5% uncertainty ? I spent a lot of time thinking about that. It could be: 2. Using same-sign lepton-lepton sample, maybe with z-distance measurement between profile-lepton vertices 3. J/  vertex xy-distance from profile. 4. Bf(B-> D +  - ), Bf(B-> D 0  - ), Bf(B-> D* 0  - ) measurements. 5. Number of slow photons from B s * decays. No one of these methods is perfect 1. CLEO method, from Bf(Y(5S)-> D s X), with better statistics. Improved measurement of f s 14

New Belle results with 23.6 fb -1 Jean Wicht First observation of B s ->   and new upper limit for B s -> . Bf (B s ->  )=( ) Bf (B s ->  ) < 8.7 x (90% CL) First measurement of Y(5S) -> Y(nS)  +  - decays (21.7 fb -1 ). 15

Is the ϒ (10860) purely ϒ (5S)?  (2S)  (3S)  (2S)  (1S) -> look for: +-h+h-+-h+h- e + e - ->  (1S)  +  - X e + e - ->  (2S)  +  - X arXiv: [hep-ex] (accepted PRL) Study motivated by observation of Y(4230) -> J/   +  - signal (analogous?). 16

4 modes seen : Expectation: Υ (5S) width comparable to Υ (2S/3S/4S) larger by > 10 2 Conclusion: not pure Υ (5S)? Energy scan: 12/07. Y(5S) ->  (nS) h + h - Is the ϒ (10860) purely ϒ (5S)? 17

New Belle results with 23.6 fb -1 Jean Wicht First observation of B s ->   and new upper limit for B s -> . Bf (B s ->  )=( ) Bf (B s ->  ) < 8.7 x (90% CL) First measurement of Y(5S) -> Y(nS)  +  - decays (21.7 fb -1 ). First measurement of B s -> X + l - decay. 18

Motivation, feasibility of B s lifetime measurement. B s and D s lifetimes can be measured using D s vertex, lepton track and beam profile. K+ K-K- ++ Ds+Ds+ ++ z This analysis requires much more work …. PDG 2007 : Bf( B 0 ->X + l -  ) = (  0.28 )% Theory predicts about 12%. It is not yet understood by theory. Some recent models predict better (dis)agreement. Calculation problems? Exotics? Maybe semilep. B s decays can shed some light. Semileptonic decays have no hadronic corrections.  (B 0 ) >  (B s )  difference (in contrast with theory). 19

Electron : Bf( B s ->X + e -  ) = ( 10.9  1.0  0.9 )% Combined fit (electron+muon) : Bf( B s ->X + l -  ) = ( 10.2  0.8  0.9 )% it can be compared to PDG 2007 : Bf( B 0 ->X + l -  ) = (  0.28 )% First measurement of B s -> X + l - decay Electron Muon MC from B s from D s,D… DATA preliminary 23.6 fb -1 Muon : Bf( B s ->X +  -  ) = ( 9.2  1.0  0.8 )% Assuming similar decay widths and  (B s )/  (B 0 )=1.00  0.01 (theory; exp.diff.~2.3  ) 20

New Belle results with 23.6 fb -1 Jean Wicht First observation of B s ->   and new upper limit for B s -> . Bf (B s ->  )=( ) Bf (B s ->  ) < 8.7 x (90% CL) First measurement of Y(5S) -> Y(nS)  +  - decays (21.7 fb -1 ). First measurement of B s -> X + l - decay. Measurement of B s -> D s +  - and B s -> D s + K - decays. Bf (B s -> D s +  - )=( ) Bf (B s -> D s +  - )=( ) R=0.066  R. Louvot, T. Aushev, J.Wicht 21

Why it is interesting? Bf (B s -> D s +  - )=( ) PDG: Bf (B-> D +  - )=(2.68  0.13) W-exchange diagram? Difference is not yet significant. 2. M(B s *)=  0.4  1.0 MeV/c 2 PDG:M(B s )=  0.6 MeV/c 2  (B s 0 )= 51.3  1.2 MeV/c 2  (B 0 )=  0.35 MeV/c 2 Very unexpected difference 3. N(B s *B s *) / N(B s ( * ) B s ( * ) ) = (90± ± 0.2)% very unexpected 4. Flat B direction angular distribution –> has to be explained. 22

1. K. Sayeed, A. Schwartz: B s - > J/   and B s - > J/  K s decays. 2. J.-H. Chen : Search for B s - > K + K - decay. Important for future CP studies. CP eigenstate, can be used in future for  s /  s measurement. Analysis started: 1. S. Esen : Measurement B s -> D s +( * ) D s - ( * ) Belle results expected soon with 23.6 fb -1 Mostly CP eigenstates, important for indirect  s /  s measurement. 23

 s /  s measurement from Bf (B s -> D s +( * ) D s - ( * ) ) M Bs = (M H + M L )/ 2  s = (  H +  L )/ 2 i d d t ( ) BsBs BsBs = ( M – / 2  ) i ( ) BsBs BsBs - Schrodinger equation Matrices M and G are t-dependent, Hermitian 2x2 matrices Assuming CPT : M 11 = M 22  11 =  22 | B H,L (t) > = exp( - ( i M H,L +  H,L / 2)t ) | B H,L > BSM  s = arg (- M 12 /  12 ) 2  s =  s  s = 2 |  12 | cos 2  s SM:  s =arg(-V ts V tb */ V cs / V cb *) =O( 2 ) - no CP-violation in mixing  m s = M H – M L  =  L -  H >0 in SM “ 24

 s = 2 |  12 | cos  s  s SM =  CP s = 2 |  12 | To prove this formula experimentally : a) Contribution of B s -> D s +( * ) D s - ( * ) n  is small b) Most of B s -> D s + D s - * and B s -> D s + * D s - * states are CP- even. Since  CP s is unaffected by NP, NP effects will decrease  s. Assuming corrections are small (~5-7%), Bf measurement will provide information about  CP s or |  12 |. B s ->D s ( * ) + D s ( * ) - decays have CP-even final states with largest BF’s of ~ (1-3)% each, saturating  s /  s.  CP s =   (CP=+) –   (CP= –)  s /  s measurement from Bf (B s -> D s +( * ) D s - ( * ) )  CP s ss ~ ~ Bf(B s ->D s ( * ) + D s ( * ) - ) 1- Bf(B s ->D s ( * ) + D s ( * ) - ) / 2 (first proposed by Y. Grossman) 25

D s + ->  +, K* 0 K +, K s K + Eff(B s ->D s + D s - ) ~ 2x10 -4 Eff(B s ->D s * + D s - ) ~1x10 -4  CP s ss ~ Bf(B s ->D s ( * ) + D s ( * ) - ) 1- Bf(B s ->D s ( * ) + D s ( * ) - ) / 2 Expected with 25 fb -1 at Y(5S): => Accuracy of Bf (B s ->D s ( * ) + D s ( * ) - ) has to be ~25%. should be compared with direct  s /  s measurement to test SM. <= ~ B s -> D s + D s - B s -> D s * + D s - B s -> D s * + D s * - Eff(B s ->D s * + D s * - ) ~5x10 -5 N ~ 10 7 x 2x10 -4 x ~ 5 ev N ~ 10 7 x x 2x10 -2 ~ 5+5 ev N ~ 10 7 x 5x10 -5 x 3x10 -2 ~ 4 ev  s /  s lifetime difference can be measured directly with high accuracy at Y(5S) and also at Tevatron and LHC experiments.  s /  s measurement from Bf (B s -> D s +( * ) D s - ( * ) ) Y(5S), 1.86 fb -1 26

1. B s -> D s +  -, B s -> D s + a 1 -, B s -> D s * +  -, B s -> D s * +  -, B s -> D s * + a 1 -. BF’s should be compared with B 0 partners to test SU(3). 2. B s -> J/ , J/   ’, J/  , J/  f 0 (980), …,B s -> J/  K + K -. What is fraction of ss component in different mesons? Further physics program with 23 fb -1 Quark model :  (  )=(uu+dd-ss)/ 3  (  ’ )=(uu+dd+2ss)/ 6 B(B s 0 ->J/   ) = 1/3 B(B s 0 ->J/   ) Mixed channels? Enhanced branching fractions? B(B s 0 ->J/   ’ ) = 2/3 B(B s 0 ->J/   ) 27

3. B s -> D sJ +  - (4 states). Interesting physics issues, critical test of D sJ nature. Inclusive D sJ production study? Further physics program with 23 fb B s -> D 0 K 0( * ). Statistically significant signals are expected with BF’s predicted at [C-K.Chua,W-S.Hou, hep-ph/ ]. Color-suppressed Bf(B s ->D 0 K 0 ) ~8 x => ~20 signal events should be seen with 23.6 fb -1 at Y(5S). 28

Color-suppressed B s -> D 0 K 0 decay Bf (B 0 ->D +  - ) Bf (B 0 ->D 0  0 ) = (2.91 ± 0.28) x (3.4 ± 0.9) x ~ 0.1 Which diagram, color-suppressed or FSI, is dominant in B 0 ->D 0  0 decay ? Decay mode B s ->D 0 K ( * )0 has no FSI diagram. If the ratio Bf(B s ->D 0 K 0 )/Bf(B s ->D s +  - ) ~0.1, then color-suppressed diagram dominates. If the ratio is significantly smaller, then FSI diagram dominates. Color-suppressed FSI Color-suppressed ~ 29

5. B s -> D s + l -, B s -> D s * + l - (B s ->K + l - ?). 7. B s lifetime measurement. Further physics program with 23 fb -1 Important SU(3) test. CDF obtained large D sJ semileptonic BF (?). Good accuracy is expected (5-10%). Important measurement. 6. B s decays with baryons (with   baryons). Largest B 0 baryonic Bf’s are ~ Is it similar in B s decays? Different samples can be used: fully reconstructed events, CP-fixed modes, two lepton events, D s + lep + events …. 30

Feasibility of B s lifetime measurement with same-sign leptons Lifetime can be measured using two fast same sign lepton tracks and beam profile. To remove secondary D meson semileptonic decays : P( l )>1.4 GeV. ++  z =  c  t ++ Y(5S) BsBs BsBs Beam profile Z beam ~ 3 mm;  z ~ mm. Y(5S) : B s ( l + ) B s ( l + ) / B s ( l + ) B s ( l - ) = 100% Y(4S) : B( l + ) B( l + ) / B( l + ) B( l - ) ~ 10% 31

Comparison with Fermilab B s studies. There are several topics, where Y(5) running has advantages comparing with CDF and D0: 1) Model independent branching fraction measurements. 2) Measurement of decay modes with ,  0 and  in final state (D s +  - ). 3) No trigger problems for multiparticle final states (like D s + D s - ). 4) Inclusive branching fraction measurements (semileptonic B s ). 5) Partial reconstruction ( Bf (D s + l - ) using “missing- mass” method). There are also disadvantages: 1) We have to choose between running at Y(4S) or Y(5S). 2) Number of B s is smaller than in Fermilab experiments. 3) Vertex resolution is not good enough to measure B s mixing (???). 32

Future physics program at Y(5S) Realistic value of 200 fb -1 Optimistic value of 2000 fb -1 Only big deals: 1.  s /  s measurement 2. Measurement of B s ->  decay Decay modes D s ( * ) D s ( * ), K + K -, , , J/    (  ) ~500 CP-fixed events with 200fb -1 => 5-10% accuracy in  s. It also requires about 1000fb -1 to measure. 3. B s mixing measurement 33

It is often postulated, that B s mixing cannot me measured at the Y(5S). Have anybody checked it? Is it correct or not? B s mixing measurement Can we measure B s mixing? Let’s check it. Distance between max and min of oscillation function:  z=   m s  c = 22.5  m with  =0.425 Can we increase  at Y(5S) runs by 50%? Probably yes. Then we need to get single vertex resolution of ~20  m. Is is planned resolution for fast (0 0 dip angle) tracks (next slide). => with high statistics we can select high vertex resolution events. 34 Yes, we can.

Impact Parameter resolution r-   direction z direction Calculated by TRACKERR Occupancy effects. Degradation of intrinsic resolution is included. Efficiency loss is NOT included Beampipe radius is important Competitive performance as the current SVD LoI ‘04 sBelle SVD2(now) For   0.2GeV 0.5GeV 1.0GeV 2.0GeV sin  [cm] 0.03 [cm] T.Kawasaki, Atami BNM2008 Jan  m

CLEO PRL 54, 381 (1985)  ( 5S ) What else can be done at Super B Factory? Rates at e+e- continuum should be similar, baryon production is large. M(  b ) x 2 = (11248 ± 18) MeV/c 2 => 6.3 % up from Y(4S) CME. M(  b ) = (5624 ± 9) MeV/c 2 Can Super B factory CM energy range be increased ? M(B c ) = (6286 ± 5) MeV/c 2 e+e- Y(6S,7S) B s B s,  b  b, B c B c,  b  b … ? PDG (Z->bb, pp at S 1/2 =1.8TeV) b hadron fraction(%) B +, B ± 1.0 B s 10.4 ± 1.4 b baryons 9.9 ± 1.7 ____ 36

Conclusions B s studies at e+ e- colliders running at Y(5S) have some advantages comparing with hadron-hadron colliders. These colliders are in some sense complementary. B s decays with branching fractions down to can be measured with statistics of ~100 fb -1 at e+e- colliders running at Y(5S). Many important SM tests can be done with statistics of the order of 1000 fb -1. It is important to have more flexibility in beam energies. 37

Background slides

 / K L detector Central drift chamber He(50%)+C 2 H 6 (50%) EM calorimeter (CsI(Tl)) Cherenkov detector n=1.015~1.030 Si vertex detector TOF counter SC solenoid 1.5T Belle Detector 8GeV e  3.5GeV e 

dz resolution dz resolutoin SuperB SVD3mod SVD3 For  0.2GeV 0.5GeV 1.0GeV 2.0GeV dz T.Kawasaki, Atami BNM2008 Jan 2008