Advanced Study Institute Physics at LHC LHC Praha2003

Advanced Study Institute “Physics at LHC”, LHC Praha-2003, B Physics Day, 11/07/2003 Galina Pakhlova, (ITEP, Moscow) for LHCb collaboration Study of exclusive radiative B decays with LHCb Pakhlova Galina , ITEP Moscow

Physics motivation v v v Constraint on CKM matrix elements |Vtd/Vts| from the ratios of decay rates B 0 r (w) g / B 0 K*0 g Bs 0 K*0 g / Bs 0 f g W Search for New Physics b v Direct CP violation ACP < 1% in SM (tiny u quark contribution in loop) t(c, u) up to ~ 10% in SM extensions B 0 K*0 g ; B+ K*+ g ; B K**g ; Bs 0 K*0 g v Mixing-induced CP violation ~0 in SM (different g polarizations in B 0 and B 0 decays ) up to ~ 50% in SM extensions Bs 0 f g ; B 0 K**0 g with K**0 , e. g. K*00 (1430) K 0 s p+p- (Dalitz plot analysis of K 0 s p+p- to extract flavour blind K 0 s r 0 ) v Exotics, super rare decays B 0(s) l+ l- g, B 0(s) g g, etc Test of QCD B K*g , K**g , K np g; Lb L g, etc g s(d) 2

Experimental overview The first exclusive radiative penguin B 0 K*0 g was observed by CLEO in 1993 B 0 K*0 g B+ K*+ g ACP(B K* g) B K* 2 (1430) g (4. 3 ± 0. 4)× 10 -5 (3. 8 ± 0. 5)× 10 -5 PDG 2003 -0. 01 ± 0. 07 PDG 2003 (1. 3 ± 0. 5)× 10 -5 PDG 2003 B+ K+ p- p+ g ; B+ K*0 p+ g ; B K p g ; B K r 0 g Belle B 0 r 0 g ; B+ r+ g ; B 0 w g ; B 0 g g Upper limit BABAR B 0 barion g Upper limit CLEO Statistics at B-factories is still not sufficient for stringent test of SM in CP violation in radiative decays and for observation of b d g 3

B B event at LHCb primary vertex K p B decay vertex Impact Parameter (IP) g B decay distance v Nb hadrons = 10 12 / year v all types of b hadrons v sbb/ sinel ~ 5× 10 -3 v high charged and neutral multiplicity v few primary vertices 4

Challenge v v Electromagnetic penguins are indeed rare B decays v inclusive Br (b s g) = (3. 3 0. 4) × 10 -4 v exclusive Br are ~10 -100 smaller v b d g are further suppressed by |Vtd/Vts|2 v f( b B 0 S) / f( b B 0) ~ 0. 26 Background sources v Huge combinatorial background from generic b b events v Minimum bias events v B X p 0 decays v g / p 0 separation using different shower shape v for X=V polarization can be exploited 5

Search for CP violation v v Direct CP violation v no tag and lifetime analysis are required v Systematic errors are of great importance: fake (non-CP) asymmetries v different B and B production rates in pp interactions v detector asymmetries Hopefully can be calibrated with other final states e. g. B 0 J/y K*0 Mixing-induced CP violation v need to tag B flavour v proper time resolution is critical especially for B 0 S due to fast oscillations (x s >19) to be resolved 6

Can we measure radiative decays at LHCb v v v B 0 K*0 (K+ p-) g Direct CP v Energetic g v huge soft g and p 0 background can be effectively removed dedicated trigger selects high ET photons at Level 0 v two charged tracks v B 0 vertex reconstruction B 0 S f (K+ K-) g Bs decay & mixing-induced CP v small Gf narrow DM(f) cut; small f yield in B decays: v background conditions are more favourable v extremely small opening angle f K+ K- v problem with B 0 S vertex (proper time) resolution B 0 w (p+ p- p 0) g b d g v small Gw narrow DM(w) cut v problem with combinatorial background from p 0 v in progress (not presented here) 7

B 0 K*0 g reconstruction Select K+ p- g combinations Tracks are consistent with K+ & p- hypothesis Reject K+ , p- tracks from all primary vertices (c 2 IP >16) K+ p- produce secondary vertex (c 2 49) | M ( K+ p- ) - M PDG ( K*0 ) | 60 Me. V/c 2 ET (g) > 2. 8 Ge. V (close to L 0 trigger requirement) 2. 2 E*T (g) 2. 7 Ge. V Select primary vertex with minimum IP of B 0 candidate Require B 0 momentum and flight direction to be consistent (q. B 6 mrad) v K*0 helicity angle: in K*0 rest frame |cos (p. B, p. K)|<0. 7 v v v v 8

Background suppression (1) v Require large IP of K+, p- to all reconstructed primary vertex (c 2 IP > 16) v suppress primary tracks v especially effective background suppression in events with multiple interactions N comb No c 2 IP cut optimized c 2 IP > 16 S/ B 9

Background suppression (2) bb inclusive E*T (g) , Ge. V E*T (g), Ge. V B 0 K*0 g ET (g), Ge. V Transverse energy E*T (g) in B 0 rest frame vs ET (g) in the lab frame ET (g) > 2. 8 Ge. V 2. 2 E*T (g) 2. 7 Ge. V powerful against low energy g and p 0 rejects also soft K*0 candidates 10

Background suppression (3) K p reconstructed B 0 momentum q reconstructed primary vertex reconstructed B 0 decay vertex g Angle q between B 0 momentum and B 0 flight direction One of the most powerful cut! esignal ~ 60% ; ebb < 1% v should be 0 for real B 0 candidates (smeared by K p vertex resolution) v randomly distributed for combinatorial background 11

p 0 suppression K*0 helicity: in K*0 rest frame cos (p. B, p. K) Significant suppression of fast p 0 is expected after application of special algorithm for g / p 0 separation based on shower shape (to be implemented soon) Contribution from B 0 K*0 p 0 is small and can be further suppressed exploiting K*0 polarization : K*0 helicity states 0 B 0 K*0 p 0 +1 -1 B 0 K*0 g 12

B 0 K*0 g: signal and background summary ~ 4 min LHCb ~ 54 hours LHCb s = 64 Me. V /c 2 0 events in (4. 5 -6. 0) Ge. V/c 2 mass window after trigger and off-line selection with all available MC statistics ~ 10. 3 M events In blue: contribution from B 0 K*0 p 0 after trigger and off-line cuts 2. 2% of B 0 K*0 g Br (B 0 K*0 p 0) 3. 6× 10 -6 PDG 2003 13

B 0 K*0 g : Annual yield and CP sensitivity Efficiency, [%] Reconstruction selection trigger & acceptance 4. 5 9. 2 38 total 0. 16 Nyear s (ACP) 35 K < 0. 01 BACKGROUND / SIGNAL < 0. 7 @ 90 % CL Assuming: Br ( B 0 K*0 g ) f ( b B 0 ) = = ( 4. 3 0. 4 ) × 10 -5 0. 39 N ( B 0 K*0 g ) - N ( B 0 K*0 g ) ACP = --------- N ( B 0 K*0 g ) + N ( B 0 K*0 g ) 14

B 0 S f g reconstruction Two different tasks v Branching ratio measurement similar to B 0 K*0 g reconstruction: v v Tracks are consistent with K+ & K- hypothesis v Reject K+ , K- tracks from all primary vertices (c 2 IP > 4) v K+ K- produce secondary Vertex (c 2 49) v | M ( K+ K- ) - M PDG ( f) | 10 Me. V/c 2 v ET (g) > 2. 8 Ge. V (close to L 0 trigger requirements) v 2. 0 E*T (g) 2. 7 Ge. V v Select primary vertex with minimum IP of B 0 S candidate v B 0 S momentum and flight direction to be consistent q. B 12 mrad ( worse B 0 S vertex resolution) v f helicity : in f rest frame |cos(p. B, p. K)| < 0. 7 Search for mixing-induced CP: selection to be re-optimized 15

B 0 s f g : background suppression M ( K+ K- ) After L 0 x. L 1 trigger & off-line cuts bb inclusive: 0 event in mass window (4. 5 - 6. 0) Ge. V/c 2 from ~ 10. 3 M events ( ~ 4 min LHCb) N(B 0 s f p 0 ) / N(B 0 s fg) 4% assuming Br (B 0 K*0 p 0) = Br(B 0 s f p 0 ) 3. 6× 10 -6 Gf GK*0 : narrow DM(f) < 10 Me. V/c 2 Background is mainly due to real f Fortunately small f yield from B decays. 16

B 0 s f g after trigger and off-line cuts ~ 87 hours LHCb Efficiency [%] reconstr. selection trigger & accept. 4. 3 15 34 BACKGROUND / SIGNAL total Nyear 0. 22 9. 4 K < 2. 4 90 % CL limited by MC statistics, expected to be better Assuming: Br (B 0 s f g ) = ( 4. 3 0. 4 ) × 10 -5 f ( b B 0 S) = 0. 10 s = 64 Me. V/c 2 17

Search for mixing induced CP in B 0 S f g Main problem: Proper time resolution is critical: v dominated by poor f K K vertex resolution kaons are almost collinear B 0 S fg proper time resolution simple vertex fit s 1 ~ 160 fs (40%) s 2 ~ 380 fs(60%) v no extra information on B 0 s vertex from g Improvement can be achieved: v Constrain the B 0 s flight direction to B 0 s momentum: (“direction” vertex fit) Improve vertex resolution by factor > 2. 5 “direction” vertex fit s 1~ 62 fs (60%) s 2~ 200 fs (40%) v Select kinematical region with the better vertex (proper time) resolution: (slow f larger K K opening angle) 18

B 0 s f g lifetime resolution cuts re-optimization for CP violation study B 0 S proper time resolution, ps Decay angle q* (between f momentum in rest frame of B 0 s and B 0 s flight direction) is convenient variable to select the kinematical region that provides the better proper time resolution No s ~ 84 fs cos q* 0, (g) > 3. 2 Ge. V, tighter cut E*T (g) Require: ET s ~ 60 fs B 0 S proper time resolution, ps Achieved lifetime resolution is close to those for charged modes. Estimation of CP violation sensitivities is in progress 19

Conclusion B 0 K*0 (K+ p-) g v record statistical sensitivity in direct CP violation s (ACP )< 0. 01 for one year LHCb. Stringent test of SM extensions! v systematic errors to be studied carefully B 0 S f (K+ K-) g v precise measurement of the branching fraction with ~2% accuracy for one year LHCb v For indirect CP violation B 0 S proper time resolution is critical v advanced fit vertex procedure and dedicated cut optimization allow to hope on reasonable sensitivity Others channels of interest v v B 0 w (p+ p- p 0) g |Vtd/Vts| B K** g direct CP violation B 0 K** (1430) ( K 0 s r 0) g mixing-induced CP violation B 0 m+ m- g why not? under study 20