1. LHCb Physics Programme
CERN Theory Institute: Flavour as a window to New Physics at the LHC: May 5 –June 13, 2008
LHCb Physics Programme
Marcel Merk
For the LHCb Collaboration
May 26, 2008
Contents:
The LHCb Experiment
Physics Programme:
CP Violation
Rare Decays

2. LHC b b
√s = 14 TeV
LHCb: L=2-5 x 1032 cm-2 s-1
sbb = 500 mb
inel / s bb = 160
=> 1 “year” = 2 fb-1 b b
CERN
LHCb
ATLAS
Lumi is a factor 50 to 20 below peak design lumi of General Purpose Detectors (GPD) – defocussing beams
CMS
ALICE
A Large Hadron Collider Beauty Experiment for Precision Measurements of CP-Violation and Rare Decays

3. Installation of major structures is complete
The LHCb Detector
VELO
Muon det
Calo’s
RICH-2
Magnet
OT+IT
RICH-1
Muon det
Calo’s
RICH-2 OT
Magnet
RICH-1
VELO
Installation of major structures is complete

4. A walk through the LHCb spectrometer…
Indicate collission point. Acceptance between 15 mrad and 200 mrad. First half serves roughly to reconstruct the trajectories of charged particles in a big dipole magnetic field. The second half are detectors to recognize the particle identity: photon, electron, muon, pion, kaon. 

5.  B-Vertex Measurement d Example: Bs → Ds K s(t) ~40 fs 144 mm 47 mm d
47 mm
144 mm
440 mm
Primary vertex
Decay time resolution = 40 fs
s(t) ~40 fs
Example: Bs → Ds K
The detector built around the interaction point is the Verex locator. It is an array of silicon detetectors that can be moved in to a distance of 8 mm of the beam line.
Vertex Locator (Velo)
Silicon strip detector with
~ 5 mm hit resolution
 30 mm IP resolution
Vertexing:
Impact parameter trigger
Decay distance (time) measurement

6. Momentum and Mass measurement
Momentum meas.: Mass resolution for background suppression
Next there is the magnetic field region. A large dipole magnet is surrounded by tracking stations in fron and behind. The Bfield integral is 4 Tm, which allows for precise momentum measurements up to high momenta. 

7. Momentum and Mass measurement
Momentum meas.: Mass resolution for background suppression bt Bs K K
p+, K  Ds
Primary vertex
Bs→ Ds K
Bs →Ds p
Mass resolution
s ~14 MeV
Next there is the magnetic field region. A large dipole magnet is surrounded by tracking stations in fron and behind. The Bfield integral is 4 Tm, which allows for precise momentum measurements up to high momenta. 

8. Particle Identification
RICH: K/p identification using Cherenkov light emission angle
The RICH detectors serve mainly to identify pions and kaons. They are not present in Atlas or CMS. The detection mechanism is based on the fact that particles traverse the detectors with velocities higher that the velocity in the detector medium (gas). RICH 1 optimized for low momentum tracks, RICH 2 for high momentum tracks. As charged particle tracks traverse the medium they radiate photons in a cone around their track direction. These photons make ring images around the track as measured in the tracking detectors. The emission angle of the cone depends on the velocity of the particle. If we know the velocity from the RICH and the momentum from the trackers, the particle mass and therefore the identity is known. Again a number of postdocs and grad students is required to make the pattern recognition work. Image of RICH2. 
RICH1: 5 cm aerogel n=1.03
4 m3 C4F10 n=1.0014
RICH2: m3 CF4 n=1.0005

9. Particle Identification
RICH: K/p identification; eg. distinguish Dsp and DsK events.
Cerenkov light emission angle bt Bs K K
,K  Ds
Primary vertex
Bs → Ds K
KK : ± 0.06%
pK : 5.15 ± 0.02%
The RICH detectors serve mainly to identify pions and kaons. They are not present in Atlas or CMS. The detection mechanism is based on the fact that particles traverse the detectors with velocities higher that the velocity in the detector medium (gas). RICH 1 optimized for low momentum tracks, RICH 2 for high momentum tracks. As charged particle tracks traverse the medium they radiate photons in a cone around their track direction. These photons make ring images around the track as measured in the tracking detectors. The emission angle of the cone depends on the velocity of the particle. If we know the velocity from the RICH and the momentum from the trackers, the particle mass and therefore the identity is known. Again a number of postdocs and grad students is required to make the pattern recognition work. Image of RICH2. 
RICH1: 5 cm aerogel n=1.03
4 m3 C4F10 n=1.0014
RICH2: m3 CF4 n=1.0005

10.  LHCb calorimeters bt e h Calorimeter system : Bs K K  Ds
Subsequently the calorimeters identify photons, elctrons, hadrons by stopping the showers. They supply the Et trigger.  bt Bs K K  Ds
Primary vertex
Calorimeter system :
Identify electrons, hadrons, neutrals
Level 0 trigger: high ET electron and hadron

11.  LHCb muon detection btag m Muon system: Bs
And finally the muon system identifies muons. Anything else is stopped. 
btag Bs K K  Ds
Primary vertex
Muon system:
Level 0 trigger: High Pt muons
Flavour tagging: eD2 = e (1-2w)2  6%

12. LHCb trigger bt L0, HLT and L0×HLT efficiency Detector 40 MHz
HLT: high IP, high pT tracks [software] then full reconstruction of event
40 MHz
1 MHz
2 kHz
L0: high pT (m, e, g, h) [hardware, 4 ms]
Storage (event size ~ 50 kB)
Detector
HLT rate
Event type
Physics
200 Hz
Exclusive B candidates
B (core program)
600 Hz
High mass di-muons
J/, bJ/X (unbiased)
300 Hz
D* candidates
Charm (mixing & CPV)
900 Hz
Inclusive b (e.g. bm)
B (data mining)
Efficiency 
Note: decay time dependent efficiency: eg. Bs → Ds K
L0 confirmation: HLT associate L0 objects with large impactr parameter tracks 2) inclusive and exclusive selctions
Efficiency in many decay modes will depend on decay time  control channels bt Bs K K  Ds
Primary vertex
Proper time [ps]  12

13. Measuring time dependent decaysbt Bs K K   Ds
Primary vertex
Measurement of Bs oscillations:
Experimental Situation:
Ideal measurement (no dilutions)
Bs->Ds– p+ (2 fb-1)

14. Measuring time dependent decaysbt Bs K K   Ds
Primary vertex
Measurement of Bs oscillations:
Experimental Situation:
Ideal measurement (no dilutions)
+ Realistic flavour tagging dilution
Bs->Ds– p+ (2 fb-1)

15. Measuring time dependent decaysbt Bs K K   Ds
Primary vertex
Measurement of Bs oscillations:
Experimental Situation:
Ideal measurement (no dilutions)
+ Realistic flavour tagging dilution
+ Realistic decay time resolution
Bs->Ds– p+ (2 fb-1)

16. Measuring time dependent decaysbt Bs K K   Ds
Primary vertex
Measurement of Bs oscillations:
Experimental Situation:
Ideal measurement (no dilutions)
+ Realistic flavour tagging
+ Realistic decay time resolution
+ Background events
Bs->Ds– p+ (2 fb-1)

17. Measuring time dependent decaysbt Bs K K   Ds
Primary vertex
Measurement of Bs oscillations:
Experimental Situation:
Ideal measurement (no dilutions)
+ Realistic flavour tagging dilution
+ Realistic decay time resolution
+ Background events
+ Trigger and selection acceptance
Bs->Ds– p+ (2 fb-1)
Two equally important aims for the experiment:
Limit the dilutions: good resolution, tagging etc.
Precise knowledge of dilutions

18. Expected Performance: GEANT MC simulation
Simulation software:
Pythia+EvtGen
GEANT simulation
Detector response
We have made full MC simulation studies to optimise the experiment and to test the reconstruction quality. More later.
Reconstruction software:
Event Reconstruction
Decay Selection
Trigger/Tagging
Physics Fitting
Used to optimise the experiment and to test physics sensitivities

19. Physics Programme LHCb is a heavy flavour precision
experiment searching for new physics
in CP-Violation and Rare Decays
CP Violation
Rare Decays

20. CP Violation – LHCb Program
Bd triangle
Bs triangle b q1
d, s q2 W− g
d (s) q
W − b
u,c,t
CP Program: Is CKM fully consistent for trees, boxes and penguins?
g measurements from trees
g measurement from penguins
bs from the box: “Bs mixing phase”
bs in penguins b q
u, c, t W+ W−
V*ib
Viq

21. 1.a g +fs from trees: Bs→DsK
Time dependent CP violation in interference of bc and bu decays: Bs
Bs/Bs →Ds–K+ (10 fb-1) Bs
Time

22. Bs→DsK s(g+fs) = 9o–12o Since same topology BsDsK, Bs Dsp
combine samples to fit Dms, DGs
and Wtag together with CP phase g+fs.
10 fb-1 data:
Bs→ Ds-p+
Bs→ Ds-K+
(Dms = 20)
Use lifetime difference
DGs to resolve some
ambiguities (2 remain).
s(g+fs) = 9o–12o
Depending on the value of the strong phase and background level
Yield & B/S : LHCb 2007 – 017 Gamma: LHCb
Channel
Yield (2 fb-1)
B/S
(90% C.L.)
BsDsK
6.2 k
[ ]
BsDsp
140 k
[ ]
Decay Time →

23. B →Dp and Bs→DsK B  D(*)p measures g in similar way as BsDsK
More statistics, but smaller asymmetry
No lifetime difference:
8 –fold ambiguity for g solutions
LHCb
Channel
Yield (2 fb-1)
B/S
BsD*(K p) p
206 k
<0.3
B->D p
210 k
0.3
Invoking U-spin symmetry (d↔s) can resolve these ambiguities in a combined analysis of DsK and D(*)pi (Fleisher)
|lambda_s| ~0.4 |lambda_d|~0.02
Method works because we measure sin(phi_d + gamma) and sin (phi_s + gamma) with same strong phase:
LHCb 2005 – 036. Also for event yields.
Can make an unambiguous
extraction, depending on
The value of strong phases:
s(g) < 10o (in 2 fb-1)
U-spin breaking

24. 1.b g from trees: B→DK D0K– fDK– B– D0K– GLW method: ADS method:
Interfere decays bc with bu
to final states common to D0 and D0
color
suppression
GLW method:
fD is a CP eigenstate common to
D0 and D0: fD= K+K-, p+p-,…
Measure: B→D0K, B → D0K, B→ D1K
Large event rate; small interference
Measurement rB difficult
bc favoured
Cabibbo supp.
D0K–
fDK– B–
Self tagging modes from kaon charge: B-K-, B+K+, B0K*(K+pi-)
Gronau London Wyler – triangle relations difficulty: rB? Semileptonic decay with huge background.
Atwood Dunietz Soni
D0K–
bu suppressed
Cabibbo allowed.
ADS method:
Use common flavour state fD=(K+ p –) Note: decay D0→K+p– is double Cabibbo suppressed
Lower event rate; large interference
Decay time independent analysis

25. B→D(*)K(*) g, rB, dB, dDp , dD3p s(g) = 5o with 2 fb-1 of data
See talk of Angelo Carbone
GLW: D KK (2 rates) ;
ADS: D Kp ( 4 rates) ;
ADS: D K3p (4 rates) ;
Channel
Yield (2 fb-1)
B/S
B → D(hh) K
7.8 k
1.8
B → D(Kp) K , Favoured
56 k
0.6
B → D(Kp) K , Suppressed
0.71k 2
B → D(K3p) K, Favoured
62k
0.7
B → D(K3p) K, Suppressed
0.8k
Normalization is
arbitrary:
7 observables for 5 unknows:
g, rB, dB, dDp , dD3p
s(g) = 5o to 13o
depending on strong phases.
GLW: Gronau, London, Wheeler ADS: Atwood, Dunietz, Soni
MM: D* difficult to reconstruct – BG!- Mitesh
s(g)
Also under study:
B± → DK± with D → Ks pp o
B± → DK± with D → KK pp o
B0 → DK*0 with D → KK, Kp, pp 6o -12o
B± → D*K± with D → KK, Kp, pp (high background)
Dalitz analyses
Overall: expect precision of
s(g) = 5o with 2 fb-1 of data

26. 2. g from loops: B(s)→hh s ( g ) ~10o
See talk of Angelo Carbone
Interfere b u tree diagram with penguins:
Strong parameters d (d’), q (q’) are
strength and phase of penguins to tree.
Weak U-spin assumption :
d=d’+-20% , q, q’ independent
Assume mixing phases known
2 fb-1
Dms=20
BsK+K- BG
s ( g ) ~10o
Channel
Yield
(2 fb-1)
B/S
Bpp
36k
0.5
BsKK
0.15
Time

27. 3. The Bd and Bs Mixing Phase
Time dependent CP violation in interference between mixing and decay Bs
fCP B
fCP
s(sin2b ) ~ 0.02
Channel
Yield
(2 fb-1)
B/S
BdJ/yKs
216 k
0.8
“Golden mode”
“Yesterday’s sensation is today’s
calibration and tomorrow’s background”
– Val Telegdi

28. Bs mixing phase 1. Pure CP eigenstates Low yield, high background
2. Admixture of CP eigenstates: Bs→J/y f
“Golden mode”: Large yield, nice signature
However PS->VV requires angular analysis
to disentagle h=+1 (CP-even) ,-1 (CP-odd)
Decay
Yield
(2 fb-1)
s (fs)
J/y hgg
8.5 k
0.109
J/yhppp
3 k
0.142
J/y h’pph
2.2 k
0.154
J/y h’rg
4.2 k
0.08
hc f
0.108
Ds+ Ds- 4k
0.133
All CP eig -
0.046
J/y f
130 k
0.023
All
0.021
Measure the decay rates….

29. Full 3D Angular analysis
cos y = cosqf
cosq f
Study possible systematics of LHCb acceptance and reconstruction on distributions.

30. Bs→ J/y f
signal
backg
sum
Simultaneous likelihood analysis in time, mass, full 3d-angular distribution
Include mass sidebands to model the time spectrum and angular distribution of the background
Model the t resolution
 s (fs)=0.02
s(t)= 35 fs
time
s(M)= 14 MeV
See talk of Olivier Leroy
mass
cosqf
cos qtr
ftr

31. 4. b & bs from Penguins
Compare observed phases in tree decays with those in penguins
Gluonic penguins. Roger forty sigma(sym = 0.04)  check. B Phi ks : LHCb 2007 – 013, Bs phi phi : LHCb
Bs  f f requires time dependent CP asymmetry
(PSVV angular analysis a la J/yf )
See talk of Olivier Leroy
Channel
Yield (2 fb-1)
B/S
Weak phase precision
B f Ks
920
0.3 < B/S < 1.1
s (sin(fdeff)  0.23
Bs  f f
3.1 k
< 0.8
s (fseff ) = 0.11

32. Physics Programme LHCb is a heavy flavour precision
Experiment searching for new physics
in CP-Violation and Rare Decays
CP Violation
Rare Decays

33. Rare Decays – LHCb Program
Weak B-decays described by effective Hamiltonian:
i=1,2: trees
i=3-6,8: g penguin
i=7: g penguin
i=9,10: EW penguin
New physics shows up via new operators Oi or modification of Wilson coefficients Ci compared to SM.
Study processes that are suppressed at tree
level to look for NP affecting observables:
Branching Ratio’s, decay time asymmetries,
angular asymmetries, polarizations, …
Ci Wilson coefficients at weak scale. Oi non-perturbative (matrix element) operators. At B-mass energy scale: Ci effective.
K*mumu : O_9L and O_10L sensitive to right handed currents
B s gamma: O_7 gamma and O_7gluon magnetic penguins
Bs mumu O_S,_P
Vanya:
1. Brho gamma and B W gamma measures |Vtd/Vts|
2. Lambda b Lambda(*) gamma, photon polarization
3. Bs phi gamma allows photon polarization due to Delta Gamma
Rk is ratio B+mumu / B+K+ee (sensitive to Higgs couplings)
LHCb Program: Look for deviations of the SM picture in the decays:
b sl+l- ; Afb(BK*mm) , B+→K+ll (RK), Bs→ fmm
b  s g ; Acp(t) Bs fg , B→K*g, Lb→Lγ, Lb→ L*γ,B →r0g, B→wγ
Bq l+l- ; BR (Bs mm)
LFV Bq l l‘ (not reported here)

34. 1. B0→K*0µ+µ- 3 angles: ql, f ,qK AFB
Contributions from electroweak penguins
Angular distribution is sensitive to NP
3 angles:
ql, f ,qK
m2 [GeV2]
AFB(m2μμ) theory illustration
AFB
Observable: Forward-Backward Asymmetry in ql
B.R is in agreement with Standard model. Contributions: O_9L and O_10L. AFB in particular sensitive to models with right handed currents (not V-A)
Zero crossing point (so) well predicted:
hep-ph: v2

35. B0→K*0µ+µ- s0 Event Selection: Afb(s)  Systematic study:
See talk Mitesh Patel
Event Selection:
Channel
Yield (2 fb-1)
BG (2 fb-1)
Bs→K*m+ m–
(BR)
AFB(s), fast MC, 2 fb–1
s = (m)2 [GeV2] 
2 fb-1
Afb(s)  s0
s(s0) = 0.5 GeV2
Remove
resonances
BG: Double semileptonic events. AT2 = theoretically clean
Systematic study:
Selection should not distort m2mm
s0 point to first order not affected

36. 2. Bs→fg Probes the exclusive b →s radiative penguin
See talk Mitesh Patel
Probes the exclusive b →s radiative penguin
Measure time dependent CP asymmetry:
b   (L) + (ms/mb)  (R)
In SM b→s g is predominatly (O(ms/mb)) left handed
Observed CP violation depends on the g polarization
Event selection:
Channel
Yield
(2 fb-1)
B/S
Bs→fg
11k
<0.55
SM:
Adir  0, Amix  sin2y sin2f , AD  cos 2y cosf
tan y = |b→s g R| / | b→s g L| , cos f  1
Direct CP asymmetry suppressed in SM. Mixing induced asymmetry. Tan psi is the fraction of “wrong”gamma
Polarization. A_delta measures this quantity directly.
Statistical precision after 2 fb-1 (1 year)
s(Adir ) = , s (Amix ) = (requires tagging)
s (AD) = (no tagging required)
Measures fraction “wrong” g polarization

37. 3. Bs →m+m- Bs mm is helicity suppressed
See talk Mitesh Patel
Bs mm is helicity suppressed
SM Branching Ratio: (3.35 ± 0.32) x 10 -9
Sensitive to NP with S or P coupling
hep-ph/ v5
Event Selection:
Main Backgrounds: Suppressed by:
bm, bm Mass & Vertex resol.
B hh Particle Identification
Channel
SM Yield
(2 fb-1)
Background
Bs mm 30 83
Scalar or Pseudoscalar
With 2 fb-1 (1 “year”):
Observe BR: 6 x 10-9 with 5 s
Assuming SM BR: 3s observation

38. Physics Programme LHCb is a heavy flavour precision
Experiment searching for new physics
in CP-Violation and Rare Decays
CP Violation
Rare Decays
+ “Other” Physics

39. “Other” Physics See the following talks for an overview:
Raluca Muresan:
Charm Physics: Mixing and CP Violation
Michael Schmelling:
Non CP Violation Physics:
B production, multiparticle production, deep inelastic scattering, …

40. Conclusions
LHCb is a heavy flavour precision experiment searching for New Physics in CP Violation and Rare Decays
A program to do this has been developed and the methods, including calibrations and systematic studies, are being worked out..
CP Violation: 2 fb-1 (1 year)*
g from trees: 5o - 10o
g from penguins: 10o
Bs mixing phase: 0.023
bseff from penguins: 0.11
Rare Decays: 2 fb-1 (1 year)*
BsK*mm s0 : 0.5 GeV2
Bs g Adir , Amix : 0.11
AD : 0.22
Bsmm BR.: 6 x 10-9 at 5s
We appreciate the collaboration with the theory community to continue developing new strategies.
We are excitingly looking forward to the data from the LHC.
* Expect uncertainty to scale statistically to 10 fb-1. Beyond: see Jim Libby’s talk on Upgrade

41. Backup

42. LHCb Detector RICH-2 PID MUON ECAL HCAL RICH-1 PID vertexing
Tracking (momentum)

43. Display of LHCb simulated event
Hope to see reconstructed events like this soon.

44. Flavour Tagging Performance of flavour tagging: Efficiency e
Wrong tag w
Tagging power Bd
~50% Bs
33%
~6%
Tagging power:

45. Full table of selections
Nominal year = 1012 bb pairs produced (107 s at L=21032 cm2s1 with bb=500 b)
Yields include factor 2 from CP-conjugated decays
Branching ratios from PDG or SM predictions

46. BsDsK 5 years data
BsDs-K+
BsDs+K-
BsbDs-K+
BsbDs+K-

47. Angle g in Summary