LHCb Upgrade – P Collins Beauty 09, Heidelberg LHCb on the eve of the LHC LHCb upgrade The physics programme NP opportunities at LHCb (10 fb-1) and beyond The data taking opportunity Increasing statistics by a factor 10-20 The hardware challenge Detector modifications needed Schedule and Conclusions Paula Collins (CERN) On behalf of the LHCb collaboration 09/09/09 LHCb Upgrade – P Collins Beauty 09, Heidelberg

LHCb: The Day 1 Experiment bb production correlated and sharply peaked forward-backward Single-arm forward spectrometer : 15-300 mrad σbb ~ 500 µb in LHCb acceptance Production of B+, B0, B, b-baryons.., LHCb runs at L=2-5 1032 cm-2s-1 by not focusing the beam as much as ATLAS and CMS maximizes probability of a single interaction per crossing LHCb will go immediately to design luminosity LHCb in numbers Mass resolution: 14 MeV Time resolution: 40 fs IP resolution: 14 + 28/pT mm PID over 2-200 GeV ~ 1012 bb pairs produced per year Trigger goes from 40→1 MHz in hardware, then to 2kHz in software 09/09/09 LHCb Upgrade – P Collins Beauty 09, Heidelberg

LHCb: in design And in reality Muon HCAL ECAL RICH2 Tracker Magnet TT VELO Outer/Inner Tracker 3 3 3

LHCb 10 fb-1 NP physics highlights Rare decays: Bs  µµ Direct search for NP 3s measurement of SM prediction Mixing phase in Bs  J/yf (tree) Sensitive to NP in mixing Measure s (fs) ≈ 0.01 Mixing phase in Bs  ff (penguin) Sensitive to NP in loops Measure phase ≠ 0 (=NP) with s ≈ 0.035 CKM angle g from Bd  D/DK, Bs→DsK, Bd->Dp (tree) Determine standard candle against which NP sensitive measurements can be compared measurements to ~ 2° degrees Santos CKM angle g from Bd(s)→pp, KK (penguin), B→hhh g_NP vs g_SM Sensitive to ≈ 3° degrees CPV in penguins Bdf Ks : b vs beff Sensitive to NP in loops Sensitive to ≈ 0.1 Search for RH currents in radiative decays Bs f g ; Asymmetry FB of Bd  K* µµ zero of AFB(s) to 7% D meson physics CPV, D-D mixing, rare decays e.g. D0→mm Carson Leroy Leroy Reece, Belyaev Ricciardi Magnin 09/09/09 LHCb Upgrade – P Collins Beauty 09, Heidelberg

Why upgrade? Or, Why Should We Fully Exploit LHCb b production? There are hints of NP being just around the corner. If they are confirmed by LHCb e.g. in Bs→J/yf then the path forward is clear and the precision measurements must be pursued If ATLAS/CMS discover NP, but the effects are more subtle in the flavour sector then we will need precise measurements to test predictions Pursue channels statistically limited at current experiment e.g. Bs→ff Keep pace with theoretical predictions which will improve with time e.g. e.g. indirect precision on g 2010: sg ~ 5o >2015: sg ~ factor 5 better thanks to significant improvements from lattice QCD? arXiv:0902.1815v33

Pushing the Precision Frontier fs=-0.0360.002 rad one of the most precisely predicted CPV quantities in the standard model If NP in the Bs sector turns out to be large, then LHCb upgrade will be needed If it turns out to be small (as appears to be the case in the Bd sector) then precise measurement is needed to match clean prediction LHCb (J/y f ) Upgrade (J/y f ) SM s 0.01 ~0.003 0.002 LHCb numbers are with J/y f alone Precision will be even better with other channels e.g. J/y fo(pp) Direct proportionality to η leads to interesting constraint on UT LHCb precision, 10 fb-1 09/09/09 LHCb Upgrade – P Collins Beauty 09, Heidelberg

b→s penguins, a very promising place for NP to lurk Bs0→ff dominated by penguins: NP can enter in mixing box and/or in penguins CP violating asymmetry is zero, due to cancellation of mixing and decay phases At upgrade 0.6M events and error of 0.01 Intriguing pattern emerging, but existing precision poor LHCb and upgrade will clarify this picture Can be complemented by B0→fKs0 and family: At the upgrade we expect 20000 events and a precision of 0.02 09/09/09 LHCb Upgrade – P Collins Beauty 09, Heidelberg

Bs →mm and Bd →mm – an exciting future SM prediction is precise, and small! Blanke et al, JHEP 0610:003, 2006 This can be reached with 10fb-1 at LHCb. If NP gives significant enhancements then will be seen sooner! However: Certain models, e.g. MCPVMFV, can give enhancements but can also suppress the branching ration < SM depending on phases Need more statistics than available at baseline LHCb 10fb-1 to approach SM 10% precision Rate of (Bd/Bs)→mm tightly constrained and can distinguish SM and MFV (Buras: hep-ph/060405). Might Bd→mm be visible at upgraded LHCb??? 09/09/09 LHCb Upgrade – P Collins Beauty 09, Heidelberg

LHCb Upgrade – P Collins Beauty 09, Heidelberg B → K0* l+l- Powerful NP laboratory Host of interesting observables. Angular distributions e.g. forward-backward asymmetry of the angle between lepton and B in the dilepton rest frame Position of zero asymmetry crossing point will be measured by LHCb as well as needed, but many other theoretically clean observables will only come into play with > 10 fb-1 0.35 M yield at upgrade 100 fb-1 Egede, JHEP11 (2008) 32 With larger statistics, study of further observables (transverse asymmetries: AT(2) , AT(3) , AT(4) ) sensitive to NP 09/09/09 LHCb Upgrade – P Collins Beauty 09, Heidelberg

LHCb 10 fb-1 NP physics highlights Rare decays: Bs  µµ Direct search for NP 3s measurement of SM prediction Mixing phase in Bs  J/yf (tree) Sensitive to NP in mixing Measure s (2bs) ≈ 0.01 Mixing phase in Bs  ff (penguin) Sensitive to NP in loops Measure 2bseff ≠ 0 (=NP) with s ≈ 0.03 CKM angle g from Bd  D/DK, Bs→DsK, Bd->Dp (tree) standard candle against which NP sensitive measurements can be compared measurements to ~ 2° degrees Santos CKM angle g from Bd(s)→pp, KK (penguin), B→hhh g_NP vs g_SM Sensitive to ≈ 3° degrees CPV in penguins Bdf Ks : b vs beff Sensitive to NP in loops Sensitive to ≈ 0.1 Search for RH currents in radiative decays Bs f g ; Asymmetry FB of Bd  K* µµ (zero of AFB(s) to 7%) D meson physics CP, D-D mixing Carson Leroy Leroy Reece, Belyaev Ricciardi Magnin 09/09/09 LHCb Upgrade – P Collins Beauty 09, Heidelberg

LHCb 10 fb-1 NP physics highlights Upgrade Rare decays: Bs  µµ Direct search for NP 3s measurement of SM prediction Mixing phase in Bs  J/yf (tree) Sensitive to NP in mixing Measure s (fs) ≈ 0.01 Mixing phase in Bs  ff (penguin) Sensitive to NP in loops Measure phase ≠ 0 (=NP) with s ≈ 0.03 CKM angle g from Bd  D/DK, Bs→DsK, Bd->Dp (tree) standard candle against which NP sensitive measurements can be compared measurements to ~ 2° degrees Santos CKM angle g from Bd(s)→pp, KK (penguin), B→hhh g_NP vs g_SM Sensitive to ≈ 3° degrees CPV in penguins Bdf Ks : b vs beff Sensitive to NP in loops Sensitive to ≈ 0.1 Search for RH currents in radiative decays Bs f g ; Asymmetry FB of Bd  K* µµ (zero of AFB(s) to 7%) D meson physics CP, D-D mixing, rare decays Carson Measure BR to ~5-10%, search for Bd  µµ Improve by factor 4-5 Leroy Improve by factor 3: Level of indirect prediction Comparison at 0.03o level Leroy Reece, Belyaev Measure to s≈0.01, pin down NP New observables to improve NP sensitivty Ricciardi Magnin Measure and characterise CPV sub-degree precision 09/09/09 LHCb Upgrade – P Collins Beauty 09, Heidelberg

To extend the physics programme, why not just turn up L? LHC is (will be) capable of delivering to LHCb ~50% of the L delivered to the GPDs – would be more than good enough! Baseline scenario: increase our luminosity from 2.1032 to 2x1033cm-2s-1 Crossings with ≥ 1 interaction 10 MHz → 30 MHz Average number of interactions per crossing 1.2 → 4.8 Spillover: increases linearly with L Compatible with SLHC Ihigh x Ilow scenario BUT, it’s all about the trigger! Level-0: Largest ET hadron, e(g) and m 1MHz read-out rate is currently the bottle neck in the system m channels: yield proportional to L Hadronic channels: If we increase L we have to increase our ET cut and there is no net efficiency gain! Our current 2.5 us latency is inadequate for a trigger decision of this complexity 09/09/09 LHCb Upgrade – P Collins Beauty 09, Heidelberg

LHCb upgrade strategy = Move to a full software trigger The idea The consequence: Read out all sub-systems at 40 MHz The consequence Aim to operate at L=2.1033 Perform entire trigger on CPU farm with input rate 30 MHz Goal is to double trigger efficiency for hadronic channels and make the yield scale with luminosity All subsystems must be fully read out at 40 MHz Replace most FE electronics, silicon modules, RICH HPDs, FE boards of calorimeter and outer tracker Replace all FE-electronics; all silicon modules, RICH-HPDs, FE boards of Calorimeter, Outer Tracker FE Trigger uses ALL event information, reconstructs all primary vertices, and can cut on pT for high i.p. Preliminary studies show that the hadron efficiency will improve by ~2, and yield will be proportional to L Increase in statistics x20 for hadronic channels and x10 for leptonic channels This gives the additional flexibility to adapt to the physics landscape in the next decade Case study: Bs  ff (after HTL1) L = 2x1032 cm-2s-1 ET> 3.5 GeV  e ≈ 22% Min.Bias retention 10 kHz L = 1033 cm-2s-1 ET> 2 GeV  e ≈ 45% Min.Bias retention 28 kHz 09/09/09 LHCb Upgrade – P Collins Beauty 09, Heidelberg

LHCb Upgrade – P Collins Beauty 09, Heidelberg LHCb Upgrade TimeLine First upgrade workshop Jan ‘07 Edinburgh EOI submitted April ‘08 Upgrade task force formed TDR planned for 2010 Plan to: Accumulate 10fb-1 @ LHCb in 5 years Install substantial upgrade in 2015-16 (in synch with machine shut downs and long GPD interventions) Extract from EOI LHCb-2008-007 09/09/09 LHCb Upgrade – P Collins Beauty 09, Heidelberg

Detector Environment @ Upgrade Radiation: Current detector designed to withstand 20 fb-1 Affects mainly large h (trackers, inner part of calorimeter) Running experience needed Note that VELO in any case will be replaced B yield from tracking efficiency (no spillover) Tracking and Occupancy: Si can be operated without spillover: giving occupancy increase of ~2 Outer tracker straws: occupancy too high: Increase area coverage of IT and use faster gas Move to scintillating fibres Material Budget an important issue (occupancy, momentum resolution) Luminosity (1032 cm-2s-1) 09/09/09 LHCb Upgrade – P Collins Beauty 09, Heidelberg

Vertexing @LHCb Upgrade Upgrading the VELO to 40 MHz implies complete replacement of all modules and FE electronics. Two major challenges Data rate of ~1300 GBit/s Radiation levels and hence thermal management of modules Important to maintain current performance by keeping material low Small modules with low power Thinning of sensor and electronics Use of CVD diamond planes for cooling and/or sensor Removal/rework of RF foil Dose after 100 fb-1 500 neqcm-2 x 1016 TID (MRad) 50 5 Radius (cm) tip of current VELO 09/09/09 LHCb Upgrade – P Collins Beauty 09, Heidelberg

Vertexing @ LHCb upgrade Candidate module designs Strip solution Small inner radius Thin sensors Diamond cooling plane New electronics needed Occupancy 1-2% strip pixel Pixel Solution Low occupancy helps for PR and timing TIMEPIX (55x55 mm) is a promising candidate for FE electronics square pixel and buttable design allows single layer construction Modification needed to architecture Testbeams this summer show excellent performances Timepix: 55 mm pitch PRELIMINARY Residual (micron) Estimated track contribution to residual 2.5 mm Track angle (degrees) 09/09/09 LHCb Upgrade – P Collins Beauty 09, Heidelberg

Tracking @ LHCb upgrade 2x1032 cm-2s-1 OT Occupancy 75 ns gate 50 ns gate High Luminosity will produce high occupancy and ghosts in the tracking system particularly in the Outer Tracker where a gas technique is used (drift time in straw tubes) and signal is collected in 50-75 ns The area coverage of the Inner Tracker has to be increased, along with new sensors, new FE chip Full silicon tracker in front of magnet must be rebuilt Finer pitch? More layers? Material budget is an important issue Critical for momentum resolution and PR Also: Radiation/cooling OT+IT 09/09/09 LHCb Upgrade – P Collins Beauty 09, Heidelberg

Alternative Solutions 19 OT IT A fiber tracker with mixed fiber dimensions (250 mm for inner part, 700-1000 mm for outer) to be readout with SiPM and/or conventional MaPMT Simplified services configuration: no cables, no cooling, frames thinning, FEE outside  gives less X/X0 Good timing performances Increased granularity in x – spatial resolution enough Problems to be addressed: SiPM readout and its optimisation radiation hardness mechanics 32 channels Si PM: 0.25 x 1 mm2 Lausanne Preliminary: 80 mm residuals

Particle ID in the RICH system Need to replace front end electronics AND photon detectors to cope with 40 MHz readout Baseline approach; keep current geometrical layout RICH1 (aerogel+C4F10) + RICH2 (CF4) OR, replace aerogel with new TOF system (TORCH) located after RICH2 Photon sensors together with their new (independent) chip are the critical item: could be MaPMT 0.3M channels R&D needed (x talk, spillover, B field, lens...) or upgraded HPDs study ion feedback rate many new components: production time an issue or MCPs B tolerance, timing, new idea... Next generation MaPMTs Example PID over full momentum range with TORCH

New concept for RICH: the TORCH Time of flight detector based on a quartz plate, for the identification of p<10 GeV hadrons (replacing aerogel) reconstruct photon flight time and direction in specially designed standoff box Clock arrival time to ~20 ps Deduce photon emission time → track ToF R&D on hardware needed: Microchannel plate PM with multianode readout MCP-PM achieved in testbeam 10-20 ps time resolution with 30-20 p.e. mechanics, electronics (Timepix?), aging etc. possible synergy with PANDA and VELO aerogel expected to give marginal photon yield and its use will become challenging with the higher occupancy of the upgrade contributes material in the tracking volume provide low momentum coverage with new torch detector decoupled from the RICH system Micro Channel Plate PMT : flat photocathode with MCP to amplify electron image before collecting charge in anode of device 25 micron pore good for us but 8x128 segmentation needed front end electronics an issue synergy with PANDA who have similar deveopments E.Ramberg /TIPP09

Calorimeters 22 September 2009: calorimeter modules placed in LHC tunnel to accumulate dose HLT seeds already provided at 40 MHz Modifications to electronics needed: Upgraded FE boards Lower PM gain (÷5) and increase preamp sensitivity and lower noise accordingly No showstopper foreseen Radiation tolerance currently an open issue Affects inner part of ECAL (would need replacement) Tested up to 2.5 Mrad (1 upgrade year) R&D underway to collect more comprehensive data Pile-up (~4 at upgrade) will affect resolution: watch carefully for low pT, g physics Resolution 2x1032 cm-2s-1 Q varying 2x1033 cm-2s-1 Energy

Muon System Maximum rates per channel at 2.1033 R 23 Muon System Maximum rates per channel at 2.1033 FEE readout already at 40 MHz Two main concerns: dead time due to high rate in inner regions aging due to radiation in inner regions An evaluation of these effects will come from first data R z Accumulated chage (C/cm) on wires for 100 fb-1 R z Possibility of replacement of the inner parts with large area GEM (70x30 cm2) or with MWPC, in any case with new, smaller pad readout M1, due to background and to upgraded L0, will be removed A transition to a fully 3D muon projective readout (and not based on strips crossing) could reduce the fake associations (detailed studies to be performed)

LHCb Upgrade – P Collins Beauty 09, Heidelberg Outlook LHCb commissioning is well underway On course to accumulate 10 fb-1 in ~ 5 years Upgrade strategy SLHC independent: Goal: 100 fb-1 in 5 years at L=2x1033 X20 statistics in hadron channels X10 statistics in leptonic channels Maintain tracking and PID performance Upgrade TDR out in 2010 09/09/09 LHCb Upgrade – P Collins Beauty 09, Heidelberg

LHCb Upgrade – P Collins Beauty 09, Heidelberg Homing in on g Interferences between the bc and bu tree transition for B-(D°/D°)K- Example : Dalitz analysis of D°/D°Ksππ to extract (rB,δB,γ) B- B+ Here D final state is common to D and Dbar, many possibilities. Want to exploit As many modes as possible but each decay brings its own strong phase Difference which needs to be known Precision of 3-3 deg with 10 fb-1? Amplitude fit, model error 7o Binned fit, CLEO-c error 2o Binned fit, CLEO-c error 5o 09/09/09 LHCb Upgrade – P Collins Beauty 09, Heidelberg