- Email: [email protected]
The LHCb experiment and its expected physics performance
Nuclear Physics B (Proc. Suppl.) 185 (2008) 213–219 www.elsevierphysics.com
The LHCb experiment and its expected physics performance. S. Vecchia∗ a
INFN, via Saragat 1, Blocco C, 44100 Ferrara, Italy The LHCb experiment is dedicated to the study of heavy flavour physics at the LHC machine. Its primary motivation is to study with high statistics and precision CP violation and rare B hadron decays, to improve the knowledge of the Standard Model or to get indirect hints of new physics contribution to the flavour physics. The main experimental features of the experiment are presented and the sensitivities on some important measurement that LHCb will perform since the first data will be available are discussed.
1. Introduction
2. The LHCb experiment
In the last decade the experiments at the Bfactories and at Tevatron contributed to an impressive improvement in the understanding of the Standard Model (SM). In particular, many different measurement allowed to check the consistency of the CKM picture of the SM and constraint its matrix elements to a good precision. Nevertheless the level of CP violation that the SM account for is insufficient to explain the amount of matter in the universe. A new source of CP violation beyond the SM is therefore needed to solve this puzzle. If New Physics (NP) exists at a not too high energy scale (∼ TeV), it could manifest itself indirectly in beauty or charm meson decays that proceed through loop diagrams, where virtual NP particles could contribute in addition to the SM ones. Since these processes are usually suppressed in the SM, the eventual NP contributions should be easier to be measured. The possible results would be unexpected CP violation effects or modified properties of rare decays. Thanks to the high statistics of b-hadrons that will analyze and the excellent expected experimental performance, LHCb will contribute to a further improvement of several interesting measurement already in the first year of physicsquality data taking in 2009.
The LHCb experiment will exploit the high b¯b cross√section (σb¯b ∼ 500μb) that is expected at the s = 14 TeV energy of the colliding protons at LHC. Given that the b¯b pairs are mostly produced in the forward or backward direction, the LHCb detector was designed as a forward spectrometer, covering a pseudo-rapidity range 1.9 < η < 4.9. Selecting B with transverse momentum pT > 2 GeV/c the b-hadron production cross section is σb¯b ∼ 230μb. In order to maximize the probability of a single interaction per beam crossing, it was also decided to limit the luminosity in the LHCb interaction region to 2 − 5 × 1032 cm−2 s−1 . This choice facilitates the study of B-physics and has the additional advantage of reducing the radiation damage due to the high particle flux at small angles. In these conditions one year (∼ 107 s) of ”normal” LHCb running corresponds to 2 fb−1 of integrated luminosity and ∼ 1012 b¯b pairs produced in the region covered by the spectrometer. The detector and the trigger of the experiment are especially designed to reach excellent performance in the specific B meson decays reconstruction that usually have small Branching Ratios (BR< 10−4 ), while rejecting the background due mostly to inclusive b¯b and inelastic pp interactions (σvisible /σb¯b ∼ 120). The LHCb spectrometer layout is shown in Figure 1 and consists of several detectors [1].
∗ on
behalf of the LHCb Collaboration
0920-5632/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.nuclphysbps.2008.10.027
214
S. Vecchi / Nuclear Physics B (Proc. Suppl.) 185 (2008) 213–219
Figure 1. Layout of the LHCb detector.
The Vertex Locator (VELO) consists of 21 modules of silicon strips that reconstruct the radial and azimuthal coordinates of the tracks closest to the interaction point. It is divided into two halves that can be moved as close as 8 mm to the beams in stable running conditions. Some modules are used at trigger level to reject events with multiple interactions. This detector allows to reconstruct primary vertices with an excellent precision: 10 (60) μm in the transverse (parallel) direction to the beam axis. The measurement of the impact parameter of tracks relative to the primary vertex, a crucial parameter for identifying particles originated by a B decay, is made with a precision of σIP =(14 + 35/pT )μm (with pT in GeV/c units). The particle’s tracking is completed by two tracking systems one upstream (Turicensis Tracker, TT) and one downstream the 4 Tm magnet (Inner and Outer Tracker: IT and OT). TT and IT consist of silicon strip detectors, while OT consist of straw tubes. They allow to measure the momentum of the charged particles with a precision of σp /p = (0.4 + 0.0015 × p)% (with p in GeV/c units) with > 95% efficiency. The overall performance of the tracking system allows to reconstruct the B invariant mass and proper time with a resolution of σm ∼ 15 − 20 MeV/c2 and σt ∼ 30 − 40 fs respectively, depending on the channel.
Two Ring Imaging Cherenkov detectors (RICH1 downstream the VELO, and RICH2 downstream the IT/OT) allow to identify efficiently pions, kaons and protons in the wide momentum range 2 − 100 GeV/c of the B decay products. They play a key role in the selection of some specific channels, the background suppression, and the determination of the B flavour at production (B-tagging). The calorimeter system of the experiment comprises a pre-shower detector (PS), an electromagnetic calorimeter (ECAL) and a hadron calorimeter (HCAL). It provides the identification of electrons, hadrons and neutrals (photons and π 0 ) as well as the measurement of their energies and positions. Its informations are used also at trigger level L0 to select particles with high ET (> 2 − 3 GeV/c) deposit coming from B decays. The muon system is designed to identify muons both at trigger level L0, selecting high pT muons, and off line for the channel reconstruction and the B flavour tagging. It consists of five stations (M1-M5) made of multi wire proportional chambers and 12 triple-GEM detectors to cover the innermost region of M1, which has higher occupancy. The overall tracking and particle identification performance of the detector allow efficient and low background reconstruction of many b-hadron exclusive decays, also with very low BR. They also allow to tag the B flavor at production with a tagging power of 4 − 5% for a Bd and 7 − 9% for a Bs , by means of the opposite and the same side tagging algorithms described in [2]. The LHCb trigger is organized in two levels: the first level trigger (L0), realized with custom electronics, selects events with relatively ”high” pT (typically > 1 − 4 GeV/c) electrons, muons, photons and hadrons that most likely come from a B decay, while rejecting the events with multiple interactions. It uses the informations of the calorimeters, the muon system an the pile-up veto modules in the VELO detector. The decision is taken within a latency of 4μs at the input frequency of 40 MHz with an output rate of 1 MHz. The L0 selected events are then read-out and examined by the High Level Trigger (HLT), which
S. Vecchi / Nuclear Physics B (Proc. Suppl.) 185 (2008) 213–219
consists of a software trigger running on a farm of ∼ 1800 CPU. First it requires a confirmation of the L0 trigger decision with the more accurate information available; after a partial track reconstruction, cuts on the impact parameter of the L0 selected tracks can be applied. Finally the event is fully reconstructed and written on tape with a frequency of 2 kHz if it fulfills the requirements of one of the possible streams of HLT. 200 Hz of the output rate is dedicated to the exclusive selections of specific B decay channels that require high trigger efficiency, typically > 70 % when normalized to the selected events). The remaining bandwidth is dedicated to inclusive selection of muons (900 Hz), di-muons (600 Hz) and D ∗ events (300 Hz). The inclusive streams will be used for calibration and to control systematics on different aspects of the analysis (trigger, tracking, tagging, particle identification) but also for physics studies. At June 2008 all the detectors, with the exception of the first station of the muon system M1 have been installed. Several cosmics runs have been performed to commission different parts of the experimental apparatus and of the trigger in order to be ready to acquire and reconstruct the first collisions. 3. LHCb physics program. In summer 2008 LHC will become operational, providing colliding beams at low energy and luminosity. The first events collected will be used for the final commissioning of the apparatus and the trigger optimization. In 2009 LHCb should run at the design luminosity and energy and after the HLT optimization, collect events useful for B-physics studies, hopefully about 0.5 fb−1 . Well known quantities and asymmetries like sin 2β or ΔMs will be measured in order to demonstrate that all the steps of a time dependent CP analysis are calibrated and well understood and the systematic effects under control. As it will be discussed in the following section, the collected statistics will be enough to perform some interesting measurement on CP violation and rare decays. From 2010 on the experiment should collect about ∼ 2 fb−1 per year. The present goal is to accumulate an integrated luminosity of ∼ 10 fb−1
215
by the end of 2013. In the following sections an overview of the expected physics performance on some selected topics that LHCb is planning to measure starting from the first B-physics run is presented. 3.1. Rare decays. The B−meson rare decays that proceed via Flavour Changing Neutral Current (FCNC) transitions are particularly interesting to search for new physics indirect evidence. Among them, the decay Bs → μ+ μ− has been identified as one of the most sensitive channel to probe SUSY models and put constraints to their parameters. In the SM the b → s transition is electro-weak loop and helicity suppressed so that the expected BR is particularly small: (3.35 ± 0.32) × 10−9 [3]. If NP contributes to the loop, the BR may be significantly enhanced. So far CDF and D0 can only set upper limits to the BR (47 × 10−9 [4] and 75 × 10−9 [5] at 90% of C.L. respectively) that are still compatible with a wide variety of SUSY models and parameter values [6]. Thanks to the high efficiency of the HLT di-muon trigger on this channel and the excellent tracking and particle identification performance, LHCb will collect about 113 events in 2 fb−1 of integrated luminosity [7]. The background (∼ 83 events) are mostly due to inclusive ¯bb → μμX events, while a small fraction is due to B → hh channel, where both hadrons are misidentified as muons. The μ+ μ− pairs that satisfiy minimal selection requirements will be classified as signal or background according to their distribution on a 3D space where the axis represent the geometrical likelihood, the particle identification likelihood and the μ+ μ− invariant mass. This approach has the advantage that efficiencies can be calibrated on real data using side bands and different control channels (like B → hh, J/ψ → μμ, B → J/ψ(μμ)X, etc). As normalization the decay B ± → J/ψ(μμ)K ± will be used, because it has a large and well measured BR. Moreover, since this channel will be triggered and selected in a similar way, some systematic errors can cancel. The main systematic error will be the uncertainty in the hadronization factors of Bs and Bu which at the moment amounts to 13%. Based on the results of the simulation studies it
216
S. Vecchi / Nuclear Physics B (Proc. Suppl.) 185 (2008) 213–219
is possible to predict the dependence of the upper limit on the branching ratio that LHCb can set as a function of the integrated luminosity, assuming that only background is observed (Fig. 2). Already with 0.05 fb−1 LHCb can reach the expected limits of the Tevatron experiments with the final data sample. With 0.5 fb−1 LHCb can push the BR limit down to the SM prediction and a 3 (5) standard deviations observation (measurement) of the SM prediction could be made with 2 (10) fb−1 of accumulated statistics.
Figure 2. Expected 90% CL upper limits on the BR of Bs → μ+ μ− a function of the integrated luminosity. Among the rare decays the B 0 → K ∗ μ+ μ− represents also an excellent channel to look for NP. This channel has also the advantage that more observables can be studied to test the dynamics: the angular distributions of the decay products and their dependence on the invariant mass s of the μ+ μ− pair. At present the experimental situation is limited by the low statistics (O(100) events) at the B-factories [8]. To study this channel efficiently LHCb will use a dedicated HLT exclusive selection. The expected signal yield in 2
fb−1 is ∼ 7200 and B/S < 0.5, mostly due to ¯bb → μμX events and the non resonant Kπ contribution under the K ∗ peak [9]. A precise study of the Forward-Backward Asymmetry (FBA) of the angular distribution of the μ+ relative to the B direction in the μ+ μ− rest frame as a function of s can be achieved already with 0.5 fb−1 . In particular the zero-crossing value s0 of the FBA, 2 4 well predicted in the SM (4.36+0.33 −0.31 GeV /c ) and particularly sensitive to NP contributions, can be determined with a precision of 0.9 (0.5) GeV2 /c4 in 0.5 (2) fb−1 of statistics. Moreover the study of the full angular distributions is in progress [10] as discussed by T. Hurth in this workshop [11]. The radiative decay Bs → φγ has high potentiality in evidencing NP contribution. In the SM the b → sγ transition produces a mostly left-handed photon with a defined CP eigenstate. As a consequence, CP violation effects are expected to be negligible. If NP contributions to the loop modify the photon polarization a mixing induced CP asymmetry Amix would originate. With a sizable value of ΔΓs also a non null value of AΔ could be measured analysing also untagged events. Given the statistics expected in 2 fb−1 (∼ 11k signal events and B/S < 0.55 at 90% C.L.) the precision on the CP asymmetries are respectively σ(Adir ), σ(Amix ) = 0.11 and σ(AΔ ) = 0.22 [12]. 3.2. CP violation studies. Time dependent CP asymmetries in B decays are originated by interference of amplitudes which have different properties through CP transformation. In Bd0 and Bs0 decays they can also be originated in the interference between the mixed and non mixed decay to a common final state. ¯ 0 mixing phase φSM is preIn the SM the Bs0 − B s s dicted to be rather small −0.0368±0.0017 rad [13] and NP can contribute to the loop affecting its P value: φs = φN +φSM s s . Recently both CDF and D0 published their results that have been interpreted by the UTfit collaboration as a first > 3σ indirect indication of non Minimal Flavour Violation NP contribution in flavour physics [14] and raised interesting discussions at the workshop[15]. The best way to measure φs is studying the time dependent asymmetry in Bs0 → J/ψφ decay channel. Since the final state is an admixture of CP
S. Vecchi / Nuclear Physics B (Proc. Suppl.) 185 (2008) 213–219
eigenstates an angular analysis is required to disentangle the odd contribution from the even ones. Thanks to the relatively large BR and the easy trigger selection of J/ψ LHCb expects to collect ∼ 131k signal events in 2 fb−1 with a low background contamination B/S=0.12. With a full three-angular analysis of the final state distributions the expected precision on φs is 0.023 rad. In addition to this channel φs can also be measured in many different CP-eigenstate modes, that in general have smaller statistics and higher background, providing a lower sensitivity to φs . All the decay channels considered so far are listed in Table 1: for each mode the expected signal yield in 2 fb−1 , the ratio B/S and the precision on φs are reported. Already with 0.5 fb−1 of statistics the combined sensitivity of 0.042 rad can be reached, improving by about a factor two the precision the Tevatron experiments expect to reach by the end of 2009. If the present Tevatron results will be confirmed, LHCb will be able to measure it with a > 3σ sensitivity.
Table 1 Expected experimental performance on different Bs decays channels that LHCb will use to measure the mixing phase φs . channel yield B/S σ(φs ) [rad] in 2 fb−1 Bs → J/ψη(γγ) 8.5k 2.0 0.109 3 0.142 Bs → J/ψη(π+ π− π0 ) 3 k 2.2 k 1 0.154 Bs → J/ψη(π + π − η) Bs → J/ψη(ργ) 4.2 k 0.4 0.08 Bs → ηc φ 3k 0.6 0.108 4k 0.3 0.133 Bs → Ds− Ds+ all CP eigenstates 0.046 Bs → J/ψφ 131k 0.12 0.023 combined 0.021
Another important contribution of LHCb to the precision tests of the SM is the measurement of the angle γ of the unitarity triangle. LHCb will perform this measurement using several channels, summarized in Table 2. The most powerful strategy to measure γ is with the family of B → DK
217
decays where the dependence on γ emerge in the interference of b → c and b → u tree-level transitions. ¯ 0 ) → D∓ K ± the two ampliIn the decay Bs0 (B s s ¯ 0 mixing. tudes can interfere through the Bs0 − B s The four possible time dependent decay rates depend on the combination of γ + φs in a clean way, so by measuring φs also γ can be determined. This method has an intrinsic 8-fold ambiguity in the γ values, that can be reduced to two in case ΔΓs has a sizable value. LHCb is planning to ana¯ 0 ) → D∓ π ± lyze this channel together with Bs0 (B s s in order to constrain common physics parameters like ΔΓs , ΔMs and the experimental ones like the tagging efficiency or the time acceptance. The expected sensitivity to γ + φs in 2 fb−1 is estimated within 9 − 12◦ , depending on the systematics (see Table 2). ¯ 0 ) → D∓ π ± channel In a similar way the Bd0 (B d can measure γ + 2β, provided that external constraint on the ratio of the two interfering amplitudes is applied. In this case, since ΔΓd ∼ 0 the 8-fold ambiguity cannot be resolved. Since this channel is governed by the same diagrams as the previous one, with the only substitution of the s quark with a d, it was suggested to use U-spin symmetry to relate the processes and extract γ without ambiguities or need of external constraints. In this case the sensitivity to γ is 10 − 15◦ depending on the assumptions on the Uspin symmetry. The angle γ can also be determined in the charged ¯ 0 )K ± by studying final states B ± decays to D0 (D ¯ 0 . In this case the common to both D0 and D interference of the b → c and b → u tree transitions determine a difference in the rates or in the kinematical distributions of the D decay products coming from the two charged B. In a combined fit of the two body D decays to π + π − , K + K − (GLW method) and K ± π ∓ (ADS method) the sensitivity to γ is estimated to be in the range 8.2 − 9.6◦ , depending on the strong phase values of the involved processes. In multi-body D decays the different kinematical distribution of the final products in B + and B − decays is due to interference effect that depend on γ. In this case the measurement precision is limited by the uncertainty of the D decay model,
218
S. Vecchi / Nuclear Physics B (Proc. Suppl.) 185 (2008) 213–219
Table 2 Expected experimental performance of different analyses to measure the angle γ of the unitarity triangle. The channel Bs → Ds K (*) measures γ + φs , so the sensitivity to γ should be obtained from the quoted value (*) subtracting the error on φs . channel D decay method σ(γ) Ref. (in 2 fb−1 ) Bs → Ds K KKπ time dependent CP analysis 9 − 12◦ (*) [16] Kπ time dependent CP anal. + U-spin 10 − 15◦ (*) [17] Bd → Dπ KK/ππ/Kπ/K3π decay rates (ADS + GLW) 8.2 − 9.6◦ [18] B + → DK + + + 0 Ks ππ dalitz (GGSZ) 7 − 12◦ + 10◦ [19] B → DK dalitz model independent 9 − 13◦ + 3◦ [20] + + KKππ dalitz 18◦ [21] B → DK KK/ππ/Kπ decay rates (ADS + GLW) 6 − 25◦ [22] B 0 → DK ∗0 combined time dep.+ADS+GLW+Dalitz 3.9 − 5.1◦ [23] Bd → ππ, Bs → KK time dependent CP analysis 10◦ [25] The range in the estimated error accounts for some systematics studies, like the variation of the strong phase involved in the processes, of the B/S ratio, of the ΔΓs value. that has to be measured independently. In the case of D0 → Ks0 π + π − this systematics amounts to 10◦ or 3◦ depending if the analysis assumes the amplitudes that fit the experimental data, or the CLEO-c data. The methods can be extended to a wider set of D decays, and to the B → DK ∗0 channel. The individual sensitivities to γ are listed in Table 2. A global fit to all the channels that involve the tree decays discussed so far give an overall sensitivity to γ of 3.9 − 5.1◦ in 2 fb−1 . In a less straightforward way γ can also be extracted from the time dependent CP analyses of Bd → π + π − and Bs → K + K − . In this case the dependence on γ comes from the interference of b → u tree process with b → d(s) penguin ones, that eventually could receive contribution from NP. Following the reasoning and the assumptions proposed in Ref.[24], the two decays can be related by U-spin symmetry and γ extracted. Allowing a 20% of U-spin symmetry breaking in the strong amplitudes of the two channels, and no constraints on the phases, γ can be measured with an error of 10◦ in 2 fb−1 . Nevertheless, given that the assumptions are rather strong and at the moment cannot be demonstrated, the method has not an unanimous consensus among the theoreticians. For this reason, before using the results of the analysis of these channels as a measurement
of γ it will be necessary to have first an experimental proof of the reliability of the hypotheses it is based on. 3.3. Charm Physics. LHCb can also study charm physics thanks to the dedicated stream of the HTL that will select D∗ → D0 π with ∼ 300Hz of bandwidth. This trigger will allow to collect a huge amount of D → hh events that will be useful to calibrate the RICH performance for the particle identification but can also be used to study charm physics. Since the D flavour is tagged by the charge of the accompanying pion in the D∗ decay and its lifetime can be precisely measured (σt ∼ 45fs), ¯ 0 mixing parameters LHCb can study D0 − D (x = ΔM/Γ and y = ΔΓ/2Γ) that are expected to be small (< 1%) in the SM [26], and were recently measured by BaBar and Belle. By studing the time dependence of the doubly Cabibbo-suppressed decay rates D0 → K + π − it is possible to determine the ”rotated” parameters x2 and y . Moreover from the lifetime ratio of D0 decays to CP (K + K − ) and non-CP eigenstates Cabibbo-favoured K − π + it is possible to measure the parameter yCP that equals y in case CP is conserved. Given the statistics that LHCb can collect [27], the statistical precision of σ(x2 ) = 0.14 × 10−3 , σ(y ) = 1.95 × 10−3
S. Vecchi / Nuclear Physics B (Proc. Suppl.) 185 (2008) 213–219
and σ(yCP ) = 1.1 × 10−3 in 2 fb−1 can be obtained. These are comparable to the B-factories final statistics precisions if the systematic errors will not dominate. In addition to the D0 mixing studies LHCb can also look for CP violation in D0 decays to π + π − and K + K − that are expected to be < 0.001 in the SM, but could be enhanced to O(0.01) in case of NP contribution. In this case the expected statistical precision in 2 fb−1 is σ(ACP ) = 1.1 × 10−3 . With the 10 fb−1 sample, depending on the size of the systematic effects that are being evaluated, LHCb could potentially measure the SM expected values. 4. Conclusions. The LHCb experiment is getting ready to collect and analyze real data. The experiment has excellent chances to contribute significantly to a better understanding of flavour-physics and possibly find indirect evidence of physics beyond the Standard Model. Already in 2009, with 0.5 fb−1 of collected statistics interesting results on the rare decays Bs → μ+ μ− and B → K ∗ μ+ μ− can be achieved and the indication of a NP contribu¯ 0 mixing coming from the retion in the Bs0 − B s cent interpretation of the Tevatron results could be confirmed with a larger significance. Within the first 2 years of data taking LHCb will be able to measure the unitarity angle γ with the same precision (∼ 6◦ ) that is now given by the global fit of all the experimental data assuming the Standard Model validity and the CKM matrix. In this way any inconsistencies due to new physics could be evident. LHCb will also contribute to charm-physics ¯ 0 mixing and the CP violation studying the D0 −D in some specific decays. The estimated statistical precision will allow to measure the SM values with the 10 fb−1 data sample. REFERENCES 1. The LHCb collaboration, ”The LHCb detector at the LHC”, in press on JNST. 2. M.Calvi et al., LHCb public note 2007-058. 3. M. Blanke et al., hep-ph/0604057v5.
219
4. T. Aaltonen et al., The CDF collaboration, Phys. Rev. Lett. 100, 101802 (2008). 5. V. M. Abazov et al., The D0 collaboration, D0 Note 5477-CONF, Phys. Rev. D 76, 092001 (2007). 6. J. Ellis et al. Journal of High Energy Phys. 0710:092, 2007 and arXiv:0709.0098 [hep-ph], G. Isidori et al. Phys. Rev. D 75, 115019 (2007). 7. D.Martinez, LHCb public notes 2008-018, D.Martinez et al., LHCb public note 2007033. 8. J. Walsh, these proceedings. 9. J.Dickens et al., LHCb public notes: 2007038, 2007-039. 10. W.Reece, LHCb public note 2008-021, U.Egede, LHCb public note 2007-057. 11. T. Hurth, these proceedings. 12. L.Shchutska et al, LHCb public notes: 2007030, 2007-147. 13. M. Bona et al., The UTfit Collaboration, http://www.utfit.org/. 14. M. Bona et al., The UTfit Collaboration, arXiv:0803.0659 [hep-ph]. 15. L. Silvestrini, these proceedings, J.Charles, these proceedings. 16. S.Cohen et al., LHCb public note 2007-041. 17. V.V. Gligorov and G. Wilkinson, LHCb public note 2008-035. 18. M. Patel, LHCb public notes 2008-011, 2007043, 2006-066. 19. V. Gibson et al., LHCb public note 2007-048. 20. J. Libby, LHCb public note 2007-141. 21. J. Libby et al., LHCb public note 2007-098. 22. K.Akiba and M.Gandelman, LHCb public note 2007-050 23. K. Akiba et al., LHCb public note 2008-031. 24. R. Fleischer, Eur. Phys. J. C52 (2007), 267281 and arXiv:0705.1121 [hep-ph] 25. A.Carbone et al., LHCb public note 2007-059. 26. I. Bigi, arXiv:hep-ph/9712475 and these proceedings. 27. P.Spradlin et al., LHCb public note 2007-049.
The LHCb experiment and its expected physics performance. S. Vecchia∗ a
INFN, via Saragat 1, Blocco C, 44100 Ferrara, Italy The LHCb experiment is dedicated to the study of heavy flavour physics at the LHC machine. Its primary motivation is to study with high statistics and precision CP violation and rare B hadron decays, to improve the knowledge of the Standard Model or to get indirect hints of new physics contribution to the flavour physics. The main experimental features of the experiment are presented and the sensitivities on some important measurement that LHCb will perform since the first data will be available are discussed.
1. Introduction
2. The LHCb experiment
In the last decade the experiments at the Bfactories and at Tevatron contributed to an impressive improvement in the understanding of the Standard Model (SM). In particular, many different measurement allowed to check the consistency of the CKM picture of the SM and constraint its matrix elements to a good precision. Nevertheless the level of CP violation that the SM account for is insufficient to explain the amount of matter in the universe. A new source of CP violation beyond the SM is therefore needed to solve this puzzle. If New Physics (NP) exists at a not too high energy scale (∼ TeV), it could manifest itself indirectly in beauty or charm meson decays that proceed through loop diagrams, where virtual NP particles could contribute in addition to the SM ones. Since these processes are usually suppressed in the SM, the eventual NP contributions should be easier to be measured. The possible results would be unexpected CP violation effects or modified properties of rare decays. Thanks to the high statistics of b-hadrons that will analyze and the excellent expected experimental performance, LHCb will contribute to a further improvement of several interesting measurement already in the first year of physicsquality data taking in 2009.
The LHCb experiment will exploit the high b¯b cross√section (σb¯b ∼ 500μb) that is expected at the s = 14 TeV energy of the colliding protons at LHC. Given that the b¯b pairs are mostly produced in the forward or backward direction, the LHCb detector was designed as a forward spectrometer, covering a pseudo-rapidity range 1.9 < η < 4.9. Selecting B with transverse momentum pT > 2 GeV/c the b-hadron production cross section is σb¯b ∼ 230μb. In order to maximize the probability of a single interaction per beam crossing, it was also decided to limit the luminosity in the LHCb interaction region to 2 − 5 × 1032 cm−2 s−1 . This choice facilitates the study of B-physics and has the additional advantage of reducing the radiation damage due to the high particle flux at small angles. In these conditions one year (∼ 107 s) of ”normal” LHCb running corresponds to 2 fb−1 of integrated luminosity and ∼ 1012 b¯b pairs produced in the region covered by the spectrometer. The detector and the trigger of the experiment are especially designed to reach excellent performance in the specific B meson decays reconstruction that usually have small Branching Ratios (BR< 10−4 ), while rejecting the background due mostly to inclusive b¯b and inelastic pp interactions (σvisible /σb¯b ∼ 120). The LHCb spectrometer layout is shown in Figure 1 and consists of several detectors [1].
∗ on
behalf of the LHCb Collaboration
0920-5632/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.nuclphysbps.2008.10.027
214
S. Vecchi / Nuclear Physics B (Proc. Suppl.) 185 (2008) 213–219
Figure 1. Layout of the LHCb detector.
The Vertex Locator (VELO) consists of 21 modules of silicon strips that reconstruct the radial and azimuthal coordinates of the tracks closest to the interaction point. It is divided into two halves that can be moved as close as 8 mm to the beams in stable running conditions. Some modules are used at trigger level to reject events with multiple interactions. This detector allows to reconstruct primary vertices with an excellent precision: 10 (60) μm in the transverse (parallel) direction to the beam axis. The measurement of the impact parameter of tracks relative to the primary vertex, a crucial parameter for identifying particles originated by a B decay, is made with a precision of σIP =(14 + 35/pT )μm (with pT in GeV/c units). The particle’s tracking is completed by two tracking systems one upstream (Turicensis Tracker, TT) and one downstream the 4 Tm magnet (Inner and Outer Tracker: IT and OT). TT and IT consist of silicon strip detectors, while OT consist of straw tubes. They allow to measure the momentum of the charged particles with a precision of σp /p = (0.4 + 0.0015 × p)% (with p in GeV/c units) with > 95% efficiency. The overall performance of the tracking system allows to reconstruct the B invariant mass and proper time with a resolution of σm ∼ 15 − 20 MeV/c2 and σt ∼ 30 − 40 fs respectively, depending on the channel.
Two Ring Imaging Cherenkov detectors (RICH1 downstream the VELO, and RICH2 downstream the IT/OT) allow to identify efficiently pions, kaons and protons in the wide momentum range 2 − 100 GeV/c of the B decay products. They play a key role in the selection of some specific channels, the background suppression, and the determination of the B flavour at production (B-tagging). The calorimeter system of the experiment comprises a pre-shower detector (PS), an electromagnetic calorimeter (ECAL) and a hadron calorimeter (HCAL). It provides the identification of electrons, hadrons and neutrals (photons and π 0 ) as well as the measurement of their energies and positions. Its informations are used also at trigger level L0 to select particles with high ET (> 2 − 3 GeV/c) deposit coming from B decays. The muon system is designed to identify muons both at trigger level L0, selecting high pT muons, and off line for the channel reconstruction and the B flavour tagging. It consists of five stations (M1-M5) made of multi wire proportional chambers and 12 triple-GEM detectors to cover the innermost region of M1, which has higher occupancy. The overall tracking and particle identification performance of the detector allow efficient and low background reconstruction of many b-hadron exclusive decays, also with very low BR. They also allow to tag the B flavor at production with a tagging power of 4 − 5% for a Bd and 7 − 9% for a Bs , by means of the opposite and the same side tagging algorithms described in [2]. The LHCb trigger is organized in two levels: the first level trigger (L0), realized with custom electronics, selects events with relatively ”high” pT (typically > 1 − 4 GeV/c) electrons, muons, photons and hadrons that most likely come from a B decay, while rejecting the events with multiple interactions. It uses the informations of the calorimeters, the muon system an the pile-up veto modules in the VELO detector. The decision is taken within a latency of 4μs at the input frequency of 40 MHz with an output rate of 1 MHz. The L0 selected events are then read-out and examined by the High Level Trigger (HLT), which
S. Vecchi / Nuclear Physics B (Proc. Suppl.) 185 (2008) 213–219
consists of a software trigger running on a farm of ∼ 1800 CPU. First it requires a confirmation of the L0 trigger decision with the more accurate information available; after a partial track reconstruction, cuts on the impact parameter of the L0 selected tracks can be applied. Finally the event is fully reconstructed and written on tape with a frequency of 2 kHz if it fulfills the requirements of one of the possible streams of HLT. 200 Hz of the output rate is dedicated to the exclusive selections of specific B decay channels that require high trigger efficiency, typically > 70 % when normalized to the selected events). The remaining bandwidth is dedicated to inclusive selection of muons (900 Hz), di-muons (600 Hz) and D ∗ events (300 Hz). The inclusive streams will be used for calibration and to control systematics on different aspects of the analysis (trigger, tracking, tagging, particle identification) but also for physics studies. At June 2008 all the detectors, with the exception of the first station of the muon system M1 have been installed. Several cosmics runs have been performed to commission different parts of the experimental apparatus and of the trigger in order to be ready to acquire and reconstruct the first collisions. 3. LHCb physics program. In summer 2008 LHC will become operational, providing colliding beams at low energy and luminosity. The first events collected will be used for the final commissioning of the apparatus and the trigger optimization. In 2009 LHCb should run at the design luminosity and energy and after the HLT optimization, collect events useful for B-physics studies, hopefully about 0.5 fb−1 . Well known quantities and asymmetries like sin 2β or ΔMs will be measured in order to demonstrate that all the steps of a time dependent CP analysis are calibrated and well understood and the systematic effects under control. As it will be discussed in the following section, the collected statistics will be enough to perform some interesting measurement on CP violation and rare decays. From 2010 on the experiment should collect about ∼ 2 fb−1 per year. The present goal is to accumulate an integrated luminosity of ∼ 10 fb−1
215
by the end of 2013. In the following sections an overview of the expected physics performance on some selected topics that LHCb is planning to measure starting from the first B-physics run is presented. 3.1. Rare decays. The B−meson rare decays that proceed via Flavour Changing Neutral Current (FCNC) transitions are particularly interesting to search for new physics indirect evidence. Among them, the decay Bs → μ+ μ− has been identified as one of the most sensitive channel to probe SUSY models and put constraints to their parameters. In the SM the b → s transition is electro-weak loop and helicity suppressed so that the expected BR is particularly small: (3.35 ± 0.32) × 10−9 [3]. If NP contributes to the loop, the BR may be significantly enhanced. So far CDF and D0 can only set upper limits to the BR (47 × 10−9 [4] and 75 × 10−9 [5] at 90% of C.L. respectively) that are still compatible with a wide variety of SUSY models and parameter values [6]. Thanks to the high efficiency of the HLT di-muon trigger on this channel and the excellent tracking and particle identification performance, LHCb will collect about 113 events in 2 fb−1 of integrated luminosity [7]. The background (∼ 83 events) are mostly due to inclusive ¯bb → μμX events, while a small fraction is due to B → hh channel, where both hadrons are misidentified as muons. The μ+ μ− pairs that satisfiy minimal selection requirements will be classified as signal or background according to their distribution on a 3D space where the axis represent the geometrical likelihood, the particle identification likelihood and the μ+ μ− invariant mass. This approach has the advantage that efficiencies can be calibrated on real data using side bands and different control channels (like B → hh, J/ψ → μμ, B → J/ψ(μμ)X, etc). As normalization the decay B ± → J/ψ(μμ)K ± will be used, because it has a large and well measured BR. Moreover, since this channel will be triggered and selected in a similar way, some systematic errors can cancel. The main systematic error will be the uncertainty in the hadronization factors of Bs and Bu which at the moment amounts to 13%. Based on the results of the simulation studies it
216
S. Vecchi / Nuclear Physics B (Proc. Suppl.) 185 (2008) 213–219
is possible to predict the dependence of the upper limit on the branching ratio that LHCb can set as a function of the integrated luminosity, assuming that only background is observed (Fig. 2). Already with 0.05 fb−1 LHCb can reach the expected limits of the Tevatron experiments with the final data sample. With 0.5 fb−1 LHCb can push the BR limit down to the SM prediction and a 3 (5) standard deviations observation (measurement) of the SM prediction could be made with 2 (10) fb−1 of accumulated statistics.
Figure 2. Expected 90% CL upper limits on the BR of Bs → μ+ μ− a function of the integrated luminosity. Among the rare decays the B 0 → K ∗ μ+ μ− represents also an excellent channel to look for NP. This channel has also the advantage that more observables can be studied to test the dynamics: the angular distributions of the decay products and their dependence on the invariant mass s of the μ+ μ− pair. At present the experimental situation is limited by the low statistics (O(100) events) at the B-factories [8]. To study this channel efficiently LHCb will use a dedicated HLT exclusive selection. The expected signal yield in 2
fb−1 is ∼ 7200 and B/S < 0.5, mostly due to ¯bb → μμX events and the non resonant Kπ contribution under the K ∗ peak [9]. A precise study of the Forward-Backward Asymmetry (FBA) of the angular distribution of the μ+ relative to the B direction in the μ+ μ− rest frame as a function of s can be achieved already with 0.5 fb−1 . In particular the zero-crossing value s0 of the FBA, 2 4 well predicted in the SM (4.36+0.33 −0.31 GeV /c ) and particularly sensitive to NP contributions, can be determined with a precision of 0.9 (0.5) GeV2 /c4 in 0.5 (2) fb−1 of statistics. Moreover the study of the full angular distributions is in progress [10] as discussed by T. Hurth in this workshop [11]. The radiative decay Bs → φγ has high potentiality in evidencing NP contribution. In the SM the b → sγ transition produces a mostly left-handed photon with a defined CP eigenstate. As a consequence, CP violation effects are expected to be negligible. If NP contributions to the loop modify the photon polarization a mixing induced CP asymmetry Amix would originate. With a sizable value of ΔΓs also a non null value of AΔ could be measured analysing also untagged events. Given the statistics expected in 2 fb−1 (∼ 11k signal events and B/S < 0.55 at 90% C.L.) the precision on the CP asymmetries are respectively σ(Adir ), σ(Amix ) = 0.11 and σ(AΔ ) = 0.22 [12]. 3.2. CP violation studies. Time dependent CP asymmetries in B decays are originated by interference of amplitudes which have different properties through CP transformation. In Bd0 and Bs0 decays they can also be originated in the interference between the mixed and non mixed decay to a common final state. ¯ 0 mixing phase φSM is preIn the SM the Bs0 − B s s dicted to be rather small −0.0368±0.0017 rad [13] and NP can contribute to the loop affecting its P value: φs = φN +φSM s s . Recently both CDF and D0 published their results that have been interpreted by the UTfit collaboration as a first > 3σ indirect indication of non Minimal Flavour Violation NP contribution in flavour physics [14] and raised interesting discussions at the workshop[15]. The best way to measure φs is studying the time dependent asymmetry in Bs0 → J/ψφ decay channel. Since the final state is an admixture of CP
S. Vecchi / Nuclear Physics B (Proc. Suppl.) 185 (2008) 213–219
eigenstates an angular analysis is required to disentangle the odd contribution from the even ones. Thanks to the relatively large BR and the easy trigger selection of J/ψ LHCb expects to collect ∼ 131k signal events in 2 fb−1 with a low background contamination B/S=0.12. With a full three-angular analysis of the final state distributions the expected precision on φs is 0.023 rad. In addition to this channel φs can also be measured in many different CP-eigenstate modes, that in general have smaller statistics and higher background, providing a lower sensitivity to φs . All the decay channels considered so far are listed in Table 1: for each mode the expected signal yield in 2 fb−1 , the ratio B/S and the precision on φs are reported. Already with 0.5 fb−1 of statistics the combined sensitivity of 0.042 rad can be reached, improving by about a factor two the precision the Tevatron experiments expect to reach by the end of 2009. If the present Tevatron results will be confirmed, LHCb will be able to measure it with a > 3σ sensitivity.
Table 1 Expected experimental performance on different Bs decays channels that LHCb will use to measure the mixing phase φs . channel yield B/S σ(φs ) [rad] in 2 fb−1 Bs → J/ψη(γγ) 8.5k 2.0 0.109 3 0.142 Bs → J/ψη(π+ π− π0 ) 3 k 2.2 k 1 0.154 Bs → J/ψη(π + π − η) Bs → J/ψη(ργ) 4.2 k 0.4 0.08 Bs → ηc φ 3k 0.6 0.108 4k 0.3 0.133 Bs → Ds− Ds+ all CP eigenstates 0.046 Bs → J/ψφ 131k 0.12 0.023 combined 0.021
Another important contribution of LHCb to the precision tests of the SM is the measurement of the angle γ of the unitarity triangle. LHCb will perform this measurement using several channels, summarized in Table 2. The most powerful strategy to measure γ is with the family of B → DK
217
decays where the dependence on γ emerge in the interference of b → c and b → u tree-level transitions. ¯ 0 ) → D∓ K ± the two ampliIn the decay Bs0 (B s s ¯ 0 mixing. tudes can interfere through the Bs0 − B s The four possible time dependent decay rates depend on the combination of γ + φs in a clean way, so by measuring φs also γ can be determined. This method has an intrinsic 8-fold ambiguity in the γ values, that can be reduced to two in case ΔΓs has a sizable value. LHCb is planning to ana¯ 0 ) → D∓ π ± lyze this channel together with Bs0 (B s s in order to constrain common physics parameters like ΔΓs , ΔMs and the experimental ones like the tagging efficiency or the time acceptance. The expected sensitivity to γ + φs in 2 fb−1 is estimated within 9 − 12◦ , depending on the systematics (see Table 2). ¯ 0 ) → D∓ π ± channel In a similar way the Bd0 (B d can measure γ + 2β, provided that external constraint on the ratio of the two interfering amplitudes is applied. In this case, since ΔΓd ∼ 0 the 8-fold ambiguity cannot be resolved. Since this channel is governed by the same diagrams as the previous one, with the only substitution of the s quark with a d, it was suggested to use U-spin symmetry to relate the processes and extract γ without ambiguities or need of external constraints. In this case the sensitivity to γ is 10 − 15◦ depending on the assumptions on the Uspin symmetry. The angle γ can also be determined in the charged ¯ 0 )K ± by studying final states B ± decays to D0 (D ¯ 0 . In this case the common to both D0 and D interference of the b → c and b → u tree transitions determine a difference in the rates or in the kinematical distributions of the D decay products coming from the two charged B. In a combined fit of the two body D decays to π + π − , K + K − (GLW method) and K ± π ∓ (ADS method) the sensitivity to γ is estimated to be in the range 8.2 − 9.6◦ , depending on the strong phase values of the involved processes. In multi-body D decays the different kinematical distribution of the final products in B + and B − decays is due to interference effect that depend on γ. In this case the measurement precision is limited by the uncertainty of the D decay model,
218
S. Vecchi / Nuclear Physics B (Proc. Suppl.) 185 (2008) 213–219
Table 2 Expected experimental performance of different analyses to measure the angle γ of the unitarity triangle. The channel Bs → Ds K (*) measures γ + φs , so the sensitivity to γ should be obtained from the quoted value (*) subtracting the error on φs . channel D decay method σ(γ) Ref. (in 2 fb−1 ) Bs → Ds K KKπ time dependent CP analysis 9 − 12◦ (*) [16] Kπ time dependent CP anal. + U-spin 10 − 15◦ (*) [17] Bd → Dπ KK/ππ/Kπ/K3π decay rates (ADS + GLW) 8.2 − 9.6◦ [18] B + → DK + + + 0 Ks ππ dalitz (GGSZ) 7 − 12◦ + 10◦ [19] B → DK dalitz model independent 9 − 13◦ + 3◦ [20] + + KKππ dalitz 18◦ [21] B → DK KK/ππ/Kπ decay rates (ADS + GLW) 6 − 25◦ [22] B 0 → DK ∗0 combined time dep.+ADS+GLW+Dalitz 3.9 − 5.1◦ [23] Bd → ππ, Bs → KK time dependent CP analysis 10◦ [25] The range in the estimated error accounts for some systematics studies, like the variation of the strong phase involved in the processes, of the B/S ratio, of the ΔΓs value. that has to be measured independently. In the case of D0 → Ks0 π + π − this systematics amounts to 10◦ or 3◦ depending if the analysis assumes the amplitudes that fit the experimental data, or the CLEO-c data. The methods can be extended to a wider set of D decays, and to the B → DK ∗0 channel. The individual sensitivities to γ are listed in Table 2. A global fit to all the channels that involve the tree decays discussed so far give an overall sensitivity to γ of 3.9 − 5.1◦ in 2 fb−1 . In a less straightforward way γ can also be extracted from the time dependent CP analyses of Bd → π + π − and Bs → K + K − . In this case the dependence on γ comes from the interference of b → u tree process with b → d(s) penguin ones, that eventually could receive contribution from NP. Following the reasoning and the assumptions proposed in Ref.[24], the two decays can be related by U-spin symmetry and γ extracted. Allowing a 20% of U-spin symmetry breaking in the strong amplitudes of the two channels, and no constraints on the phases, γ can be measured with an error of 10◦ in 2 fb−1 . Nevertheless, given that the assumptions are rather strong and at the moment cannot be demonstrated, the method has not an unanimous consensus among the theoreticians. For this reason, before using the results of the analysis of these channels as a measurement
of γ it will be necessary to have first an experimental proof of the reliability of the hypotheses it is based on. 3.3. Charm Physics. LHCb can also study charm physics thanks to the dedicated stream of the HTL that will select D∗ → D0 π with ∼ 300Hz of bandwidth. This trigger will allow to collect a huge amount of D → hh events that will be useful to calibrate the RICH performance for the particle identification but can also be used to study charm physics. Since the D flavour is tagged by the charge of the accompanying pion in the D∗ decay and its lifetime can be precisely measured (σt ∼ 45fs), ¯ 0 mixing parameters LHCb can study D0 − D (x = ΔM/Γ and y = ΔΓ/2Γ) that are expected to be small (< 1%) in the SM [26], and were recently measured by BaBar and Belle. By studing the time dependence of the doubly Cabibbo-suppressed decay rates D0 → K + π − it is possible to determine the ”rotated” parameters x2 and y . Moreover from the lifetime ratio of D0 decays to CP (K + K − ) and non-CP eigenstates Cabibbo-favoured K − π + it is possible to measure the parameter yCP that equals y in case CP is conserved. Given the statistics that LHCb can collect [27], the statistical precision of σ(x2 ) = 0.14 × 10−3 , σ(y ) = 1.95 × 10−3
S. Vecchi / Nuclear Physics B (Proc. Suppl.) 185 (2008) 213–219
and σ(yCP ) = 1.1 × 10−3 in 2 fb−1 can be obtained. These are comparable to the B-factories final statistics precisions if the systematic errors will not dominate. In addition to the D0 mixing studies LHCb can also look for CP violation in D0 decays to π + π − and K + K − that are expected to be < 0.001 in the SM, but could be enhanced to O(0.01) in case of NP contribution. In this case the expected statistical precision in 2 fb−1 is σ(ACP ) = 1.1 × 10−3 . With the 10 fb−1 sample, depending on the size of the systematic effects that are being evaluated, LHCb could potentially measure the SM expected values. 4. Conclusions. The LHCb experiment is getting ready to collect and analyze real data. The experiment has excellent chances to contribute significantly to a better understanding of flavour-physics and possibly find indirect evidence of physics beyond the Standard Model. Already in 2009, with 0.5 fb−1 of collected statistics interesting results on the rare decays Bs → μ+ μ− and B → K ∗ μ+ μ− can be achieved and the indication of a NP contribu¯ 0 mixing coming from the retion in the Bs0 − B s cent interpretation of the Tevatron results could be confirmed with a larger significance. Within the first 2 years of data taking LHCb will be able to measure the unitarity angle γ with the same precision (∼ 6◦ ) that is now given by the global fit of all the experimental data assuming the Standard Model validity and the CKM matrix. In this way any inconsistencies due to new physics could be evident. LHCb will also contribute to charm-physics ¯ 0 mixing and the CP violation studying the D0 −D in some specific decays. The estimated statistical precision will allow to measure the SM values with the 10 fb−1 data sample. REFERENCES 1. The LHCb collaboration, ”The LHCb detector at the LHC”, in press on JNST. 2. M.Calvi et al., LHCb public note 2007-058. 3. M. Blanke et al., hep-ph/0604057v5.
219
4. T. Aaltonen et al., The CDF collaboration, Phys. Rev. Lett. 100, 101802 (2008). 5. V. M. Abazov et al., The D0 collaboration, D0 Note 5477-CONF, Phys. Rev. D 76, 092001 (2007). 6. J. Ellis et al. Journal of High Energy Phys. 0710:092, 2007 and arXiv:0709.0098 [hep-ph], G. Isidori et al. Phys. Rev. D 75, 115019 (2007). 7. D.Martinez, LHCb public notes 2008-018, D.Martinez et al., LHCb public note 2007033. 8. J. Walsh, these proceedings. 9. J.Dickens et al., LHCb public notes: 2007038, 2007-039. 10. W.Reece, LHCb public note 2008-021, U.Egede, LHCb public note 2007-057. 11. T. Hurth, these proceedings. 12. L.Shchutska et al, LHCb public notes: 2007030, 2007-147. 13. M. Bona et al., The UTfit Collaboration, http://www.utfit.org/. 14. M. Bona et al., The UTfit Collaboration, arXiv:0803.0659 [hep-ph]. 15. L. Silvestrini, these proceedings, J.Charles, these proceedings. 16. S.Cohen et al., LHCb public note 2007-041. 17. V.V. Gligorov and G. Wilkinson, LHCb public note 2008-035. 18. M. Patel, LHCb public notes 2008-011, 2007043, 2006-066. 19. V. Gibson et al., LHCb public note 2007-048. 20. J. Libby, LHCb public note 2007-141. 21. J. Libby et al., LHCb public note 2007-098. 22. K.Akiba and M.Gandelman, LHCb public note 2007-050 23. K. Akiba et al., LHCb public note 2008-031. 24. R. Fleischer, Eur. Phys. J. C52 (2007), 267281 and arXiv:0705.1121 [hep-ph] 25. A.Carbone et al., LHCb public note 2007-059. 26. I. Bigi, arXiv:hep-ph/9712475 and these proceedings. 27. P.Spradlin et al., LHCb public note 2007-049.