LHCb-PROC-2013-005 Future prospects at LHCb Marie-H´el`ene Schune1 LAL, Universit´e Paris-Sud 3 1 CNRS/IN2P3 0 Orsay, FRANCE 2 n a J 1 Abstract 2 The LHCb experiment is running at the Large Hadron Collider to study ] CP violation and rare decays in the beauty and charm sectors. The motivation x e and the strategy of the upgrade envisaged for the long shutdown LS2 (2018) is - presented. The current results for some exemplary physics analyses are given p e and the expected performances foreseen for 2018 and for the LHCb Upgrade h project with an integrated luminosity of 50 fb−1 are summarized. [ 1 v 1 Introduction 1 1 8 The LHCb experiment, located on the Large Hadron Collider (LHC), has been suc- 4 . cessfully taking data during the last three years in running conditions very encour- 1 0 aging for the future. The general context of the LHCb Upgrade project is presented, 3 followed by a short overview of the main changes foreseen. Finally, using some exem- 1 plary physics analyses, the current results are shown and the expected precisions are : v given. i X r a 2 General context 2.1 Introduction Studies of CP violation and more generally flavor changing neutral currents were es- sential in establishing the Standard Model (SM). Today, they appear as a powerful tool to reveal processes beyond the SM and to understand their nature. In this con- text, LHCb plays a key role. The LHCb detector[1], shown in Fig.1, is a single-arm 1On behalf of the LHCb collaboration. 1 forward spectrometer covering the pseudorapidity range 2 < η < 5, designed for the study of particles containing b or c quarks. The detector geometry is driven by the kinematics of the bb pair production at the LHC energy where both b and b quarks mainly fly in the forward or backward direction. The detector includes a high pre- cision tracking system consisting of a silicon-strip vertex detector surrounding the pp interaction region, a large-area silicon-strip detector located upstream of a dipole magnet with a bending power of about 4Tm, and three stations of silicon-strip de- tectors and straw drift tubes placed downstream. The combined tracking system has a momentum resolution ∆p/p that varies from 0.4% at 5GeV/c to 0.6% at 100GeV/c, and an impact parameter resolution of 20µm for tracks with high transverse momen- tum. Charged hadrons are identified using two ring-imaging Cherenkov detectors. Photon, electron and hadron candidates are identified by a calorimeter system con- sisting of scintillating-pad and preshower detectors, an electromagnetic calorimeter and a hadronic calorimeter. Muons are identified by a system composed of alter- nating layers of iron and multiwire proportional chambers. The trigger consists of a hardware stage, the Level-0 (L0) trigger based on information from the calorimeter and muon systems, followed by a software stage the High Level Trigger (HLT) which applies a full event reconstruction. 2.2 Current data taking and longer term plans During2011,boththeLHCmachineandtheLHCbdetectorperformedsuperbly. This √ hasallowedLHCbtoaccumulate1.0 fb−1 of s=7TeVppcollisionsthatisavailable for physics analysis. The detector was running at an instantaneous luminosity of about 3.5 × 1032cm−2 sec−1, and with a number of pp interactions per crossing of ∼ 1.4. The values of these running parameters are above the design values by a factor 1.8 and 3.5 respectively. The large integrated luminosity was collected thanks to the luminosity leveling schema set in place by the LHC team. Instead of being defocused, the beams are displaced from head-on collision in order to achieve the required luminosity. This offset is reduced with time in order to follow the loss in beam intensity throughout the fill. This procedure allows a constant luminosity to be maintained during the whole period of the fill, in contrast to ATLAS and CMS which experience an exponential decrease. During 2012 data taking, the center of mass energy has been increased by ∼ 15% √ ( s = 8 TeV) which corresponds to an increase of about 15 % in the number of bb events. The instantaneous luminosity is also slightly larger : 4 × 1032cm−2 sec−1. About 1.4 fb−1 of pp collisions have been already recorded and three months of data taking are still planned.2 In addition, two changes have been made for 2012 data taking : the HLT output rate has been increased from 3 kHz in 2011 to 4.5 kHz √ 2At the end of the 2012 run 2 fb−1 of pp collisions at s = 8 TeV have been recorded. 2 Figure 1: Vertical view of the LHCb detector. 3 which mainly allows to record more charm events and to cope with the increase in cross-sections. A deferred trigger schema has also been put in place. In this schema about 20 % of the events are written to disk before the HLT decision which is then made during the inter-fill gap periods. It results in an increase of about 20% in the data sample. These two novelties ensure a larger data sample to work with during the two years shut-down period which will take place in 2013-2014. After this shutdown period, a three years data taking period is foreseen at a center √ of mass energy of s = 13-14 TeV which corresponds approximately to a doubling of the bb cross section compared to 2011. It is expected that the recorded integrated luminosityduringthisperiodshouldbeoftheorderof5 fb−1. Afterthesethreeyears, according to the LHC planning, a second shutdown (LS2, starting in 2018) will take place and the upgraded LHCb detector will be installed. Very important changes in the detector will be made (see Sec. 3) and it is foreseen to be able to record 5 fb−1 per year for a total of 50fb−1 at a center of mass energy of 14 TeV. 2.3 Why do we need to upgrade the LHCb experiment ? Using the 2011 dataset, the LHCb experiment has already been able to perform extremely precise measurements. In most cases the precision obtained is the world’s best. Unfortunately, all the measurements are consistent with the SM and New Physics (NP) has not shown up. However, even if the measurements are much more precise than before, there is still plenty of room for NP. As an example, let’s take the global fits as performed by the CKMFitter [2] or UTFit [3] collaborations. In these fits they allow for the presence of generic NP signals in the B0 or B0 mixing. s Whatever the statistical treatment, there is a good agreement with the SM but, due to the size of the uncertainties, the presence of NP at the level of 20 to 30% is still possible. It is thus highly desirable to largely increase the available data samples and this can only be achieved through redesign of the LHCb experiment both from the detector point of view, to be able to cope with a larger instantaneous luminosity, but also from a trigger point of view as will be explained in the next part. 3 The LHCb upgrade The aim of the upgraded LHCb experiment is to reach experimental sensitivities comparable to or better than the theoretical ones and thus to record an integrated luminosity of about 50 fb−1. It implies to run at L = 1×1033 cm−2 sec−1 with a fully flexible software trigger running at a readout rate of 40 MHz [4, 5]. This increases the annual signal yields, as compared to those obtained by LHCb in 2011, by a factor around ten for muonic B decays and twenty or more for heavy-flavour decays to hadronic final states. For reason of flexibility, and to allow for possible evolutions 4 of the trigger, it was decided to design those dectectors that need replacement such that they can sustain a luminosity of L = 2 × 1033cm−2sec−1. The challenge of this update is related to the number of pp interactions per crossing. The average number of interactions per crossing is about 2.3 at L = 1 × 1033cm−2sec−1 and reaches more than 4 interactions per crossing at L = 2 × 1033cm−2sec−1. In the latter case, all crossings have at least one interaction. These extreme conditions put heavy requirements on the tracking as well as on the trigger algorithms. 3.1 The current LHCb trigger ThecurrentLHCbtrigger[6]containstwostages,theL0andtheHLT.TheL0reduces the rate from 40 MHz down to 1 MHz. It is based on custom electronics receiving dedicated information from the calorimeters and from the muon detector. It looks for lepton and hadron candidates with a high transverse momentum. The HLT trigger reduces the rate down to 3 to 4.5 kHz. The HLT is a software trigger running on a dedicatedCPUfarmandreceivingthefulldetectorinformationat1MHz. Byrunning trackingandvertexingalgorithmsitselectsleptonsandhadronswithahightransverse momentum as well as a high impact parameters. More elaborate algorithms, close to the off-line selections, are then applied to select inclusive or exclusive heavy-quark decays. The L0 saturates for hadronic channels when the luminosity increases [4]. At high luminosity we cannot rely only on the transverse momentum cut for efficient triggering. On the contrary, a software based trigger allows use of many discriminants including track impact parameters and combinations of different criteria. 3.2 The upgraded LHCb trigger The main point in the upgraded LHCb trigger is to remove the 1 MHz bottleneck at the output of the L0 trigger which is determined by the front-end electronics rate. These front-end electronics will be upgraded to allow reading events at the LHC clock rate. In principle, in the upgrade schema, the data acquisition (DAQ) and event building could be performed in a large CPU farm at the full rate of 40 MHz. However the upgrade is designed to be able to cope with a staged DAQ system which cannotyethandlethefullrate, occupancyfluctuationswhichpreventthefullreadout, and insufficient CPU power in the CPU farm. Hence the upgrade will also contain a Low Level Trigger (LLT), which, similarly to the current L0 trigger, should not just pre-scale to a rate acceptable, but enrich the selected sample with interesting events. The LLT corresponds closely to the current L0, but with a tunable output rate higher than the current 1 MHz limit. In the case of hadronic channels, such as for example the B0 → φφ decay mode, with a LLT output rate of 10 MHz, the trigger efficiency s should be about 4 times larger than with the current 1 MHz rate [4]. 5 3.3 Sub-detectors upgrades In addition to the change of most of the front-end electronics, several parts of the current LHCb detector will have to be replaced in order to cope with the increased occupancy [4]. The open geometry of the LHCb detector will allow portions of the upgraded detector to be installed in any reasonably long shutdown. The VELO is a silicon strip detector with r and Φ geometry. It will be replaced by either a pixel or a strip detector. The pixel version provides a very low occupancy for each channel, reducing the combinatorial for the tracking algorithm. A decrease of the inner radius is also studied in order to improve the impact parameter resolution which is one of the key ingredients for the HLT. The TT stations are a silicon-strip based detector. They will be replaced using the same technology with an enlarged acceptance and an improved granularity. The T stations are composed of an OT with straw tube detectors and an IT with silicon-strip detectors to cover the high occupancy area near the beam pipe. To account for the higher occupancy due to the increase in luminosity, two options are studied, a large area silicon-strip IT completed by OT straw tubes, or a Central Tracker made from scintillating fibers read by silicon photomultipliers. The current RICH photo detectors are HPDs in which the readout chip, operating at 1 MHz, is encapsulated within the vacuum tube. The candidate replacement technology is a multi-anode photomultiplier, read out at 40 MHz. In addition, due to the increased occupancy in the detector, the aerogel part of the RICH-1 will be removed. Studies are on-going to use a novel detector based on time-of-flight to identify low momentum particles below ∼ 10GeV/c [7]. The calorimeter upgrade consists in replacing the current front-end electronics by a new system able to send data to the DAQ at 40 MHz. Moreover, the new electronics must have a gain five times higher than the present system, in order to compensate for a gain reduction that will be imposed on the photomultipliers so that the mean anode current remains at an acceptable level during high-luminosity running. The muon systems will almost remain unchanged, but the M1 station will be removed. 4 Upgrade prospects for some selected physics chan- nel The current precisions for few selected channels are given and further extrapolated to the expected precision in 2018 (before the LHCb Upgrade) and after the LHCb Upgrade phase when an integrated luminosity of 50 fb−1 will have been recorded. Obviously, because of time constraints, many interesting analyses are not discussed. In particular the whole field of CP violation in the B0 has been omitted. The φ s s and ∆Γ observables have been determined simultaneously from B0 → J/ψφ decays s s 6 using using time-dependent flavour tagged angular analyses [8]. It is expected that the determination of φ will remain statistics limited, even with the data samples s available after the upgrade of the LHCb detector [9]. 4.1 The B0 → µ+µ− and B0 → µ+µ− decays s Precise measurements of very rare decays can give some insights of the presence of NP particles present in the loop or box diagrams. A very good example of such type of decays is B0 → µ+µ− which, in the SM, occurs due to very suppressed s loop diagrams. Its branching fraction is expected to be extremely small due to the valuesoftheCKMmatrixelements, theGIMmechanismandthehelicitysuppression. However NP models can enhance it very significantly. Similarly one can also search for B0 → µ+µ− which is even more suppressed due to the relevant CKM matrix elements. ThesummaryofthereachoftheLHCexperiments[10]isshowninFigure2. The obtained 95 % CL limits obtained using the CL method [11] are 4.2 10−9 for s B0 → µ+µ− and 8.1 10−10 for B0 → µ+µ−. This corresponds to about 1.2 times the s SM value for the B0 meson and still 8 times the SM value for the B0 meson. In case s of the B0 meson the number of observed candidates is consistent with the expected s number of events taking into account the background and assuming the SM value for the branching ratio. This already put severe constraints on New Physics models, such as the constrained MSSM for large values of the tanβ parameter as can be seen for example in [12]. Very recently the LHCb experiment has announced the first evidence for the B0 → µ+µ− decay [13] leading to : s BR(B0 → µ+µ−) = (3.2+1.5)×10−9 (1) s −1.2 in agreement with the SM prediction. The invariant mass distribution of the B0 → s µ+µ− candidates in a signal enriched region is shown in Fig. 3. The expected precisions on BR(B0 → µ+µ−) and on BR(B0 → µ+µ−)/BR(B0 → s s µ+µ−), which is an interesting probe of Minimal Flavor Violation models, are sum- marized in Table 1 and compared to the theory uncertainties. Table 1: Comparison of the expected precisions as obtained or foreseen by the LHCb experiment on BR(B0 → µ+µ−) and on BR(B0 → µ+µ−)/BR(B0 → µ+µ−). s s LHCb 2012 LHCb 2018 LHCb-Upgrade Theoretical (50 fb−1) uncertainty BR(B0 → µ+µ−) 1.5 10−9 0.5 10−9 0.15 10−9 0.3 10−9 s BR(B0→µ+µ−) - 100 % ∼ 35% ∼ 5% BR(B0→µ+µ−) s 7 s 1 s 1 L L C ATLAS+CMS+LHCb C CMS+LHCb 0.8 0.8 0.6 0.6 0.4 0.4 0.2 0.2 0 0 0 1 2 3 4 5 6 7 0 01.1 02.2 03.3 04.4 05.5 06.6 07.7 08.8 09.9 B(B0 (cid:65) µ+ µ-) × 10-9 B(B0 (cid:65) µ+ µ-) × 10-10 s Figure 2: CLs as a function of the assumed branching fraction for (left) B0 → µ+µ− s and (right) B0 → µ+µ− decays. The long dashed black curves are the medians of the expected CLs distributions for B0 → µ+µ−, if background and SM signal were s observed, and for B0 → µ+µ−, if background only was observed. The yellow areas cover, for each B, 34% of the expected CLs distribution on each side of its median. The solid blue curves are the observed CLs. The upper limits at 90% (95%) C.L. are indicated by the dotted (solid) horizontal lines in red (dark gray) for the observation and in gray for the expectation. Measuring the branching fractions are not the only interesting physics results one can get from the analysis of the B0 → µ+µ− and B0 → µ+µ− decays. For example s the measurement of the B0 → µ+µ− effective lifetime is envisaged as it is a clean test s of NP. With the upgraded LHCb detector this appears to be feasible [14]. 4.2 The B0 → K∗0µ+µ− decay The decay B0 → K∗0µ+µ− is a flavour changing neutral current process which pro- ceeds via electroweak loop and box diagrams in the SM. Physics beyond the SM, such as SUSY, can contribute with a comparable amplitude via gluino or chargino loop diagrams and modify observables in the three-body final state. A number of angular observables in B0 → K∗0µ+µ− decays can be predicted as functions of the dimuon invariant mass squared, q2. These include the forward-backward asymmetry of the muons, A or the fraction of longitudinal polarization. The CP asymmetry is also FB an interesting probe of NP [15]. The B0 → K∗0µ+µ− decay is normally treated using theframeworkoftheoperatorproductexpansion, wheretheWilsoncoefficientsC 7,9,10 dominate [16]. These have right-handed partners C(cid:48) , that are highly suppressed 7,9,10 in the SM and in minimal flavour-violating models [17]. In the presence of NP, the values of these coefficients may change due to new heavy degrees of freedom in the loops. Measuring the Wilson coefficients then allows for entire classes of NP to be observed or excluded. 8 ) 2c 14 / LHCb V e 12 1.0 fb- 1(7TeV) +1.1 fb- 1(8TeV) M BDT > 0.7 0 10 5 ( / 8 s e t 6 a d i d 4 n a C 2 0 5000 5500 6000 m [MeV/c2] m +m - Figure 3: Invariant mass distribution of the selected B0 → µ+µ− candidates (black s dots) in a signal enriched region. The result of the fit is overlaid (blue solid line) and the different component detailed : B0 → µ+µ− (red long dashed), B0 → µ+µ− (green s medium dashed), B0 → h+h(cid:48)− (pink dotted), B0 → π−µ+ν (black short dashed) (s) µ and B0(+) → π0(+)µ+µ− (light blue dot dashed), and the combinatorial background (blue medium dashed). 9 The measurement of several angular variables as a function of q2 has been per- formed recently by the LHCb experiment using 1.0 fb−1 of data [18]. As an example the q2 value at which A is 0 can be seen on Fig. 4 with the theory prediction FB from [19] underlayed. This is of particular interest since it is both sensitive to NP and is cleanly predicted in the SM. The measured value is q2 = (4.9+1.1) GeV2/c4 (2) 0 −1.3 in good agreement with the SM. In 2018 the expected precision from LHCb is of the orderof6%andshouldreach2%usingtheLHCbUpgradedatasample(50 fb−1). But Theory Counting Experiment Unbinned 1 B F A LHCb Preliminary 0.5 0 -0.5 -1 2 4 6 q2 (GeV2/c4) Figure 4: A as a function of q2, that comes from an unbinned counting experiment FB (blue dashed line) overlaid with the theory prediction. The data-points are the re- sult of counting forward- and backward-going events in 1 GeV2/c4 bins of q2. The uncertainty on the data points is statistical only. The red-hatched region is the 68% confidence interval on the zero-crossing point observed in the data. theforward-backwardasymmetryisnottheonlyinterestingobservableandimportant theoretical work has been going on to design variables which are at the same time free of hadronic form factor uncertainties and experimentally accessible. Some examples 10