CMS Conference Report CMS - Concept and Physics Potential

CMS (Compact Muon Solenoid) will be one of two general purpose detectors at the CERN Large Hadron Collider. Its main feature is a strong solenoidal magnetic field ensuring high momentum resolution for charged particles. The detector consists of an inner tracker with an embedded pixel detector, a crystal electromagnetic calorimeter, a copper-scintillator hadron calorimeter and a dual muon system made up of tracking chambers and special trigger chambers. Forward calorimetry is also foreseen. The discovery potential of CMS for the Standard Model Higgs, the SUSY Higgses and other supersymmetric particles is presented. Abstract. CMS (Compact Muon Solenoid) will be one of two general purpose detectors at the CERN Large Hadron Collider. Its main feature is a strong solenoidal magnetic (cid:12)eld ensuring high momentum resolution for charged particles. The detector consists of an inner tracker with an embedded pixel detector, a crystal electromagnetic calorimeter, a copper-scintillator hadron calorimeter and a dual muon system made up of tracking chambers and special trigger chambers. Forward calorimetry is also foreseen. The discovery potential of CMS for the Standard Model Higgs, the SUSY Higgses and other supersymmetric particles is presented.

INTRODUCTION CMS (Compact Muon Solenoid) is a general purpose experiment designed to explore physics at the planned Large Hadron Collider (LHC) at CERN. It is expected to go into operation in 2005. Proton-proton collisions as well as heavy ion collisions will be available. More than 150 institutions with 1700 physicists and engineers are presently taking part in the collaboration.
The design concept of CMS was rst presented at the LHC Workshop at Aachen in 1990 1]. It is based on a strong solenoidal magnetic eld of 4 Tesla generated by a superconducting coil. The inner tracking system, the electromagnetic calorimeter, and the hadron calorimeter with the exception of a tail catcher in the central region, are inside the magnetic eld volume. The muon chambers are embedded in the return iron yoke. Forward and Very Forward Calorimetry complete the apparatus in order to detect non-interacting particles.
A perspective view of CMS is shown in Fig. 1.

DETECTOR SETUP
Following the completion of the Technical Design Reports 2{6] of all subdetectors the CMS layout has been essentially nalized in April 1998.
The long superconducting coil is the heart of CMS. It provides a solenoidal eld o f 4 T esla parallel to the beam direction. It is essential that the coil be completed before most other detector parts. The design is well advanced and real construction has started. The nished magnet is expected to be tested in 2003.

Inner Tracking
The active v olume of the CMS inner tracker is a cylinder with a radius of 115 cm and a length of 270 cm on each side of the interaction point. Three di erent detectors well suited to the high, medium and low occupancy regions have been chosen to satisfy the stringent resolution and granularity requirements: a silicon pixel detector up to a radius of approximately 20 cm, a silicon strip detector in the region between 20 and 60 cm, and Micro Strip Gas Chambers (MSGC's) from 70 to 120 cm. The setup is shown in Fig. 2. The tracker geometry has been chosen such that typically 13 high resolution measurement planes for high-p T tracks are available up to j j 2, gradually falling o to a minimum of 8 planes at j j 2 : 5. Overall, the silicon and MSGC trackers consist of more than ten thousand independent detector modules instrumented with 12 10 6 channels. The occupancy of each channel will be about 1 to 2 percent at high luminosity.
High-p T isolated tracks are reconstructed with a transverse momentum resolution of better than p T =p T (15p T 0:5)%, with p T in TeV, in the central region of j j 1 : 6, degrading to p T =p T (60p T 0:5)% as j j approaches 2.5. Electromagnetic Calorimeter CMS has chosen an electromagnetic calorimeter made out of scintillating lead tungstate crystals (PbWO 4 ) because it o ers the best prospects of identifying and measuring precisely the energies of photons and electrons in a hostile environment with a magnetic eld of 4 Tesla, a time of 25 ns between bunch crossings and radiation doses of 1 to 2 kGy peryear at maximum LHC luminosity. The choice was based on the considerations that PbWO 4 has a short radiation length of 0.89 cm and Moli ere radius of 2.19 cm and a short light decay time. The initial drawback of low light yield has been overcome by progress in crystal growth and through the development of large-area silicon avalanche photodiodes. The geometrical coverage extends to j j = 3 . Precision energy measurement of photons and electrons will be carried out to j j = 2 : 63. A total thickness of about 26 radiation lengths at j j = 0 is required to limit the longitudinal shower leakage of high-energy electromagnetic showers to a reasonable level. This corresponds to a crystal length of 23 cm in the barrel region. In the endcap region a , 0 separating preshower detector corresponding to 3 X 0 of lead allows the use of slightly shorter crystals.
For the energy range of about 25 to 500 GeV, typical for photons from the H ! decay, the energy resolution can beparametrized as: ( = E) 2 = ( a= p E) 2 + ( n =E) 2 + c 2 where a is the stochastic term, n the noise, and c the constant term. Fig. 3 shows the di erent contributions to the energy resolution. Depending on luminosity a

Hadron Calorimeter
Together with the electromagnetic calorimeter the hadron calorimeter will be essential to measure jets and missing energy, crucial for the discovery of many new particles or phenomena. The targeted energy resolution is: In addition to the barrel and endcap hadron calorimeters (HB, HE) extending to j j 3 a separate forward calorimeter (HF) covering the region 3 < j j < 5 is foreseen to maximize hermeticity (Fig. 4). HB and HE are sampling calorimeters consisting of 4 mm thick plastic scintillator tiles read out with wavelength-shifting plastic bres inserted between copper plates. The barrel hadron calorimeter has only 5.15 nuclear interaction lengths at = 0 . T o ensure adequate sampling depth the rst muon absorber layer is instrumented with scintillator tiles to form a tail catcher.
The HF calorimeter which has to withstand high radiation doses uses quartz bres as the active medium, embedded in a copper absorber matrix. It is not only important for the measurement of missing energy as required for example in Standard Model and SUSY Higgs searches or top quark physics, but also for forward jet detection needed in the search for the heavy Higgs boson in the TeV mass region.
In order to get an idea of the physics performance the process t ! W b with W decaying into jets was simulated. The obtained dijet mass resolution was 12 GeV with pileup and 8 GeV without. , y , y , y , y

Muon System
In the barrel region the muon system consists of drift chambers with bunch crossing identi cation capability (DTBX) to reconstruct muon tracks and of resistive plate chambers (RPC) to detect muon hits for trigger purposes. In the forward region the cathode strip chambers (CSC) perform the tasks of reconstructing the tracks. RPC's are also available.
The chambers are arranged in four stations interleaved with the iron return yoke plates as shown in Fig. 4. They are arranged in concentric cylinders around the beam line in the barrel region, and in disks perpendicular to the beam in the endcaps. The momentum resolution of charged tracks for the muon system alone and combined with the inner tracker at = 0 : 1 is shown in Fig. 5.

Trigger and Data Acquisition
At LHC the collision rate will be40 MHz. 16 million channels will have to be processed in total. One event is expected to contain 1 Mbyte of data on average. Filtered events will be written to a storage medium with a frequency of 100 Hz. The trigger system has therefore to perform a sizable reduction of data. The rstlevel trigger, a partly programmable hard-wired system, will run at a frequency of  We will concentrate here only on Standard Model Higgs 7] and supersymmetry searches 8]. CMS's B-physics capabilities are described in another contribution to these proceedings 9]. Heavy ion physics will not be dealt with either.

Standard Model Higgs searches
In the framework of the Standard Model particles acquire mass through the interaction with the Higgs eld. This implies the existence of the Higgs boson. Theory does not predict its mass, but it does predict production rates and decay modes as a function of the Higgs mass. CMS has been optimized to detect the Higgs over the It should be noted that in the lower mass region (m H < 130 GeV) the branching ratio for H ! b b is close to one, but due to the large dijet background this channel seems only usable together with an additional lepton signature (e.g. pp ! W H ! l b b ). Fig. 6 depicts Higgs signals for the di erent mass ranges.
For the H ! c hannel the diphoton mass resolution is essential. Calorimeter granularity is crucial for photon isolation measurements to suppress the 0 ! background. The mass resolution at m 100 GeV is better than 1%, resulting in a signal to background ratio of approximately 1/20.
In the mass range 130 MeV < m H < 700 GeV the most promising channel is the Higgs decay t o t w o Z's, one of them being o -shell for masses smaller than 200 GeV. The detection relies on the excellent performance of the muon chambers, the tracker and the electromagnetic calorimeter. For m H < 170 GeV a mass resolution of about 1 GeV should be achieved.
For the highest Higgs masses, in the range 0.5 to 1 TeV, high luminosity is needed. One also has to exploit decays of Z's and W's into jets and neutrinos. Hadron calorimeter performance is very important. At the very highest masses, above 800 GeV, the signal to background ratio has to beimproved by requiring a central jet veto. Thus the t t and Z,W to jets backgrounds are reduced considerably. At the highest luminosities one must also take i n to account pile-up from minimum bias events. Double forward jet tagging will be necessary.
To summarize, CMS can detect a Standard Model Higgs boson in the entire mass range, from the LEP2 limit up to approximately 1 TeV with a signi cance of at least 5 .

Supersymmetry searches
Supersymmetry predicts a number of particles in addition to the Standard Model ones. Fermions have boson super-partners and bosons have fermion super-partners. We use the minimal supergravity-inspired standard model (mSUGRA) with a stable lightest supersymmetric particle (LSP) as a benchmark model 10]. The particle spectrum one expects consists of squarks ( q ), gluinos (g), sleptons (l), neutralinos ( 0 i i=1,4]) and charginos (~ j j=1,2]). There is also a Higgs sector with ve SUSY Higgses, three neutral (h 0 ; H 0 ; A 0 ) and two charged ones (H ). mSUGRA is determined by only ve parameters, the universal scalar (m 0 ) and gaugino masses (m 1=2 ), the SUSY breaking universal trilinear coupling A 0 , the ratio of the vacuum expectation values of the Higgs elds tan and the sign of the Higgsino mixing parameter sign( ).

Squarks and Gluinos
The total SUSY particle production cross-section is dominated by strongly interacting gluinos and squarks, which through their cascade decays can produce many jets and leptons with missing energy due to escaping LSP's and possibly neutrinos. Due to these escaping particles a complete mass reconstruction of squarks and gluinos is impossible. However, the presence of SUSY can be established by an excess of events of a given topology over known Standard Model backgrounds such a s t t , W + jets, Z + jets, WW, ZZ, ZW, Zb b and QCD. In order to establish the limits of the SUSY reach in the (m 0 ; m 1 = 2 ) parameter space the signal was generated at more than 100 points. tan = 2, A 0 = 0 and < 0 have been assumed. Fig. 7 shows the expected sparticle reach in various channels, for signatures containing leptons in di erent charge combinations. For 10 5 pb ,1 integrated luminosity the ultimate mass reach for gluinos would bemq 2.5 TeV for small m 0 (below 400 GeV) and up to 2 TeV for any reasonable value of m 0 (below 2000 GeV). Squark masses can be probed for values in excess of 2 TeV. The cosmologically interesting region within the relic neutralino dark matter density contour of h 2 1 can be probed entirely already with an integrated luminosity around 10 3 pb ,1 . Chargino/neutralino pair production Direct production of~ 1~ 0 2 with leptonic decays of both sparticles gives three high-p T isolated leptons accompanied by missing energy. These events have no jet activity except from initial-state QCD radiation. A central jet veto is therefore appropriate. WZ, ZZ and Zb b backgrounds can be removed by a Z mass cut. Other backgrounds are t t, b b and SUSY channels (g,q,l,~ 0 ,~ ). From Fig. 8  concluded that~ 1~ 0 2 direct production can be investigated up to m 1=2 170 GeV for all m 0 with 10 5 pb ,1 and m 1=2 150 GeV with 10 4 pb ,1 . With 10 5 pb ,1 the discovery region extends up to m 1=2 420 GeV for m 0 < 120 GeV. It is possible to measure the mass of the lightest neutralino for m 0 > 160 GeV by using the fact that the dilepton mass distribution has a sharp cuto which is approximately equal to the mass of~ 0 1 for the three-body decay process~ 0 2 ! ll~ 0 1 11].

Sleptons
To search for direct slepton production the most appropriate signature is 2 leptons + missing energy + no jets. Backgrounds are expected to come from , WW, t t, b b and other SUSY channels. Fig. 9 shows the slepton mapping of the mSUGRA parameter space. With 10 4 pb ,1 luminosity, CMS is sensitive u p t o m l L 160 GeV. With 10 5 pb ,1 the reach extends up to ml L 340 GeV for all allowed LSP masses (< 200 GeV), and up to ml L (340...440) GeV if m LSP (0.45...0.6) ml L for a given ml L . ; tan β= 2, A 0 = 0, µ < 0 5 σ contour, reach at 10 4 pb -1