OPAL

The Omni Purpose Apparatus for LEP Detector is designed to "provide precise measurements of charged particles and of electromagnetic energy over nearly the full solid angle" [1]. During LEP1 (1989-1995), OPAL collected millions of events to make precision measurements of the Z[2]. During LEP2 (1996-2000), the physics goal was to search for new physics through W⁺W^- pair production.

Detector Design

The detector provides acceptance for Z⁰ decays over 4π in solid angle. The main detector components are a system of central tracking chambers inside a solenoidal field of 0.435 T, a time-of-flight counter, a lead glass electromagnetic calorimeter with a presampler, a hadron calorimeter in the form of an instrumented magnet return yoke, and an outer muon chamber. A forward calorimeter serves as a luminosity moniter.

Central Tracking Detector

The Central Tracking Detector is divided into a precision vertex chamber, a large jet chamber, and a z chamber.

The vertex detector is a cylindrical drift chamber that surrounds has a length of 1 m, an inner radius of 88 mm, and an outer radius of 235 mm. There are two layers of 36 cells each: an inner layer of axial wires and an outer layer of stereo cells. Both types of cells have radial wire planes consisting of 200 μm gold plated Cu-Be potential wires and 20 μm gold plated W-Rh andoe wires. In the axial cells, there are 12 anode wires with radial spacing of 5.3 mm. The stereo cells have 6 anode wires with a spacing of 5 mm with a stereo angle of 4 °. The cathode planes use 125 μm Cu-Be wires with a 1 mm spacing in both cells. In the r-φ plane, the vertex chamber provides a resolution of 55 μm.

The jet chamber is absed on the design of the jet chamber in the JADE experiment at PETRA. The inner diameter is 0.5 m and the outer diameter is 3.7 m. The chamber is divided into 24 sectors divided by cathode wire planes. Each sector has a radial wire plane with 159 sensing wires parallel to the beam. For 43° < θ < 137°, 159 points will be measured along each track, and at least 8 points are achieved over 98% of the solid angle. The anode wires are spaced 10 mm apart and alternated with potential wires which are set at -2.38 kV. The jet chamber provides a resolution of 135 μm in the r-φ plane and 6 cm in the z direction. The dE/dx resoultion is 3.8% in a diumuon sample with at least 130 points per track.

The z chamber covers 44° < θ < 136° and 94% of the azimuthal angle. The purpose is to make a precision measurement of the z coordinate of charged particles that leave the jet chamber. There are 24 drift chambers, each 4 m long, 50 cm wide, and 59 mm thick. Each chamber is divided into 8 cells in z with six anode wires separated by 4 mm in the radial plane but perpendicular to the beam direction. The chamber provides a resolution of 300 μm in z and 1.5 cm in r-φ.

All three chambers use the same gas mixture of argon (88.2%), methane (9.8%) and isobutane (2.0%) at 4 bar. The gas system has a recirculation and purifation system to remove oxygen to a level of a few ppm. A laser system monitors drift velocity at an accuracy of &sigma/v < 0.1%.

Combined, three chambers provide a resolution of 75 μm in the r-φ plane and 2mm in the r-z plane. The invariant mass resolution for K⁰ goes to π⁺π^- is 8 MeV.

Time-of-Flight System

The time-of-flight (TOF) system covers the region |cosθ| < 0.82 with the purpose of providing triggers, aiding in particle identification for charged particles with momentum in the range 0.6-2.5 GeV, and rejecting cosmic rays.

The TOF is a barrel of radius 2.360 m made of 160 scintillation counters. Each counter is 6.840 m long with a trapezoidal cross section. The maximum gap between counters is 2.6 mm. Light from each counter is collected at both ends by 300 mm light guides that go directly to PMTs, which have a gain of 3 x 10⁷ at 1850 V.

Trigger signals are generated within 50 ns from the arrival time. The mean speed of scintillation was measured to be 0.582 c from test beam data at CERN-PS. The mean attenuation length was 2.33 m. The timing resolution from the center was 280 ps and 350 ps from the ends, with a z resolution of 5.5 cm. If additional z information is provided, the time resolution at the ends improves to 220 ps.

Electromagnetic Calorimeter

The electromagnetic calorimeter was designed to measure the energy of positrons, electrons and photons from a few MeV to 100 GeV. The barrel and the endcap together cover 98% of the solid angle, and the forward calorimeter extends the hermeticity even farther. The barrel and endcap are made of lead glass because of its resolution of 5%/sqrt(E).

Because there are approximately 2 X₀ in front of the calorimeters, both the barrel and endcap have presamplers to measure the position and energy of electromagnetic showers. The barrel presampler consists of 16 chambers covering the cylinder of radius 2388 mm and 6623 mm long. Each chamber has two layers with four sections of 24 cells each. Anode wires of stainless steel run down the center of each cell and are 75 μm in diameter. The gas in each cell is a mixture of n-pentane (32%) and CO₂ (68%). Charge is collected from cathode strips that cross both top and bottom of each layer offset from the wire direction by 45°, with top and bottom strips orthogonal. The resolution for a single particle is 1-2 mm depending on angle of incidence and 4-6 mm for showers depending on energy. Resolution in z is 10 cm for a single particle.

The barrel calorimeter is made of 9,440 lead glass blocks of 24.6 X₀ at a radius of 2455 mm. It covers |cosθ| < 0.82 and 2π in azimuthal angle. The blocks are pointed at the interaction point with a tilt slightly away from perfect to prevent neutral particles from escaping along gaps. There are 59 segments in the z direction and 160 in the azimuthal direction. There are 16 different shapes. PMTs are attached to the back of each block and can be operated in an external field of up to 100 G with <1% effect on the gain. For calibration, every counter was exposed to a 50 GeV electron beam twice at CERN-SPS. The gains were measured to 0.1% and nonlinearity was measured to be less than 1%. A Xenon lamp is used to provide ongoing calibration to every counter.

The endcap presampler is 16 wedge shaped sectors with a total of 32 chambers covering 0.83 < |cosθ| < 0.95. Thin multiwire chambers are used in the design. Each sector has a large and a small trapezoid shaped chamber and neighboring sectors overlap. Readout is done simultaneously for groups of 4 wires and strips. The resolution is equal to the intrinsic chamber resolution.

The endcap calorimeter consists of two dome shpaed arrays of 1,132 lead glass blocks each, covering 0.83 < |cosθ| < 0.98. The blocks are mounted parallel to the beam axis and instrumented with special vacuum phototriodes (VPTs) which work in the full field of the magnet. The blocks have a depth of at least 20.5 X₀. The VPTs have a gain of 12.3 and quantum efficiency of 26%. Calibration is done with a laser system to provide equivalent light for 10 and 20 GeV electrons.

Overall, the calorimeter provides a resolution of 0.2% + 6.3%/sqrt(E) before material is added in front. When 2.08 X₀ of aluminum is added in front, the resolution degrades by about 50%. The presampler and lead glass can achieve electron identification efficiency of 80-90% with a pion rejection on the order of 10^-3.

Hadron Calorimeter

The hadron calorimeter covers 97% of the solid angle to measure the energy of hadrons that pass through the electromagnetic calorimeter using the iron of the magnetic return yoke as an absorber. This material provides 4 interaction lengths or more and is segmented into 8 layers of 100 mm thickness, which are alternated with 9 layers of chambers in the barrel. In the endcap, there are 8 layers of chambers and 7 layers of iron.

The active layers are similar to those used in the presampler. Signals are read through large pads on one side and through aluminum strips that run parallel to the anode wires on the other side. The 57,000 strips help provide single particle tracking for muon identification and shower profiles. The pads are grouped into towers for energy measurements. The chambers use a mixture of isobutane (75%) and argon (25%), which is changed daily. The anode voltages are kep between 4.65 kV and 4.85 kV.

Calibration was done on prototype modules in a pion test beam. The energy resolution measured was 120%/sqrt(E) for incident energy of 10 GeV.

The pole tip hadron calorimeter covers 0.91 < |cosθ| < 0.99 with 9 iron layers of 80 mm thickness. The gaps between the samples had to be made smaller to prevent perturbations to the magnetic field, so the sampling rate was increased to compensate. The active chambers are similar to the endcap presampler with anode wires at 3.5 kV and a gas mixture of CO₂ (55%) and n-pentane (45%). The chambers are also read out with a combination of pads and strips. One module was placed in a hadron test beam at CERN with 6-50 GeV incident energy. Below 15 GeV resolution was 100%/sqrt(E) and higher for higher energyies because of leakage.

Muon Detector

The muon detector covers 93% of solid angle with at least one layer. Over this area, there is a 0.001 probability for a pion to not interact before reaching the muon detector. Muon identification is done by matching a track from the central tracker to a track in the muon detector. The single muon efficiency over the detector is nearly 100% for muons above 3 GeV, while 5 GeV pions will be misidentified 1% of the time.

The barrel muon detector is made of 110 large drift chambers, which are 1.2 m wide, 90 mm deep, and come in lengths of 10.4, 8,4 and 6.0 m. The chambers each have 2 cells with a 50 μm anode wire running the entire length. The gas is ethane (10%) and argon (90%) with a typical flow rate of 200 ml/min and a drift velocity of 38 mm/μs. The maximum drift time is 8 μs. The φ coordinate is determined to an accuracy of 1.5 mm and the z coordinate to a resolution of 2 mm.

In the endcap, which covers 0.67 < |cosθ| < 0.985, there are four layers of tubes perpendicular to the beam in 8 quadrants and 4 ptach chambers. The 100 μm anode wires run along the chambers in a gas of argon (25%) and isobutane (75%) at a voltage of 4.3 kV.

Forward Detectors

The forward detectors serve as luminosity monitors by measuring small-angle Bhabha scattering. The acceptance is 39 mrad < θ < 120 mrad on either side and full azimuthal acceptance. The calorimeter has 35 layers of lead-scintillator (24 X₀) with readouts with wavelength shifter to VPTs. The front 4 X₀ serves as a presampler with readouts on only the outer edge, but the rest of the calorimeter has readouts on both the inner and outer edges. The detector provides an energy resolution of 17%/sqrt(E) and a radius resolution of 2 mm. Three planes of proportional tube chambers in front of the calorimeter provide radial resolution of 0.5 mm.

There are drift chambers in front of the presampler with two layers each and two sense wires per layer. There is also a gamma catcher, which is a lead scintillator sampling calorimeter, to provide a veto on on the neutrino counting channel. The far forward luminosity monitor provides coverage for electrons scattered between 5 and 10 mrad.

The luminosity was measured in the forward detectors corresponding to a 24.23 nb cross section for Bhabha scattering.

Trigger

LEP bunch crossings occure every 22.2 μs or at 45 kHz. The trigger system reduces the rate to 5 Hz for the data acquisition system.

The detector is divided into 144 overlapping bins, with 6 in θ and 24 in φ. Subdetectors produce trigger signals for each of these bins and also independent signals for each detector. Trigger signals are correlated in the trigger logic to create a θ-φ matrix which is compared to conditions for acceptance in each event.

The vertex chamber and jet chamber provide the track trigger, which consists of signals for the 144 bins and 6 signals for ≥ 1, 2, 3 barrel tracks and ≥ 1, 2, 3 total tracks. The TOF provides a single signal telling if ≥ 6 sectors (out of 24 in φ with no θ divisions) have fired. There is a threshold for the number of counters in each sector to fire which is between 2 and 5.

The electromagnetic calorimeter provides 200 signals from the barrel and 24 from each endcap to make the 144 signals and total sums for all three detectors. There are two thresholds for the totals at 6 and 4 GeV and two thresholds for the θ-φ bins at 2.5 and 1 GeV. The hadronic calorimeter provides 92 signals which are discriminated at three levels and then mapped to the 144 matrix signals.

The muon detector barrel provides a single signal if three out of four layers have fired in at least one φ bin (with no θ divisions). The endcap muon detector produces 4 x 24 θ-φ signals by summing over adjacent strips. There are also signals for total hits in each endcap and for coincidence hits.

OPAL achieved a typical trigger rate of roughly 1 Hz at the beginning of LEP1 running in 1989 with a readout deadtime of 50 ms per event. About 1.5 million events were recorded with a multi-hadronic Z⁰ sample of about 30,000. Event sizes were on average 100 kbyte. At design luminosity, a trigger rate of 4 Hz and data rate of 400 kbyte/s are expected.

Experimental Results

OPAL produced its first result in 1989 using 190 nb^-1 collected during the first 15 days of running at LEP [3]. Using 4350 multi-hadronic events from seven energy points between 89.26 GeV and 93.26 GeV, the collaboration reported a mass m_Z = 91.01 ± 0.05 ± 0.05 GeV with a width Γ_Z = 2.60 ± 0.13 GeV and found that there should be 3.1 ± 0.4 neutrino generations.

Hadronic events for this sample were triggered by using the electromagnetic calorimeter, the TOF, and the jet chamber. An energy sum in the calorimeter of at least 6 GeV in either the barrel or one of the endcaps is required with TOF hits in at least three nonadjacent countes. There must be at least two tracks from the interaction vertex with at least 450 MeV/c transverse momentum. The forward calorimeter could trigger if there was at least 12 GeV in both sides back-to-back or at least 15 GeV in both sides without being back-to-back. The trigger efficiency was found to be > 99.5% using a Monte Carlo simulation.

Offline cuts were made using the calorimeter and the TOF. A multihadron candidate requires at least 10 different calorimeter clusters with more than 100 MeV in each cluster, the total energy in the calorimeter is greater than 10% of the collision energy, and the ratio | R_bal | = Σ(E_clus•cosθ) / ΣE_clus ≤ 0.65. Monte Carlo studies found that the acceptance for these cuts was 98.2%.

Monte Carlo calculations were made to estimate the background contamination of the sample for two photon processes, electron-pair and τ-pair decays of the Z⁰. The calculation found these backgrounds to be < 0.1%, <0.1%, and 0.3%, respectively. Using a sample of 819 events from the mass peak, the τ decay, beam-gas, beam-wall, and cosmic ray backgrounds were estimated. Another sample with reconstructed vertex >50 from the interaction vertex was used for just beam-wall and beam-gas backgrounds. The overall background contamination of the sample was 0.3% from the τ-pair decays.

The dominating systematic error of the measurement is the luminosity determination. Position calibration uncertainties and resolution effects are the main sources of this error, which results in a systematic error of 5% in the cross section measurement.

The mass of the Z⁰ is determined by fitting a Breit-Wigner lineshape to the cross section vs. collision energy spectrum with 3 free parameters: m_Z, Γ_Z, and a normalization factor. The three parameters were varied independently.

To constrain the number of light neutrino flavors, the fit is recalculated by replacing Γ_Z with Γ_Z^SM + (N_ν - 3)•Γ_ν^SM and varying m_Z and N_ν, but constraining the normalization constant. From the fit, the probability of N_ν = 3 is 47% and the probability of N_ν = 4 is 8% or that N_ν < 3.88 at a confidence level of 90%.

References

1. The OPAL Collaboration. "The OPAL Detector at LEP." CERN-PPE/90-114. August 14, 1990.

2. http://opal.web.cern.ch/Opal/

3. The OPAL Collaboration. "Measurement of the Z⁰ Mass and Width with the OPAL Detector at LEP." CERN-EP/89-133. October 13, 1989.