DETECTORS, COLLIDER

The experimental study of the basic constituents of matter and their interactions requires detectors that are able to measure the important characteristics of particle interactions, whether they are produced in large accelerators, such as the Fermilab Tevatron, the Stanford Linear Accelerator Center (SLAC) Asymmetric B Factory (PEP-II), the Relativistic Heavy Ion Collider (RHIC), or the future CERN Large Hadron Collider (LHC), or whether they are produced in high-energy cosmic ray interactions or in the rare (and as yet unseen) decays of the proton in a large tank of water.

Detectors installed at large high-energy accelerators can be further divided into two categories: fixed-target and colliding beam. In both types of detectors, the aim is the same: to identify as many of the characteristics of the products of a high energy collision as possible. The two types differ mainly in their geometrical layout. In a fixed-target experiment, an accelerator beam of protons, electrons, neutrinos, or pions, which may have energy as high as several hundred GeV, impinges on a target (typically liquid hydrogen, although solid targets such as carbon are also employed) in the laboratory. The detector is arrayed upstream of the target, where it can intercept the products of the beam-target interaction. Because of the relativistic motion of the center of mass of the collision, the products of the collision are thrown forward and can be intercepted with high efficiency in this arrangement.

In a colliding beam detector, two high-energy beams of various combinations of protons, antiprotons, electrons, or positrons are brought into a direct collision. If the beams have equal energy, as in most colliders, the collision center of mass is stationary in the laboratory. In several installations, such as the HERA electron-proton collider or the PEP-II and KEK-B electron-positron colliders, the two beams have unequal energy, and the collision products are boosted in the direction of the motion of the collision center of mass in the laboratory. In both types of colliders, the experimental challenge is to detect and measure the properties of as many of the particles produced in the collision as possible.

The products of a high-energy collision consist of charged particles (π^± mesons, K^± mesons, protons, electrons, and muons) and neutral particles (primarily photons from π⁰ decay, but also including neutrinos, neutrons, and mesons). The aim of detector design is to produce an instrument capable of measuring, with the highest possible efficiency and precision, the direction and momentum (or energy) of each collision product and identifying the particle species. These functions are performed by a variety of devices. Some typical approaches are discussed below.

In many cases, the particles actually detected were not produced directly in the collision but resulted from the decay of unstable particles produced in the primary interaction. These unstable particles typically fall into two classes: those that decay in a very short time (∼10^-21 to 10^-15 second) via strong or electro-magnetic interactions, and those that decay more slowly (∼10^-13 to 10^-12 second) via weak interactions and can travel a measurable distance (typically 100 μm to 1 cm) within the detector before decay.

Short-lived particles are identified by constructing a quantity called the invariant mass, which combines the measured momenta and directional information of putative decay products in such a way as to isolate individual parent particles, such as π⁰ or ρ mesons. Longer-lived particles are isolated by reconstructing their decay vertex, which is displaced from the primary interaction point. Long-lived charged particles (D^±, B^±, …) produce detached vertices with an odd number of prongs, whereas neutrals (D⁰, B⁰, …) produce an even number of prongs. These prongs may be the tracks of pions, kaons, protons, electrons, or muons. In order to ascertain that this decay vertex is detached from the interaction point, it is necessary to reconstruct the origin and direction of each track of the vertex with sufficient spatial resolution to distinguish its origin from the interaction point. These measurements are often made in a silicon vertex detector, which typically consists of three to five planes of thin (∼300 μm thick) high-resistivity silicon on which a series of fine lines (typically of 25 ∼m pitch and several centimeters long) form a series of diodes. When the charged particles pass through the junction of the diode, energy deposited by ionization is collected and amplified to produce a signal that can be used to locate the position at which the particle passed through the silicon plane to a precision of the order of 5 to 10 μm. Often, the two sides of the silicon wafer have orthogonal diode structures, allowing the simultaneous measurement of two coordinates in the plane for each particle. A computer is used for pattern recognition, that is, to associate the numerous measurements indicating the passage of particles though the series of precisely positioned silicon planes into a series of tracks and then to associate the tracks into a vertex. Pixilated silicon planes are now coming into use. In these devices, a single electronics channel measures both coordinates of the track, eliminating ambiguities that can arise in high multiplicity situations when each coordinate is measured separately.

With knowledge of the initial direction of each of the charged particles produced in the collision, the momentum of each particle can be measured. This is done by surrounding the interaction point with a strong magnetic field of 1 to 3 Tesla, causing each charged particle to bend in the field, with a radius of curvature in the plane perpendicular to the field that is proportional to its momentum. It is then necessary to determine all particle trajectories and to measure their radii of curvature and thus their momenta. This is often done in a drift chamber. A drift chamber can be built in planar (for fixed-target experiments) or cylindrical (for colliding beam experiments) geometry. Each plane or layer of the chamber is comprised of a set of individual cells a few centimeters in diameter. The cell perimeter may be defined by an array of fine wires (composed of, e.g., 80 μm gold-plated aluminum) or by a very thin mylar tube with a layer of aluminum deposited on its inner surface. In the center of each cell is a very fine (25 μm) gold-plated tungsten wire kept at a positive potential with respect to the perimeter. This array is placed in a volume of gas (80% helium + 20% isobutane in a typical gas mixture). The charged particles produced in the collision pass through the silicon vertex detector and then into the drift chamber, ionizing the gas along their trajectory. Electrons thus liberated in the gas then drift onto the fine central wire of each cell under the influence of the carefully calibrated electric field in each cell; the drift time is proportional to the drift distance within the cell. When the drifting ionization reaches the region immediately surrounding the high electric field near the central wire, an avalanche is created, producing a sufficient number of secondary electrons on the central wire to allow the recording of an electrical signal with well-defined amplitude and time characteristics. This allows the reconstruction of the particles' position with respect to the central wire to a precision of ∼150 μm in each cell. A typical drift chamber has fifty layers, allowing the reconstruction of the individual particle trajectories in the magnetic field to high precision and the measurement of particle momenta to within a few percent accuracy.

With particle momenta measured, identifying the particle species remains. Since a particle's momentum is the product of its mass and velocity, the particle mass and thus its species may be identified if the particle's velocity can be independently determined. Several approaches are in common use. The first uses the details of the ionization left by the particle trajectory in a drift chamber. The ionization energy loss per unit length dE/dx in a gas is an essentially universal function of the particle velocity, independent of mass. Thus, the sum of all the ionization left by a track in the drift chamber, normalized to the length of the trajectory, provides a measure of a particle's velocity. If plotted against the momentum of a given particle, each species produces a characteristic energy loss in the gas at a given momentum. As a result of the shape of the universal dE/dx curve, however, there are regions of ambiguity that effectively limit this method of particle identification to momenta below 0.5 GeV/c .

Another method of particle identification is to measure the time of flight of the particle over a known distance with high resolution. A typical time of flight from the point of creation to a point at which a time can be recorded is 5 nsec, which can be measured to a precision of ∼100 psec. This precision suffices to distinguish pions from kaons up to momenta of 0.6 GeV/c .

A third class of techniques makes use of the phenomenon of Cherenkov radiation, a shock wave emitted when a particle's velocity exceeds the speed of light in a medium. Cherenkov radiation devices come in several forms. There are threshold counters, in which the index of refraction of a high-pressure gas is chosen so that, for example, electrons emit Cherenkov radiation while pions do not, and there are devices that measure the angle of the Cherenkov shock cone with respect to the particle direction, since the cosine of this angle is proportional to the reciprocal of the particle's velocity. The latter device, known as a DIRC, can, using quartz as a Cherenkov medium, distinguish pions from kaons up to 4 GeV/c .

The identification of muons makes use of the fact that these particles do not have strong interactions and are thus able to more readily penetrate material, whereas pions, although close in mass, interact strongly and are absorbed. Detectors with solenoidal magnetic fields require a steel flux return of the order of a meter thick to control the field distribution. Muon detectors are thus typically made by distributing the flux return into several layers and inserting devices that can record the passage of a particle between the layers. These devices may be large arrays resembling crude drift chambers, large planar chambers with a well-defined gap maintained at high voltage and filled with gas, called resistive plate chambers (RPCs), or they may be made of plastic scintillators. In such arrays, a muon is identified as a particle penetrating many layers, while a pion typically does not survive all the way through the flux return.

The detection of high-energy photons presents unique challenges. Photons do not ionize the media through which they pass. Photons in the energy range of interest in elementary particle physics, from tens of MeV to tens of GeV, lose energy in matter through pair production, the photoelectric effect, and Compton scattering. They are detected in devices called calorimeters, which come in many varieties, but which have certain common characteristics. When a high-energy photon enters matter it creates an electro-magnetic shower, a cascade of electrons, positrons, and photons. The electrons and positrons are ionizing particles. If the material in which the shower is initiated is large enough to contain the shower products, then the charged particle component of the shower deposits an amount of energy in ionization that is closely proportional to the energy of the initiating high-energy photon. This ionization energy can be detected in several ways. The highest-quality devices are arrays of large high-atomic-number crystals, such as CsI, PbWO₄, or LSO that emit scintillation light, which is detected by photomultiplier tubes or silicon photodiodes. Calorimeters that collect the ionization energy deposited by electromagnetic showers in noble liquids such as krypton or xenon have also been employed. Other devices consist of alternating layers of high-atomic-number material, such as lead, with a layer of material, such as plastic scintillator or liquid argon, that is sensitive to the ionization produced in the showers initiated in the high Z material.