The injector system for typical particle accelerators consists of a source of charged particles, a DC electric field to give the charged particles an initial kinetic energy, followed by a radio frequency (rf) acceleration stage that prepares the charged particles for injection into the main accelerator system. The main accelerator system accelerates the charged particles to their final energy. These particles are then used either for injection into a storage-ring collider or to provide a beam to produce secondary particles for a variety of scattering or particle production experiments.
Charged particle sources vary from simple thermionic emission from hot tungsten filaments for low-current electron sources to carefully engineered solid-state photo cathodes for high-pulsed-current polarized electron beams. Low-energy beams of protons are normally produced using an rf plasma discharge with either magnetic confinement of the plasma (magnetron) or electric field confinement (penning trap) of the discharge. Several methods are used to convert these beams to H- beams via a charge exchange reaction in a low-pressure gas with suitable characteristics.
In the case of electrons, the initial acceleration mechanism can either be a DC electrostatic field or a very-high-gradient rf field, used in conjunction with a pulsed laser illuminating a photocathode timed so that the accelerating rf electric field is at a maximum. With a DC electrostatic field as the initial accelerator, a combination of rf fields is then used to bunch the beam so that it can be captured and accelerated by the rf system. In modern electron accelerators, the rf accelerator usually consists of a disk-loaded circular waveguide, with the disks providing a propagating rf wave with a phase velocity matched to the velocity of the electrons. These accelerators typically operate with an rf frequency of 2,856 MHz, and the rf power is provided by pulsed klystron amplifiers capable of an output of up to 100 MW peak with pulse lengths of a few microseconds. Typical electron energies at the end of this initial rf acceleration section are a few hundred MeV.
Because protons or H- ions are much heavier, more elaborate initial acceleration systems are required to increase their velocity to the point where they can be accelerated by an rf disk-loaded waveguide structure. In the first method, they are accelerated by a DC electric field of several hundred kilo-volts. This is then followed by a linear accelerator consisting of an rf tank, in which the beam passes through a series of drift tubes of increasing length to shield the particles from the rf field when it is of the wrong phase to accelerate the particles. Usually, these drift tubes also contain DC magnetic fields to provide focusing for the beam. At the end of this structure, the particles have an energy of approximately 200 MeV and a velocity of 0.2 c (c = speed of light). This velocity is sufficient so that the beam can be captured in a disk-loaded waveguide structure. This system typically provides an additional 200 to 400 MeV of energy to the particles. With a final energy ranging from 400 to 600 MeV, the ions are sufficiently relativistic so that they can be injected into a rapid cycling synchrotron and accelerated to their final energy for injection into the main accelerator. With the invention of the rf quadrupole accelerator in the early 1980s, a lower DC accelerating field can be used, and the initial part of the drift-tube linac can be replaced by a compact and efficient acceleration system. The rf quadrupole accelerator consists of a tank with four precisely machined vanes orientated at 90° to one another, extending toward the center of the tank. The inner edges of the vanes are machined with a wave shape that increases in wavelength as the particles gain energy. The rf field inside the tank induces an accelerating gradient via the vanes along the axis of the cylinder that then accelerates the particles. An rf quadrupole can provide energies of up to 2 MeV. Most modern proton accelerators incorporate one of these devices as part of the initial acceleration chain.
For electrons, the main acceleration stage is either a synchrotron (described below) or more of the same disk-loaded waveguide to accelerate the electrons to their final energy. The nominal accelerating gradient in these disk-loaded structures is 15 MV/m. The Stanford Linear Accelerator is the highest-energy accelerator of this type, achieving 50 GeV with a length of 3,200 meters.
For protons or H- ions, the next element following the linear accelerator in the injection system is a circular accelerator. It accelerates the particles by confining them to a circular orbit using electromagnets that provide both bending and focusing and then passing the particles through an rf-accelerating cavity system many times. These machines are called synchrotrons since the magnetic fields have to increase in synchronism with the energy increase of the particles. A system of pulsed electric and magnetic fields deflects the incoming protons onto the stable orbit. In the case of H- ions, injection into the circular accelerator is accomplished by passing the ions through a thin foil that strips the two electrons from each of the ions, leaving protons that are then guided by the magnetic field of the synchrotron. Modern proton accelerators use H- injection since much higher beam intensities can be achieved using this technique. Electron synchrotrons use a system of pulsed elements to inject onto the stable orbit.
Synchrotrons are necessarily cyclic machines. The ratio of peak energy of the particles to the injection energy of the particles is typically between 10 and 20. Remnant field effects in the iron-based electromagnets limit the dynamic range of synchrotrons to this range. The power supplies that provide the current for the guide field magnets can be either programmable supplies or be configured as a resonant circuit with a DC bias. Usually, lower-energy synchrotron magnet systems are configured as resonant circuits and high-energy synchrotrons use programmable power supply systems. An example of the former is the 8-GeV booster at Fermilab, whereas the new main injector at Fermilab uses a programmable power supply system for its magnets. The 8-GeV booster beam is injected into the main injector, which then accelerates the protons to 120 GeV. These 120-GeV protons are subsequently injected into the Tevatron that then accelerates them to 980 GeV. The cycle time for synchrotrons varies from 1/60th of a second to minutes, depending on available power for the magnets and the rf acceleration system. When the particles reach their full energy at the end of the chain of accelerators, they are either stored for use in colliding-beam experiments, or they are extracted from the accelerator by the extraction system. The extracted beam is then used to produce secondary particle beams to carry out scattering and particle production experiments.
Chao, A., and Tigner, M. Handbook of Accelerator Physics and Engineering (World Scientific, Singapore, 1998).
Edwards, D. A., and Syphers, M. J. An Introduction to the Physics of High Energy Accelerators (Wiley, New York, 1993).