ACCELERATORS, COLLIDING BEAMS: HADRON

Accelerators providing colliding beams of hadrons have become one of the major tools of high-energy physics research. Hadrons, from the Greek word for "thick," or "heavy," are those elementary particles which feel the strong nuclear interaction. The proton and its antimatter counterpart, the antiproton, are the only stable hadrons. All other hadrons ultimately decay, yielding protons and other, nonhadronic, particles. Because protons are electrically charged and stable, they can be made into beams (beam optics) and controlled by electric and magnetic fields, for example, in particle accelerators such as the cyclotron, synchro-cyclotron, and synchrotron. Although they are held together by the strong nuclear force, atomic nuclei (positively charged ions) strictly speaking are not hadrons but are stable and can be controlled and accelerated in the same way as protons.

Fixed Targets versus Colliding-Beam Machines

In particle accelerators, beams of particles such as protons are taken to high energy by suitable arrangements of electric and magnetic fields. When the resultant beams hit a fixed target (a block of material), energy is released in the collisions between the incident particles and the constituent protons and neutrons of the target nuclei. According to Einstein's equation E = mc², this released energy can produce additional particles that were not present initially. From the late 1940s, these fixed-target accelerators enabled physicists to discover many new kinds of elementary particles. The greater the energy of the incoming particle, the greater the energy that can be liberated for making new particles. To calculate how much collision energy is thus made available means viewing the collision in a reference frame which moves with the center of mass of the colliding particles. When a high-energy proton collides with a stationary proton, the resultant collision energy E_cm in the center-of-mass frame is E_cm = mc²√2 + 2E /mc² where m is the proton mass, E the energy of the moving particle, and c the velocity of light. If E is very large, this can be approximated to E_cm = mc²√2E /mc² This shows that the collision energy which can be exploited for making new particles increases only as the square root of the energy of the incident proton. The majority of the incident energy is "lost" simply in making the target particle recoil.

This energy loss can be avoided by colliding two particle beams together, when a greater share of the energy of the two colliding particles becomes available for the production of new particles. In 1943, Rolf Wideröe, who built the first linear accelerator in 1927 and subsequently developed the betatron, a machine for accelerating electrons, proposed such a colliding beam idea, which he patented in 1953. In 1956 Gerald O'Neill of Princeton took the idea further by proposing the use of a standard synchrotron to accelerate the particles and then hold them in two rings which met at a common tangent, where the stored beams would collide. Soon such electron-electron colliders were built at Stanford, California, by a Stanford-Princeton collaboration, and in Novosibirsk, Russia. These developments soon led to the first electron-positron colliding beam accelerators, in which a beam of electrons and a beam of positrons (the antimatter counterpart of electrons) can be held in adjacent orbits in a single ring before being made to collide inside the ring.

As charged particles are accelerated in a ring, any particles with velocities directed out of the beam oscillate around the stable beam orbit, and these oscillations have to be controlled if the circulating beam is to remain stable. Being very light particles, electrons and positrons when accelerated radiate a considerable amount of electromagnetic energy in the form of "synchrotron radiation." This radiation emission naturally damps the oscillations of electrons and positrons, ensuring that the circulating beams are narrow. It also means that a storage ring for electrons and positrons has to replenish continuously the energy of the circulating beams to compensate for these energy losses.

For protons, several major obstacles had to be overcome before the colliding beam idea could be put to work. Beams of protons are nowhere near as dense as even the lightest of natural materials, so most of the time the colliding beams would simply slide past each other, with no collisions taking place. To overcome this, the particle beams would have to be held in storage rings, where they could cross and hopefully collide over and over again. In these rings, the particle beams would have to circulate stably, so that they could be held over much longer time periods than a normal fixed-target machine. But protons, being heavy, do not lose much energy through

FIGURE 1

synchrotron radiation, and other techniques are required to ensure that the beams become sufficiently concentrated.

This can be achieved using the idea of beam stacking, invented in 1956 by the Mid-Western Universities Research Association (MURA), based in Madison, Wisconsin. For beam stacking, the injected particles are stored over as wide a kinematic range in the machine as possible. At the time, CERN, the European Laboratory for Particle Physics in Geneva, Switzerland, was well on the way to building its first high-energy synchrotron, the 25-GeV (giga electron volt) Proton Synchrotron (PS), which supplied its first beams in 1959. But CERN was already looking further ahead and wanted to use the PS to feed new physics facilities and reach higher collision energies. Protons, unrestricted by synchrotron radiation losses, attain much higher collision energies than an electron-positron collider, and in much smaller rings.

At CERN, one possibility was to build two rings to hold contra-rotating 25-GeV proton beams from the PS and bring them into collision, thereby exploiting the full 50 GeV of two colliding particles. This scheme, the Intersecting Storage Rings (ISR), with two interlaced rings 300 meters in diameter crossing at eight different points, was approved for construction in 1965. Mindful that its performance would depend on tight control of the proton beams, tight tolerances were imposed everywhere—magnets, power supplies, vacuum. Special attention was paid to beam diagnostics and control systems. Although it was designed primarily as storage rings, the ISR had its own radio frequency system, to stack the proton pulses arriving from the PS, and this radio frequency power later allowed the ISR protons to be accelerated to 31 GeV.

Many pessimists predicted that the ISR would not work. Unlike the spectacularly successful electron-positron colliders, proton machines at these energies are not subject to synchrotron radiation. This threatened that every ripple and field error would impress itself on the circulating particle beams, and these effects would accumulate, destroying the beam. But no such thing happened. Beams in the ISR did grow slowly, mainly due to scattering on residual gas, and to minimize these effects the ISR pioneered new ultrahigh-vacuum techniques, reaching 10^-12 torr. The ISR produced its first collisions on January 24, 1971, and the machine subsequently ran for thirteen years. Beams were routinely stored for physics runs of fifty hours or more without replenishment.

In 1968, even before the ISR was running, Simon van der Meer at CERN proposed an idea for beam cooling, controlling the spread of the transverse beam momentum in a beam's particles. Van der Meer's stochastic beam cooling scheme would monitor the fluctuations of a circulating beam and transmit a suitable correction signal across the diameter of the ring to meet the same particles as they came round. Over a period of time, this would beat down the statistical fluctuations and increase the beam density. Stochastic cooling was first demonstrated at the ISR in 1974.

Meanwhile another technique for beam cooling had appeared. At Novosibirsk, where one of the first electron colliders had been built, Gersh Budker had the vision of building a proton-antiproton collider. Instead of having two rings, like CERN's ISR, such a machine could hold protons and antiprotons in a single ring before colliding them together, analogous to an electron-positron collider. But antiprotons are difficult to produce and then even more difficult to control. To handle such unruly beams, Budker proposed the idea of electron cooling, surrounding the particles with a sleeve of well-behaved (cold) electrons which would absorb transverse motion from the enclosed beam. In this way, the core beam would become better behaved. Electron cooling was first demonstrated at Novosibirsk by Alexander Skrinsky, a colleague of Budker, in 1974.

Proton-Antiproton Colliders

In particle-antiparticle colliders, the colliding particles, with their mutually opposite quantum numbers, can annihilate. In this way, all the shared collision energy becomes available for producing completely new kinds of particle. This was soon exploited in electron-positron colliding beam machines. One of the earliest of such machines, SPEAR, at the Stanford Linear Accelerator Center (SLAC), transformed electrons and positrons into new varieties of quark-antiquark bound states in 1974.

As well as quark-antiquark bound states, particle-antiparticle collisions could in principle also furnish other particles, such as the tau lepton discovered at SLAC in 1975. The hope had been that electron-positron colliders could also find the long-awaited W and Z particles, respectively the electrically charged and neutral carriers of the weak interaction. But in the mid-1970s, experimental results from a new generation of synchrotrons at Fermilab, near Chicago, and at CERN began to suggest values for the hitherto unknown masses of the W and Z particles. These masses, about 100 GeV, were out of reach of any contemporary electron-positron collider.

In 1976, David Cline, Peter McIntyre, and Carlo Rubbia proposed building a proton-antiproton collider, using the new beam cooling techniques to control the antiprotons. The higher energies available from proton-antiproton colliders with multihundred-GeV beams could be used to look for the W and Z particles. At this time, Fermilab's new 500-GeV synchrotron was furnishing its first beams, and CERN's new 400-GeV Super Proton Synchrotron (SPS) was nearing completion. Fermilab was not immediately interested in the proton-antiproton collider proposal, being already committed to building a more powerful synchrotron, the Tevatron, using super-conducting magnets cooled to 4.2 K.

After constructing a special ring to test the new electron cooling and stochastic beam cooling techniques, CERN saw that the latter route was best suited for high-energy antiprotons, and in 1979 CERN decided to convert its SPS synchrotron into a proton-antiproton collider, building special experiments around the beam collision points. This first meant constructing an ambitious antiproton supply system using stochastic beam cooling, and with this in place the new SPS proton-antiproton collider duly delivered its first collisions of 270-GeV protons and 270-GeV antiprotons in the summer of1981. In 1983, the experiments at the collider discovered the W and Z particles, with masses near 80 and 90 GeV, respectively. The following year, Rubbia, who had pushed the proton-antiproton collider scheme from the beginning and had led the major experiment, and van der Meer, the architect of stochastic beam cooling, were awarded the Nobel Prize in Physics.

Bigger Colliders

Even while these developments were taking place, physicists were looking further ahead at the subsequent generation of machines to explore higher collision energies. In 1962, the United States decided to begin work on a "Super-ISR" at Brookhaven, near New York City, where the Alternating Gradient Synchrotron (AGS) would supply 30-GeV proton beams to a new two-ring arrangement, called ISABELLE, to collide together 200-GeV proton beams in a 3.8-kilometer circumference tunnel. Meanwhile Fermilab was pushing ahead with its superconducting Tevatron project, to supply 1-TeV (1,000 GeV) proton beams. Once the Tevatron was available, then this could be adapted, CERN SPS style, into a proton-antiproton collider, but at higher energies than had been available at CERN. The Tevatron went on to provide its first proton-antiproton collisions in 1985, and in 1995 experiments there discovered the sixth (top) quark. The Tevatron collider initially supplied two 800-GeV beams, and this beam energy was subsequently increased to 980 GeV. The Tevatron pioneered the use of superconducting magnets in a hadron collider, making it possible to achieve collision energies in the TeV range.

CERN's discovery of the W and Z particles in its proton-antiproton collider wrested the crown of particle physics from the United States, which had monopolized particle physics discoveries since the post–World War II introduction of high-energy accelerators. To re-establish its position, the United States proposed an ambitious new machine, the Superconducting Supercollider (SSC). A plan emerged for an 84-kilometer oval racetrack at a totally new laboratory at Waxahachie, Texas, to collide 20-TeV proton beams. With such a large machine, the advantages of using protons and antiprotons in the same ring were outweighed by the difficulties of pumping as many particles as possible into the ring. Even with beam cooling, the antiproton supply was necessarily limited. The SSC was therefore designed as a two-ring proton-proton collider. In the initial SSC approval package, one condition had been that work would stop on Brookhaven's ISABELLE, which meanwhile had been renamed the Colliding Beam Accelerator (CBA).

SSC construction began in Texas in the early 1990s, but in the decade since the project's initial approval, the financial climate had changed. In 1992, SSC funding was drastically reduced, and in 1993 the decision came from Washington, D.C., to scrap the project completely. Of what would have been the world's largest hadron collider, there remained a few huge superconducting magnets and 23 kilometers of empty tunnel in Texas. In the wake of the SSC cancellation, the superconducting CBA scheme at Brookhaven was resurrected and transformed into the Relativistic Heavy Ion Collider (RHIC). Instead of protons, RHIC's initial aim was to collide beams of heavy ions, such as gold nuclei, at energies of up to 200 GeV per nucleon. RHIC went on to collide its first ion beams in 2000. RHIC also collides beams of polarized (spin-oriented) protons.

Several laboratories had occasionally proposed colliding protons with electrons. At the Deutsches Elekronen-Synchrotron Laboratory (DESY) in Hamburg, Germany, a large electron-collider, HERA, was built. Using 30-GeV electrons and 850-GeV protons in a 6.3-kilometer tunnel, using superconducting magnets for the proton ring, HERA provided the world's first high-energy electron-proton colliding beams in 1991.

The LHC

Even while CERN was still building its proton-antiproton collider, a longer-term plan was emerging for a 27-kilometer-circumference tunnel housing the LEP electron-positron collider, to mass-produce Z particles. LEP construction began in 1983, and the machine began operating in 1989. However, with a 27-kilometer-circumference ring available, a hadron collider—either for protons on protons or for protons on antiprotons—eventually could be built in the LEP tunnel.

Plans for this machine, the Large Hadron Collider (LHC), initially emerged in the mid-1980s in parallel with the United States's SSC scheme. Like the SSC, the LHC would be a proton-proton collider, but with its smaller circumference it needed powerful magnets to hold its 7-TeV proton beams. The LHC design thus uses superfluid helium at 1.9 K as the superconducting cooling medium. To optimize the difficult cryogenics, the LHC design evolved into a scheme with the two separate proton rings held inside a single, common cryostat.

The LHC beams will emit small, but measurable, quantities of synchrotron radiation. This is useful for beam diagnostics but insufficient to damp beam oscillations, so traditional beam stacking procedures still have to be relied on. But for a highly cryogenic machine, even a small amount of heating produced by synchrotron radiation has to be kept under tight control. LEP was decommissioned in 2000 so that LHC construction work could begin. The LHC is scheduled to produce its first proton-proton collisions in 2007. It will also sometimes be used to collide beams of heavy nuclei. The large-scale experiments at the LHC attract research physicists from all over the world.

While LEP used 50- to 100-GeV electron and positron beams, the LHC will operate with 7-TeV protons in the same 27-kilometer circumference tunnel, illustrating well the different constraints on accelerating and storing protons and electrons in circular machines. With heavy particles such as protons, the challenge is to provide a strong magnetic field to enable the particles to reach as high an energy as possible in the ring, while with electrons, which are very light particles, the beams have to be gently constrained to minimize synchrotron radiation losses in the ring.

In broad terms, hadron colliders provide high collision energies and are best suited for an initial exploration of a new physics regime. Protons contain three valence quarks together with an attendant cloud of quarks, antiquarks, and gluons, and the collision energy is smeared out among these constituent particles. Electrons and positrons, on the other hand, contain no constituents, and in electron-positron colliders the resultant collision energy can be more sharply focused.

The experiments built for the initial generation of hadron colliders (the ISR and the SPS collider at CERN) revolutionized the design of large-scale particle detectors, with a powerful magnet and with tracking and energy measurement (calorimetry) surrounding the point where the beams collide so as to intercept as many as possible of the emerging particles. This approach is used in all major experiments at colliders. Hadron colliders and planetary systems share the distinction of being the only large-scale systems that have stable orbits for 10¹² and more revolutions.

TABLE 1

See also:Accelerator; Beam Transport; Cooling, Particle; Extraction Systems; Injector System

Bibliography

Bryant, P. J., and Johnsen, K. The Principles of Circular Accelerators and Storage Rings (Cambridge University Press, Cambridge, UK, 1993).

Fraser, G. The Quark Machines (Institute of Physics Publishing, Bristol and Philadelphia, 1997).

Johnsen, K. "The CERN Intersecting Storage Rings: The Leap into the Hadron Collider Era" in The Rise of the Standard Model, Particle Physics in the 1960s and 1970s, edited by L. Hoddeson, L. Brown, M. Riordan, and M. Dresden (Cambridge University Press, Cambridge, UK, 1997).

Richter, B. "The Rise of Colliding Beams" in The Rise of the Standard Model, Particle Physics in the 1960s and 1970s, edited L. Hoddeson, L. Brown, M. Riordan, and M. Dresden (Cambridge University Press, Cambridge, UK, 1997).

Gordon Fraser