* Up until the early 1950s, physicists learned about elementary particles primarily by observing cosmic rays. During that early era, experimental physicists had been developing accelerators to support studies of particles. In the last half of the century, this exercise would accelerate into a global "arms race" to develop more powerful accelerators and more sophisticated detectors to go along with them.
* The structure of the atom had been revealed through the use of low-cost experimental gear, such as ingenious and deceptively simple experiments performed by Rutherford and his people, as well as observations of cosmic rays falling to the Earth from space using electroscope, Geiger counter, cloud chamber, and stacks of photographic plates. However, by the 1930s the need to take a more pro-active approach was beginning to become apparent. Physicists needed some way to accelerate particles to high speeds, or equivalently high energies, for example to accelerate alpha particles -- helium nuclei -- so they could penetrate heavy nuclei. A helium nucleus is positively charged; it has to be given considerable energy to make a dent in a nucleus with a large number of protons, since they repel the alpha particle.
The first step down this road was taken by two researchers, John (later Sir John) Cockcroft (1897:1967), an Englishman, and his Irish colleague Ernest Walton (1903:1995), who developed their "voltage multiplier" at Rutherford's Cavendish lab in 1929. Their device used a transformer to provide a high voltage step-up and then "rectified" the alternating voltage to a high direct voltage, ultimately up to 800 kilovolts. This high voltage could be used to accelerate charged particles to relatively high velocities; the device was focused on accelerating protons so that they could "tunnel" into nuclei.
Cockcroft and Walton had invented the first "particle accelerator", what would become popularly known later as an "atom smasher", though physicists never liked the phrase. The two researchers got the Nobel Prize in physics in 1951 for their work; Cockcroft is also remembered for being a major figure in the development of radar in World War II, and is memorialized by a crater named after him on the far side of the Moon.
At roughly the same time, an American researcher named Robert J. van de Graaff (1901:1967) developed a very simple scheme for producing high voltages that could be used to accelerate particles. His "van de Graaff" generator consisted of a hollow metal sphere mounted on a post, in which a moving belt mounted on the post and rubbing on a brush produced a static electric charge.
* Although the van de Graaff generator would become a staple of high-school physics labs, not to mention the labs of mad scientists in old horror movies, physicists quickly moved on to more powerful particle accelerators.
The problem with the Cockcroft-Walton voltage multiplier and the van de Graaff generator was that they only had two electrodes; the amount of potential that could be placed between two electrodes was limited, since at a high enough potential an arc would jump between the electrodes or punch through electrical insulation.
The solution was to use multiple electrodes, resulting in the "linear accelerator" or "linac", based on concepts outlined in the late 1920s. As originally built, a linac consisted of an evacuated glass tube, with a set of tubular electrodes running down its centerline. The electrodes were linked in pairs to a high frequency alternating current source so that they alternate in electrical polarity from electrode to electrode. A charged particle is injected at one end, and accelerated each time it passes from one electrode to the next.
A linac designed to accelerate electrons (or positrons) has a different design than one intended to accelerate protons (or antiprotons). An electron, being a light particle, is relatively easy to accelerate to near the speed of light, with further attempts to increase its velocity running into diminishing returns. Since the speed of the electron is relatively constant, the electrodes can be all the same length. A proton, however, is much heavier, and so accelerating it is a more gradual proposition. That means that the electrodes in a proton linac have to start out short and gradually get longer.
In 1931, an American physicist named Ernest O. Lawrence (1901:1958) and his colleagues demonstrated a linac with 30 electrodes at a potential of 42 kilovolts alternating at 10 megahertz. It could accelerate positive mercury ions to energies of up to 1.26 million (mega) electron volts (MeV).
After World War II, in 1946, the American physicist Luis Alvarez (1911:1988), who had worked on radar during the conflict, came up with an improved linac design, known as a "drift tube linac (DTL)". The DTL had a general configuration along the same lines as the original linac, with a tube containing a set of electrodes of increasing length, but the DTL had a metal external tube and there were no direct electrical connections to the electrodes, which were called "drift tubes". The DTL was driven by a high-power radio transmitter, like those developed for radar, which filled the tube with a resonant pattern of radio waves at a frequency of a few hundred megahertz. The radio waves induced rapidly alternating voltages in the drift tubes to accelerate a charged particle, which effectively "surfed" on the waves down the tube.
The world's biggest linac is at the "Stanford Linear Accelerator Center (SLAC)", operated by Stanford University in California. Stanford was home to a series of linacs, developed under the direction of Stanford physicist William Hansen (1909:1949). His first linac came online in 1947; it was all of 3.6 meters long and could push electrons to energies of 6 MeV. By 1953, after Hansen's death, the series had evolved to a linac 63.6 meters long that could get electrons up to energies of 600 MeV.
A number of Stanford physicists, most significantly Wolfgang "Peif" Panofsky (1919:2007), thought that the next step should be a vastly bigger linac, 3 kilometers long. US government funding was provided in 1959, and the 3 kilometer SLAC came online in 1967. It originally had 82,650 electrodes and could drive electrons to energies of 22 billion (giga) electron volts (GeV). SLAC has been repeatedly upgraded, and is now capable of pushing particles to 50 GeV.
BACK_TO_TOP* The linac was and remains a highly effective design, but it has one unavoidable drawback: length. While Lawrence was tinkering with linacs in the early 1930s, he was also considering a more compact particle accelerator, which would emerge as the "cyclotron".
A cyclotron consists of two vacuum chambers in the form of shallow cans split topside down the middle, and called "dees" for their appearance. The dees face each other across their open side across a gap. A voltage is placed across the two dees that alternates in a regular cycle to accelerate a charge particle when it passes over the gap. To make the particle curve around so that it passes repeatedly over the gap, the dees are sandwiched top and bottom by an electromagnetic that forces the particle to curve around in an ever-widening spiral, until it reaches the edge of the dee and is emitted as a high-energy beam.
Lawrence's first cyclotron was only about 13 centimeters in diameter and could accelerate particles to 80 thousand (kilo) electron volts (keV); he quickly came up with one twice as wide, and by 1939 the Radiation Lab at the University of California at Berkeley had one 1.5 meters in diameter. Lawrence got the Nobel Prize in that year, 1939, for his invention.
Although Lawrence thought the cyclotron could be scaled up considerably and was planning to build an even bigger cyclotron, the device has an inherent limitation: it can't accelerate particles to high velocities. This is because as the velocity of the particle reaches an appreciable fraction of the speed of light, by relativistic physics its mass increases and that throws off the neat cycle of the synchrotron. Lawrence was aware of the issue, but he planned to get around it by accelerating the particle very quickly. The modern consensus is that his approach would have been limited in the energies it could attain.
War intervened, with much of the US physics community put to work on the Bomb; the big magnet system that Lawrence had acquired to build his super cyclotron was put to work separating fissionable isotopes for the US atomic bomb project. The war years did give physicists time to think things over in their spare time, coming up with new ideas that could be put into practice once the conflict ended. An improved derivative of the cyclotron, known as the "synchrocyclotron", was independently proposed in 1945 by the American physicist Edwin McMillan (1907:1991) and the Soviet physicist Vladimir I. Veksler (1907:1966).
The fix seems somewhat obvious in hindsight: instead of exciting the dees at a fixed frequency, the frequency is dropped slowly as the particle gains mass. The synchrocyclotron is sometimes called the "frequency modulated cyclotron". Lawrence modified his big cyclotron design appropriately, bringing it online in 1946.
The synchrocyclotron works well enough for ions and protons, but it doesn't work so well for electrons, since they are comparatively light and can be accelerated to relativistic speeds very quickly -- too quickly to allow the synchrocyclotron to stay in step. Another particle accelerator, the "betatron", is used to accelerate electrons; it dispenses with the separate dees of the cyclotron and synchrocyclotron, replacing them with a doughnut-shaped chamber sandwiched between an electromagnet. The alternating magnetic field generated by the electromagnet is carefully controlled to bring the electrons up to high speeds.
* Lawrence's synchrocyclotron used magnets 4.6 meters in diameter, and scaling up magnets further didn't seem very promising. The solution was to build a particle accelerator known as a "synchrotron", which was in the form of an evacuated donut or "toroid".
The toroid was spliced with electrical accelerating elements at intervals, powered by RF energy, as with the elements of a DTL. It was also surrounded with a set of electromagnets that would bend the particle beam to follow the curve of the toroid. Magnets cannot be used to increase the velocity of the particle, since a magnetic field exerts a force at a right angle to the direction of a charged particle's motion. The magnetic field has to be increased as the particle beam zips around the toroid to keep it on track, which is the reason the device was named the "synchrotron". Once the beam reaches the desired energy, an electromagnet system shunts the beam off to collide with a target or another beam, generating a shower of particles for analysis. The particles produced by the collision could be sorted by a magnetic "switchyard" into positive, neutral, and negative streams if needed.
Typically, a relatively small linac is used to inject charged particles into a synchrotron. A synchrotron can be used to accelerate protons or electrons, though most of the bigger synchrotrons accelerate protons. However, the synchrotron has an inherent limitation: a charged particle beam moving in a circle -- that is, being radially accelerated -- will emit electromagnetic energy, or "synchrotron radiation", robbing it of power. This is a worse problem for lightweight electrons than it is for heavy protons; linacs have an advantage for accelerating electrons and positrons, and in fact most big synchrotrons are proton accelerators. Synchrotron radiation can be reduced by making the toroid as wide as possible, since that minimizes the radial acceleration of the particle.
The first particle accelerator to break the 1 GeV limit was a proton synchrotron called the "Cosmotron" at the US Brookhaven National Laboratory on Long Island, New York, built under the authority of the US Atomic Energy Commission (AEC, now the Department of Energy / DOE). It had a diameter of 18.3 meters and started out accelerating particles to 2.3 GeV; by 1953 was pushing protons to energies of 3 GeV. For the first time, an artificial particle accelerator was able to top the energies of typical cosmic rays, and from that time on the hunt for new particles switched from trapping cosmic rays to building ever bigger particle detectors.
The Cosmotron was followed in 1954 by another American proton synchrotron, the "Bevatron" at the University of California at Berkeley, also built under AEC authority, which could reach 6.4 GeV. The USSR, not to be outdone, inaugurated a 10 GeV proton synchrotron at Dubna in 1955. There matters rested for a few years. One of the limitations of the original synchrotron was the fact that the particle beam tended to disperse, weakening the particle flux; a new scheme, known as "strong focusing", was invented, in which there were two sets of magnets -- one to bend the beam, the other to focus it.
Since a magnet could only focus the beam in one plane, the focusing magnets alternated between a magnet that focused the beam horizontally and a magnet that focused the beam vertically. The concept of strong focusing was invented independently in the early 1950s by a team at Brookhaven led by Stanley Livingston, and by Nicholas Christofilos (1916:1972), an unorthodox Greek experimental physicist who later signed on with UC Berkeley. The first strong focusing synchrotron came online at Cornell University in 1954; it was an electron synchrotron, capable of driving electrons to energies of 1.5 GeV.
Using strong focusing technology, a much more powerful set of particle accelerators came online. A European collaborative organization named CERN (the French acronym for "Council / Center for European Nuclear Research") had been formally set up in 1954 after four years of negotiations and studies, with the initial objective of setting up a strong focusing synchrotron with a diameter of 100 meters at a site on the Swiss-French border near Geneva. It was a 28 GeV proton synchrotron, just known as "PS", and came online in 1959. Incidentally, CERN was later renamed the "European Organization for Nuclear Research", but the CERN acronym has stuck.
There was a generally friendly rivalry between Europe and the US on building particle accelerators, and in response in 1960 a 33 GeV proton synchrotron, the "Alternating Gradient Synchrotron (AGS)", was brought online at Brookhaven. The Soviets inaugurated a 76 GeV proton synchrotron at the Institute for High Energy Physics at Serpukhov, about a hour's drive south of Moscow, in 1967, grabbing the title for the world's most powerful accelerator for a few years. Western physicists had reasonable access to this facility by the standards of the Cold War.
* The Soviets didn't hang on to their lead for very long. The 1970s brought another generation of powerful accelerators. In 1972, a proton synchrotron 6.3 kilometers in circumference was brought online at the Fermi "National Accelerator Laboratory (Fermilab)" near Batavia, Illinois. It initially could push protons to energies of up to 200 GeV, which was boosted to 500 GeV in 1975.
Development of the Fermilab accelerator was pushed by Robert R. Wilson (1914:2000), who when he found that the available funding for the proposed design was much less than anticipated, announced that the machine would ultimately produce energies two and half times greater than initially proposed. He drove the whole thing through, though the project was delayed for a year by various problems -- worst being when the underground tunnel in which the accelerator was built became damp, resulting in almost half of the magnets shorting themselves out the first time they were fired up.
The Fermilab accelerator was 2 kilometers in diameter and 6.3 kilometers around. Protons in the Fermilab synchrotron were boosted by a Cockcroft-Walton accelerator; then a linac; then a small synchrotron, known as the "Booster", before being shoved into the main ring. The order of the boost stages mirror the history of particle accelerators, and collectively kick up proton energies by orders of magnitude at each stage.
CERN replied in 1976 with the "Super Proton Synchrotron (SPS)", which had a circumference of 7 kilometers and straddled the Swiss-French border in an underground tunnel. The SPS could drive protons to 400 GeV.
The Fermilab accelerator was uprated to 1 trillion (tera) electron volts (TeV) in 1983 by installing a new ring with superconducting magnets in the tunnel underneath the original ring, with the new accelerator becoming the "Tevatron" and taking the distinction of being the most powerful particle accelerator in the world -- for a while.
BACK_TO_TOP* As a rule, these large particle accelerators pushed particles up to high speeds and then slammed them into a fixed target to generate a shower of particles for observation by a detector system -- a complicated topic in itself, discussed below. This isn't very efficient, since the velocity of the secondary particles carries off a high proportion of the energy that should preferably go into creating particles. A better approach is to slam two particles of equal mass directly into each other, bringing them to a dead halt with a tremendous burst of energy. By the same logic, rear-end automotive collisions tend to be less disastrous than head-on collisions.
The idea of using two particle accelerators to generate colliding beams had been floating around since the mid-1940s, but work on "colliders" didn't really start to pick up momentum for another decade. Two American physicists, Gerard K. O'Neill (1927:1992) of Princeton -- later to become one of the patron saints of space exploration for his energetic promotion of space colonies -- and Wolfgang Panofsky (1919:2007) of Stanford, collaborated on development of an electron collider, with ground broken on the project in 1959.
One of the problems with a collider is that energetic beams may be tenuous, meaning that even if they are shot straight at each other, they're not likely to score hits. To ensure more collisions, the particle beam could be circulated in a "storage ring" to build up a greater particle density. The Princeton-Stanford collider consisted of two storage rings, built alongside each other and intersecting at a junction between them. First results were obtained in 1965. Each storage ring could only provide an energy of 500 MeV, resulting in collision energies of 1 GeV, but if an electron beam were fired at a stationary target, it would have had to have an energy of 1 TeV to have the same effect.
In the meantime, a group of Italian researchers under an Austrian physicist named Bruno Touschek (1921:1978) was taking a different approach, working on a demonstrator named the "Annelo d'Accumulazione (ADA / Accumulation Ring)" at Frascati that performed electron-positron collisions. Since electrons and positrons curve in opposite ways in a magnetic field, they could both be accelerated in different directions in the same synchrotron. The magnetic field could then be tweaked slightly to bring the two streams of particles into a head-on collision.
An operational electron-positron collider followed at Frascati, which was followed in turn by other electron-positron colliders at Cambridge, Massachusetts, in the USA; Orsay near Paris, France; and Novosibirsk in central Siberia. These machines had energies of up to 7 GeV, with an effect far greater than that of a single beam striking a target.
One of the most significant of the early colliders was the "Stanford Positron Electron Acceleration Ring (SPEAR)". It was the brainchild of Burton Richter (born 1931), who had worked on the Princeton-Stanford collider with O'Neill and Panofsky. SPEAR was completed at SLAC in 1972 and consisted of an oval ring with dimensions of 63 x 80 meters built in a parking lot near the end of the SLAC linac. Electron beams were tapped off from the linac to be fed into the SPEAR ring directly, and to collide with a copper target for the production of positrons, that were then fed into the SPEAR ring to circulate in the opposite direction. Each beam had, in the maturity of SPEAR, a maximum energy of 8 GeV. The two beams could collide at two points along the ring, with the collisions detected by a large, advanced wire-chamber-based detector system, the Mark 1.
In the meantime, a comparable collider, the "Double Storage Ring Facility" or "DORIS" in its German acronym, was being built at the German Deutsches Elektron Synchrotron (DESY) lab at Hamburg, coming online in 1974. As the name implied, DORIS was (originally) a two-ring collider, designed so that it could be used as either an electron-electron collider or an electron-positron collider. In maturity, each beam could obtain an energy of 7 GeV. In 1977, DORIS was rebuilt as a single-ring electron-positron collider.
Towards the end of the decade a "third generation" of even more powerful colliders appeared, including the "Positron Electron Tandem Ring Accelerator (PETRA)" at DESY (38 GeV initially, uprated to 46 GeV); the "Cornell Electron Synchrotron Ring (CESR)" or "Ceasre" at Cornell University in New York state, USA (16 GeV); and the "Positron Electron Project (PEP)" at CERN (36 GeV), with a circumference of 2 kilometers.
The fourth generation appeared later in the 1980s. The CERN "Large Electron-Positron (LEP)" ring came online in 1989: it had a circumference of 27 kilometers and provided energies of 100 GeV. In contrast, SPEAR had a circumference of only 250 meters and the CERN PEP a circumference of 2 kilometers. The Japanese built a similar collider named TRISTAN at the KEK laboratory that was capable of 60 GeV.
* These systems were electron-electron or, more generally, electron-positron colliders. During the 1960s, CERN engineers worked on a proton-proton collider, which emerged as the "Intersecting Storage Rings (ISR)" in 1971. It consisted of two rings on top of each other that intersected at eight locations, and used a scheme known as "stochastic cooling", devised by a Dutch researcher named Simon Van der Meer (1925:2011), to concentrate the proton beams. By itself, the proton beam tended to smear out, but in stochastic cooling, magnetic pulses were used to "herd" the protons together in bunches.
The ISR obtained protons at 26 GeV from the CERN PS synchrotron, with some of the protons also being smashed into a target and fed into an intermediate "antiproton accumulator (AA)" ring. The protons and antiprotons were then boosted to 31.5 GeV. The 63 GeV collisions were equivalent to a 1.8 TeV impact with a stationary target.
In 1976, the Italian physicist Carlo Rubbia (born 1934) of CERN and two American physicists, David Cline and Peter McIntyre, proposed that the CERN SPS and the Fermilab Tevatron be modified into proton-antiproton colliders. CERN built an "Antiproton Accumulator" subsystem that used stochastic cooling to herd antiprotons obtained from the old PS for injection into the SPS. The SPS began to produce collisions between protons and antiprotons in 1981, reaching collision energies of 540 GeV -- equivalent to a 150 TeV impact with a stationary target. The modified SPS with the Antiproton Accumulator was named the "Super Proton-Antiproton Synchrotron". The Fermilab Tevatron was similarly modified in 1986 into a proton-antiproton collider.
* These monster colliders were followed by several interesting variations on the collider concept. At SLAC, Burton Richter pushed through a scheme in which the old three-kilometer linac was used to drive positrons and electrons into a ring in different directions, so they could then collide head-on. The modified hybrid system became known as the "Stanford Linear Collider (SLC)".
DESY also brought another unusual collider online in 1992, the "Hadron-Electron Ring Accelerator (HERA)", basically a proton-electron collider consisting of two rings built through the same loop tunnel, 6.3 kilometers in circumference. The rings intersected at eight locations.
The Americans planned a huge "Superconducting Supercollider (SSC)" to be sited in Texas, with a circumference of 80 kilometers and superconducting magnets, but it was too much for the US Congress to swallow and was cancelled in 1993. Some of the money for the SSC was then funneled off to Fermilab and SLAC to upgrade their accelerator systems. CERN did move forward on a "Large Hadron Collider (LHC)", which is now online and the preeminent particle physics facility in the world.
The LHC is 8.6 kilometers in diameter, 27 kilometers around, straddling the Swiss-French border; the ring is tilted underneath the Earth by 1.4 degrees, varying in depth from 50 meters on the Swiss side to 175 meters on the French side. The LHC ring is supported by almost 7,000 superconducting magnets, chilled by liquid helium. It is so big that tides change its diameter by a millimeter, with compensation required to keep it working right. The ring is fed by the SPS.
An international group of physicists has been promoting a linear electron-positron collider, the "International Linear Collider (ILC)", to complement the LHC. It would have two linear accelerators firing into each other, with one for electrons, the other for positrons. Each linac would be 11.3 kilometers long and use superconducting accelerators to accelerate their particles to 250 GeV. Cost estimates are in the billions of dollars, with the earliest date for coming online being about 2020. No site has been chosen yet -- candidates include CERN, Fermilab, and a site in Japan.
* Some believe that the development of bigger and bigger particle accelerators is reaching diminishing returns, but new technology may be able to provide a big step up in accelerator energies. The new "plasma accelerator" techniques being investigated involve generating a plasma and then setting up a high-speed wave by shining a high-power laser through it. The high-speed plasma wave can provide accelerators far greater than those provided by RF accelerators.
Beam controllability is a problem, but a demonstrator plasma-wave accelerator is now being fitted to SLAC. If the technology can be brought into practical use, it will not only boost the power of existing large accelerators, but permit the development of low-cost "tabletop" accelerators for academic and industrial uses.
BACK_TO_TOP* Along with machinery to accelerate particles, physicists also developed new means to detect them. Before the time of the big particle accelerators, physicists had used:
Analysis of cloud chamber and photographic stack particle traces was performed by a team of "scanners", almost always women, who looked over the traces to see if something unusual happened.
* In the postwar period, "scintillation counters" came of age. Rutherford and his students had used scintillation screens, in which the impact of a particle produced a "scintillation" or burst of light, but these were strictly "eyeball" instruments. A modern scintillation counter consists of a block of plastic, zinc sulfide, sodium iodide, anthracene, or other material that scintillates when hit by a high-energy particle. The scintillating material is mated to a "photomultiplier tube", a type of vacuum tube that greatly amplifies light and produces an electrical signal at its output.
Scintillation counters have replaced the Geiger-Mueller tube in many applications, and they are also used in mass as detectors in particle physics experiments. One well-known scintillation detector design, named the "Crystal Ball", consists of a hollow crystal sphere about 2.1 meters wide surrounded by 730 sodium iodide scintillation detectors. It went into service at SLAC in 1979, to be passed on to DESY in Germany in 1982.
Another class of detector is based on "Cerenkov radiation", a phenomenon discovered by Soviet physicist Pavel Alexeyvich Cerenkov (1904:1990) in 1934, which would win him a share in the 1958 Nobel Prize for physics. Light will of course slow down when passing through a medium instead of a vacuum, with the rate of decrease corresponding to the index of refraction of the medium. If a particle passes through a nonconductive material at a velocity higher than that of light in the medium, it will generate pale blue light, Cerenkov radiation, in the direction of its travel. This radiation can be observed to trace the path of the particle. Selecting a material with the right index of refraction will also allow a Cerenkov detector to be built so that particles below a given velocity will not be observed.
* Scintillation and Cerenkov detectors could detect particles and give some indication of their energy, but they couldn't provide particle tracks like those produced by the cloud chamber. The cloud chamber had proven extremely valuable and had been greatly refined, but it really wasn't adequate to track particles once the energies went above a certain level. The particles moved so fast that a really long cloud chamber would have been required to observe their decay processes; in addition, a cloud chamber could only provide about one trace a minute, which was much slower than a modern particle accelerator could provide them, which was wasting the capability of the accelerator. Something better was needed.
In 1952, the American physicist Donald Glaser (1926:2013) invented the "bubble chamber", in which a liquid is heated to the point where it is just about to boil; charged particles will trigger the boiling, leaving a track of bubbles behind them that can be photographed for analysis. Placing the bubble chamber in a magnetic field will cause charged particles to curve, allowing their electrical polarity and energy to be measured. The bubble chamber is much more sensitive than the cloud chamber.
Glaser invented the bubble chamber after noticing the effervescence of a bottle of beer after he popped the cap, building a demonstrator prototype in which the "chamber" was a small glass vial. Nobody took his idea very seriously until he met Nobelist Luis Alvarez in 1953, who realized that the bubble chamber was exactly what was needed for his own work. Alvarez threw his weight behind Glaser, resulting in the construction of a series of prototypes that resulted in the installation of the "72-inch" bubble chamber at the UC Berkeley Bevatron in 1959. This instrument was built around a chamber measuring 1.8 x 0.5 x 0.4 meters (71 x 20 x 16 inches), accommodating 17 liters (4.5 US gallons) of liquid hydrogen. It was installed in its own building, sandwiched by a magnet system that drew enormous amounts of power.
A few years later, Brookhaven followed with an "80-inch" bubble chamber linked to the AGS accelerator. In operation, a large piston, almost a meter in diameter, was lifted 15 milliseconds before the particle beam was scheduled to arrive. Once the particle beam arrived, after a delay of a millisecond to allow bubbles to form, an arc flash illuminated the chamber for an instant to allow a set of fast film cameras to take pictures. The piston was then pushed back, with the film cameras winding to their next frame. The bubble chamber could cycle in a few seconds. Modern bubble chambers generally use liquid hydrogen, and can be very large; since hydrogen tends towards the explosive, bubble chambers can also be a bit risky to deal with.
Of course, neutral particles don't leave a direct trace in any kind of detector, but they can be detected from collisions with atoms or other particles that generate charged particles in turn. The general indicator of a neutral particle is a trace in a bubble chamber or other detector that disappears and then reappears farther on from a collision or the production of charged breakdown particles.
* One of the problems with the bubble chamber was that, unlike the cloud chamber, it had to be activated before the arrival of the particles to be observed. The answer was a new class of detector, the "spark chamber".
A spark chamber is something of an evolution of the concept of a proportional counter. The Italian physicist Marcello Conversi (1917:1988) invented the ancestor of the spark chamber, the "flash tube", in the mid-1950s. These were sealed glass tubes filled with neon gas, arranged in stacks between metal electrodes. A high voltage was placed across the electrodes, alternating in polarity between layers; if a charged particle passed through the stack of flash tubes, it would ionize the neon gas in the tubes, causing a spark to jump between the electrode plates. The stack could be photographed, with the sparks providing a trace of the path of the particle.
Researchers in the UK and Japan looked at this scheme and wondered: why were the glass tubes needed? Why not simply put the plates in a single enclosure, filled with neon or some other inert gas? And so the spark chamber was born. The basic spark chamber consists of metal electrode plates a few millimeters apart, with a high voltage of 10 to 20 kilovolts placed across the plates in an alternate fashion, and a charged particle ionizing the gas between the plates to leave a trail of sparks behind. As with the cloud chamber, the spark chamber could be activated after the passage of a charged particle to reveal the ionization trail left behind. Scintillation counters surrounding the spark chamber could indicate the passage of the particle and activate the chamber.
In the 1960s, a CERN engineer named Frank Kriernan came up with an improved spark chamber, in which the plates were replaced by arrays of wires. The "wire spark chamber" had the advantage that particle traces no longer needed to be photographed: "sense amplifiers" connected to the wires could tell which wire pairs generated a spark, allowing the trace to be recorded electronically. The alternating planes of wires are at right angles to each other, providing "row" and "column" lines to give a three-dimensional particle track. Computers had been used beginning in the late 1950s to analyze photographs of particle traces, with the level of automation steadily increasing over time. With the wire spark chamber, eventually the computer was able to do it all, ignoring events known to be irrelevant, to record events that were relevant and display them on a computer workstation. The scanners were out of a job.
The fact that the traces didn't have to be photographed also meant that the voltages could be lowered, which implied a faster cycle time. A wire spark chamber can obtain traces a thousand times a second, a thousand times faster than a bubble chamber. Spark chambers have been flown on a number of spacecraft to observe cosmic high-energy particles, and even gamma rays -- of course, a photon is a neutral particle and can't be tracked by a spark chamber, but a gamma ray photon can generate electron-positron pairs that can be tracked.
A number of variations on wire spark chambers have been developed, such as the "multiwire proportional chamber", which has closer spacing, lower voltages, and faster cycle times; and the "drift chamber", which provides higher resolution by timing the voltage pulses from the wires relative to a trigger provided by external scintillation detectors.
Bubble chambers still managed to compete, though ironically they became smaller, designed to be operated at rapid cycle times. The rapid cycles meant that the bubble traces were faint, but the small size of these bubble chambers meant that a high-resolution camera would be able to pick them up.
* The detector systems used with modern particle accelerators are just that, systems, with multiple layers of elements. One interesting example was the "Underground Area One (UA1)" detector system at CERN, built for the SPS in the early 1980s. It was built in four layers:
UA1 was supported by a sophisticated "trigger system" that waited for the occurrence of events meeting certain preprogrammmed criteria before actually recording measurements. UA1 was a huge system, weighing thousands of tonnes; it has led to more sophisticated detector systems.
The LHC, which is state of the art at present, has four primary detector systems: the "Compact Muon Solenoid (CMS)", "LHC bottom (LHCb)", "A Toroidal LHC Apparatus (ATLAS)", and "A Large Ion Collider Experiment (ALICE)". All four are elaborate systems whose description would necessarily be lengthy. They are complemented by smaller, specialized instruments.
100 million channels of data are produced by each of the two largest detectors, churning out enough data to fill up 100,000 CD-ROMs every second. Storing all that data being somewhat impractical, the channels are filtered by trigger & data-acquisition systems that discard all the routine data and pass on the interesting events. A CERN "cloud computing" system then organizes the data into manageable data sets, which are in turn analyzed at scientific "cloud computing" centers at a dozen sites on three continents.
* There are a number of other types of particle detectors. One interesting variation on the proportional counter is the "neutron monitor", invented by physicist John A. Simpson (1916:2000) of the University of Chicago in the late 1940s. Although a normal proportional counter cannot pick up neutral particles, the neutron monitor uses a proportional counter tube filled with the helium-3 isotope; a neutron will interact with the helium-3 isotope to produce more normal helium-4, with the interaction creating an ionization path through the tube. Other fill gases have been used, working on much the same principle. The neutron monitor is shielded by polyethylene to slow down neutrons (and, it seems, screen out charged particles).
A solid-state detector is simply a semiconductor crystal with a PN junction. The impact of a charged particle generates electron-hole pairs, resulting in a current across the junction. The crystal is generally cooled to reduce thermal noise. The common charge-coupled device (CCD), generally associated with digital cameras, can also be used to detect charged particles.
Neutrinos and other weakly-interacting particles are of course hard to detect by definition, and systems designed to detect them are often based on interactions with atomic nuclei. The actual detection of the ghostly neutrino in 1956 by Fred Reines (1918:1998) and Clyde Cowan (1919:1974) under "Project Poltergeist" used a detector based on such principles. They had originally come up with the idea while they were on the Bomb at Los Alamos during the war, initially thinking they could set up a detector system to pick up neutrinos from a nuclear explosion.
Nuclear weapons were a highly intermittent source of neutrinos, so the Reines-Cowan detector was built next to a nuclear reactor, which should have been dumping out a flood of antineutrinos, with a tank of cadmium chloride solution as the antineutrino trap. It detected reverse beta decay, in which a proton absorbs an antineutrino and turns into a neutron, emitting a positron:
proton + antineutrino --> neutron + positron
The positron quickly runs into an electron, resulting in mutual annihilation, producing two gamma rays of distinctive energies. The neutron produced by the neutrino is absorbed by a cadmium nucleus, which also then emits distinctive gamma rays. The tank was surrounded by a liquid scintillating material, which was monitored by arrays of photomultiplier tubes.
The Poltergeist detector was buried under heavy layers of dirt and old battleship armor to block out other stray events that could confound the observations. It only picked up about three antineutrinos an hour; however, but that rate went to near zero when the reactor was shut down, meaning that the detections were not likely to be confounding events. They sent a telegram to Wolfgang Pauli, who had come up with the idea of the neutrino in the first place and was then in Zurich:
WE ARE HAPPY TO INFORM YOU THAT WE HAVE DEFINITELY DISCOVERED NEUTRINOS.
Reines shared the 1995 Nobel Prize in physics for this work.
BACK_TO_TOP