Particle physicists at Germany’s national accelerator laboratory in Hamburg, DESY, are eagerly awaiting the fruits of eight years’ hard work. On 15 May 1984, excavations began at the laboratory on a new machine called HERA, the Hadron Electron Ring Accelerator. Now HERA is ready, and the detectors lie in wait to capture the particles produced in the collisions of high-energy protons and electrons. The machine is more than just another accelerator – it presents particle physicists with a new type of tool to investigate the subatomic world, and promises them an exciting journey into the heart of matter.
Accelerators, which take beams of subatomic particles to very high energies, are a vital component of research in particle physics. The collisions of energetic particles with other particles can produce new, short-lived forms of matter, which are important to our understanding of fundamental subatomic processes. The best way to exploit the energy of the collisions is to allow two particle beams to collide head on. HERA is a particle collider, but with a difference.
In the past, most particle colliders have been based on a beam of particles and a beam of antiparticles, which have the same mass as particles but opposite electrical charge. This means that the two beams can travel in opposite directions around a single ring-shaped accelerator. At CERN, the European Laboratory for Particle Physics near Geneva, physicists have made important discoveries by colliding protons with antiprotons and, more recently, by colliding electrons with antielectrons (positrons).
Advertisement
Where two paths cross
The new approach with HERA is to collide two beams of quite different particles – electrons and protons. These are vastly different in mass – protons are nearly 2000 times as heavy as electrons. So HERA consists of two different accelerator rings, one for each type of particle, which cross at two points and could eventually be made to cross at up to four.
Why go to all this trouble to collide electrons and protons? What the physicists at HERA hope to do is to use the electron to explore the internal structure of the proton. We know the electron has mass, but as far as we can tell it has no size: we say it is ‘point-like’. On the other hand, the proton has mass and a size of around one millionth of a millionth of a millimetre (10-12 millimetres). We also know that the proton is not a truly fundamental particle, but is built from particles known as quarks.
During the past few decades, particle physicists have developed the Standard Model of the particles of matter and the forces that act on them (see ‘The origin of mass’, Â鶹´«Ã½, 18 April). According to this model, matter falls into two categories: quarks and leptons (the latter group includes the electron). We believe that quarks, like leptons, are point-like. They appear to come in six types: up, down, strange, charm, bottom and top. (The top quark has been inferred theoretically, but has yet to be found.) The proton is made from two up quarks and one down quark, held together through the exchange of massless particles called ‘gluons’. These are the carriers of the strong force, one of the four forces in nature that particles feel. Because the electron is point-like it is an ideal tool to ‘probe’ in detail the quark and gluon structure of the proton, and this is what the physicists at HERA aim to do. The experiments will be the latest in a long line of successful investigations that go back 40 years.
In the early 1950s Robert Hofstadter at Stanford University in California fired a beam of electrons with an energy of 400 million electronvolts (MeV) at a target of liquid helium, and observed the electrons that emerged at angles of 45 degrees and 60 degrees to the beam direction. In ‘elastic’ scattering (that is, with no new particles created) we would expect to observe electrons with a unique energy, just as we would expect from the scattering of two billiard balls at a fixed angle. (Energy and momentum are conserved, so particles scattering at a particular angle have a particular energy.) Hofstadter saw such a peak, but he also saw a second peak at a lower energy. This corresponded to the elastic scattering of electrons from the constituents of the helium nuclei – two neutrons and two protons. The peak was broad, indicating that the neutrons and protons were moving around inside the nucleus. This motion was expected because in quantum theory, particles confined to a small volume must have some spread in their momentum, as a consequence of Heisenberg’s uncertainty principle.
The next generation of this kind of experiment came in the late 1960s at the Stanford Linear Accelerator Center. A team from SLAC and the Massachusetts Institute of Technology used liquid hydrogen as as target (that is, nuclei of single protons) and an electron beam with an energy of up to 16 gigaelectronvolts (GeV), equivalent to 16 thousand million electronvolts. This was an energy 40 times greater than Hofstadter had used and, once again, electrons emerged at large angles to the beam direction. To everyone’s excitement, at a given fixed angle, not only did they see a sharp peak corresponding to the elastic scattering of electrons from the protons, but there was also a broad peak at lower energies. This provided the first direct evidence of structure within the protons. The broadness of the second peak showed that whatever the particles making up the proton, they were moving about.
Experiments carried out in the 1970s clearly identified quarks as the constituents of the proton. But they produced other surprising results too. From measurements of the energies and angles of the scattered fragments, the physicists could infer the momentum of the quark relative to that of the proton as a whole. It was reasonable to suppose that each quark should have exactly one-third of the proton’s momentum, since a proton is made from three quarks. But the experiments showed that the three quarks carried only roughly half of the proton’s momentum. This was the first evidence for gluons, the particles that are exchanged between quarks. It is the gluons that carry the remaining half of the proton’s momentum.
Protons and predictions
The structure of the proton in terms of quarks and gluons has now been confirmed by a wide variety of experiments at different laboratories. In these investigations, physicists have used not only electrons but also other leptons, namely muons (which are like electrons but 200 times heavier) and neutrinos (which are like uncharged electrons, with little or no mass). These experiments have shown strong support for quantum chromodynamics (QCD), the theory that describes the interactions of quarks and gluons. One of the aims at HERA will be to test the predictions of QCD in a new, higher energy domain.
HERA will collide electrons at an energy of 30 GeV with protons at an energy of 820 GeV. At these energies physicists will be able to probe distances as small as 10 -15 millimetres, that is, a thousandth the ‘size’ of a proton. This is because the higher the energy of the particle beams, the greater the detail they reveal. What will lie in store? We will be certain to observe in minute detail the behaviour of quarks and gluons within the proton. HERA might even hint at internal structure of the quarks themselves. This could show up in subtle deviations in the data from the predictions of the Standard Model and QCD, which treat quarks as point-like.
Physicists are hoping that HERA will also be the showcase for exotic objects that until now have existed only theoretically. One such object is the ‘leptoquark’. The particles that build up matter seem to fall neatly into the two categories of quarks and leptons. But is there a direct connection between the two groups? Some theorists have proposed the existence of leptoquarks – leptons and quarks bound together. The search for leptoquarks at other accelerators has proved fruitless, but this could be because leptoquarks are so heavy that these machines do not have enough energy to produce them. At HERA, with its high-energy electron-proton collisions, physicists will be able to search for particles with more than twice the mass that they could hope to detect elsewhere. The existence of leptoquarks – and hence a direct link between quarks and leptons – would be one of the most exciting discoveries of modern particle physics.
HERA, the machine that will open up these possibilities, is based on an approximately circular ring with a circumference of 6.3 kilometres. The two accelerators, one for protons and one for electrons, sit in a single tunnel, 5.2 metres in diameter, which lies between 10 and 20 metres below ground, and extends out from the DESY site under suburban Hamburg. For much of the ring, the electron accelerator nestles under the proton accelerator. At two points the beams are brought into line so that the two types of particle can collide head on.
In any ring-shaped accelerator, as the particles gain energy, the magnetic fields that steer the beams must increase, to keep the particles circulating on the same path. The maximum energy a machine can reach is therefore dictated by the maximum field created in the electromagnets through which the beams of particles must pass. To reach as high an energy as possible, HERA’s proton ring is built from electromagnets made from superconducting, rather than conventional, wire. When it is cooled to a temperature only a few degrees above absolute zero, electric currents flow through it with negligible resistance. This means superconducting magnets can run at higher currents than conventional electromagnets, producing stronger fields.
HERA’s proton ring, which accelerates protons to 820 GeV, contains nearly 1400 superconducting magnets – 416 ‘bending magnets’ to steer the proton beam, 246 ‘quadrupole magnets’ to focus it, and 712 ‘correction magnets’ to ensure that it remains stable. Half the bending magnets have been built by industry in Italy, half the quadrupoles in France, while the correction magnets have been built in the Netherlands. The remainder were made by industry in Germany.
Superconducting materials are also important in the electron ring, in the devices that accelerate the electrons. Electrons lose energy in the form of electromagnetic radiation as they travel on curved paths. The higher the energy of the electrons, or the tighter the curve, the more ‘synchrotron radiation’ they emit. So the electron accelerator is a compromise: it is designed to accelerate the electrons to as high an energy as possible without wasting too much power in synchrotron radiation. In practice this means a relatively low energy of 30 GeV, which is low enough for non-superconducting electromagnets to steer the electrons to their maximum energy.
In both rings the particles receive accelerating boosts each time they pass through box-like resonant cavities installed around the ring. Here they encounter powerful radio-frequency standing waves set up inside the cavities, rather like the sound waves in an organ pipe. However, energy can be lost in the cavities through electric currents set up in the metal walls. In the electron accelerator, the energy transferred from the radio waves to the electron beam must more than compensate for the energy lost as synchrotron radiation if the particles are to gain energy. To improve their performance, some of the cavities in HERA’s electron ring are made from superconducting material. The currents still form in the cavity walls, but they encounter no electrical resistance and therefore lose no energy.
So how will physicists observe electron-proton collisions? There are large complex detectors, called H1 and ZEUS, at two of the points where HERA’s beams can collide. Each detector is the responsibility of physicists and engineers from many countries. Components of both detectors were built at locations around the world, then assembled at DESY. Each detector surrounds the collision region with different layers to detect the different types of particle produced when the beams collide.
HERA poses many new technical challenges for the experimenters. Particle accelerators are usually designed to achieve the maximum number of collisions in the detectors. This gives the greatest scope for rare but interesting ‘events’ such as wide-angle electron-quark scattering, or for producing heavy new particles. However, interesting events will occur at a manageable rate of a few per second, whereas uninteresting ‘background’ events, such as collisions between particles and residual gas molecules in the vacuum pipes in which the beams travel, could happen as often as 100 000 times a second. The detectors at HERA must be able to select or, ‘trigger on’ the interesting events, while rejecting the background ones.
After the event
In colliders the beams usually come in ‘bunches’, and in HERA, there are 210 bunches each of electrons and protons. A bunch of electrons and a bunch of protons cross each other at the centre of the detectors every 96 nanoseconds (96 thousand-millionths of a second). So there are 10.4 million crossings a second, and an interesting event could happen at any of these crossings. The problem is that 96 nanoseconds is not enough time to decide if an event is interesting because there is too much information to process for each interaction. For example, to select events that have greater-than-average total energy, even the most advanced electronics used in the detectors at HERA could need up to 5000 nanoseconds.
Experiments at HERA overcome this problem by ‘pipelining’ the information. The detectors produce about two million bytes of data for each interaction. (This is approximately equivalent to a quarter of a million words of ordinary text.) All this information must be stored in electronic ‘pipeline’ memories for up to 5 microseconds until ‘trigger’ electronic devices make a decision about whether to keep an event or not. The information includes the vital statistic of the ‘beam-crossing number’. This identifies which crossing the information came from. While the decision to keep an event is being made, information from subsequent crossings is stored in adjacent locations in the pipeline and is updated at least every 96 nanoseconds.
Once the ‘trigger’ electronics has decided to keep an event, data for the appropriate beam-crossing are retrieved from the pipeline memories and passed to the next stage of the processing chain. Here, an array of microprocessors further reduces the rate of background events, and compresses the data within each event. There are three or more levels of decision making, which turn 100 000 interactions a second into around five potentially interesting interactions a second. Even so, this produces as much as a million bytes of data each second. Once recorded on magnetic tape, physicists can start analysing the events.
Last autumn, HERA was put through its preliminary paces and monitors registered the effects of collisions between protons at 480 GeV and electrons at 26.5 GeV. Since then the teams preparing the H1 and Zeus detectors have been working endlessly to have them ready for this month’s deadline. HERA soon begins in earnest, and the goal of colliding electrons with protons at full energy will open a new chapter in the unfolding story of particle physics.
Neville Harnew and Christine Sutton are members of the physics department at the University of Oxford. Both work on the ZEUS experiment at HERA.
* * *
THE HISTORY OF HERA
The decision to build HERA at DESY, the German national accelerator laboratory, was taken by the federal government of West Germany in April 1984. Even then, a lot of work had already been done towards building this world-class machine.
DESY was founded in 1959 to house a 6 gigaelectronvolt accelerator for particle physics in West Germany. At that time, several European countries had accelerator laboratories. They also contributed to CERN, the European Laboratory for Particle Physics in Geneva (founded in 1954 and conceived as a joint European venture).
As CERN, with its larger machines, became well-established, the general policy in countries such as France, Italy and Britain was not to expand research at their national accelerators. West Germany, however, managed not only to make the largest financial contribution to CERN (subscriptions are in proportion to GNP) but also to enlarge the facilities at DESY by building DORIS and PETRA, two electron-positron colliders.
DORIS and PETRA offered opportunities that were not available at CERN, and so attracted physicists from outside Germany to participate in the experiments. In 1979 physicists at PETRA were rewarded with the discovery of the gluon.
When physicists at DESY turned their thoughts to what they should do next, building an electron-proton collider was a natural choice. This machine would be unique, and would complement the facilities at CERN. But it was clear from the start that such a complex and costly project would have to be an international venture.
After several years of discussions, the West German government agreed to build HERA on the understanding that contributions would also come from other countries. The total cost came to DM 1050 million ( £358 million). Canada, France, Italy, the Netherlands and Britain agreed to provide components and manpower. The City of Hamburg also agreed to pay 20 per cent of the cost of civil engineering and 10 per cent of the cost of the new machine. Now HERA is ready to serve around 750 physicists from all over the world.