TWENTY metres beneath suburban Hamburg lies the world鈥檚 most unusual microscope. For the past three years, HERA, the Hadron Electron Ring Accelerator, has been bringing high-energy beams of protons and electrons into head-on collision with each other. The results are allowing physicists to peer into the complex world deep inside the proton, where a mysterious object called a pomeron may lurk.
HERA is the latest in a series of experiments that have delved deeper and deeper into the proton. Back in 1969, physicists from the Massachusetts Institute of Technology and the Stanford Linear Accelerator Center (SLAC) in California used electrons from the 3-kilometre machine at SLAC to bombard protons in liquid hydrogen. From this they made a dramatic discovery: the proton鈥檚 charge appeared to be concentrated in lumps, rather than spread across the particle.
This was the first direct evidence for the existence of quarks, which the American theorist Murray Gell-Mann had proposed a few years earlier to be the building bricks of protons, neutrons and other, shorter-lived particles. In Gell-Mann鈥檚 model, the proton consists of three quarks: two 鈥渦p鈥� quarks with charge +鈪� each and one 鈥渄own鈥� quark with a charge of 鈭掆厯, making a net charge of +1. In 1973, researchers working with beams of neutrinos at CERN near Geneva confirmed Gell-Mann鈥檚 model of three lumps of charge in the proton, and showed that the charges really are +鈪� and 鈭掆厯.
Advertisement
There were also signs from these early experiments that the proton must contain more than just the three quarks. The quarks accounted for half the proton鈥檚 momentum at most, indicating that the proton must harbour something else as well. That 鈥渟omething鈥� turned out to be gluons, the carriers of the strong force that binds quarks to each other, and prevents quarks of the same electric charge from flying apart. But the picture is more complicated still, because a gluon can temporarily transform itself into a quark-antiquark pair, before turning back into a gluon.
Strange seas
During the 1980s, experiments at Fermilab near Chicago and at CERN probed the proton with higher-energy neutrinos and also with muons 鈥� particles just like the electron but 200 times as heavy. These experiments explored the interior of the proton in more detail, and confirmed that things there really are complex, with the three 鈥渧alence鈥� quarks of the original model caught in an ephemeral sea of quarks, antiquarks and gluons.
Now, at the DESY laboratory in Hamburg, HERA is exploring still deeper into the proton. The particle has a diameter of about 10鈭�15 metres, and the experiments at Fermilab and CERN reached down to distances of a little less than 10鈭�16 metres. But HERA can achieve far higher collision energies than before, and because particles with higher energies have shorter associated wavelengths, the experiments can probe features almost 100 times smaller, down to 10鈭�18 metres. It reaches these high energies by accelerating protons and electrons within separate rings of magnets, and then making them collide head-on (鈥淭urning the proton inside out鈥�, 快猫短视频, 30 May 1992).
At two of the collision points, large detectors, called H1 and ZEUS, surround the beam pipe to collect and record the particles thrown out from the collisions. Each detector was built by a collaboration of 300 or so physicists and engineers from a variety of institutions around the world, and the two teams are now busy analysing the flood of data they have collected. So far, the results from both detectors are showing the proton to be more complicated than ever. Paradoxically, it seems that the closer you look, the more complex the proton becomes.
One key measurement is the fraction of a proton鈥檚 momentum carried by the quark that is struck when a proton and electron collide. (Because gluons are uncharged, charged particles such as electrons, muons or neutrinos do not interact with them directly.) Different quarks may be involved in different collisions, and adding up their individual contributions can tell us just how many quark-antiquark pairs, and hence how many gluons, there are in a proton.
Low momentum
The previous experiments had shown that each of the 鈥渟ea鈥� quarks and antiquarks has relatively little momentum, no more than 1/10th of the proton鈥檚 total and possibly much less than this. If the quark-antiquark pairs have low momentum, this implies that their parent gluons also have low momentum. This in turn suggested that large numbers of 鈥渟ea鈥� quarks and antiquarks 鈥� and consequently large numbers of gluons 鈥� are involved, since they have to account for as much as half the proton鈥檚 momentum. But the surprise from HERA is just how many gluons there seem to be. HERA can measure momentum fractions more than 10 times smaller than before, down to less than 1/1000th the momentum of the proton, and the results show that there are in the region of 100 gluons overall 鈥� many more than simple extrapolations of data from the previous experiments suggested.
Is the number of gluons going to go on increasing as physicists look closer and closer? Probably not. Ultimately, at high enough gluon density, there should be a balance between the processes that form low-momentum gluons (say when a higher-momentum gluon splits into two) and the ones that form high-momentumgluons (say when the two low-momentum gluons recombine). But so far, the data from the H1 and ZEUS experiments have shown no signs that the number of low momentum gluons is levelling off in this way. Nor are there any indications of the existence of 鈥渉ot spots鈥� 鈥� small regions of high gluon-density where the recombination processes might begin.
The company of quarks
Results from HERA are extending our understanding of quantum chromodynamics (QCD), the quantum theory of the strong force (鈥淗ow to glue the quarks together鈥�, 快猫短视频, 4 December 1993). A specific challenge for QCD has come from an unexpected effect that indicates the existence within the proton of an object with properties quite different from a quark or gluon. When a quark is knocked out of a proton, it never emerges alone. Shortly after it separates from its neighbours it 鈥渃lothes鈥� itself with additional quarks and antiquarks created from the energy of the interaction. The net result is one or more particles made up of quarks and antiquarks, and it is these particles that are detected in experiments at HERA or elsewhere.
This reluctance of quarks to appear alone has to do with a property known as 鈥渃olour鈥�, which gives rise to the strong force in the same way that electric charge gives rise to the electromagnetic force. But whereas there is one electric charge (negative, say) and one anticharge (positive), there are three colours for the strong force, and three anticolours. And just as electric charges give rise to an electric field, these 鈥渃olour charges鈥� give rise to a 鈥渃olour field鈥� associated with the strong force.
It turns out that when the quarks combine to form particles they can do so only in combinations that have a net colour of zero. One way they can do this is to group together in threes, each quark with a different colour, so that the colours in effect neutralise each other, rather as the three primary colours make white. Three-quark particles of this kind, such as protons and neutrons, are called baryons. Alternatively a quark can pair with any antiquark with the appropriate anticolour, so that the colour and anticolour cancel, although other properties such as electric charge need not cancel. Such particles are known as mesons. Gluons, meanwhile, carry a combination of a colour and a different anticolour, which endows them with net colour 鈥� indeed there are eight different types of gluon each with a different colour combination. The upshot is that neither a quark nor a gluon can travel far alone before somehow surrounding itself with quarks and antiquarks to make a colourless particle.
In the high-energy collisions at HERA that are probing the proton, what usually happens is this. An electron strikes a proton and violently knocks out a quark, which then emerges in the guise of a 鈥渏et鈥� of many new particles, at a relatively large angle to the proton鈥檚 direction. The electron, meanwhile, in effect ricochets off the quark to emerge at a relatively large angle to its original direction. This so destabilises the remainder of the proton that it breaks apart, and its constituent quarks and gluons also form particles, generally travelling more or less in the original direction of the proton. However 鈥� and this is the important point 鈥� before the quark that is knocked out of the proton has formed new particles, a strong colour field must exist between the struck quark (which has colour) and the quarks left in the proton鈥檚 remnant (which must also have net colour). As the struck quark and the remnant separate, this field is in effect 鈥渟tretched鈥� until it breaks up. The energy of the field then materialises as new, colourless particles. So while new particles are formed mainly in the directions of the struck quark and the proton鈥檚 remnant, a few are also formed in the space in between (see Diagram).
But at HERA something different happens in about 10 per cent of these high energy collisions: no new particles are created in the region where the colour field should be. In other words, whatever is knocked out of the proton must have no net colour, and the residue must also be colourless. A possible explanation involves an almost forgotten object called the 鈥減omeron鈥�. This appeared on the scene in 1958, the idea of the Russian physicist Isaac Pomeranchuk, who predicted the existence of an object that has no properties except energy and momentum.
During the 1970s, as the quark model was developed into the theory of QCD, the interactions of protons and other hadrons came to be thought of as involving the interactions between individual quarks and gluons. But theorists found it difficult to fit pomerons into this picture. Although pomerons have a particularly simple set of properties 鈥� zero electric charge, no net colour and so on 鈥� they seem to behave like complicated clusters of gluons. Without a suitable experimental test bed, it was difficult to distinguish between different theoretical possibilities, and the study of pomerons declined.
But the pomeron could explain the mysterious lack of a colour field in some of the collisions at HERA. Perhaps the target in the collision is a quark (or antiquark) that has been momentarily transformed from a gluon inside the pomeron. What could be happening is that the quark is knocked violently out of the proton, but stays tied by a colour field to the rest of the pomeron 鈥� in effect a colourless 鈥渋sland鈥� within the proton (see Diagram). Because the pomeron as a whole has no net colour, the struck quark pulls the remainder of the pomeron with it, leaving behind a destabilised but still colourless proton. HERA is proving to be fertile new ground for answering questions about whether the pomeron really does exist, what exactly it is and how it fits in with the theory of QCD. Now that the H1 and ZEUS researchers know what to look for, they plan to study more closely the interactions that may involve pomerons.
Spinning tops
Meanwhile, a third experiment is due to begin collecting data at HERA later this year. This one should help to unravel yet another puzzle about the proton: where does its intrinsic angular momentum or 鈥渟pin鈥� of 陆 come from? By analogy with electric charge, it seemed reasonable that the three valence quarks should be responsible for the proton鈥檚 overall spin, and at first sight this did seem to be the case. The quarks each have spin 陆, but two of them can align like tops spinning in opposite directions, clockwise and anticlockwise. The net spin of these two quarks is zero, while the third quark has a spin of 陆, leading to an overall spin of 陆 for the proton.
However, experiments in the late 1980s, first at CERN and later at SLAC, showed that only a few per cent of the proton鈥檚 spin can be due to the valence quarks. Like the quarks, electrons and muons also have spin 陆, and the CERN and SLAC experiments involved firing beams of electrons or muons, aligned to spin in one direction, with a target of protons whose spins were also lined up, either with or against those in the beam. The results showed that 70 per cent of the proton鈥檚 spin must come from somewhere other than the valence quarks. This missing spin must somehow come from the gluons, which have spin 1, and the quark-antiquark sea. But the data available so far are not precise enough to reveal how this happens.
The new experiment at HERA is HERMES, built by a team from 10 countries and located at a third position around the accelerator ring. Like H1 and ZEUS, HERMES will study collisions between electrons and protons, but in this case the electrons and the protons will be partially polarised: a proportion of them will have their spins aligned. HERA itself will provide electrons with a polarisation greater than 60 per cent, thanks to a self-aligning effect that occurs naturally in the stored electron beam. However, to obtain polarised protons, the HERMES team will have to do without HERA鈥檚 protons, and instead use a special target at the heart of the HERMES detector. This will be a small tube containing polarised hydrogen (or deuterium or helium) gas. HERMES will obtain high polarisations in both the electron beam and the proton target and the researchers are optimistic that within a year or so they will have enough data to resolve the puzzle about the origin of the proton鈥檚 spin.
Meanwhile ZEUS and H1 continue to amass more data. The insights they are yielding look set to allow scientists one day to discover fully the nature of one of the Universe鈥檚 most ubiquitous particles 鈥� the humble proton.