¿ìè¶ÌÊÓƵ

If the Universe was born out of a hot fireball of energy, as the widely
accepted big bang theory tells us, how did that energy get converted into
the matter that we see around us today? The standard theory of matter says
that ordinary hadrons – particles such as the protons and neutrons that
make up atomic nuclei – are composed of fundamental entities known as quarks.
These are held together by the strong force carried by the whimsically named
gluons. The exchange of gluons produces a force so strong that no individual
quark can ever escape from a hadron. But under the conditions of extreme
pressure and temperature during the first split second of the Universe,
15 billion years ago, individual hadrons could not have existed. Instead,
according to standard theory, the Universe consisted of a soup of quarks
and gluons – a ‘quark-gluon plasma’ (see Figure overleaf).

The quark-gluon ‘era’ ended about one hundred-thousandth of a second
after the big bang. At this critical time, a phase transition took place,
equivalent to the way steam changes into liquid water, and hadrons were
formed. Physicists on both sides of the Atlantic have now started experiments
to probe this quark-hadron transition, providing ways of testing the theories
on which our understanding of the early Universe are based.

To get a feel for just how extreme the conditions involved are, we need
to look at temperature and density in terms rather different from those
of everyday life. Confusingly, physicists measure both quantities in terms
of what seems to be the same unit – the electronvolt. Strictly speaking,
this is a measure of energy, with 1 electronvolt equivalent to 1.6 x 10
^-19 joules, so it is a perfectly good measure of temperature.
Particles which collide with one another with kinetic energies of a few
electronvolts have a temperature equivalent to a few tens of thousand degrees
on the kelvin scale. These are the energies and temperatures associated
with ordinary chemical reactions.

Energy can be converted into a mass equivalent by dividing by the square
of the velocity of light, c ², in line with Einstein’s famous
equation E = mc², and when electronvolts are used as units of
mass, or in expressing densities, the division by c ² is taken
as read. In these terms, the mass of an electron is 500 kiloelectronvolts,
and the mass of a proton is 1 gigaelectronvolt. The neutron has almost the
same mass (actually slightly more), and the packing of neutrons and protons
together in atomic nuclei provides the greatest density of matter that can
exist in the Universe today (except for a faint possibility that hadrons
may be squeezed so hard in the centres of some neutron stars that they are
squashed into a quark-gluon soup). The radius of a proton is about 8 x 10
^-16 metres, which is near enough, for our present needs, to one
femtometre (1 fm = 10 ^-15 metres). So the density of a proton
– the ultimate density of everyday matter – is, in round terms, 1 giga-electron-volt
per cubic femtometre.

Theorists have calculated the way matter behaves at the quark-hadron
transition using a relatively new technique called lattice gauge theory,
using powerful computers (see ‘All the world’s a lattice’, ¿ìè¶ÌÊÓƵ,
5 January 1991). The critical temperature is in the range from 150 to 200
mega-electronvolts (corresponding to the critical temperature at which water
boils), according to these calculations. This corresponds to an energy density
of 2 to 3 gigaelectronvolts per cubic femtometre – enough pure energy present
in the volume of a single proton to create three protons in line with Einstein’s
equation. How can physicists create such extreme conditions?

The line of attack which is now being followed at the European particle
physics laboratory, CERN, in Geneva, and at the Brookhaven National Laboratory,
in the US, is to force beams of heavy ions to collide. Particle accelerators
routinely carry out experiments in which beams of protons or electrons (or
their antimatter counterparts) are smashed into targets containing nuclei
of heavier elements, or into opposing beams of elementary particles. But
now researchers are developing the techniques required to introduce beams
containing nuclei of very heavy elements into particle accelerators. These
heavy nuclei will then be accelerated and smashed into either stationary
targets or opposing beams containing the same type of heavy nuclei. To get
a picture of the kind of collision that will result when two such nuclei
meet head on, consider the (as yet hypothetical) example of what happens
to a gold nucleus accelerated to 0.999957 times the speed of light.

A gold nucleus contains 118 neutrons, and 79 protons; so it has 79 units
of positive charge. The charge provides the means by which the nucleus may
be accelerated to such speeds using magnetic fields. At this speed, effects
due to relativity will make the mass of the nucleus increase, while it shrinks
in the direction of motion to become flattened like a pancake. The two effects
involve the same relativistic factor, so the mass increases to 108 times
its rest mass while its thickness along the line of flight shrinks to less
than 1 per cent the thickness of a stationary gold nucleus. In round terms,
it is 100 times heavier (with a mass of more than 100 gigaelectron-volts
per nucleon), but the thickness of the pancake is now only 1 per cent of
its diameter measured across the line of flight.

If such a relativistic nucleus meets an identical nucleus travelling
the opposite way, the results will be spectacular. With (just over) 100
times as much mass in 1 per cent of the original volume for each nucleus,
the density of matter in the colliding nuclei at the moment of overlap is
more than 20 000 times the density of an ordinary gold nucleus. The same
kind of densities would be achieved in collisions involving nuclei of other
heavy elements, such as lead or uranium. And as the two nuclear pancakes
try to pass through one another, there will be repeated collisions between
protons and neutrons meeting head on, and between nucleons and the wreckage
produced by collisions that have taken place just in front of them. The
best picture of what happens then comes from physicists’ standard model
of nucleons as composed of quarks.

Each nucleon – proton and neutron – contains three quarks. But quarks
cannot exist in isolation. They come either in triplets or in pairs, and
the best way to understand this is to think of them as being held together
by a piece of elastic (in reality, an exchange of gluons) holding two quarks
together. If you tried to separate two quarks this elastic would stretch
and energy put in to separating the two quarks would be stored in a way
similar to the way energy is stored in a stretched spring.

This means that two quarks joined in this way are held together more
tightly the further they are apart – the opposite of the way in which familiar
forces like magnetism or gravity operate. But eventually, the stretched
‘elastic’ will snap – but only when enough energy has been put in to the
system to create two ‘new’ quarks, one on each side of the break.

The process is reminiscent of trying to separate a north magnetic pole
from a south magnetic pole by sawing a bar magnet in half. Every time you
break the two poles apart, you find you are left with two new bar magnets,
each with a north pole and a south pole, instead of two separated poles.

So the picture of the collision between two heavy ions moving at speeds
close to that of light is one in which quarks are ripped out of individual
nucleons, stretching the elastic joining them to other quarks until it breaks,
creating new combinations of pairs and triplets of quarks, with an over-lapping
tangle of breaking and rejoining elastic, like high-energy spaghetti. Tangled
elastic may end up joining two quarks moving in opposite directions at close
to the speed of light, absorbing large amounts of the kinetic energy of
the collision and snapping to produce a host of new particles at the site
of the collision, after what is left of the original nuclei has moved away.
This is the quark-gluon plasma that physicists are eager to study – the
‘little bangs’ which reproduce the conditions that may not have existed,
since the big bang itself.

Because particles are being manufactured out of energy (the high kinetic
energy of the colliding nuclei), it is easy to produce a greater mass of
particles from this mini-fireball than the mass of the two original nuclei.
Such particle collisions are not simply a question of breaking the incoming
nuclei apart to release their components, but are a means of creating the
high-energy densities out of which new particles can be formed. The energy
to make the new particles has come from the electromagnetic fields used
to accelerate the original nuclei.

How close are the experimenters to achieving the quark-gluon plasma?
And how will they analyse the little bangs if they do succeed in creating
them?

Existing particle accelerators were simply not built to do this kind
of experiment, and quite apart from the energy required there are other
constraints that limit the kind of nuclei that can be used. At CERN, for
example, the Superproton Synchrotron accelerator (SPS) can be operated only
with nuclei that have equal numbers of protons and neutrons, while very
heavy nuclei always have many more neutrons than protons. Working with nuclei
of sulphur-32, which has 16 protons and 16 neutrons, the SPS can reach 19
gigaelectronvolts per nucleon. This leads to energy densities which, on
average, fall below or barely reach the range needed for the formation of
a quark-gluon plasma. The Alternating Gradient Synchrotron at Brookhaven
can reach 5 gigaelectronvolts per nucleon, with silicon-28; here the average
density is most likely well below the critical value. With new booster systems
(sometimes known as pre-accelerators), both laboratories will soon be able
to handle all heavy nuclei but still at the same values of the energy per
nucleon. This means that there will be a good chance for CERN to have a
first try at making a quark-gluon plasma. Experimentalists will be able
to single out particularly violent collisions leading to a higher than average
density.

But in 1997 or 1998 Brook-haven’s Relativistic Heavy Ion Collider (RHIC)
and CERN’s new accelerator, the Large Hadron Collider (LHC) should both
become operational. Although the LHC is being principally designed to accelerate
protons, it will eventually be used to accelerate heavy ions to high energies.
RHIC will run at about 200 gigaelectronvolts per nucleon, while the LHC
should reach 6300 per nucleon. These are equivalent to energy densities
of 3 gigaelectron-volts per cubic femtometre at a temperature of 200 mega-electronvolts
for RHIC, and 5 gigaelectronvolts per cubic femtometre at 220 megaelectronvolts
for the LHC, both well in the range where, theory says, the quark-gluon
plasma should form.

The way to test whether the quark-gluon plasma has formed is to look
for a ‘signature’ that could not be produced in any other way. The particles
that are detected emerging from the little bang will, in general, be everyday
hadrons because by the time anything reaches the detectors, the original
fireball has long turned into conventional matter. Experimenters are interested
in all products of these reactions, anticipating some surprises as they
probe into the unknown. The hadrons may retain some imprint of the conditions
under which they were ‘born’. This will shed new light on the way in which
the kind of matter we are made of was created in the big bang itself.

Of particular interest, however, are particles that were emitted before
the fireball had cooled off enough to undergo the quark-hadron phase transition.
These particles would carry with them information about the hot early time
of the little bang. One such signature is very similar to what astrophysicists
use to measure the temperature and composition of distant stars. The light
that reaches us from stars is emitted from the atoms that make up the stellar
matter, and such atoms emit and absorb light of well-defined frequencies.
The hotter the star is, the more high-frequency and the less low-frequency
light it emits. The fireball produced in the high-energy collision of two
nuclei is comparable to such a star: it also emits ‘light’ which can be
observed in the form of pairs of electrons and positrons or muons and antimuons.
This light also contains spectral lines, and the analysis of such spectra
can tell us how hot the original was. Researchers at CERN have, in fact,
seen the first evidence for spectral changes with increasing temperature
in their experiments.

Another intriguing possibility is that the mini-fireballs ofthe little
bangs may produce a different kind of stable matter. Although protons and
neutrons each contain three quarks, only two types of quark are present
in everyday matter – the up and down quarks (and their antiparticles). A
proton consists of two up quarks (each with a charge of +2/3) and one down
quark (with a charge -1/3) bound together by gluon elastic, while a neutron
contains two down quarks and one up quark.

There are, however, other types of quark which take part in interactions
at high energies and combine to form more exotic, usually short-lived, particles.
One of these additional quarks, the so-called ‘strange’ quark should, according
to theory, be able to combine with up and down quarks to form droplets of
‘strange matter’ (or ‘quark nuggets’) containing roughly equal numbers of
up, down and strange quarks. There have been claims that such strange matter
has been detected in cosmic rays (‘The quarks that fell to Earth’, ¿ìè¶ÌÊÓƵ,
20 July 1991), but as yet no unambiguous evidence for the existence of such
droplets has been found. But a quark-gluon plasma could produce strange
droplets, as well as ordinary hadrons. They would have roughly zero electric
charge (the strange quark has a charge of -1/3), but masses comparable to
those of a large atomic nucleus, and would be very easy to detect, providing
a striking signature for the quark-gluon fireball.

Some researchers have suggested that it may even be possible to measure
the size and shape of the mini-fireball itself, using a technique originally
developed to measure the sizes of distant stars. The technique is named
after the two astronomers who developed it, R. Hanbury-Brown and R. Q. Twiss.
As used by astronomers, the technique involves making simultaneous observations
of the same star using two telescopes up to 200 metres apart, and combining
the light signals from the two telescopes to produce an interference pattern.
The interference pattern then provides information about the size of the
star that the light is coming from.

The Hanbury-Brown-Twiss method might be adapted to measuring the properties
of the mini-fireballs produced in heavy ion collisions, using pions instead
of light as the probe. The quark-gluon plasma should produce copious quantities
of pions, and by studying the way in which these particles interfere with
each other (remember that in the quantum world every particle is also a
wave) it should be possible to infer the geometrical properties of the fireball
they are emerging from.

It is worth spelling out just how breathtaking a leap in scale this
involves. Hanbury-Brown and Twiss developed the technique to measure the
sizes of stars (typically a billion metres across) across distances of several
light years. It is now being adapted to measure the sizes of mini-fireballs
typically less than 10 femtometres (less than 10 ^-14 metres)
across, at a distance of a metre or so. This involves a scaling factor of
23 orders of magnitude. In the same vein, the spectral analysis of stellar
matter was developed to study the properties of matter many light years
away from us, but which exists in the same form over very long times. We
now want to use this technique to measure the temperature of a mini-fireball
which lives at best some 10 ^-23 seconds.

Such a trick will not be easy, but if it succeeds it will neatly close
the circle of this investigation. Interest in creating the quark-gluon plasma
comes from its importance in cosmology (the big bang) and possibly in astronomy
(the insides of neutron stars); it will be appropriate if a technique developed
by astronomers can be adapted to provide insights into the nature of the
bubbles of quark-gluon plasma that the experimenters confidently expect
to be manufacturing before the decade is out.

This article is based on information supplied by Helmut Satz, of CERN.