ÎçÒ¹¸£Àû1000¼¯ºÏ

EVERYTHING astronomers have learnt about the Universe they have squeezed
painstakingly from the electromagnetic radiation that pours down to our
planet from space. But astronomers have come to realise that the Universe
has played a cruel trick on them. There is much more out in the Universe
than meets the eye. We now know that luminous matter – the matter that emits
electromagnetic radiation which we can detect with optical telescopes, radio
dishes and so on – makes up a mere 10 per cent or less of the total mass
of the Universe. No one knows the nature of the ‘dark matter’ that makes
up the other 90 per cent.

In laboratories all over the world, the search for dark matter is hotting
up. For the group that first identifies the Universe’s major constituent
there will be instant fame and the chance to shake physics and cosmology
to their foundations. For one thing, the dark matter will have been crucially
important in the formation of the Universe’s large-scale structures, such
as galaxies. There are many possible candidates for the dark matter, such
as faint stars called brown dwarfs, black holes and ‘jupiters’. But it is
the possibility that the dark matter consists of exotic subatomic particles,
such as photinos or neutrinos or monopoles – that has excited experimental
physicists. Any search for these particles is a huge gamble, but it is a
gamble many are willing to take.

A striking feature of many experiments looking for dark matter is that
they are small, relatively cheap and can be carried out in a laboratory
by a small team in a few years. This is at a time when big particle experiments
can involve hundreds of physicists and cost hundreds of millions of dollars.

At Brookhaven National Laboratory in New York, our money is on a particle
called the axion . Our team, led by Adrian Melissinos, is hoping to catch
the enigmatic particle with the aid of a strong magnet and a series of conducting
cylinders. In 1987, our group, which numbers just 10 and also includes physicists
from the University of Rochester in New York State and Fermilab near Chicago,
was the first to publish results of a search for axions in our Galaxy.

We are looking for the axion in an Earth-bound laboratory because theorists
believe there may be a lot of dark matter all around us in space. They think
that, as the Universe cooled in the aftermath of the big bang, axions would
have naturally clumped around our Galaxy. The Milky Way may, therefore,
be surrounded by an extensive halo of dark matter.

Whenever astronomers talk about dark matter in the Universe, they discuss
a parameter, V. This represents the density of mass of the Universe. Astronomers
define V as the ratio of the density of the observed Universe to the so-called
‘critical density’. This is the density at which there would be just enough
matter in the Universe to allow gravity to brake its expansion. It works
out at about 10-29 grams per cubic centimetre or about 10 hydrogen atoms
per litre. If V were equal to 1, the Universe would sit precisely on the
dividing line between an ‘open’ universe, expanding forever, and a ‘closed’
universe, which will eventually begin to collapse.

When astronomers add up all the luminous matter that they can see with
their telescopes, the value of V they obtain is only about 0.1. So, at first
sight, the Universe seems to be open. When astronomers look at how fast
individual galaxies rotate about their axes and whole galaxies orbit within
clusters of galaxies, however, they find that these speeds are too great
for the amount of matter they appear to contain. They can explain their
high speeds only if there is a lot of hidden, or dark, matter, providing
additional gravitational force. When astronomers take this new matter into
their calculations, their estimate of V moves into the range of 0.1 to 0.3.

But observations are not alone in pointing to a value of V near 1. There
are strong theoretical arguments that V should be exactly 1. The most compelling
is the so-called inflationary picture of the expansion of the early Universe.
In this scenario, the Universe increased its size enormously in a minute
interval shortly after the big bang. If, at the beginning of the so-called
inflationary phase, V had differed even marginally from unity, the inflation
would have magnified this difference, implying that the present value of
V really is 1.

The most popular explanation for the anomalous motion of galaxies is
that there is an extensive halo of invisible dark matter clumped around
each galaxy. Because the dark matter is concentrated so close to each galaxy,
say the proponents of this idea, the local density of matter within the
halo is much greater than the value in the Universe at large. In the neighbourhood
of the Earth, the local density of dark matter would be 5X10-25 grams per
cubic centimetre, or about 10 000 times the critical density.

It is quite possible, however, that another component of dark matter
is necessary to make V equal 1. This additional component need not clump
in the same manner as the gravitating material in galaxies. But what could
the dark matter be? Is it baryonic – in other words, matter made up of familiar
protons and neutrons – or do we need something altogether more exotic to
fill the Universe? Detailed calculations of how the elements built up in
the heat of the big bang can shed some light on the question. Astronomers
believe that they know accurately the amounts of the lightest elements after
hydrogen: deuterium, helium-3, helium-4 and lithium-7 that were created
in the heat of the big bang. If they insert these quantities into the theoretical
models of how elements were produced in the early Universe, the calculations
serve to constrain the amount of baryonic matter present in the Universe.
The results of the calculations force theorists to conclude that protons
and neutrons contribute less than 0.15 to V.

While this just about permits a baryonic universe, there are problems
with many of the proposed forms of baryonic matter because these would emit
more radiation than astronomers observe with their telescopes. Although
some baryonic candidates, such as brown dwarf stars or ‘jupiters’ may yet
be shown to exist in large numbers, many astronomers believe it most likely
that the dark matter is not made of baryons.

You can divide the candidates for this more exotic dark matter into
two general classes. On the one hand, there is hot dark matter, composed
of particles that are moving at or close to the speed of light, or are ‘relativistic’.
Then there is cold dark matter which is not relativistic. A prime candidate
for the hot dark matter is the neutrino. There is a snag here. Experiments,
so far, seem to show that the neutrino has no mass. Nevertheless, it may
have a mass which is too small to be detected at the moment. This form of
dark matter has problems, however. Theorists cannot understand how structures
on the scale of galaxies could form in a universe dominated by neutrinos.
For this reason, they lean towards a universe filled with cold dark matter,
instead.

Particles of cold dark matter would be moving slowly at the epoch after
the big bang when they became separated from ordinary matter. They would
also be able to form clumps on scales the size of galaxies. Cold dark matter
would interact with ordinary matter only very weakly, perhaps solely by
gravity.

Theorists have proposed many possible candidates for cold dark matter.
Some are exotic particles required by theories which attempt to unify the
fundamental forces of nature. The leading contenders are the heavy particles
known as weakly interacting massive particles, or WIMPs, and the extremely
light axions. Axions are required to patch up the theory of quantum chromodynamics,
or QCD, which explains the strong nuclear force between protons and neutrons
. WIMPs include heavy leptons and particles predicted by so-called supersymmetric
theories which attempt to unify gravity with the other forces of nature.

Detecting dark matter directly poses an enormous challenge to experimenters.
If it can be done, it will have important consequences for understanding
cosmology and the early Universe, as well as the formation of galaxies,
and whatever new physics may also arise. The technical problems are formidable
because many of the candidate particles interact very weakly with ordinary
matter and, of course, no one has even detected them before. Experimenters
must also build different detectors for each possible type of dark matter.
So far, experimenters have concentrated on either heavy WIMPs or on the
light axions.

Theorists expect WIMPs to have a mass in the same range as atomic nuclei.
This suggests that the best way to detect them is to wait patiently for
an atomic nucleus to suffer an unexpected recoil. This is because kinetic
energy is most efficiently transferred between colliding particles when
their masses are equal. The problem is that when theorists take into consideration
estimates of how strongly WIMPs interact and their local abundances, they
predict that such events will be rare: around one event in every gram of
material being used as a detector per day. Any experiment, therefore, has
to be extremely good at filtering out confusing signals from the background,
such as cosmic rays.

One major type of WIMP detector to spot the recoil due to WIMPs measures
changes in heat. This detector is called a cryogenic bolometer and works
just like a calorimeter in a school laboratory. Near absolute zero, the
specific heat (that is, the amount of heat required to raise by 1 degree
the temperature of 1 gram of material) of a very pure material varies with
the cube of the temperature. This means that even tiny amounts of heat energy
produce large and, therefore measurable, changes in temperature. Silicon
is a promising material for a detector because the semiconductor industry
is geared up to producing large quantities of extremely pure material.

An advanced variation of the cryogenic bolometer is the ballistic phonon
detector. With this, experimenters can locate the precise position of the
signal inside the detector, and so rule out background signals.

Another class of detector is the solid-state detector. A passing WIMP
excites an electron in the detector into a higher band of energy, leaving
behind a positively charged ‘hole’. When the electron and hole recombine,
this produces a photon which can be detected. Physicists originally used
detectors of this type, made of germanium, to measure double-electron decays
in atomic nuclei. Several research teams have already used them to set limits
on some dark matter candidates, such as heavy neutrinos.

The problem of detecting axions is a challenge of quite a different
kind from that of detecting WIMPs. This is because theorists expect that
the mass of the axion, and, therefore, its energy is many times smaller
than that of any WIMP.

Observations of Supernova 1987A, the star of the Magellanic Cloud that
exploded two years ago, imply that the axion, if it exists, must have a
mass less than a thousandth of an electronvolt, about a million-millionth
of the mass of a proton. If it has a mass greater than this, then axions
would have cooled the supernova so much that a burst of neutrinos would
not have been seen in underground detectors. Cooling would have occurred
because massive axions would have carried away energy from the centre of
the star very efficiently. There is another limit on mass of the axion.
This comes from the requirement that, in the big bang, axions should not
have been produced in such numbers that they would cause the Universe to
re-collapse prematurely. This argument predicts that mass of the axion must
be greater than a hundred-thousandth of an electronvolt. This is the mass
that the theorists prefer, because it exactly closes the Universe. It is
also equivalent to the energy of a photon in the middle of the microwave
region of the electromagnetic spectrum, and this has important implications
for the type of experiment needed to detect axions.

Today, theorists think that axions should be clumped in the haloes of
galaxies. They ought to be very cold, with a range of velocities, peaked
around the average velocity in the Galaxy.

Working at the University of Florida, Pierre Sikivie has invented an
elegant scheme for detecting galactic axions. When an axion passes through
a strong magnetic field, it should convert into a photon. For the case of
an axion mass of a hundred-thousandth of an electronvolt, the photon will
be in the microwave region, close to 2 gigahertz (a wavelength of about
15 centimetres).

With the aid of a conducting cavity which resonates with the microwaves,
Sikivie pointed out, you should be able to detect an axion. When the resonant
frequency of the cavity matches that of the photons, it greatly enhances
the strength of the signal, just as a radio station blares out loud and
clear when it resonates with the frequency of the radio’s tuning circuit.

The major difficulty of a search for galactic axions is that theorists
are not quite sure just how heavy the axion is. It could vary in mass by
a factor of 100, which means, of course, that the frequency of the converted
photon is equally uncertain. Also, the signal will be very narrow, with
a sharp peak around the frequency of the photon, making the search similar
to looking for the proverbial needle in a haystack. So, to do the experiment,
it is necessary to build a detector with a narrow-band and ‘sweep’ it through
a large range of frequencies by gradually changing, or tuning, the resonant
frequency of the cavity. To make the task even more formidable, the power
of the signal is likely to be only 10-24 watts, about the level astronomers
expect from radio sources out in space. In fact, searching for galactic
axions with this method is rather like doing indoor radioastronomy.

Two years ago, our group at Brookhaven used a cylindrical cavity, 40
centimetres tall and 20 centimetres in diameter. It was made of very pure
copper, electrochemically polished to reduce all electrical losses. We placed
the cavity inside the bore of a superconducting solenoid magnet, a left-over
from bubble chamber experiments carried out at Brookhaven 20 years ago.
The field inside the magnet was 6 tesla, which is about 120 000 times as
strong as Earth’s magnetic field. To tune the cavity, we slowly drove a
single-crystal sapphire rod into the cavity with the aid of a motor-driven
screw. We read the signal out through the top of the cavity, amplifying
it and detecting it with a homebuilt detector with 64 frequency channels.
We read the data into a computer and reconstructed the power spectrum off
line. In total we ‘swept’ through more than 600 000 frequency bins, each
only 200 hertz wide.

Although we found a few narrow frequency bins for which the power level
was significantly above background level, we found no signals that could
have resulted from the decay of an axion. A good means of checking was to
make the same measurement with the magnetic field turned off; if the signal
remained, it could not be due to an axion. With this first experiment, we
could put limits on the coupling and abundance of axions in the small band
of frequencies that we covered (1.1 to 1.2 gigahertz, which corresponds
to 4.5 to 5.0 10-6 electronvolts for the mass of the axion). Although the
limit was 300 times larger than the most popular theoretical prediction,
we were the first to make such a measurement.

Since that time, our group has boosted the strength of our magnet to
8 tesla and extended the search to a range of between 1.1 and 4.5 gigahertz.
The sapphire rod we use for tuning has its limitations; we can tune a cavity
over only a frequency range of 10 to 20 per cent. For this reason, we need
a total of seven cavities of differing diameters to cover different parts
of the frequency spectrum. This is rather like the way in which pipes of
different diameters are needed to give a musical organ a range of tones.

This month, a conference will be held at Brookhaven in which researchers
from around the world will discuss the future of cosmic axion experiments.
The goal is to build a detector that can truly test the existence of axions
at a coupling strength given by the axion model of Michael Dine, Willy Fischer
and Mark Srednicki, and an abundance equal to the local density of the galactic
halo.

Even if no axions are found, it will have profound consequences for
both cosmology and for particle physics. Experimental physicists have been
searching for axions the past 10 years, and they would like to know whether
or not they have been searching in vain. Astronomers need to understand
how galaxies form, and why 90 per cent of the Universe is invisible. Any
simple experiment that could unite our understanding of nature on both such
small and large scales as these must be considered elegant and beautiful.

* * *

What is an axion?

AXIONS are subatomic particles which are required to make the theory
of quantum chromodynamics (QCD) consistent with experiment. QCD is the highly
successful theory of the strong nuclear force, which holds together nuclear
particles, such as the proton and neutron. The theory suffers from problems,
however. QCD predicts a ‘violation’ of parity ( P ) when particles interact.
This means that although the theory may allow a particular interaction between
particles, it does not always permit the mirror, or parity-reversed, image
to occur. Also, QCD does not permit an interaction in which a combination
of charge-conjugation times parity ( CP ) is reversed. Charge conjugation
simply involves reversing the positive and negative electric charges. The
strong force breaks neither of these ‘symmetries’.

In 1977, Roberto Peccei and Helen Quinn offered a way out of the difficulty
by adding a new variable in the equations of QCD which leads to the cancellation
of the term that broke the symmetry. Soon after, Steven Weinberg and Frank
Wilczek noted independently that the new field should be associated with
a new particle, which was named the ‘axion’. The axion is a light particle
which interacts weakly with normal matter.

Bruce Moskowitz is an assistant professor at Brookhaven National Laboratory,
Upton, New York.