IF YOU want to outwit a neutrino, you have to think up some cunning tricks.
These baffling particles are extraordinarily difficult to detect, and lighter
than any weight you can measure. What鈥檚 more, it has come to light over the
past decade that they may lead double lives, turning into their sibling
particles in an instant.
To find out for sure what neutrinos are up to, physicists at the world鈥檚
most powerful particle accelerators now plan to send streams of neutrinos
skimming under our feet for hundreds of kilometres 鈥� across Japan, from
Illinois to Minnesota, and from Switzerland to Italy. If neutrinos do turn out
to metamorphose from one type to another, it will prove that at least one type
of neutrino has finite mass. This could help to relate all the known forces of
the Universe in one grand unified theory, and could also throw some light on
one of cosmology鈥檚 greatest puzzles.
Advertisement
Tricky trio
Nature appears to supply three types of neutrino, all of which belong to
the lepton family of elementary particles. The electron, the muon and the tau
are the charged members of the lepton family, and the three type of neutrinos
are their uncharged partners.
Neutrinos created in reactions either with electrons or when electrons
disappear are 鈥渆lectron neutrinos鈥�, which can in turn give rise only to
electrons. Similarly, muon neutrinos occur only with muons. A similar close
relationship should also hold between the tau neutrino and the charged tau
particle.
Like all the leptons, neutrinos appear to have no internal structure. But
neutrino masses, if they have mass at all, have so far proved too small to
measure. Neutrinos are also very hard to trap because they only respond to the
weak force 鈥� the force responsible for radioactive beta decay and the basic
nuclear reactions that fuel the Sun. Indeed, the interaction between neutrinos
and other matter is so slight that they can pass right through the Earth
without leaving a trace.
This is why detecting such particles seems like such an impossible quest.
But if a large enough number of neutrinos pass through a big enough detector,
there is a good chance that a very small but predictable fraction of them will
interact with other matter. This becomes more likely as the neutrino energies
increase. In the process, electron and muon neutrinos can produce particles
that leave recognisable tracks. Tau neutrinos have so far slipped the net,
however. This is mainly because their cousins, the tau leptons, are much
heavier than the electron and the muon. They appear only in higher energy
collisions, so the large numbers of high-energy tau neutrinos needed for just
one to register in a detector are not so easily created.
There鈥檚 certainly no shortage of neutrinos, however. Every cubic centimetre
of space contains around 300 left over from the big bang, to say nothing of
the multitudes of neutrinos produced by stars as they burn and when they die
in supernovae. Each second, about a million electron neutrinos escape from
each square centimetre of the Earth鈥檚 surface, produced in the natural
radioactivity of the rocks beneath our feet. Muon neutrinos are born when
cosmic rays 鈥� high-energy particles from outer space 鈥� smash into the upper
atmosphere. About 30 muon neutrinos per square centimetre arrive at ground
level every second with the debris of these collisions.
Experiments with these electron and muon neutrinos have uncovered two
puzzling paradoxes over the past two decades. First, experiments that detect
neutrinos from the Sun have shown that fewer arrive at Earth than we would
expect (鈥淗igh noon for solar neutrinos鈥�, 15 August 1992, p 28).
Astrophysicists predict that the Sun should yield an enormous 60 billion or so
electron neutrinos at each square centimetre of the Earth鈥檚 surface every
second. Yet solar neutrino experiments detect only 60 to 70 per cent of that
number.
Cosmic mystery
A second mystery comes from experiments that seek out muon neutrinos
produced by cosmic rays. The cosmic rays generate showers of short-lived
particles including the charged pion, which decays to produce a muon and a
muon neutrino. The muon then decays to give an electron, along with a muon
neutrino and an electron neutrino. So by simple arithmetic, the muon neutrinos
reaching the Earth鈥檚 surface should outnumber electron neutrinos by 2 to 1.
Research groups in Japan and the US, however, have found a smaller ratio of
muon-to-electron neutrinos, near to 1 to 1.
Both of these mysteries could easily be resolved if somehow some of the
neutrinos change type 鈥� from electron to muon, or muon to tau 鈥� as they
travel. Some of the solar neutrinos would then fail to register in traps set
on Earth, which so far have been designed to capture only the anticipated
electron neutrinos. Similarly, some of the atmospheric muon neutrinos could
change into electron or tau neutrinos before they reach the detectors.
This behaviour would not be as strange as it sounds. Physicists already
know that particles called neutral kaons regularly turn into their mirror
image particles, or antiparticles, and then switch back to their original
stage.
But how could neutrinos change type? The idea is that they may not be
distinct particles with a single mass, but are instead mixtures of at least
two, or perhaps all three different neutrino quantum states. Quantum mechanics
describes a particle as if it were a wave where the frequency of the wave
depends directly on the mass of the particle. So a mixture of two different
mass states would contain two frequencies, and would continuously change as it
travels. This is similar to 鈥渂eating鈥� of sound waves, where two waves at
different frequencies combine to give a noise that pulsates between loud and
soft.
Altered states
In this way, different types of neutrinos could 鈥渙scillate鈥� from one to
another as a mixture of neutrino states travels.(see
Diagram) Provided the basic states have different masses, a muon
neutrino produced high in the atmosphere could transform into an electron
neutrino after a certain distance. And after the same distance again, the
mixture would change back into a muon neutrino. Or perhaps all three neutrinos
could oscillate from one to another.FIG-20213801.jpg
If neutrino oscillations do occur, this would prove that at least one of
the three basic neutrino states has a small but significant mass. This in turn
will have important implications for attempts to construct a grand unified
theory that will blend together the strong, electromagnetic and weak forces of
nature as different manifestations of the same basic force. Most of these
theories require neutrinos to have some mass, although the masses suggested
are usually very small.
On top of that, proof that neutrinos have a finite mass could help to solve
a big cosmological problem. Certain models of the Universe, in which it has
just enough mass to stop it expanding forever, can only be correct if most of
the mass consists of exotic particles (鈥淪ome of our Universe is missing鈥�, 8
July 1995, p 30). If neutrinos do have mass, then the huge numbers of
neutrinos surviving from the big bang could account for some of the invisible,
exotic matter that some physicists are eager to find.
To find out whether oscillations really are at work, researchers have to
try to match the experimental data with the theoretical picture. They know
that the distance over which a neutrino changes from one type to another must
depend on the masses of the basic neutrino states. The bigger the difference
in masses between the neutrinos in the mixture 鈥� that is, the bigger the
difference in frequencies of their associated waves 鈥� the more rapid the
oscillation between types. Strictly speaking, the oscillation distance would
depend on the difference in the squares of the masses. The rate of oscillation
must also depend on the speed at which the particle mixture is travelling: the
faster the particle (the higher its energy), the slower the oscillation
between types.
So what do solar neutrinos reveal about neutrino oscillations? From basic
nuclear physics, researchers know the energies that neutrinos should have when
they are born in the Sun. They also know that if oscillations are taking
place, the data imply that around 35 per cent of the neutrinos that arrive at
Earth have changed from an electron-type. Slot these figures into the
equations and everything seems fairly consistent, provided that the difference
in mass squared between the electron and muon neutrinos is around 0.00001 or
less in the traditional units of electronvolts squared (for comparison, an
electron weighs 511 000 electronvolts).
If other experiments suggested the same mass difference, it would be
powerful evidence that occillations are taking place. So, since 1992, several
research teams made similar measurements on atmospheric neutrinos. One such
team comprises the Japanese and American physicists who run the Kamiokande
detector in the Kamioka metal mine in Japan (鈥淐atch an exploding star鈥�, 26
February 1994, p 40).
Light shows
Kamiokande detects electron and muon neutrinos from cosmic rays by
registering the different patterns of flashes of light produced in water when
energetic muons or electrons emerge from neutrino interactions. As the
neutrinos can pass easily through the Earth, they arrive at the detector from
all directions. Those from cosmic rays directly above the mine come from
altitudes of around 20 kilometres. Those formed in the atmosphere on the
opposite side of the Earth travel up through the Earth over a total distance
of about 13 000 kilometres.
The Kamiokande team finds that for low-energy neutrinos, there are roughly
equal numbers of muon and electron types from all directions. For high-energy
neutrinos, however, the ratio is 2 to 1 looking up to the sky and 1 to 1
looking down into the Earth. The team concludes that this is consistent with
the idea of oscillations between two states, provided that the difference in
the square of the masses of the two neutrinos is at least 0.01 electronvolts
squared. This is a thousand times the mass difference that the solar neutrino
results imply. The difference can be explained, however, if electron neutrinos
from the Sun are changing to muon neutrinos, while muon neutrinos that form in
the atmosphere are changing to tau neutrinos. The two mass differences would
then correspond to two different pairs of neutrinos.
There are many uncertainties lurking in the data, however. For example, it
is difficult to prove just how many neutrinos of the original type come from
natural sources. The number of muon neutrinos produced by a cosmic ray depends
partly on the energy of the trigger cosmic ray, which is not known. This means
that the experimenters cannot be sure whether they are detecting too few muon
neutrinos, too many electron neutrinos, or a mixture 鈥� all could give a ratio
of less than 2 to 1. The number of solar neutrinos depends precisely on the
temperatures and pressures deep in the heart of the Sun, and again it cannot
be known exactly.
To test the idea of neutrino oscillations without these uncertainties,
physicists have turned to particle accelerators, where they can produce
neutrinos under known conditions. They can orchestrate high-energy collisions
of protons with a target to produce streams of pions. These in turn decay to
give beams of high-energy muon neutrinos. Alternatively, nuclear reactors can
be used as a rich source of electron neutrinos at lower energies. The
neutrinos can be fired at a detector some distance away, and the number of
neutrinos registered is measured.
Over the past 15 years, experiments at accelerators and nuclear reactors
have found little convincing evidence for oscillations. One seemingly positive
result came last year from a team of physicists working at the Los Alamos
National Laboratory in New Mexico, where the researchers were using a beam of
muon neutrinos. They looked for the tiny flashes of light that would be
produced if the muon neutrinos changed into electron neutrinos, which then
interacted in the detector. Sure enough, the detector did register light
patterns like those they expected.
Unanswered questions
However, the results turned out to be highly controversial. One of the
researchers in the team, James Hill from the University of Pennsylvania, made
a convincing case that other, irrelevant particle interactions produced the
flashes (快猫短视频, Science, 23 September 1995). Many other accelerator
and reactor experiments have ruled out the idea that electron-muon
oscillations take place within the limited length scales of laboratories. But
they did leave the question of oscillations between muon and tau neutrinos
unanswered. Now, two new European experiments could soon lay this matter to
rest.
Research teams at CERN, the European centre for particle physics near
Geneva, are carrying out two experiments to look for signs of muon to tau
neutrino oscillations. A detector called CHORUS sits 600 metres downstream
from a muon neutrino source. The heart of the detector consists of an 800
kilogram 鈥渢arget鈥� made of a special emulsion, akin to that on photographic
film. The task of CHORUS is to find the short-lived tau particles that would
appear if, and only if, the muon neutrinos in the beam transformed into tau
neutrinos. These would interact in the target to produce their charged tau
partners.
The lifetime of a tau particle is a fleeting tenth of a trillionth of a
second, so it can travel only 0.1 millimetres or so in the emulsion before
decaying. But the tiny tracks will nonetheless be visible in the developed
emulsion under magnification. A second detector, called NOMAD, does not detect
the tau particles directly, but can infer their presence from the decay
products.
Both the CHORUS and NOMAD teams will operate the detectors until 1997, and
are already analysing their data. They would expect to find evidence of the
oscillations, provided the mass squared difference for the two neutrinos is
greater than about 0.1 electronvolts squared. So far, however, they have found
no new evidence for neutrino oscillations. And if a team of three British
physicists is correct, then these two experiments at CERN will never see a
positive result.
Last year, Paul Harrison of Queen Mary and Westfield College, London, Don
Perkins of Oxford University and Bill Scott of the Rutherford Appleton
Laboratory made a new interpretation of all the data available. They found
that they can reconcile all the data for terrestrial, atmospheric and solar
neutrinos from 17 experiments on three continents, provided that all three
neutrinos are equally mixed together. Their results suggest that one of the
mass states weighs in at about 85 millielectronvolts, while the other two
states have masses smaller than 3 microelectronvolts. 鈥淚t鈥檚 very simple,鈥� says
Scott. 鈥淭he states couple democratically with each other, all in just the same
飞补测.鈥�
If Harrison, Scott and Perkins are correct, the neutrino paths in the CERN
experiments are not long enough to allow oscillations to take place. 鈥淭hat
will be sad in some ways for those people,鈥� says Scott. 鈥淏ut it will be great
news for our theory.鈥� The new theory, like the Kamiokande results, suggests
that oscillations of neutrinos from accelerators would take place over tens to
hundreds of kilometres, far longer than the largest laboratory on Earth.
So to test these ideas, physicists have started to think even bigger. Why
not send neutrinos through the Earth to detectors that are hundreds of
kilometres away? Three international teams are now hoping to put this idea
into action in Japan, the US and Europe.
Thinking big
The Japanese plan is to fire neutrinos into the Kamioka mine from a
distance of 250 kilometres. Along with their American colleagues, the Japanese
have already built a much bigger, improved detector in the mine. The new
detector, known as Super Kamiokande, was filled with around 50 000 tonnes of
water last autumn, and will be ready to start collecting solar neutrinos in
April. In about two years鈥� time, a neutrino source at KEK, the national
laboratory for particle physics near Tokyo, will be ready to shower the
detector with neutrinos.
A team of American and British physicists are working on a similar project.
They plan to direct neutrinos from Fermilab, Illinois, to the Soudan mine 710
kilometres away in Minnesota. Here, a new detector (MINOS) will operate with a
neutrino beam generated at a new proton accelerator at Fermilab. This is
scheduled for completion in 1999, and the MINOS team expects to start
collecting data in 2001.
The idea has also caught on in Europe. Several groups of particle
physicists are proposing to fire neutrinos on a 730-kilometre journey from
CERN to the Gran Sasso Laboratory, northeast of Rome. This underground
laboratory, near the road tunnel under the Gran Sasso massif, is already home
to cosmic-ray and solar neutrino detectors. At a workshop at Gran Sasso last
November, researchers working with two of these detectors seemed keen to study
neutrinos from CERN. Three other teams described proposals for new detectors
for the long-range experiments.
Researchers believe that these giant laboratories could resolve the
question once and for all. The experiments will be sensitive to neutrino
oscillations if the difference in mass squared is roughly 0.001 to 1
electronvolt squared, which conveniently straddles the value suggested by the
atmospheric neutrino mystery. The answers to the neutrino mass problem could
be tantalisingly close. By the end of the century, some of the world鈥檚 most
intriguing particles may have given up their baffling reputation for
good.