board77

An hour north of Duluth, Minn., and a half-mile down, the dim tunnels of the Soudan mine open up to a bright, comfortably warm cavern roughly the size of a gymnasium, 45 feet high, 50 feet wide, 270 feet long.

Well hidden from the lakes, pine forests and small towns of northern Minnesota, the mine churned out almost pure iron ore until it closed in 1962. Today, it is a state park, and it houses a $55 million particle physics experiment that is part of a worldwide effort to unravel the secrets of the neutrino, one of the least known and most common elementary particles.

Because of discoveries over the past decade, the ubiquitous neutrino, once a curiosity in a corner of particle physics, now has the potential to disrupt much of what physicists think they know about the subatomic world. It may hold a key to understanding the creation of hydrogen, helium and other light elements minutes after the Big Bang and to how dying stars explode.

The experiment at Soudan will measure the rate that neutrinos seemingly magically change their types, giving physicists a better idea of the minute mass they carry. An experiment at Fermilab outside Chicago is looking for a particle called a "sterile neutrino" that never interacts with the rest of the universe except through gravity.

Astrophysicists are building neutrino observatories in Antarctica and the Mediterranean, which will provide new views of the cosmos, illuminating the violent happenings at the centers of galaxies, distant bright quasars and elsewhere.

The particle is nothing if not elusive. In 1987, astronomers counted 19 neutrinos from an explosion of a star in the nearby Large Magellanic Cloud, 19 out of the billion trillion trillion trillion trillion neutrinos that flew from the supernova. The observation confirmed the basic understanding that supernovas are set off by the gravitational collapse of stars, but there were not enough data to discern much about the neutrinos.

The much larger detectors in operation today, Super-Kamiokande in Japan, filled with 12.5 million gallons of water, and the Sudbury Neutrino Observatory in Canada, would capture thousands of neutrinos from a similar outburst.

Because neutrinos are so aloof, successful experiments must have either a lot of neutrinos, produced en masse by accelerators or nuclear reactors, or a lot of matter for neutrinos to run into. Given the cost of building huge detectors, scientists are now turning to places where nature will cooperate.

In Antarctica, the IceCube project will consist of 80 strings holding 4,800 detectors in the ice, turning a cubic kilometer of ice into a neutrino telescope. Fourteen European laboratories are collaborating on a project called Antares that will similarly turn a section of the Mediterranean off the French Riviera into a neutrino detector.

The Soudan experiment takes the other approach, using bountiful bursts of neutrinos generated by a particle accelerator. Shoehorned into the back of the underground cavern is a detector of modest size, a mere 6,000 tons, consisting of 486 octagonal steel plates standing upright like a loaf of bread. Each plate, 1 inch thick and 30 feet wide, weighs 12 tons.

On a visit to the cavern last month, William H. Miller, the laboratory manager, pointed at the far rock wall. "Fermilab, that way," he said. This experiment is intended to catch just a few of the neutrinos created at Fermilab, 450 miles away, which gush out of the rock wall, through the cavern, through the steel plates and then through another several miles of rock before emerging out of the earth and continuing into outer space, having no effect on Dr. Miller or the reporter interviewing him.

"You need a light-year of lead to reliably stop" a neutrino, said Dr. Alec T. Habig, a professor of astrophysics at the University of Minnesota at Duluth and the operations manager for the neutrino detector.

Only occasionally, a neutrino runs into a proton or neutron among the many atoms in the steel plates, and the wreckage of that collision is recorded as tiny bursts of light careering through the detector.

When the experiment begins running at full speed later this year, Fermilab will send trillions of neutrinos every couple of seconds flying toward Soudan. The beam will spread out to half a mile wide by the time it reaches Minnesota, so most of the neutrinos will miss the cavern entirely. But even among those that strike the bull's-eye, only one every few hours will actually hit something in the detector and be detected.

So far, in the testing phase in the past two months, the Soudan detector has seen just three, maybe four, neutrinos from Fermilab. But then, Wolfgang Pauli thought physicists would never see any. Pauli, a pioneer of quantum theory, contrived the notion of neutrinos in 1930 to explain the disappearance of energy when unstable atoms fell apart. Pauli said the missing energy was being carried away by an unseen particle.

"I've done a terrible thing," Pauli wrote. "I have postulated a particle that cannot be detected." Pauli even wagered a case of Champagne that his particle would not be detected. In 1956, Pauli sent a case of Champagne to Clyde L. Cowan Jr. and Frederick Reines, two American physicists who proved him wrong using the flood of neutrinos produced in a nuclear reactor.

Physicists later discovered that neutrinos come in three types, whimsically called flavors. The flavor seen first was the electron neutrino, which interacts only with electrons. Heavier electronlike particles known as muons and tau particles are accompanied by their own flavors of neutrinos.

In 1998, an experiment at Super-Kamiokande showed that neutrinos change flavors as they travel along. For that to occur, the laws of physics dictate that the neutrinos, which had been thought be massless, must actually carry along a smidgen of weight, less than a millionth as much as an electron, the next lightest particle. Each flavor also has a slightly different mass.

In the Fermilab-to-Soudan experiment, the neutrinos are generated from a beam of protons, which are directed down a newly built $125 million tunnel, focused to a very narrow width with powerful electric fields and then slammed into a piece of graphite. That produces short-lived particles called pions, which in turn generate muon neutrinos as they decay.

The beam passes through a smaller version of the Soudan detector, allowing the physicists to verify the number of neutrinos. The tunnel, sloped downward three degrees, ends just beyond the detector. The neutrinos keep going, into the earth, to emerge in the Soudan cavern one four-hundredth of a second later.

By using neutrinos created in an accelerator, physicists will be able to vary the energy of the neutrinos and see how that changes the number detected at Soudan.

"This wiggle has not really been seen," said Dr. Boris Kayser, a physicist at Fermilab. "It is one of the central expectations of our picture of neutrino oscillations."

The data from Soudan is expected to refine the Super-Kamiokande results, not overturn the prevailing wisdom. Another experiment at Fermilab may do just that.

A decade ago scientists at Los Alamos National Laboratory in New Mexico looked at neutrinos traveling a short distance from a nuclear reactor and saw indications of a large oscillation, suggesting a relatively large mass gap between two of the neutrino flavors. The gap was large enough that it could not fit into any theory consisting of just three neutrinos. But other experiments showed that only three flavors of neutrinos that interact with ordinary matter exist.

That has led to speculations of a new class of particles called sterile neutrinos. These particles would exert a force on other matter through gravity but would otherwise be completely inert.

"If it's really due to oscillations, then it implies physics way beyond the Standard Model," said Dr. William C. Louis of Los Alamos, who worked on the experiment.

Dr. Kayser said most theorists "are skeptical, because it doesn't fit," yet no one can point to obvious flaws in the Los Alamos work, either. The new experiment at Fermilab, called MiniBooNE, is looking for the same effect but with a different setup, firing neutrinos into a spherical tank containing 250,000 gallons of baby oil.

(The name BooNE is an awkward contraction of Booster Neutrino Experiment. Booster refers to Fermilab's booster ring that accelerates protons, and the project's leaders added the prefix "mini" because they imagined a second, larger stage with a second detector if the current "mini" run confirms the Los Alamos findings.)

Dr. Louis, who is also one of the spokesmen for MiniBooNE, said that initial answers could be out by fall and insisted that he was not betting either way. "We're just concentrating on getting the correct result," he said, "and we'll worry about the consequences later."

The consequences may include the understanding of atom production in the aftermath of the Big Bang and in supernovas. Because neutrinos are essential to the nuclear reactions that change protons to neutrons and vice versa, they influence which elements form in what relative proportions.

"That would have profound implications for our models of the early universe and for supernovas," said Dr. George M. Fuller, a professor of physics at the University of California, San Diego. "It could change everything."

The problem is that current models that include three flavors of neutrinos do a good job of explaining the amount of hydrogen and helium in the universe. The existence of sterile neutrinos would send astrophysicists scurrying to come up with new calculations to produce the same answers.

On the other hand, current models of supernovas have trouble producing enough neutrons to form the heavier elements like uranium, and Dr. Fuller said sterile neutrinos could shift the reactions toward producing more neutrons.

Future experiments should aim at understanding other aspects of neutrinos, said Dr. Kayser, who was co-chairman of a committee that just released recommendations for future neutrino study.

For one, the neutrino oscillation findings say only that a difference in mass between the different flavors exists, but not the exact mass of any of them. The presumption is that because an electron is lighter than a muon and a muon is lighter than a tau that the same pattern should be true of the three neutrino flavors, with the electron neutrino the lightest and the tau neutrino the heaviest. But that does not have to be the case.

"It could be the other way around," Dr. Kayser said.

Physicists are also trying to learn whether an antineutrino is actually a neutrino. (Other antiparticles have opposite electrical charge. Because neutrinos are electrically neutral, nothing would prevent a neutrino from being its own antiparticle.)

Another open question is whether neutrinos play a role in the imbalance of matter and antimatter. If the early universe had contained equal amounts of the both, everything would have been annihilated, leaving nothing behind to form stars and galaxies.

Among quarks, which form protons and neutrons, physicists have observed a subtle matter-antimatter imbalance, called CP violation, in the behavior of particles known as mesons. "That CP violation is completely inadequate to explain the universe that we see," Dr. Kayser said.

So physicists suspect that there must be CP violation elsewhere and that the oddity of neutrinos suggest they could be a source. That, in turn, leads to speculation of yet more new types of neutrinos - very heavy ones that existed only in the very early universe - and the decay of those heavy neutrinos created the preponderance of matter.

Then come even wilder ideas - that neutrinos play a role in the mysterious dark energy that is pushing the universe apart or that neutrinos could be used for interstellar communication.

"Most of these ideas are of course probably wrong," said Dr. Louis of Los Alamos. "But if even one of them is right, it would be a tremendous breakthrough."

I wanted to be a physicist once. But writing fiction is so much more of an exact science...

The stream of neutrinos from that experiment passes almost directly beneath my house. So if I start acting strange, blame the neutrinos!