The Soudan Mine opened 123 years ago
to unearth iron ore. Today physicists
have converted it into a lab where
they prospect for exotic particles.
On a gray midwinter dawn, a dozen scientists and technicians in hard hats gather outside the entrance to the Soudan Mine. Wearing heavy coats and clutching lunch bags, they step expectantly into the open cage of an 80-year-old hoist. An ancient engine chuffs as it lowers a steel gondola down the mine’s 2,341-foot-long shaft. The trip takes three minutes, but the passengers do not talk much. Instead, they listen idly to the icy air swooshing upward as the cage plunges into the guts of northeastern Minnesota’s Iron Range, away from the snowdrift-filled landscape and into a world of darkness.
The scene echoes an ordinary day at the mine more than a century ago. The Soudan Mine opened in 1884 atop one of the richest iron deposits in the world—hard hematite, about 65 percent pure. The ore quality kept Soudan going until 1962, when U.S. Steel decided that the mine had become too deep and hence too expensive to operate. New techniques made it cheaper to process lower-quality ore somewhere else. But the great depth that condemned the mine as a commercial venture has more recently given it new life as a research outpost, the Soudan Underground Laboratory. The half-mile layer of overlying rock creates a shield of peace and quiet unattainable anywhere on the surface of the earth. Physicists now burrow down deep here, oddly enough, to mine the heavens.
The gondola reaches level 27—the very bottom—and opens onto a state-of-the-art particle physics lab carved from two high-ceilinged caves. This is the home of the Cryogenic Dark Matter Search, inevitably abbreviated as CDMS.
The goal of CDMS is to hunt down another physics acronym: WIMPs, or weakly interacting massive particles. These enigmatic, hypothetical particles are the leading suspects in the search for dark matter, the unseen bits of whatever that are thought to make up the bulk of the matter in the universe. Ordinary atomic matter may seem like everything. After all, we are talking about all the stars as well as planets, comets, moons, the Crab nebula, black holes, brown dwarfs, the Pacific Ocean, you, me, cans of soup, and the family dog—all of it. Seven decades of astronomical research says that we are missing the big picture, however. There is four times as much dark matter as there is so-called ordinary matter, and we know hardly anything about it.
Today there is little doubt that dark matter exists. “It was very difficult at the beginning to imply that there were things out there that you couldn’t see,” says CDMS project coleader Bernard Sadoulet, an astrophysicist at the University of California at Berkeley. “Modern measurement techniques have authenticated the observations, but the problem is to detect whatever it is.” Whoever finds some is a virtual lock for the Nobel Prize. Actually uncovering a sample is . . . well . . . another matter. That is what draws researchers into the bowels of an abandoned Minnesota iron mine, working long shifts in what is surely one of the most tedious, yet potentially rewarding, jobs in all of physics.
A thick metal door leads to the CDMS lab, which is built into two connected caves carved out of solid rock, each about 100 yards long, 45 feet wide, and 45 feet high. The overall effect is James Bond modern—a high-tech workshop in the middle of nowhere, complete with exotic machinery and electronics, the brilliant light of scores of fluorescent lamps, and cranes in the ceiling to ferry heavy components from one place to another. The walls have been coated with sprayed concrete, making them look something like papier-mâché. The walls don’t sweat, though, and the air is dry enough to encourage chapped lips. There is no cafeteria, which explains the lunch bags: Everyone down here is here to stay until 3:30 p.m., when the first of two hoist trips takes people back to the surface.
The space may be cavernous, but by big-science standards, CDMS is a decidedly modest project. It has the participation of about 46 scientists from 12 universities and the U.S. Department of Energy’s Fermi National Accelerator Laboratory, or Fermilab. Project manager Dan Bauer, of Fermilab, says it cost about $30 million to set up and operate the project, bare-bones money compared with, say, the $8 billion that the Europeans are spending on their new particle accelerator, the Large Hadron Collider. The current version of CDMS went online last October and will continue quietly collecting data for the next two years.
The CDMS experiment is, expressed simply, nothing more than a cylindrical icebox covered with shielding and placed in a clean room to keep naturally occurring radioactivity from contaminating it. Hidden inside are the detectors, designed to track the particles that are invisible to us but which are meandering through every inch of the universe. The detectors, too, are unexpectedly simple things, disks of germanium about the size of hockey pucks that are placed together in five stacks. The guts of the experiment are no bigger than a piano stool. Padded with insulation, the whole dark-matter detector is about 10 feet high and looks like an outsize water heater.
In theory, millions of WIMPs pass through the detector stacks every second. By their very nature, these dark-matter particles barely interact with ordinary matter, but in some rare instances one should collide in just the right way to make its presence known. Germanium, which is chemically similar to silicon, makes a particularly good target because its nucleus—a clump of 32 protons and 38 to 44 neutrons—is close to what theorists say is the lower size limit for a WIMP. As in billiards, the object being hit responds especially effectively if the cue ball (the WIMP) and the target ball (the germanium nucleus) are roughly the same size.
Work and play inside the physics laboratory
If a passing WIMP bumps directly into a germanium atom, the nucleus should vibrate, producing a tiny amount of heat. Each CDMS detector is outfitted with sensitive layers of aluminum and tungsten designed to record that minuscule heat signal. At the same time, the dark particle should also jar loose some electrons from the germanium atoms, triggering an electric charge that will be recorded by an electrode. The ratio of charge to heat tells researchers whether the particle struck the nucleus, and therefore might be a WIMP, or if it is just a rogue electron or some other familiar particle that is simply stirring up the atomic neighborhood.
This whole process is unimaginably sensitive. That is why the detectors are kept in an icebox chilled to just 70 thousandths of a degree Fahrenheit above absolute zero. To understand how cold this is, remember that even in the depths of space, residual radiation from the Big Bang keeps things at a relatively toasty 4.86 degrees Fahrenheit above absolute zero, making the inside of the CDMS detector one of the coldest places in the entire universe. At that temperature, vibrations in the germanium crystals virtually cease, eliminating jiggles that could interfere with a WIMP signal. “The WIMP is like a pebble dropping into a pond,” Bauer says. “It’s hard to see the ripples on a windy day, so you want the water to be as quiet as possible.”
The need for quiet also explains all the shielding around the detectors: successive blankets of polyethylene and lead. The lead—15 tons of it—blocks 99.995 percent of gamma rays that would normally overwhelm the detectors. In fact, even ordinary lead generates its own radiation. Instead, much of the detector’s flak vest is made of ancient lead obtained from a 2,000-year-old Roman shipwreck. It takes a few hundred years for radioactive leftovers in the metal to fully decay, so there is a brisk market in old lead for experiments like CDMS. Ancient lead sells for $40 per pound, about 40 times the going rate for ordinary lead.
As the data accumulate at roughly 10,000 particle detections per day, the team has a physicist on-site 24-7 whose job it is to watch the refrigerator, make sure nothing breaks down, and monitor the experiment for the one telltale event that could finally prove WIMPs really exist. The crew also swaps out empty liquid nitrogen and liquid helium bottles for new ones every other day, a nerve-racking job that, if botched, can end up bleeding the cryogenics more quickly than the new bottles can replace them. The icebox takes weeks to rechill. “I wouldn’t call it tedious; we’re certainly excited about the big picture,” Bauer says. He acknowledges that “the hardware stuff” also appeals to the inner Erector set in all physicists. Nevertheless, he concedes the job can be wearing. It carries the full burden of command without the promise of any interim eureka moments. To blow off steam and pass the time, staff members play rousing bouts of Ping-Pong on the lab floor in the shadow of inflatable palm trees.
Based on earlier experiments and a preliminary run in 2003, the CDMS team hopes to detect between 6 and 15 WIMPs over the next two years. At the same time, the workers know full well that they may find nothing at all. WIMPs are still a theoretical construct—but right now, they are physicists’ best guess about what glues the universe together.
One of the hocky-puck-size
detectors used in the CDMS
experiment, where it is shielded
from radiation and cooled to
near absolute zero.
”What we’re doing is high risk, but it’s a good place to look,” says Stanford University physicist Blas Cabrera, one of the team leaders. “Of course, it’s always possible that nature chose something else.”
The first hints of dark matter were found as far from the depths of the Soudan Mine as you can imagine: in the Coma galaxy cluster, about 320 million light-years away. In 1933 Caltech astrophysicist Fritz Zwicky observed that outlying galaxies in the cluster are moving so rapidly that the whole group should fly apart. The gravity of all the stars in the cluster is not enough to hold it together, so something else had to be binding the cluster. It didn’t show up in photographs, even in silhouette, but there had to be a lot of this mysterious dark stuff—more than 10 times the mass of all the stars—to keep the Coma cluster from spraying galaxies all over the cosmos.
Zwicky’s explanation was not very popular at the time. “Scientists don’t like things they can’t see,” says the CDMS’s Sadoulet. There was also the problem of Zwicky himself, known for contemptuous sarcasm and vituperative rages that alienated students and colleagues alike. He reputedly derided the astronomers at Mount Wilson Observatory as “spherical bastards” because they looked like bastards from any side. “Fritz Zwicky was ahead of his time,” says University of Chicago astrophysicist Michael Turner, a leading dark-matter researcher. “But he did not play well with others.”
Still, no one could disprove or resolve Zwicky’s “missing matter” problem. And when Vera Rubin, an astronomer at the Carnegie Institution of Washington, showed in the 1970s that there was a matter deficit not only in galaxy clusters but also in individual galaxies, interest perked up. “Today there’s no lack of evidence for dark matter, but not then,” Turner says. “Vera Rubin really made a strong case, and that’s when people began seriously to look at it. Individual galaxies brought the problem much closer to home.”
The next step was to figure out what dark matter was. The only thing scientists knew about it was that it neither emitted nor absorbed light. Maybe it was just large accumulations of dim but familiar objects, like extremely faint red stars or white dwarfs, some astronomers speculated. Then around 1980 a team of Soviet researchers claimed that neutrinos, evanescent particles ubiquitous throughout the universe, might have enough mass to make up the matter deficit. Perhaps this was dark matter.
Further research showed that neutrinos were not the answer, but the Soviet experiment had “changed the conversation,” Turner says. Scientists began to regard dark matter not only as a dilemma for cosmologists but also as a problem in particle physics. During the 1980s cosmologists used other techniques to compute the amounts of matter in the universe more precisely, leading to a growing realization that there were not enough atoms in any form to provide the needed gravity. Whatever dark matter was, it did not consist of stars, gas, or any kind of known object.
In the late 1990s this picture grew even more confusing, when scientists found evidence of another dark entity, called dark energy, that seems to be causing the universe to expand at an ever-accelerating rate. Fortunately, a pioneering space mission called the Wilkinson Microwave Anisotropy Probe (WMAP) clarified this muddle by delivering the first accurate account of the overall makeup of the universe. The answer is decidedly strange. Dark energy makes up 73 percent of the universe, dark matter another 23 percent. Atomic matter—everything around us and everything astronomers have ever seen—accounts for just 4 percent.
“We have an extraordinary claim—that the universe is not primarily star stuff. We want to know what it’s made of.”
Cabrera compares the makeup of the cosmos to a lit Christmas tree. “The lights are the stars,” he says, but the tree—representing dark matter and dark energy—is unseen behind the lights, and no one knows whether it’s a white pine, a spruce, or a cedar. "So here we are with a big problem,” Turner says. “We have an extraordinary claim—that the universe is not primarily star stuff. We want to know what it’s made of.”
Today cosmologists are convinced that the answer lies in physics theory, which predicts the existence of fundamental particles that have not yet been discovered, many with the right attributes for dark matter. Sadoulet became interested in the dark-matter hunt after reading a paper in the mid-1980s suggesting that the particles were massive. If so, he realized, “we could detect them by elastic scattering” of a target nucleus. “That was the first part of the revolution.” A second paper suggested that these particles would have a very slow interaction rate—weak interactions, in other words. “Then the whole thing came together,” he says: weakly interacting massive particles, or WIMPs.
So what exactly is a WIMP? In the current physics models, these particles interact through gravity but have little response—or none at all—to the other forces. There are a few dark-matter candidates besides WIMPs, but all of them have to be similarly slow-moving to produce the kind of clumpy distribution that astronomers infer from dark matter’s gravitational pull. Because these particles rarely interact with ordinary matter, they would have to be electrically neutral and hence invisible to light, X-rays, or any other form of electromagnetic radiation. Theorists suggest that WIMPs are anywhere from 40 to 1,000 times as massive as protons. Clouds of these particles seem to embrace galaxy clusters in a large sphere, and they seemingly move unimpeded through the universe. Billions may be passing through your body every instant, but since they seldom hit anything and have no charge, they would provide no evidence of either their arrival or their departure. They are called “dark” for a reason.
What makes WIMPs the most attractive of the dark-matter candidates is that their existence would elegantly resolve not only the missing mass problem but also a shortcoming in the standard model of particle physics. That model describes all the particles and forces in the universe but does not adequately explain the theory of mass. To fix this, theorists have introduced the concept of supersymmetry, in which each matter particle, known as a fermion, and each force particle, known as a boson, have large-mass counterparts, called superpartners.
Click image to enlarge
Collisions between dark-matter particles
known as WIMPs, and ordinary atoms
should be detectable because only WIMPs
and neutrons can pierce an atom's cloud
of electrons with enough momentum to make
nucleus vibrate, creating a tiny amount
of heat. A neutron will rebound hitting
multiple atoms, while a WIMP
will interact with just one atom
producing a telltale signal.
Image courtesy of Michael Attisha
In this framework the lightest neutral superpartner particle—with a mass at least 50 times that of a proton—is the neutralino. Particle physicists predict that the number of neutralinos left over from the Big Bang roughly matches the expected amount of dark matter, making the neutralino the prime WIMP candidate. “In the 1980s you show that dark matter is really there,” Turner says. “Then you describe the particles that could explain the discrepancies in the standard model, and then you find that those particles have the right properties to be dark matter. That can’t be a coincidence.”
There are now about 10 teams of scientists huddled in underground labs around the world trying to catch a WIMP. One of them, at the University of Rome, even claimed success in 2000, sparking a brief outbreak of elation in the physics community—until no one else could confirm the result. All the experiments rely on the same principle: detecting WIMPs on the rare occasions when they collide with an atomic nucleus. The hunt for the WIMP is a global competition, and competition being what it is, some other physicists think they have come up with a better WIMP trap. Most promising are detectors that use liquid noble gases—unreactive gases like xenon and argon—that will emit a flash of light when a WIMP strikes.
Richard Gaitskell of Brown University, a former CDMS team member who currently participates in XENON, a liquid gas experiment in an underground laboratory in Italy, says that inert gases have a couple of advantages over CDMS. Since the experiment doesn’t rely on detecting vibrations, the XENON team doesn’t need to worry about quieting the target atoms by cooling them to near-absolute zero. Xenon, a gas at room temperature, turns to liquid at a relatively balmy –148°F. There is no need to go lower. Also, as a single nearly spherical liquid target, the xenon system has a much smaller surface-to-volume ratio than multiple germanium detectors and therefore is less susceptible to noise. Finally, scaling up the size of the detector, and thereby enhancing the probability of success, is much easier with a liquid. “We’re capable of not only equaling but also surpassing CDMS,” says Gaitskell. “You need multiple technologies so you can check each other’s results.”
The CDMS researchers are conceding nothing. “Right now we are better than anyone else,” says Sadoulet, although he acknowledges, “maybe not for long. At the moment, the noble liquids are breathing down our necks.” And Soudan will not be the last word in the CDMS team’s dark-matter search. Sadoulet and company are already hatching plans to build larger detectors and take them even deeper—6,800 feet down in the Sudbury Neutrino Observatory in Ontario, Canada.
“Every time we have a new stage, we’re hoping to find WIMPs, but it’s really only in the last year that we have really enhanced our chances,” Sadoulet says. If CDMS finds dark-matter particles, “there will be a rush to increase the mass of detectors and cross-check with other technologies to make sure they are WIMPs.”
Even with just a few detection events—“maybe five,” Sadoulet suggests—CDMS should be able to make a lower limit estimate of WIMP mass and get an idea of the interaction rate, giving us a first glimpse at the properties of the cosmic dark stuff. Above all, a detection would prove that physicists’ dark-matter theories are on track. Sadoulet sees good odds that it will happen within the next couple of years.
Then a new hunt will begin. By building larger detectors to get more hits, researchers will be better able to characterize WIMPs. They will also want to know how dark-matter particles move through our planet and our galaxy. “You’ll want to look at directionality, see where the WIMPs are coming from,” Sadoulet says.
The practical value of finding dark matter would probably be minimal, he admits. The philosophical implications, on the other hand, would be mind-boggling. “If you found one, you’d want to find more,” Sadoulet says. “I would like very much to know what makes up most of the universe.”