The vacuum inside KATRIN's spectrometer must be kept as airless as the surface of the moon.


Weighing the universe's most elusive particle

The silvery vacuum chamber resembles a zeppelin, the vaguely Art Deco lines of the welds between its stainless steel panels looking at once futuristic and old-fashioned. One tenth the size of the Hindenburg—but still as big as a blue whale—the vessel looms in a hangarlike building here at the Karlsruhe Institute of Technology (KIT), seemingly ready to float away. Although it is earthbound, the chamber has an ethereal purpose: weighing the most elusive and mysterious of subatomic particles, the neutrino.

Physicists dreamed up the Karlsruhe Tritium Neutrino (KATRIN) experiment in 2001. Now, the pieces of the €60 million project are finally coming together, and KATRIN researchers plan to start taking data early next year. "This is really the final countdown," says Guido Drexlin, a physicist at KIT and co-spokesperson for the roughly 140 researchers working on the project.

It might seem absurd that physicists don't know how much neutrinos weigh, given that the universe contains more of them than any other type of matter particle. Every cubic centimeter of space averages roughly 350 primordial neutrinos lingering from the big bang, and every second, the sun sends trillions of higher-energy neutrinos streaming through each of us. Yet no one notices, because the particles interact with matter so feebly. Spotting just a few of them requires a detector weighing many tons. There's no simple way to weigh a neutrino.

Instead, for the past 70 years, physicists have tried to infer the neutrino's mass by studying a particular nuclear decay from which the particle emerges—the beta decay of tritium. Time and again, these experiments have set only upper limits on the neutrino's mass. KATRIN may be physicists' last, best hope to measure it—at least without a revolutionary new technology. "This is the end of the road," says Peter Doe, a physicist and KATRIN member from the University of Washington (UW) in Seattle.

KATRIN physicists have no guarantee that they'll succeed. From very different kinds of experiments—such as giant subterranean detectors that spot neutrinos from space—they now know that the neutrino cannot be massless. But in recent years, data from even farther afield—maps of the cosmos on the grandest scales—suggest that the neutrino might be too light for KATRIN to grasp. Still, even cosmologists say the experiment is worth doing. If the neutrino mass does elude KATRIN, their current understanding of the cosmos will have passed another test.

If KATRIN finds something, cosmologists will be left scratching their heads and saying, ‘Where did we go wrong?'

Licia Verde, University of Barcelona

A definitive measurement, on the other hand, would be potentially revolutionary. "If KATRIN finds something," says Licia Verde, a cosmologist at University of Barcelona in Spain, "cosmologists will be left scratching their heads and saying, 'Where did we go wrong?'"

Neutrinos first betrayed their existence through an absence. In 1914, U.K. physicist James Chadwick was studying beta decay, a form of radioactive decay in which a nucleus emits an electron, transforming a neutron into a proton. Conservation of energy suggested that the electrons from a particular nucleus, say lead-214, should always emerge with the same energy. Instead, Chadwick showed that they emerge with a range of energies extending down to zero, as if energy were disappearing.

That observation caused a minor crisis in physics. The great Danish theorist Niels Bohr even suggested that energy might not be conserved on the atomic scale. However, in 1930, the puckish Austrian theorist Wolfgang Pauli solved the problem more simply. In beta decay, he speculated, a second, unseen particle emerges with the electron and absconds with a random fraction of the energy. The particle had to be light—less than 1% of the mass of a proton—and, to avoid detection, uncharged.

Three years later, Italian physicist Enrico Fermi dubbed the hypothetical particle the neutrino. It would elude detection for another 23 years. But, in developing a fuller theory of beta decay, Fermi immediately realized that the electrons' energy spectrum holds a clue to a key property of the neutrino: its mass. If the particle is massless, the spectrum should extend up to the same energy the electron would have if it emerged alone—corresponding to decays in which the neutrino emerges with virtually no energy. If the neutrino has mass, the spectrum should fall short of the limit by an amount equal to the mass. To weigh the neutrino, physicists had only to precisely map the upper end of the electron spectrum in beta decay.

That measurement requires exquisite precision, however. For decades, physicists have striven to achieve it with tritium, the simplest nucleus to undergo beta decay. In 1949, a first study concluded that the neutrino weighed less than 500 electron volts (eV), 1/1000 the mass of the electron. Since then, successive experiments have cut the upper limit in half roughly every 8 years, says Hamish Robertson, a KATRIN physicist at UW. "There's a sort of Moore's Law for the neutrino mass," he says, referring to the trend that, for many years, described the regular shrinking of transistors on microchips. The upper limit now stands at about 2 eV—two-billionths the mass of the lightest atom—as experimenters in Mainz, Germany, and Troitsk, Russia, independently reported in 1999.

In 2001, those teams and others gathered in a castle high on a hill in the hamlet of Bad Liebenzell in Germany's Black Forest and decided to push further, by mounting the definitive tritium beta decay experiment. "That was the point of origin, the big bang for KATRIN," KIT's Drexlin says. KATRIN experimenters hope to lower the mass limit by a factor of 10, to 0.2 eV—or, better yet, to come up with an actual measurement of the neutrino mass.

Weighing a will-o'-the-wisp

Neutrinos vastly outnumber all the atoms in the universe. Yet physicists still don't know how much they weigh. For 70 years they have tried to measure the neutrino's mass indirectly, by studying a process called beta decay in tritium, a radioactive isotope of hydrogen. The Karlsruhe Tritium Neutrino (KATRIN) experiment in Germany may be physicists' last, best hope to weigh the neutrino in this way.

Tritium Energy of electron(kilo-electron volts) Mass of neutrino 0 18.6 Tritium beta decayIn tritium beta decay, a tritium nucle-us turns into a helium-3 nucleus while ejecting an electron and a neutrino. A race track for electronsIn KATRIN a high-tech tritium source produces 100 billion electrons per second and capturesthem in a magnetic field. The electrons move into a large spectrometer, which, in concert witha detector, measures their energies with exquisite precision. The heart of a giantWithin the 23-meter-long, 10-meter-wide spectrometer, the magnetic field lines guiding the electrons spread out, sculpted by magnetic coils surrounding the tank. The spreading field coaxes all the electrons to flow in very nearly the same direction. Energy of the electronThe three final particles share the energy of the decay, so the electron can emerge with any energy, ranging from zero to a high value. Neutrino massIf the neutrino were massless, then the electrons’ energy would reach a known limit set by the masses of the tritiumand helium-3 nuclei. But the electrons’ energy will fall short of that limit, and the difference reveals the neutrino’s mass. Helium-3 Electron Neutrino Electron atrest Neutrino atrest Limit Neutrinohas mass Neutrino ismassless Typicaldecay Tritium source Transportsection 1.8 m Pre-spectrometer Main spectrometer Detector Magnetic coils Vacuum chamber Stainless steel skin Magnetic field Vacuum pump Source DetectorBy varying the strength of the electric field and counting the number of electrons that make it to the detector, physicists can precisely measure the endpoint of the electrons’ spectrum. That gives the mass of the neutrino. ElectrodesThe vacuum chamber is lined with high-voltage electrodes. They shape an electric field that opposes the motion of the electrons. Those with less energy flow back toward the source. Higher-energy electron Lower-energy electron Electric force Neutron Proton

To perform the experiment, scientists need a supply of tritium—a highly radioactive isotope of hydrogen produced in certain nuclear reactors that's tightly regulated because of its potential health hazards and weapons applications. The search for it brought them to KIT, which already had a facility, unique in the Western Hemisphere, for processing and recycling tritium.

With tritium in hand, physicists then have to collect the beta electrons it emits without altering their energies. They cannot, for example, put tritium gas in a container with a thin crystalline window, because passing through even the thinnest window would sap the electrons' energy enough to ruin KATRIN's measurement.

Instead, KATRIN depends on a device called a windowless gaseous tritium source: an open-ended pipe 10 meters long that tritium enters from a port in the middle. Superconducting magnets surrounding the pipe generate a field 70,000 times as strong as Earth's. Beta decay electrons from the tritium spiral in the magnetic field to the pipe's ends, where pumps suck out the uncharged tritium molecules. Set it up right, with not so much tritium that the gas itself slows the electrons, and the source should produce 100 billion electrons per second.

Finally, physicists must measure the electrons' energies. That's where KATRIN's zeppelinlike vacuum chamber comes into play. Still riding the magnetic field lines from the source, the electrons enter the chamber from one end. The magnetic field, now supplied by graceful hoops of wire encircling the blimp, weakens to a mere six times Earth's field as the field lines spread out. That spreading is key, as it forces the electrons to move along the lines, and not around them.

Once the electrons are moving in precisely the same direction, physicists can measure their energies. Electrodes lining the chamber create an electric field that pushes against the onrushing electrons and opposes their motion. Only those electrons that have enough energy can push past the electric field and reach the detector at the far end of the chamber. So by varying the strength of the electric field and counting the electrons that hit the detector, physicists can trace the spectrum. KATRIN researchers will concentrate on the spectrum's upper end, the all-important region mapped out by just one in every 5 trillion electrons from the decays.

Everything has to be tuned perfectly. Additional coils of wire around the spectrometer must precisely cancel Earth's magnetic field, or else the electrons will run into the zeppelin's wall. The specific voltages of the myriad electrodes must be stable to parts per million. The vacuum within the spectrometer must be held at 0.01 picobar, a pressure as low as at the moon's surface and an unprecedented level for such a big chamber. And the temperature of the tritium source must be kept at a frigid 30 kelvins to slow the molecules so their motion doesn't affect the energy of the ejected electrons.

KATRIN physicists have run into some nettlesome surprises. For example, to avoid stray magnetic fields, they had the concrete floor below the 200-metric-ton chamber reinforced not with rebar of ordinary steel, which is magnetic, but with nonmagnetic stainless steel. Still, magnetic fields from the ordinary steel in the concrete walls played havoc with the spectrometer, says Kathrin Valerius, a physicist at KIT. "We had to demagnetize the building," she says, a painstaking process that required passing an electromagnet over every square meter of the walls.

Working out the kinks took longer than expected, putting the experiment roughly 7 years behind original plans. No single issue slowed it down, says Johannes Blümer, KIT's head of physics and mathematics. "Things turned out to be much more complex than we thought initially," he says. "Everything has to be perfect and perfectly stable."

The wait is almost over. Last October, physicists fired electrons from an electron gun through the spectrometer. This summer, they will calibrate it with a sample of krypton-83, which emits electrons of a fixed energy. Later this year, they will connect the tritium works, ready for next year's data taking. In a single week KATRIN should outperform all previous experiments, Drexlin says, but researchers will still need to take data for at least 5 years to make their ultimate measurement.

The long way home

The Karlsruhe Tritium Neutrino (KATRIN) experiment in Germany is the culmination of a 70-year-old quest to measure the mass of the neutrino, but the construction of the experiment involved an odyssey of its own. The heart of the experiment, a stainless steel vacuum chamber 23 meters long and nearly 10 meters wide, was fabricated by a company in Deggendorf, just a 400-kilometer drive from the KATRIN site at the Karlsruhe Institute of Technology (KIT). Yet with no direct road route that could accommodate such a big load, scientists couldn't bring their baby straight home.

Instead, in September 2006 the 200-metric-ton apparatus began a much longer journey, first sailing down the Danube River to the Black Sea. From there a ship carried it through the Aegean and Mediterranean seas, up Europe's Atlantic coast, and up the Rhine River. Finally, after 2 months, it came ashore at Leopoldshafen, a village 7 kilometers north of KIT, where a special truck inched it through the center of town, at one moment with just 5 centimeters to spare. The 8800-kilometer journey cost €600,000, says Thomas Thuemmler, a physicist at KIT, and "the last 7 kilometers cost as much as all the previous stages."


Physicists already know that neutrinos have some mass. For decades, they have known that neutrinos come in three types, or "flavors"—electron, muon, and tau—depending on the particle interactions in which they are born. In 1998, experimenters with Super-Kamiokande, a gigantic subterranean neutrino detector in western Japan, proved that muon neutrinos created when cosmic rays strike molecules in the atmosphere could change flavor as they zipped through Earth to the detector.

Such flavor changing shows that neutrinos cannot be massless. Otherwise, according to relativity, they would have to travel at the speed of light, like photons. In that case, time for them would stand still, making change impossible. But such flavor changing depends on differences in the three neutrinos' masses, not the exact values. So although physicists have studied the phenomenon in detail, they can say only that one of the neutrino types must have a mass of at least 0.05 eV. And they don't know which type is heaviest or lightest.

KATRIN may not be sensitive enough to provide an answer, results from cosmology suggest. The big bang created a sea of neutrinos that shaped the evolution of cosmic structures, slowing the formation of galaxies and galaxy clusters. By studying the cosmic microwave background, emitted when matter in the newborn universe was just beginning to clump, and the distribution of galaxies, cosmologists have inferred that the sum of the masses of all three types of neutrinos is less than 0.2 eV, right at the limit of KATRIN's sensitivity. If so, the electron neutrino alone—the kind emitted by beta decay—is likely to be too much of a featherweight for KATRIN.

That estimate comes with a caveat: It depends on the validity of cosmologists' standard model, which assumes that the universe contains ordinary matter, mysterious "dark matter" that interacts only through gravity, and bizarre, space-stretching "dark energy," which also slows the formation of galaxy clusters. So, turning things around, by measuring the neutrino mass, KATRIN physicists can test that model, particularly its assumptions about dark energy, says Kevork Abazajian, a cosmologist at the University of California, Irvine. "If they find something that's not in agreement with what's found in cosmology, they're saying that something is broken in cosmology," he says. "That would be profound."

And if the neutrino proves too light for KATRIN to measure? Physicists envision an upgrade to boost the experiment's sensitivity to 0.1 eV. Beyond that, they would have to replace the ordinary tritium gas, which forms diatomic molecules, with atomic tritium, Drexlin says, because the splitting of the molecules when the tritium beta decays creates the biggest uncertainty in the electrons' energies. But such an atomic source doesn't yet exist and would be a major challenge to invent.

Others are exploring even more radical approaches. At UW, Robertson is testing out a different way to measure the energies of electrons from beta decay. When an electron circles in a magnetic field it radiates energy—such as the x-ray synchrotron radiation that scientists use at accelerator light sources the world over. In KATRIN, the electrons are much lower energy, so they radiate radio waves—although not enough to spoil KATRIN's energy measurements.

The frequency of those radio waves corresponds to an electron's energy, a signal that Robertson and his colleagues plan to take advantage of. In what they call Project-8, they have built a thumbnail-sized chamber that can measure the radiation from a single electron. They hope to scale it up to measure the radio emissions of billions of electrons at once. That would require a 200-cubic-meter chamber studded with a slew of small radio antennas.

Others hope to pry out the neutrino's mass by studying nuclei of holmium-163, a synthetic radioisotope, created with particle accelerators, that undergoes a process nearly the opposite of beta decay called electron capture. In it, the nucleus gobbles up one of the atom's own electrons while spitting out a neutrino. By tallying up all the heat and other energy from such captures and looking for a small shortfall at the end of its spectrum, physicists should be able to infer the mass of the neutrino, just as in a tritium experiment.

That's the goal of the Electron Capture on Holmium (ECHo) experiment at the University of Heidelberg in Germany, which embeds holmium atoms in tiny gold "calorimeters." Experimenters have made working arrays of dozens of detectors, says Loredana Gastaldo, an ECHo physicist. To produce enough decays to measure the neutrino mass, she says, "we need to go from 100 detectors to 100,000 detectors."

If KATRIN falls short of a definitive measurement and neither of the new approaches pans out, then directly weighing the neutrino could remain an unreachable goal. Physicists may have to content themselves with the indirect and uncertain mass estimates that come from the heavens. And the shy and retiring particle will have preserved its shroud of mystery once again.