Nuclear Decays a Thousand Meters Underground
In 1994 I was thrown into a world of madmen, dreamers, and visionaries, a heterogeneous bunch of nuclear and particle physicists, astrophysicists, chemists, and engineers who, far away from their original areas of expertise, were setting off on a quest for the neutrino. In the fall of that year I had passed my diploma exam (the German equivalent of the MSc) more poorly than well and was now looking for an interesting research topic for my dissertation. I knew more or less what I wanted: I was fascinated by the possibility of uncovering symmetries in the fundamental laws of nature. And I wanted to work on theoretical physics, because I was more interested in the beauty of the laws of physics than in the machines employed to wrest these laws from nature. After I couldn’t find an inspiring research topic at Heidelberg University’s Institute of Theoretical Physics, I recalled something I had read when I was still studying for my exam, about a specific nuclear decay that could shed light on the masses of neutrinos. Neutrino masses that do not occur in the Standard Model are predicted in GUT theories and could yield information about the most fundamental facts of nature. This was a topic that I found enthralling even back in high school. So now, in my search for a dissertation topic, I replied to an advertisement by a research group at the Max Planck Institute for Nuclear Physics (MPIK), looking for students to work on the theoretical background of such an experiment. MPIK—technically not a part of the university—was placed by its founders on a forested hill just outside Heidelberg, in order to pacify the city population’s fear of hazards related to nuclear research.
When I first drove up that hill, the vast majority of particle physicists still assumed that neutrinos were massless. But there was noticeable unrest—stimulated mainly by GUT theorists whose models predicted neutrino masses—by astrophysicists who were looking for a suitable candidate for the dark matter in the universe, and by nuclear physicists who needed neutrino masses to understand the hints of neutrino oscillations they were finding in their underground experiments. Around this time, a poll initiated by the German Ministry for Research, looking for the most vexing problems in nuclear and particle physics, ranked as number one the question about the existence of neutrino masses.
My prospective PhD advisor at MPIK, Hans Volker Klapdor-Kleingrothaus, had, in a startling coup, gotten hold of almost eighteen kilograms of nearly pure germanium 76 from centrifuges of the Soviet nuclear weapons program. But he didn’t want to build weapons or reactors; all he had in mind was to learn about neutrino mass by observing something called neutrinoless double beta decay, in which two electrons and no neutrinos are emitted. And, unlike the weapons-capable materials produced in comparable centrifuges, this material wasn’t highly radioactive. On the contrary: It would take a hundred billion times the age of the universe before even half of the germanium nuclei would have decayed.
A decay occurring so rarely is extremely difficult to observe. The major problem here is that the natural radioactivity caused by radioactive isotopes in the materials of the environment is many times stronger than the radiation of the double beta decay itself. Without taking extreme measures to suppress it, this “background” would bury the measurement signals under piles of natural radiation. So the rare, exciting observation one is looking for has to be filtered out from uninteresting signals. It is indeed like looking for a needle in a haystack. Several strategies exist to solve this problem: One can improve the search itself; physicists do this by developing more and more advanced computer programs for their data analysis. Or one can shrink the haystack. The radioactivity of the environment is being renewed continually by the bombardment of cosmic radiation—very energetic particles that streak in from the depths of the universe and can transform stable nuclei into unstable ones. The higher the altitude the greater is this effect, since less protective atmosphere means less screening against the bombardment. So if you want to keep the precious germanium powder as pure as possible, you better not ship it by plane.
Herbert Strecker, the technician of the group, had the delicate job of carrying a few suitcases with white powder worth millions of dollars through the Eastern Bloc and across the Iron Curtain from Russia to Heidelberg. One can easily imagine the lively discussions Herbert must have had with customs officers at various borders: “Is this cocaine?” “Oh no, this is just highly enriched nuclear material”!
From Heidelberg the material was shipped on a container vessel—always stored as deep in the hull as possible below the waterline—to the United States, where it was converted into crystals usable in the experiment that was to be situated in the tunnels of the Italian Gran Sasso Laboratory. The experiment itself is a nice example of how sophisticated laboratory equipment has to be in order to reduce the background haystack and attempt to see through the neutrino’s game.
Whoever wanted to visit the Heidelberg-Moscow experiment first had to drive into the ten-kilometer-long highway tunnel connecting the Adriatic coast with Rome (see Fig. 8.1). The tunnel extends under Europe’s southernmost glacier in the Abruzzo mountains, on a peak of which Mussolini was imprisoned before he was freed by German paratroopers. During the drilling of the tunnel, seven people died when an aquifer was encountered. Half way through the tunnel, now 1,400 meters below blue sky and snowy peaks with grazing sheep and sleepy villages where Pecorino cheese is manufactured, you take a turn and come to a stop in front of a large metal gate. You feel just like James Bond when you say the password into the intercom, the gate opens slowly with an orange warning light flashing, and armed security personnel ask for your ID card. Then you enter the largest underground lab in the world, made possible only by the unflagging commitment of Italy’s particle godfather, Antonino Zichichi, descendent of a 700-year-old Sicilian family and on friendly terms with the pope and the Italian Christian Democratic party and, according to rumors, even having admirers in the Mafia. There are three huge halls, each a hundred meters long, eighteen meters high, and twenty meters wide, connected by long corridors. It is cold and dark, and water drips from the rock walls. In almost twenty different experiments, separately enclosed, researchers here look for neutrinos, rare nuclear decays, and dark matter. The reason for this whole endeavor being underground is, again, the cosmic radiation: While at the earth’s surface 400,000 muons per hour and per square meter volley down on you, inside the Gran Sasso lab, shielded by 1,400 meters of rock, there is only one muon per hour per square meter left.
Figure 8.1. Map of the Gran Sasso lab. (Courtesy Gran Sasso National Laboratory)
In comparison with some of the other experiments—for example the MACRO experiment searching for magnetic monopoles and for supernova neutrinos from the explosions of burned out stars, with its enormous dimensions of seventy-six by nine by twelve meters—the Heidelberg-Moscow experiment was not very impressive on first sight. The detectors had the approximate size of beer cans and were hidden in boxes about a cubic meter in size. But as boring as the experiment seemed to be from the outside, the details of the setup were exceedingly sophisticated: In order to get rid of the effects of even the one muon hitting the detector every hour or so, the boxes were covered with detectors of a special kind, whose function was to indicate when a muon penetrated the setup so that the data from that moment could be discarded. The source material, as I’ve already described, was ultra-pure germanium, as free of contamination as anything ever manufactured. The purpose of its purity was to assure that it contained almost no radioactive nuclei of other elements that could trigger the detector with confusing signals. To avoid contamination from the detector material, the scientists used another trick. Just as the ice-cream cups at McDonald’s can be eaten along with the ice cream to avoid unnecessary waste, the detectors were identical to the source, meaning that they would detect their own decay. Then the detector was surrounded by lead shielding in order to screen out the radioactivity of the surrounding material—and not normal but ultra-pure lead. (Some competing experiments even used lead from the keels of ancient Roman galleys that had lain for more than 2,000 years on the bottom of the sea, largely unaffected by the activating effect of cosmic radiation.) Finally, each experimental box was connected to several nitrogen gas bottles that “rinsed” the whole setup and thus prevented radioactive gas particles that escaped from the tunnel’s walls from migrating into the detector box. As you can well imagine, the entire procedure yielded a massive reduction of environmental radioactivity: Only 0.2 events per year and per kilogram of detector material were eventually observed in the energy region of interest in the super-sensitive device. In fact, the detectors were so sensitive that as a by-product they could even indicate if they were penetrated by dark-matter particles which, according to astrophysicists, are supposed to populate the universe. These particles could hit atomic nuclei inside the detector, causing them to recoil and rip loose some electrons in the atom, which would then be recorded by the detector. This allowed the Heidelberg-Moscow experiment to also provide, apart from its quest for the neutrino mass, what was at that time the most sensitive limit on the existence of dark-matter particles.
Klapdor-Kleingrothaus was understandably proud that his experiment—the Heidelberg-Moscow search for neutrinoless double beta decay—was the most sensitive one in the world, and that it had the best chance to reveal the neutrino mass. This was true, however, only if the neutrino is a Majorana particle, as I shall now explain.
In general, a double beta decay is nothing but two radioactive beta decays occurring simultaneously in one nucleus. During ordinary (“single”) beta decay, a neutron is transformed into a proton, an electron—the “beta radiation”—and an antineutrino. Consequently, in a double beta decay, two neutrons are turned into two protons and two electrons. Typically also two antineutrinos are emitted, but possibly not in all cases. For if the neutrino is a Majorana particle, meaning identical to its own antiparticle, and if it has a mass, the following decay process is also possible: A neutron decays into a proton, an electron and—as dictated by the weak interaction—a right-handed antineutrino. Being a Majorana particle, the right-handed antineutrino is the same as a right-handed neutrino, and via its mass (see Chapter 6) it can be metamorphosed into a left-handed neutrino. This left-handed neutrino can be gulped by another neutron in the nucleus and in this way trigger the second decay (see Fig. 8.2). Consequently the entire decay proceeds without the emission of any neutrino, and the bigger the neutrino mass the more likely is such an event. And since no antineutrino or neutrino can carry any energy away, the entire decay energy flows into the two emitted electrons that are detected in the experiment. The experiment then yields a pileup of signals, each with the total decay energy; such a sharp peak at the endpoint of the energy spectrum is the smoking gun the experimenters are looking for.
But what if the neutrino is not a Majorana particle? In that case, one possible way to extract information about neutrino masses is to look for the cosmological consequences of massive neutrinos (see Chapter 10). Or one can go back to Pauli’s original motivation for proposing the neutrino in the first place: the missing energy in the nuclear beta decay. According to Pauli, this energy is carried away by the emitted neutral particle. If the neutrino, as we now call it, is massless, this energy can be arbitrarily small. But if the neutrino possesses a mass, the emitted neutrino has a minimum energy—just the energy mc2 of its mass, which it possesses even if it is not moving. In that case, it is not possible for the entire decay energy to be supplied to the emitted electron. As a consequence the energy spectrum of the electrons should not terminate exactly at the total decay energy but somewhat below it. The corresponding kink in the spectrum could be observed, and if one can determine the electron energy with sufficient accuracy, it provides information about the mass of the neutrino—whether it is a Majorana particle or not.
Figure 8.2. The different double beta decay processes. Upper picture: the decay mode allowed in the Standard Model with the emission of two antineutrinos. Lower picture: the neutrinoless decay where a massive Majorana neutrino is being exchanged between the two decaying neutrons.
In order to reach the necessary sensitivity, however, both a proper source and a powerful spectrometer are necessary. A good option for the source is the heavy hydrogen isotope tritium, which decays sufficiently rapidly (its half-life is 12.3 years) and whose decay energy is so small that if the neutrino has mass, the spectrum gets distorted significantly. The needed spectrometer is a device that uses magnetic and electric fields to measure a charged particle’s energy; first a magnetic field is used to collect and guide the electrons emitted from the source into the spectrometer and deflect them in a way that they run almost parallel to the magnetic field. The resulting parallel beam of electrons then runs against the repulsion of an opposed poled electric field. As electrons with smaller energies get reflected by the electric field and don’t reach the detector by varying the electric field, the electron energy can be determined. Searches for neutrino mass using tritium had been conducted over the years, all without finding any sure evidence of a non-zero mass. (The sensitivity of these earlier experiments was, in fact, less than was available in the more recent double beta decay experiments and from cosmology; see Chapter 10.)
To search for a very small neutrino mass using tritium decay, a much-improved spectrometer was needed. Such a machine was built between 2004 and 2006 for the KATRIN experiment in Karlsruhe, and did indeed break all records. It was ten meters high, ten meters wide, and twenty-four meters long. Within its stainless-steel hulk it contained the biggest vacuum cavity in the world. Not surprisingly, it also broke the size restrictions set by the height of German highway bridges. This meant that there was no way to transport the device by road from where it was manufactured in Bavaria to where it was needed in Karlsruhe. Instead of travelling 400 kilometers on German highways, the 200-ton colossus went on a 9,000-kilometer odyssey by boat, first down the Danube river into the Black Sea, then on three different vessels through the Bosporus, Adriatic Sea, Mediterranean Sea, and Atlantic Ocean, rounding the whole of Europe to reach the mouth of the Rhine, from where it went upstream to the small village of Leopoldshafen. It was finally loaded, with Europe’s biggest heavy-duty crane, onto a fourteen-axle flatbed trailer, which crawled at two to three kilometers per hour to the lab at the Karlsruhe Institute of Technology (KIT). Traffic islands, traffic lights, and street lighting, as well as the trolley wires of two tram lines had to be taken down to make way for the detector. Even then, only a few centimeters of space were left for the gigantic device to maneuver around obstacles (see Fig. 8.3). After several years of measurement, KATRIN is supposed to reach a neutrino mass sensitivity of 0.2 eV—a number that is interesting both in view of cosmological implications and in view of a controversial result of the double beta decay search.
Figure 8.3. Like a spaceship between the houses: transport of the gigantic KATRIN spectrometer. The village of Leopoldshafen is in a state of emergency. (Courtesy Karlsruhe Institute of Technology)
In the spring of 2004, four years after I had left the research group at MPIK Heidelberg and at about the time I received my job offer from Hawai‘i, Klapdor-Kleingrothaus was convinced that he had found the signal for neutrinoless double beta decay in his data records. The particle physics community remained skeptical, though. This was partly because the new signal was found in a largely old data set from which, in a previous analysis, a signal of this strength had been excluded. So the discovery was due not so much to new data as to a new method of analyzing old data. Moreover, there existed several pileups of signal events—so-called peaks—in the interesting part of the energy spectrum, not only the one at maximum decay energy. While the other peaks remain unexplained and thus suggest that unknown radioactive background exists, the one at the total decay energy could be explained by neutrinoless double beta decay, but not definitively so. It could also be produced in the same way as the other peaks, by the radioactive background. Moreover, at this time members of the collaboration were quarreling. The Russian members didn’t agree with Klapdor’s interpretation, and also the authors of the article announcing the discovery didn’t even agree among themselves. Klapdor-Kleingrothaus himself maintains that the change in the analysis was justified and natural because better statistics allowed a better fit to background lines. He also points to independent analyses that argue against an interpretation of the signal peak as background. Meanwhile first results of the EXO-200 experiment in New Mexico—performed by, among others, Stanford University’s Giorgio Gratta and Klapdor-Kleingrothaus’s former PhD student at the University of Alabama, Andreas Piepke—were published in May 2012. These results do not confirm the discovery claim but also don’t unambiguously rule it out. As of this writing, the situation is still unclear. There certainly are good reasons to be skeptical, while, on the other hand, no experiment is sensitive enough yet to refute Klapdor’s claim.
Major competitors in the race to confirm or disprove Klapdor-Kleingrothaus’s claim are EXO-200 in the United States, KamLAND-Zen in Japan, and CUORE in Italy, as well as the GERDA experiment in Italy, which started to take data in 2011 and will operate for some five to ten years. GERDA was inspired by the GENIUS proposal, which rests on an idea of the MPIK underground expert Gerhard Heusser and was developed in Klapdor-Kleingrothaus’s group by Jochen Hellmig, Laura Baudis, and Bela Majorovits. The basic idea of their proposal is that the major part of the background radiation does not originate in the detectors themselves but in their environment. Since the detectors have to be cooled down in order to operate at all, the GENIUS proposal advocated the bold idea of cooling the detectors not through some separate cooling system, but by hanging them directly inside the cooling liquid. Either liquid nitrogen, as in the GENIUS proposal, or liquid argon, as in the GERDA realization, can then act simultaneously as cooling device and as shielding against external background sources. Within a few years of measurement, GERDA will probably be able to test Klapdor-Kleingrothaus’s claim of discovery.
In parallel with these developments, Klapdor-Kleingrothaus initiated a lively research activity in theoretical physics within his group, where I, as a budding theorist, could make some contributions and gain some valuable experience. A major topic that required calculations was how contributions of new physics might reveal themselves in neutrinoless double beta decay. After all, not just Majorana neutrinos but also every other new species of particle could be exchanged between the two decaying neutrons and in this way trigger the decay, so long as lepton number is changed by two units, just as it is for Majorana neutrinos. Actually Jose Valle, now at the University of Valencia, Spain, and his PhD advisor Joel Schechter of Syracuse University had shown in 1982 that neutrinoless double beta decay always implies a non-vanishing neutrino mass. Their theorem relies on the fact that in this case a neutrino can always fluctuate into its right-handed antiparticle by means of a neutrinoless double beta decay. The mass resulting from this fluctuation, however, can be tiny and need not constitute the dominant contribution to the decay.
So Martin Hirsch, the postdoc of our theoretical group and, for me, an inexhaustible source of knowledge, and Sergey Kovalenko of the Russian nuclear research center JINR, located in the once secret city of Dubna, had a feast analyzing various contributions to the decay: contributions from R-parity violating supersymmetry, originally proposed by Rabi Mohapatra and Kaladij Babu; contributions from leptoquarks; and, with Orlando Panella of Istituto Nazionale di Fisica Nucleare in Perugia, Italy, contributions of composite neutrinos. Hirsch and Kovalenko also worked out that neutrinoless double beta decays are possible in SUSY models without R-parity violation, if the SUSY partner of the neutrino, the sneutrino, has itself a Majorana property. In 2006 Hirsch and Kovalenko, in collaboration with Ivan Schmidt, generalized the Schechter-Valle theorem with the consequence that the discovery of neutrinoless double beta decay would imply also the existence of arbitrary processes violating lepton number conservation, changing lepton number by two units, processes that might be accessible at colliders such as the LHC. It was a fantastic time to work on neutrino physics in this environment. In my dissertation, drawing on an idea of Kovalenko and in collaboration with Hirsch, I developed a general classification of all possible contributions to neutrinoless double beta decay. In two further pieces of work I collaborated with Klapdor-Kleingrothaus, Alexei Smirnov, and Tom Weiler to analyze the mutual relations of double beta decay, tritium beta decay, and cosmology.
In subsequent years, MPIK, the Max Planck Institute in Heidelberg where I worked, became one of the most exciting places in the world for a young physicist who wanted to learn about the most fundamental theories of nature. In the adjacent building—with an all-too-evident spirit of competition with our boss—resided Till Kirsten, who was the spokesman of the GALLEX experiment, which searched for oscillations of solar neutrinos. Till Kirsten had already been the first to detect the neutrino-emitting double beta decay—the Standard Model version of the decay his opponent Klapdor-Kleingrothaus was now after—in subterranean reservoirs containing a radioactive isotope of selenium, using chemistry technology. Now Kirsten was using his expertise in nuclear chemistry for solar neutrino detection. Together with his collaborators, he had constructed an experiment that used chemical elution to extract sparse atoms of germanium from gallium, the germanium having been created from gallium atoms by neutrino capture. This experiment was sensitive to the lower energy neutrinos emitted by the sun. Since the flux of these low-energy neutrinos does not depend sensitively on exact details of the solar model that is used, the researchers hoped that they could finally find out from the comparison of theory and experiment whether solar neutrinos are indeed oscillating.
This was the situation in February 1996, when members of our research group were hanging around together in the MPIK coffee room and Jochen Hellmig slammed a Physical Review Letters article onto our coffee table and proclaimed: “The coming year will be the year of neutrinos.” In the article a research group called the LSND Collaboration, which was producing neutrinos in an accelerator located in the US nuclear weapons laboratory in Los Alamos, reported a hint of neutrino oscillations. But Jochen wasn’t right. There wasn’t a year of neutrinos about to come. Instead it was to be an entire decade of neutrinos.