New Physics Is Falling from the Skies
Beyond any doubt this experiment was a monster. A thousand meters below the rock of a holy mountain in Japan, more than ten thousand electronic eyes would peer into a tank as broad as a ballroom and as high as a ten-story building, filled with 50,000 tons of ultra-pure water (see Fig. 9.1). What the eyes were looking for were bluish flashes of light. These flashes occur when a neutrino from the sun or the atmosphere collides with an electron in one of the water molecules, accelerating it to a speed greater than the speed at which light travels in water, thus breaking the optical analog of the sound barrier.
As soon as the eyes record such a flash, they transform it into an electrical signal and send it for processing into five big containers of electronic gadgetry adjacent to the tank. After about two years of gathering data with this monster, Takaaki Kajita stood up to give his talk on the measurements of atmospheric neutrinos at the Neutrino ’98 conference. When he had finished—when he had said the words, “the Super-Kamiokande experiment has evidence for a non-vanishing neutrino mass”1—his audience was stunned. As Berkeley physicist Hitoshi Murayama (who, by the way, is responsible for the idea that the whole universe could have been created by neutrinos) remembers this moving moment: “Uncharacteristically for a physics conference people gave the speaker a standing ovation. I stood up too. Having survived every experimental challenge since the late 1970s the Standard Model had finally fallen. The results showed that at the very least the theory is incomplete.”2
Figure 9.1. The inside of the monster: the 11,200 electronic eyes of the Super-Kamiokande experiment in Japan. (Courtesy Kamioka Observatory, Institute of Cosmic Ray Research, University of Tokyo)
One day later, US President Bill Clinton delivered a speech in front of graduates of the Massachusetts Institute of Technology (MIT) in Boston: “Just yesterday in Japan, physicists announced a discovery that tiny neutrinos have mass. Now, that may not mean much to most Americans, but it may change our most fundamental theories—from the nature of the smallest subatomic particles to how the universe itself works, and indeed how it expands.”3 The fall of the Standard Model was the end of a thirty-year-old search for confirmation of Pontecorvo’s idea that neutrinos could oscillate back and forth among the three flavors of electron neutrino, muon neutrino, and tau neutrino—a fervent race for the discovery of neutrino mass.
When this story started, it was about neither atmospheric neutrinos nor a search for neutrino masses. It started with a scientist who wouldn’t give up after losing one race and looked to the sun as a means of winning another. Ray Davis had competed with Reines and Cowan for the discovery of the neutrino and, as it turned out, he had bet on the wrong horse. While Reines and Cowan were detecting antineutrinos by their ability to turn protons into neutrons and positrons, Davis relied on an idea of Pontecorvo, namely to detect the antineutrinos produced in a nuclear reactor by having them transform nuclei of chlorine atoms into nuclei of the noble gas argon and electrons. What Davis didn’t know was that this process is possible only for left-handed neutrinos and not for the right-handed antineutrinos that are produced in a nuclear reactor. And even if the neutrino were its own antiparticle, as proposed by Majorana, the transition from left-handed to right-handed would depend on the neutrino mass, and be very slow for a tiny mass. Although Davis experimented with buried tanks filled with several thousand liters of the cleaning and fire-extinguishing agent carbon tetrachloride and achieved a sensitivity twenty times better than Cowan and Reines, he didn’t measure anything. But Davis, a bulldog and a maverick, wouldn’t give in that easily. If chlorine wasn’t able to detect reactor neutrinos, perhaps it would work with neutrinos from a different source? The sun seemed to be a promising possibility.
In his Caltech lectures Richard P. Feynman liked to tell his students the story of the British astrophysicist Arthur Eddington, who, according to legend, was sitting one evening with his girlfriend on a bench watching the night sky.4 When his girlfriend exclaimed, “Look at the beautiful shining stars!” he simply replied, “Yes, and right now I am the only man in the world who knows why they are shining.” What Eddington had correctly surmised was that nuclear fusion reactions proceed inside the sun and other stars, uniting lighter atomic nuclei into heavier ones and in this way producing the energy that is radiated as starlight. But these fusion reactions produce not only energy; they also spew countless neutrinos into space, billions of which bombard each square centimeter of earth each second. And, in contrast to the antineutrinos from nuclear fission processes in nuclear weapons or reactors, in stars we are really dealing with neutrinos and not with antineutrinos. This means that the same experiments with which Davis failed to detect reactor antineutrinos could now be used to detect solar neutrinos—particles produced 150 million kilometers away. In the end, it was the exploration of solar as well as atmospheric neutrinos that would shatter the cozy world of the Standard Model of particle physics.
At first, however, it didn’t look as if Davis’s luck would change. When he submitted his first paper containing an estimate of an upper bound for the solar neutrino flux, the referee wrote that the entire endeavor reminded him of an experimenter standing on a mountain and reaching for the moon, concluding “that the moon was more than eight feet from the top of the mountain.”5 Davis’s task seemed more than Herculean.
In the early 1960s Davis found an ally, the former Louisiana tennis champion and later Princeton theorist John Bahcall, who had devoted his life to understanding the sun.6 While Bahcall’s first estimates of neutrino capture rates in chlorine did not make Davis’s experiment look promising, it soon turned out that once contributions of higher-energy neutrinos that would bring the nucleus into an excited state were taken into account, the predicted rates of argon formation were almost twenty times as big. A year later Davis was able to confirm a crucial part of Bahcall’s predictions experimentally. When he was phoned and told the result, Bahcall said later, it was the best moment of his entire career. The road to the detection of solar neutrinos seemed to be open, and Davis started to build a 400,000-liter tank to be filled with tetrachlorethylene, in America’s biggest underground gold mine, in South Dakota.
But when the first new results came in, they were disappointing. In April 1968, Davis and collaborators published an upper bound on the solar neutrino flux that was a factor of seven below the theoretical predictions. Finally in 1970 the first neutrinos from the sun could be measured. This was an enormously important result, as it finally confirmed Eddington’s idea that stars produce their energy by nuclear fusion. But the neutrino fluxes themselves remained puzzling: Even after Bahcall and collaborators had improved their calculations, a discrepancy of a factor of three remained. If Davis and Bahcall were right about this, this result could turn into a sensation: The missing electron neutrinos from the sun could be explained—corresponding to Pontecorvo’s idea—with neutrinos oscillating into other flavors. At first the physics community was skeptical. Was it really possible that Davis measured the decay rates corresponding to neutrino capture this accurately? A fraction of a nuclear decay per day in a huge tank volume? And did Bahcall really understand the fusion reactions within the sun well enough to make such an accurate prediction? This was particularly questionable because the neutrinos captured in the chlorine experiment were not directly created in the fusion of protons into deuterons (nuclei of a heavy hydrogen isotope), which is the process responsible for the major part of the energy production in the sun and thus directly related to the sun’s temperature.
Instead they originated from a subsequent reaction, whose rate was much more difficult to estimate. The only way out seemed to be an experiment capable of detecting the lower-energy neutrinos that came directly from the proton fusion. Such an experiment had been proposed as far back as 1965 by the Russian theorist Vadim Kuzmin, but it wasn’t tackled seriously until the end of the 1980s. Yet even if Davis and Bahcall’s results were confirmed, there still remained another problem: In order to explain the deficit of solar neutrinos in Davis’s experiment by neutrino oscillations, the distance of earth and sun had to be “fine-tuned” exactly to correspond to the neutrino mass differences. This was an explanation that most physicists considered to be rather unlikely.
However, in parallel with the efforts of the experimenters and the nuclear physicists, the particle theorists had also made progress.
First, Lincoln Wolfenstein, a theorist at Carnegie Mellon University in Pittsburgh, Pennsylvania, had discovered in 1977 that the mixing of neutrinos gets altered if they propagate not in empty space but in matter. The reason is that neutrinos get decelerated by interactions with matter. From outside, this looks as if the neutrino had more mass; a smaller part of its total energy can appear as kinetic energy. Now, Wolfenstein is an extremely modest and unselfish person. As Palash Pal, now a scientist at the Saha Institute in Kolkata told me, Wolfenstein applauds any colleague who solves an interesting problem, even if he himself or one of his students (like Palash) had worked on the same problem, competing for a solution, and now would get no credit for that work (Palash later wrote about this experience in a humorous essay on the subjectiveness of joy and sorrow). To push or even to influence someone else goes so much against Wolfenstein’s nature that when students ask him for a deadline for submitting their work, he tells them, after a long hum and haw, to set the deadline themselves.7 So naturally he didn’t consider his work on neutrinos to be anything more than a pastime, not worth looking into more deeply. Three years later, Vernon Barger, Kerry Whisnant (both at the University of Wisconsin, Madison, by then), Roger Phillips (Rutherford Appleton Laboratory, UK), and my later boss Sandip Pakvasa of the University of Hawai‘i worked on the mixing of neutrino flavors within matter. They found out that specific combinations of neutrino energies and matter densities can amplify the mixing significantly: The matter effects can cancel the mass differences between single neutrinos, with the result that wave packets with equal momentum run next to each other with the same velocity and the mixing becomes maximal.
The real breakthrough, however, was provided in 1986 by two Russians. In the mid-1980s, Alexei Smirnov was a young researcher at the Institute for Nuclear Research of the Soviet Academy of Sciences in Moscow. One day his colleague Stanislav Mikheev walked into his office and proposed that they study the interactions of atmospheric neutrinos with matter. “It was an obvious thing to apply the approach also to solar neutrinos,”8 Alexei told me years later when I met him at a workshop in Beijing, China. What Alexei and Stanislav did was to look at how a neutrino born in the sun’s core evolves. They made the stunning discovery that the electron neutrino—which, because of its large effective mass, is created in the heaviest state in the solar core—loses mass as it propagates toward the sun’s surface through matter of decreasing density, turning slowly into a maximal superposition with the muon and tau neutrinos. The particle remains in the heaviest state even when it mixes more and more strongly with muon and tau neutrinos, which, in vacuum, are heavier. Once the neutrino leaves the sun it is therefore in what is called a pure mass eigenstate consisting predominantly of the muon and tau flavors; it doesn’t oscillate any more until it reaches the earth. On arrival it can then, with a certain probability, be detected as a muon or a tau neutrino corresponding to the flavor composition of the heavy state in vacuum.
As Alexei told me, he immediately realized how important their discovery was. But it was a long time before the two Russians could convince the scientific community. First they sent their results to Wolfenstein, but he didn’t believe them. When they submitted their work to a Russian journal, it was rejected. Only on their second try were they successful, getting it published in the Italian journal Nuovo Cimento. Next Alexei wanted to talk about the groundbreaking result at an international workshop in Finland, but the organizers wouldn’t allocate him any time. He got a hearing only after he showed his calculations to Nicola Cabibbo, the Italian physicist who had introduced quark mixing in the first place. Cabibbo immediately understood the result and lobbied for Alexei to get a slot for his talk. The discovery became known as the MSW effect (for Mikheev, Smirnov, and Wolfenstein). It predicts a flavor conversion of solar neutrinos that is independent of the distance between the sun and earth. But of course this still didn’t prove that neutrinos really were changing their flavor.
In the early 1990s, the solar neutrino experiments that were sensitive to the direct neutrinos from proton fusion delivered their first results. These were the GALLEX experiment led by Till Kirsten at MPIK in Heidelberg and the Soviet-American SAGE experiment located close to Mount Elbrus, the highest peak of the Caucasus. (This is a region where it is not uncommon to see dead horses rotting at the sides of the roads, where public transport buses stop until the passengers have collected money for gas, and where a village with the name Neutrino exists, with its own anthem.) Both GALLEX and SAGE used gallium as a neutrino detector, either in the form of an aqueous gallium chloride solution or in metallic form, and both of them employed quantities of it that only three decades before would have exceeded the world’s production by a factor of ten. This was possible only because of the important role gallium played in the developing semiconductor industry. The price of gallium decreased greatly, but it nevertheless remained attractive for thieves, who broke into the Russian laboratory twice by using the railborne carriage leading into the underground lab (the rumbling was actually detected by the other experiments), and who stole more than two tons of the material.
The results, however, were anything but clear: While the count recorded by SAGE lay far below the expectation of solar neutrino fluxes and thus seemed to support the neutrino oscillation hypothesis, the first GALLEX result appeared to observe all neutrinos from the proton fusion and, in addition, some neutrinos produced in the subsequent solar reactions. Later the SAGE result confirmed the GALLEX result so that no clear picture emerged. Whether the earlier neutrino deficit was due to neutrino oscillations or whether it was a consequence of the limited understanding of the working of the sun could not be decided. In GALLEX, for example, the chemistry of the experiments was extremely complicated: In several steps the germanium resulting from neutrino capture had to be precipitated and extracted from the huge tanks, until the subsequent radioactive decays could be measured in tiny counting tubes and thereby indicate the capture of solar neutrinos. Finally, the MSW effect predicted a flavor conversion of neutrinos that depends on the neutrino energy. Thus even a complete observation of proton fusion neutrinos would not invalidate the MSW effect as an explanation for the disappearance of higher-energy neutrinos. In this blurry situation, experiments looking for neutrinos from a different astronomical source eventually played the decisive role.
Apart from sunlight and solar neutrinos, the earth is being bombarded also with other projectiles from outer space. To this day, the origins of the cosmic radiation are not completely understood. Most physicists believe that the most likely candidates are supernovae, the gigantic explosions accompanying the collapse of massive stars, and active galactic nuclei, black holes millions of times more massive than the sun located at the centers of galaxies, whose gravitational attraction converts their direct neighborhood into a boiling inferno of extremely energetic particles. When the quanta of the cosmic radiation (mostly protons) hit the upper atmosphere and collide with nuclei of nitrogen and oxygen atoms, medium-weight mesons—such as pions and kaons, which are bound states of quarks and antiquarks—are produced. These mesons decay quickly, creating electron and muon neutrinos in the ratio of 1:2.
Such atmospheric neutrinos were first discovered in underground experiments in South Africa and India in 1965, which provided additional hints that the observed fluxes did not correspond to the theoretical predictions. As for solar neutrinos, the situation remained unclear for a long time.
In the early 1990s several experiments, originally constructed to search for the proton decay predicted in GUT theories, delivered their most important results by measuring atmospheric neutrinos. Two of these experiments, which detected neutrinos by measuring the heat produced by them within the detector material, could find no hint of neutrino oscillations. Two others (IMB and Kamiokande), looking for the blue flashes that arise in water tanks when neutrinos hit charged particles and accelerate them to speeds greater than the speed of light in water, reported a flux of muon neutrinos about one-third less than was expected. Somewhat earlier, in 1988, just after Kamiokande had confirmed the neutrino deficit reported by IMB, John Learned of the University of Hawai‘i, in collaboration with my later bosses Sandip Pakvasa (also from Hawai‘i) and Tom Weiler (Vanderbilt University, Nashvillle, Tennessee), had found that the result could be interpreted as neutrino oscillations between muon and tau neutrinos with almost maximal mixing.
The final clarification came only with a rash of new experiments, and then everything happened almost at once. A third experiment that used the heat-detection method confirmed the atmospheric neutrino deficit. Then, in 1996, the gigantic Super-Kamiokande started to take data. Super-Kamiokande—whose name, an English-Japanese amalgam, can be pronounced in Japanese to mean super-bite-into-god—was ten times as big as Kamiokande, and after only one year, Masato Takita reported at the Beyond the Desert conference organized by Klapdor-Kleingrothaus and me that Super-Kamiokande also seemed to confirm the atmospheric neutrino deficit. After another year the Kamiokande researchers felt confident enough to announce the discovery of neutrino mass: “This is not a small effect, and we have found no way to make it go away or even be severely distorted,” explained John Learned, member of the Super-Kamiokande Collaboration and, along with Fred Reines, a founder of the IMB experiment.9 In addition to neutrino fluxes, Super-Kamiokande was able to reconstruct the directions from whence the neutrinos came. One could actually see the position of the sun in the neutrino “snapshot” of the heavens. And this measurement provided yet another hint for the oscillation hypothesis: Only those neutrinos oscillated that had been produced in the atmosphere on the other side of the world and thus had to penetrate the entire earth on their way to the Japanese lab. The neutrinos produced above Japan had a travel distance too short for complete oscillation.
Next, experiments were built that sound like science fiction (Fig. 9.2). Neutrino beams were produced in the large accelerator labs KEK (Japan), Fermilab (United States), and CERN (Switzerland) and directed over distances of 250 km, 730 km, and 732 km, respectively, through the earth onto the underground labs Super-Kamiokande, MINOS (Minnesota), and Gran Sasso (Italy). Again the Japanese were ahead in this game by a nose. In 2006, they could observe the disappearance of muon neutrinos due to their oscillations into tau neutrinos in an earthbound experiment. And in June 2010, the OPERA experiment also detected the first tau neutrino in a muon-neutrino beam and in this way finally confirmed the atmospheric neutrino oscillations.
Figure 9.2. Modern long-baseline experiments fire neutrinos hundreds of kilometers through the earth aiming at distant underground labs (even one in a different country). Shown here is the beam line of the CERN–Gran Sasso experiment starting in Geneva, Switzerland, and arriving in the Abruzzo mountains, Italy, east of Rome. Here researchers could see the oscillations of muon into tau neutrinos for the first time with an earthbound source.
In 2002 the conversion of electron neutrinos into muon and tau neutrinos through the MSW effect in the sun was confirmed. It had taken fifteen years to finally complete the SNO (Sudbury Neutrino Observatory) experiment, located in Sudbury, Ontario, Canada, the delay being caused at least partly by its enormous technical complexity. To minimize the background, the physicists had to build an experiment under clean-room conditions, with the whole thing inside the caverns of a nickel mine still running at peak activity, drawing out 5,000 tons of ore every day. But when the experiment, with a thousand tons of heavy water—worth 300 million Canadian dollars and borrowed from a state-owned Canadian nuclear power company—was finally running, it nailed down the solution of the solar neutrino problem. After Kamiokande and then Super-Kamiokande had confirmed the “missing” solar neutrinos, SNO could observe not only the deficit of missing electron neutrinos but also the excess of the oscillation products, the newly born muon and tau neutrinos. This is because SNO was sensitive not only to the W-boson exchange with electron neutrinos but also to the exchange of Z bosons that is relevant for the interactions of all neutrino flavors. The solar neutrino problem was finally solved, and in the same year this result was confirmed in a terrestrial experiment when the Japanese KamLAND Collaboration measured the neutrino fluxes that emanated from several reactors about 150 kilometers away.
Again, Tom Weiler and Sandip Pakvasa, this time in collaboration with Tom’s PhD advisor Vernon Barger (a PhD great-great-grandson of Enrico Fermi), were among the first to realize that the results of solar and atmospheric neutrinos require two large mixings—Tom christened this pattern bi-maximal—and that neutrino mixing is therefore completely different from the better-known mixing of quarks, which mix only weakly.
This time, after the experiments of the 1990s had demonstrated that neutrinos really possess masses, the particle physicists immediately faced the next set of puzzling questions: Why is the neutrino so light? Why is its mixing so totally different from that of quarks? How can the neutrino be described successfully in a single theory together with quarks and charged leptons if it is at least a million times lighter? In the words of Art McDonald of Queens University in Canada, spokesman of the SNO Collaboration that finally proved neutrino flavor conversion: “What we need is a mechanism explaining neutrino masses and in particular why neutrino masses are so much smaller than the masses of the other particles.”10 Whatever these mechanisms may be, they make neutrinos ideal probes for the new physics beyond the Standard Model: for GUT theories, extra dimensions, and maybe even time travel. But before we go there, we first make a small detour into the depths of the universe.