Science Potential of a Deep Underground Laboratory
Our understanding of the physical world of quarks and leptons and of their relation to astrophysics and the evolution of the so-far-visible universe is extensive and profound. However, we know this understanding is very incomplete. The questions scientists would like to answer include these:
Why do neutrinos have tiny masses, and how do they transform into one another?
Are the existence and stability of ordinary matter related to neutrino properties?
Are there additional types of neutrinos?
What is the mysterious dark matter, and how much of it consists of neutrinos?
What role do neutrinos play in the synthesis of the elements in the periodic table?
Is there a deeper simplicity underlying the forces and particles we see?
These are important and very basic questions whose resolution will have a major impact on physics and our knowledge of nature. A common element in answering these questions involves the study of rare processes.
A clean, quiet, and isolated setting is needed to study such rare phenomena free from environmental background. Such a setting can be obtained only deep underground, where we can escape the rain of cosmic rays from outer space. The
cosmic rays create background events that mask the critical events being searched for. It takes 2 miles of rock to absorb the most energetic of the muons created by cosmic-ray protons striking Earth’s atmosphere.
At such great depths, the only backgrounds are made by neutrinos (which easily penetrate the whole Earth but, by the same token, interact very seldom) and by local radioactivity in the rock itself. The latter can be shielded by the use of specially purified but otherwise ordinary materials, such as water. For instance, the Sudbury Neutrino Observatory (SNO) in Canada is built as a high-tech clean room 10 stories tall and more than a mile underground. Only in this laboratory could the collaboration achieve an experiment that is 10 billion times cleaner than our typical living room in terms of natural radioactivity. SNO is the most background-free environment ever achieved on Earth.
Some experiments do not require the greatest depths and can tolerate less stringent conditions either because the process being sought has a higher signal rate or because some special experimental tag can be used to identify the important events even in the presence of background. For other experiments, however, there is no option but depth and extreme cleanliness. Only in such an isolated environment can we hope to detect the faintest signals of our universe.
Scientists addressing issues of intense international interest—solar neutrinos, double beta decay, and dark matter—are poised to develop next-generation detectors that require low background, and they need an underground facility for technology development in the next few years. Once the neutrino mixing and mass parameters have been measured with some accuracy, a long-baseline experiment should be developed. The KamLAND, Borexino, MiniBooNE, and MINOS experiments are expected to lead—over the next 5 years—to the synthesis necessary for the long-baseline program. A long-baseline target detector is likely to also carry out a proton decay experiment and serve as a supernova neutrino telescope, as well as many other purposes.
The neutrino has had a very rich history. As described in the science overview (Chapter 2), the neutrino was postulated to preserve important conservation principles in the decay of nuclei and, as a consequence, had to possess novel properties: zero charge, zero mass, spin 1/2, and very weak interactions with other particles. It took the advent of nuclear reactors, which were able to produce neutrinos in profusion, to clearly demonstrate that the neutrino indeed existed. Furthermore, not one but three distinct types of neutrinos exist: an electron, muon, and tau type of neutrino, each coupled to its respective electrically charged partner. After intensive efforts to directly measure neutrino masses, an upper limit of 1–3 eV has
been established. This can be compared to the electron’s mass of 511,000 eV. The Standard Model of particle physics was therefore built with the key assumption that neutrinos are massless.
This tidy picture has been dramatically changed by recent experimental discoveries. For both atmospheric and solar neutrinos, there is now strong evidence that they change from one type to another (oscillate) as they travel through space. Because, according to Einstein’s theory of relativity, particles with no mass, such as photons, do not sense time, any change in their neutrino character signals that neutrinos do, in fact, experience time and hence must have a mass, which challenges one of the assertions of the Standard Model. Observations of the oscillation, however, determine only mass-squared differences rather than masses themselves; that is, they measure the absolute value of the difference in the squares of the neutrino masses. The mass-squared differences inferred from the data are very tiny, tenths of electron volts and less. This is very small compared with the typical masses of quarks and other leptons, and it is more than 10 orders of magnitude lighter than the top quark.
Why are the neutrino masses so tiny? Another new puzzle uncovered since these discoveries is that the compositions of neutrinos with definite mass values are highly mixed up, as shown in Figure 4.1, with large fractions of electron, muon, and tau types in a given neutrino. This must be compared with the situation among quarks, where the amount of mixture is very small, 0.01 to 5 percent. The aim of the next-generation experiments is to establish the newly emerging picture and to determine yet unknown parameters in the neutrinos, and then to understand how the Standard Model must be revised. The mixtures can be quantified in terms of angles, with an angle of 0 degrees signifying no admixture and 45 degrees the maximum admixture of a second flavor. With three flavors, there are three angles, θ12, θ23, and θ13. The mixture of the electron type in the third neutrino is related to θ13 and is known to be small, but scientists do not know how small. There may be additional neutrino species (sterile) beyond those we currently know. If so, how many are there?
We do not know if the neutrinos are their own antiparticles. If the answer is yes, they may have played a crucial role in creating a tiny imbalance between matter and antimatter in the universe, so that some matter survived the annihilation and led to our existence. For this to be the case, there must be a subtle difference between the behavior of neutrinos and antineutrinos, called charge-parity (CP) violation.1
As discussed in the subsequent sections, these issues can be studied in a variety of experiments involving more accurate studies of solar and atmospheric neutrinos, double beta decay, and accelerator-based neutrino experiments, especially those with long baselines. A deep underground laboratory will play a crucial role in these proposed experiments.
The Sun, powered by nuclear fusion, is an abundant and pure source of electron neutrinos. Trillions of solar neutrinos pass through our bodies every second. Solar neutrinos were first detected in an experiment in the Homestake Gold Mine in South Dakota by Raymond Davis, Jr. That experiment also gave the earliest indication for a finite neutrino mass when only a third of the expected number of neutrinos was seen.
This shortfall is now understood quantitatively. The SNO in Canada has recently shown that all the neutrinos are there as expected, but two-thirds have changed from their original electron flavor to flavors not detectable in the Homestake experiment—muon and tau neutrinos. Strong indications of this conversion were already apparent when data from the Super-Kamiokande, SAGE (Soviet American Gallium Experiment), Gallex, and Homestake solar neutrino experiments were considered together. A new solar neutrino experiment, Borexino, and a reactor antineutrino experiment, KamLAND, are now being used to provide tighter constraints on the neutrino mass and mixing parameters responsible for flavor conversion.
The dominant mechanism of neutrino production is referred to as the pp (proton-proton) neutrino reaction: In the standard solar model the flux from the pp reaction is predicted to an accuracy of 1 percent. Further, the total flux is related directly to the measured solar optical luminosity. Such a copious and well-understood source of neutrinos is ideal for precisely determining the neutrino masses and mixings where accelerator techniques are limited. It also affords a way to search for hypothesized sterile neutrinos as much as a million times lighter than those explored by present experiments, provided they mixed sufficiently with the active neutrinos. Unfortunately, the pp neutrinos have very low energies (see Figure 4.2).
A program of low-energy solar neutrino measurement is straightforward in concept but difficult to carry out in practice. Two types of experiment are required, both sensitive to the lowest-energy neutrinos. One experiment measures the electron-flavor component by the charged-current (CC) reaction, while the other measures a combination of electron, mu, and tau neutrinos via elastic scattering (ES) from electrons. Taken together, these measurements provide model-independent determinations of the electron and nonelectron neutrino flux components at each energy and solar-model-dependent determinations of the sterile components. Because the electron and nonelectron rates are similar, a good measurement of the difference places great demands on the quality of the CC and ES experiments.
At these low energies, backgrounds become formidable. The background
problem can be to some degree circumvented in CC experiments by selecting target nuclei (100Mo, 115In, 176Yb, etc.) that provide a “tag” for neutrino capture— that is, a subsequent decay at the same position and almost the same time that specifically identifies the neutrino event in a welter of irrelevant background events. Such tags cannot be arranged for ES experiments, but the rates are higher and the targets simpler.2 Good ideas exist for both types of experiment. They are cur-
rently in an R&D phase—some of them will need underground space for tests soon, and some are already taking place (i.e., the Low Energy Neutrino Spectroscopy (LENS) project at Gran Sasso).
Clever experimental strategies and extreme measures to remove radioactive contaminants are only a part of a successful response to the low-energy challenge. Cosmic rays continuously create new radioactivity in the detector, and the only remedy is to place the detectors beneath hundreds or thousands of meters of rock. While every experiment will have a different tolerance to this type of activation, a conservative estimate can be made by comparing the rate of solar neutrino interactions with the background rate of nuclear transmutations caused by muons.
The ES experiments expect a signal event per day in roughly 2 tons of detector, which equals the transmutation rate at a depth around 3,000 mwe. Since no tag exists for the ES experiments (other than that the scattered electrons point away from the Sun), a substantial margin of signal over background is desirable, which could be achieved at depths of 6,000 mwe. The CC experiments expect rates 10 to 100 times smaller (for unenriched isotopic material) but, in most of the cases proposed, have a tag that helps with background rejection. The signal rate equals the transmutation rate at a depth of about 6,900 mwe.
Although current-generation solar neutrino observatories could significantly advance the state of the science, there is still much to learn, especially in low-background, high-precision physics. It cannot be said that these difficult experiments would be impossible at depths less than 6,000 mwe, but it is clear that at such a depth a major background source is under control, whereas at lesser depths it remains uncertain. A laboratory sited at 6,000 mwe thus offers a unique and powerful advantage to physicists seeking to observe low-energy solar neutrinos.
Particle accelerators can provide a precisely understood source of neutrinos. In the study of neutrino properties, neutrino beams from particle accelerators can provide information complementary to that from future solar neutrino experiments that address measurements not accessible to accelerator experiments (see Figure 4.3). Protons from accelerators produce an almost pure beam of muon neutrinos, while solar neutrinos are purely electron neutrinos. With such beams, researchers may even observe CP violation, a possible subtle difference in properties between neutrinos and antineutrinos that may be fundamentally related to the matter-antimatter asymmetry of the universe. Solar neutrino beams are generated at the far end of an extremely long baseline; thus the arriving beam of neutrinos at Earth has gone through many oscillations.
The dramatic discovery of neutrino oscillation was made using a natural source
of neutrinos. When cosmic rays hit Earth’s atmosphere, neutrinos are created that enter underground experiments. Even those created on the other side of Earth easily penetrate Earth and reach these experiments. These atmospheric neutrinos were studied in great detail by the Super-Kamiokande experiment that provided convincing evidence for neutrino mass. The oscillation effects in this case occur only over distances comparable to the size of Earth. For quantitative measurements of neutrino mass and mixing parameters, however, accelerator-based neutrino oscillation experiments are crucial. With particle accelerators, scientists can control the energy, direction, flux, and even the composition of neutrino beams. To study this phenomenon and make accurate measurements requires a long distance between the accelerator, where neutrinos are produced, and the experiments, where they are detected. Funded or newly operating experiments designed to more accurately determine them include KEK to Kamioka (K2K) in Japan, ICARUS/OPERA (Imaging Cosmic and Rare Underground Signals/Oscillation Project with Emulsion Tracking Apparatus) in Europe (on hold), and NuMI/ MINOS (Neutrinos as the Main Injector/Main Injector Neutrino Oscillation Search) in the United States, all relatively long-baseline experiments (200–800 km) using neutrinos from accelerators. These are expected to provide data over the next decade that should corroborate the specific qualitative description of neutrinos and should measure some parameters to about 10 percent. It is expected that the evidence for oscillations in atmospheric neutrinos will be found by these experiments to be primarily mixing between the muon and tau types. However, the remaining critical mixing parameter, known as θ13, will be poorly determined at best. This parameter is different from zero if each of the three neutrino types (electron, muon, and tau) mixes with all the others. The value of θ13, now known to be less than 10 degrees, will be measurable by currently planned experiments only if its value is larger than 1 degree.
Though the entire picture could be changed as a result of these experiments, such as the U.S. MiniBooNE effort, it is more likely that they will reinforce the need for accurately determining all the mixing parameters. Extrapolating to the time frame of a U.S. facility for underground experiments, accurate measurement of θ13 will be the critical goal and the gateway to exciting new issues, like CP violations and establishing the neutrino family mass hierarchy (see Figure 4.2). Both can be addressed if the value of θ13 is large enough. Sensitivity to all these questions depends on many factors, but mostly on (1) the neutrino energy, (2) the distance between neutrino production and observation, (3) the neutrino source flux, and (4) the detector mass and sensitivity.
The most likely route to determining θ13 is measuring the (small) oscillations that take place between muon and electron neutrinos in an experiment designed with specific combinations of energy and distance. For accelerator-produced neu-
trino energies higher than 1 GeV, the optimal distances are longer than about 500 km, depending on the exact value of the mass splitting, which will be determined in the next few years. Since the probability of oscillation is small and the fraction of the neutrino beam intercepted by the target decreases with distance, very high fluxes of neutrinos will be required. If measuring θ13 goes reasonably well, measuring the mass hierarchy and the CP properties of the neutrino admixtures will be compelling. For these goals, the massive target/detectors and high-flux sources will have to be more substantial. It has been shown that it is not easy to disentangle effects of θ13, different mass hierarchies, and CP violation, because all of them affect the oscillation probabilities simultaneously. Researchers will need at least two different baselines and/or energies to resolve each of them separately. In Japan and Europe, the baselines currently envisaged are relatively short. Therefore it makes sense to develop plans for experiments with baselines longer than 1,000 km in the United States in the context of the international program. Indeed, distances from the two major proton laboratories (Fermilab and Brookhaven) range from 1,200 to 2,600 km for the several proposed underground sites.
Already-planned long-baseline experiments involve neutrino energies of the order of gigaelectron volta, beam powers of ~100 kW, detectors of 5–50 kt, and distances of 200–700 km. Future experiments to explore the longer-term issues will require similar neutrino energies but higher beam power (megawatts) and larger detector masses (megatons). They will also likely be planned for modestly longer distances (~2,000 km). An important issue to be resolved for such experiments is the high-power source, whether it is more intense (a superbeam) or supplies a storage ring serving as a neutrino factory. Superbeams are being considered in many parts of the world. The neutrino factory concept is undergoing substantial accelerator R&D, and demonstrating its feasibility will take time.
Such experiments in the United States are probably still a few years away. It may be wisest to finalize plans after the mixing parameters are better known. More important, the neutrino source requires careful planning and has to be coordinated with optimization of a large, well-instrumented detector. An operating underground laboratory would facilitate this planning. Also, laboratory infrastructure and staff would greatly expedite the installation, commissioning, and operation of large detectors. Positioning at even a modest depth will reduce the cosmic-ray-induced backgrounds. However, the neutrino beam energy (high) and the duty cycle possible from accelerators (short) should reduce the need for great depth to keep backgrounds to acceptable levels, although some overburden is desirable. (In fact, the more critical detector feature for a successful long-baseline program is the size of the detector, as this directly affects the flux of incoming accelerator neutrinos that can be analyzed.) It should be noted that the large detector might well serve other scientific functions, such as searching for proton
decay and/or observing supernova neutrinos. The depth-location of such a detector may be determined by these considerations.
The neutrino mixing parameters form a gateway to understanding fundamental features of matter and energy unanticipated in the Standard Model. Measuring them, in long-baseline experiments, represents an opportunity for which the United States has important (and somewhat unique) historical, scientific, and geographical advantages. If the United States is to capitalize on this opportunity to lead in such experiments, planning should start in the near future, take into account forthcoming experimental results, and be finalized in about 5 years. Such plans are likely to be much more reliable if an underground facility is available to house the target detector.
Double Beta Decay
The major discovery of the past decade regarding the properties of elementary particles has been the confirmation that neutrinos, the most elusive of the known elementary constituents of the world, have mass. Oscillation experiments have shown that there are nonzero differences between the squares of the masses of different kinds of neutrinos and therefore prove that neutrinos have a finite mass. However, the absolute value of the mass and whether the neutrino is distinct from its antiparticle are still open questions. If the neutrino is distinct from its antiparticle, it is a Dirac particle, as are all the other known elementary particles with spin 1/2. If it is indistinguishable, it is a Majorana neutrino. The search for neutrinoless double beta decay is motivated by the need to determine the mass and antiparticle nature of the neutrino.
In most nuclei found in nature with even numbers of protons and neutrons, simple beta decay (with the emission of an electron and a neutrino) is energetically forbidden. However, the simultaneous emission of two electrons with a daughter nucleus differing by two charges (double beta decay) can be possible. This process is expected and observed within the Standard Model of particle physics when it occurs with the emission of two neutrinos in addition to the two beta particles, and it is called two-neutrino double beta decay. In this type of decay, since the neutrinos go undetected, one observes a spectrum of the sum of the energies of the two beta particles that extends up to the total energy available for the decay. A more interesting process is that of neutrinoless double beta decay, in which no neutrinos are emitted and the two beta particles share the total energy. If neutrinoless double beta decay exists, it implies that neutrinos are Majorana particles, and its rate is proportional to the square of the Majorana mass. Should the existence of neutrinoless double beta decay be convincingly proven, the resulting qualitative physics conclusion regarding neutrino properties would have an ex-
tremely important impact on our understanding of the fundamental properties of nature.
There is a consensus that the atmospheric neutrino measurements by the Super-Kamiokande collaboration can only be interpreted as a consequence of the nearly maximal mixing between the muon-like and the tau-like neutrinos with the corresponding mass-squared difference Δmatm2~3×10−3 eV2. Thus at least one neutrino has a mass greater than 50 meV, and this value sets the goal for the next generation of double beta decay experiments. Although the oscillation properties can be pinned down better by further oscillation experiments, to determine the neutrino mass requires either a direct mass measurement or an observation of neutrinoless double beta decay. Furthermore, only double beta decay has the potential to elucidate the antiparticle nature of the neutrino. Hence, measurement of a nonzero rate would be truly unique and truly spectacular.
For an effective mass of 50 meV, the predicted half-lives of the various neutrinoless double beta decay candidates are about 1027 years. To reach this level of sensitivity in a few years of running, an experiment requires approximately 1 ton of a particular isotope. In addition, since the discovery of neutrinoless double beta decay requires observation of a peak superimposed upon a continuous background, the background must be very low in the peak region. The various causes of background can be sorted into three classes: two-neutrino double beta decay, natural (or sometimes manmade) radioactivity, and radioactivity induced by cosmic rays. Since the two-neutrino double beta decay rate is at least 107 times faster than neutrinoless double beta decay and its energy spectrum extends up to the peak region, good energy resolution is essential. Certain members of the naturally occurring uranium and thorium chains are radioactive and have large enough decay energies to pollute the peak region. Since the half-lives of the parent uranium and thorium isotopes (~1010 years) are much shorter than that of double beta decay, very low uranium and thorium impurities in a detector are critical. In addition, the detector must be shielded from the ambient radiation of the surrounding environment by shielding that is also free of activities, and any volume near the detector must be purged to prevent the ingress of radon. The activities of cosmogenic isotopes produced while the critical materials reside on the surface can be especially high, so it will be necessary to purify, fabricate, and assemble parts underground. Cosmic rays can also either directly produce background in the experiment or give rise to radioactivity in situ through their secondaries. The former can be completely mitigated with overburden, and the latter requires depth that depends on the material in question. Most of the proposed experiments will require a depth of 4,000 mwe or deeper to mitigate the cosmic-ray-related background.
There are several proposals in the United States and around the world for next-generation double beta decay experiments with sensitivity to 50 meV. The Enriched Xenon Observatory (EXO), Majorana, and the Molybdenum Observatory of Neutrinos (MOON) proposals have U.S. participation and will very likely be carried out in the United States if acceptable underground sites can be found. These three and the Germanium Nitrogen Underground Setup (GENIUS) and the Cryogenic Underground Observatory for Rare Events (CUORE) proposals are the most advanced efforts, in the committee’s opinion, for reaching the 50-meV target. There are a large number of experienced double beta decay experts in the United States and an even greater number of low-background experts. Many of these experts involved are involved in emerging U.S. collaborations. These collaborations are currently expanding their personnel and performing research and development measurements. They have obtained seed money and are writing proposals. Two of these experiments (EXO and Majorana) will be looking for an underground site for initial detector systems later this year. There are also U.S. collaborators on the CUORE project, which is being assembled in Europe. All in all, the United States is in a strong position to make a significant contribution to this next generation of experiments. Deep underground laboratory space is essential to success, and a national underground facility is a natural way to meet this need.
Over the past decade, strong evidence has led to the conclusion that neither ordinary matter nor even massive neutrinos can account for most of the missing matter. A strong case has been made for a new form of elementary particle as the resolution of the dark-matter problem, and this idea has become the working hypothesis of both astronomers and particle physicists. One of the leading particle candidates is a hypothetical particle motivated by supersymmetric theories, the neutralino. The committee focuses on the neutralino here since its study is within the purview of underground experiments. This type of candidate is often termed a weakly interacting massive particle, or WIMP for short.
The planned direct detection experiments are particularly well suited to complement IceCube since they search for relatively light WIMPs using neutralino-heavy nuclei interactions. (Recall that IceCube’s primary sensitivity to WIMPs is through detection of neutrino decay products when pairs of WIMPs trapped in the Sun annihilate.) As such, the two methods are sensitive to neutralinos in different environments—Earth and heavy nuclei, the Sun and ionized hydrogen—and thereby provide complementary sensitivity to different types
of neutralinos through so-called spin-independent and spin-dependent interactions.3
In the mid-1980s, new calculations showed that WIMP detection is a tractable though difficult experimental challenge. More recent calculations predicted that neutralinos in the galactic halo could be expected to interact in terrestrial detectors at most about once per day for every kilogram of detector material. Ordinary detectors under ordinary laboratory conditions would be overwhelmed by the rates of natural radioactivity and cosmic rays compared with this signal. The general methods for defeating these backgrounds are careful screening and cleaning of materials, active and passive shielding, protection against contamination, and siting in underground locations to reduce cosmic-ray-related activity. Additional tools include detectors with intrinsic rejection of background events, such as recoil discrimination or insensitivity to low-energy-density depositions, and approaches that take advantage of secondary features such as directional or seasonal modulation.
At this time, no conclusive evidence exists for WIMPs. Early WIMP detection experiments in the late 1980s using existing detector techniques were able to make progress, but it was clear that new technologies with the capabilities outlined above were required. Following a decade of successful development, new techniques have led to greater sensitivity and hold promise for significant experimental improvements. New technologies will give their promised sensitivities only with larger detectors, with improved background suppression, and preferably with the means to distinguish terrestrial backgrounds from galactic WIMP sources. Technologies that scale to 1 ton will be needed to fully explore the range of currently favored models. Internationally, groups with strong U.S. representation are advancing this broad array of technologies.
Two broad classes of background events must be addressed. First, electromagnetic backgrounds, which generally have their origin in radioactive contamination or cosmic rays, give rise to events that can be distinguished from WIMPs because they recoil from electrons, whereas WIMPs recoil from nuclei. Since the tools to distinguish these backgrounds are never perfect, it is crucial to reduce the level of contamination through screening and fabrication techniques, as with the double beta decay experiments. Second, neutrons, which come from a variety of sources, are not directly distinguishable from WIMPs because they, too, recoil from nuclei. Most sources of neutrons—for example, natural radioactivity in rock—are relatively low energy and can be effectively shielded against. A troublesome source is
Recall the discussion in the Chapter 3 section entitled “Signatures of New Physics.”
more penetrating high-energy neutrons produced by cosmic rays interacting in the rock. These neutrons are difficult to shield or detect.
There are several experiments under way in the United States and worldwide to detect the dark matter. The leading efforts include the French Edelweiss experiment in Fréjus and the U.S. Cryogenic Dark Matter Search (CDMS) at the Soudan Mine, both using low-temperature detectors; the Italian experiment Dark Matter (DAMA) in the Gran Sasso laboratory using a large array of sodium iodide detectors; and the U.K./U.S. Zoned Proportional scintillation in Liquid Noble Gases (ZEPLIN) experiment using liquid xenon in the Boulby Mine in the United Kingdom. The challenge for these collaborations is to avoid claims of a false signal by fully understanding each experiment’s systematic effects. The other experiments mentioned, along with a range of other approaches—including gas detectors with directional sensitivity (the U.S.-led Directional Recoil Identification From Tracking (DRIFT) experiment); a competing liquid xenon experiment (the U.S.-led XENON collaboration); ultrapure germanium diode detectors (the U.S.-led Majorana and German-led GENIUS efforts); and superheated droplet detectors (U.S.-led Superheated Instrument for Massive Particle Experiments (SIMPLE) and Canadian-led PICASSO efforts)—are in various stages. Each of the approaches attacks the electromagnetic backgrounds with a variety of methods, with a common aspect being that some suppression and screening of radioactive contamination are required. Common to all is the need to suppress the muon-induced neutron background by operating deep underground.
At present, the efforts range widely; some teams are using fully instrumented small-scale prototypes, and others are already working with several-kilogram experiments of proven technology. The field is very active, and there will be a worldwide race to prove approaches that support up to 1 ton of detector mass. While 1- to 10-kilogram-scale experiments get under way, the various investigators will be learning what is required to realize ton-scale detectors. Proposals for such experiments, which typically aim for 100-kg submodules, are expected in the next 2 to 4 years and could run concurrently with the Large Hadron Collider, which will explore in a complementary fashion the same underlying physics of supersymmetry or other new physics at the electroweak interaction scale.
To a degree, improvements in detector technology or assaying and screening of detector components can reduce the level of electromagnetic backgrounds at any given depth. However, neutron backgrounds are very difficult to reduce once the depth is set and the experiment is in place.
An important question is, How deep underground? The answer depends on the relative immunity to electromagnetic and neutron backgrounds and is therefore experiment-specific. In general, the committee anticipates reduction of electromagnetic backgrounds by further technological improvements that are inde-
pendent of depth. Such improvements will increase the overall sensitivity to WIMPs only if there is a concordant reduction in the neutron background. With only a modest reduction in electromagnetic backgrounds—which looks achievable in the coming decade—siting at less than 4,500 mwe could leave neutrons as the dominant source of background. By siting these next-generation ton-scale experiments at depths of 4,500 mwe or greater, the risk of being background limited by neutrons will be considerably reduced. A depth of 6,000 mwe may be required for more sensitive subsequent-generation experiments. In general, siting the lab as deep as possible will extend its ultimate reach or, alternatively, the time it will operate at the forefront of the field. With kilogram-scale experiments already under way, work on the underground lab must begin as rapidly as possible to allow the R&D for ton-scale experiments to get started. The creation of a unique, well-equipped, deep underground lab will maximize the chance for the United States to play a major role in dark-matter detection.
When astronomer Fritz Zwicky found the first evidence for dark matter many decades ago, little did he realize that the answer to his mystery would involve not faint stars but—most likely—a new form of matter whose existence is key to understanding the union of the basic forces of nature.
It is an important question whether the kinds of matter we are made of, ordinary atoms with ordinary nuclei and electrons, are stable. In the Standard Model of particle physics, the so-called baryon number is conserved. The proton that makes up atomic nuclei is the lightest particle with nonzero baryon number and, hence, is absolutely stable. In most extensions of the Standard Model, however, baryon number is not conserved, so the proton is predicted to decay. Ultimately all known forms of matter would decay, albeit with lifetimes many orders of magnitude longer than the age of the universe. The discovery of proton decay would have an enormous impact on the understanding of nature.
There are many arguments for why the proton should decay. The simplest one is that our universe consists of matter only, with no antimatter. When the universe was born, both matter and antimatter were created in equal amounts. If it had stayed that way, all matter and antimatter would have annihilated each other by now and we could not exist. The matter we are made of has survived this great annihilation because a tiny fraction of antimatter (1 part in 10 billion) has transmuted to matter. This implies that baryon number is not conserved, in turn implying that the proton must decay.
Einstein dreamed of a simplicity underlying the diverse phenomena we see. The recent discovery of a tiny neutrino mass strongly suggests that such a unified
simple description of nature exists. If all forces of nature are indeed unified at extremely short distances, such as in so-called grand unified theories, then quarks (constituents of protons) and leptons (such as electrons and neutrinos) are ultimately the same objects. These particles appear distinct because we usually study them at large distances, where the forces between them behave differently for the various particles. However, in this picture, quarks in the proton sometimes approach each other within the very short distances where the forces are unified, permitting the conversion from quark to lepton and, hence, a decaying proton.
Another argument is based on the marriage of the theory of the microscopic world—quantum mechanics—with the theory of the macroscopic world, Einstein’s theory of gravity. Because a proton could be sucked into a virtual black hole and quantum gravity is believed to violate any law of conservation not associated with long-range forces, protons will decay. However, within the context of the Standard Model, proton stability arises because no known particle species can mediate the process for the proton to decay. So, researchers expect that particles in nature that have yet to be discovered could mediate proton decay.
Probably the most important aspect of the search for proton decay is that it is a unique probe for the shortest distance scales available, with the possible exception of the neutrino mass. Past experiments have already shown that the proton lifetime must be greater than 1032 years for many of the possible decay modes. If the proton decays at all, it must be an extremely rare phenomenon. The current limit implies that the constituents of a proton, distributed over 10−13 cm, must come as close as 10−29 cm for the reaction to occur. In other words, the search for proton decay provides a unique opportunity to probe physics to very small distances, where forces may be unified and the physics is simplified.
A long series of experiments were constructed to search for proton decay, such as Fréjus, IMB, Kamiokande, Soudan, and Super-Kamiokande, all situated underground. Because proton decay, if it occurs at all, must be an extremely rare phenomenon, the only way to find it is to amass a large number of protons and watch them carefully over a long period of time, looking for them to decay. The most recent experiment, Super-Kamiokande, houses 50 kt of water. Watching carefully for a proton to decay in this tank of water over many years has enabled setting the best lower limits on proton lifetime so far, 1.6×1033 years for p → e+π0 and 6.7×1032 years for p → vK+. These lifetimes may be compared with the age of the universe, which is about 1.4×1010 years. This very important result has excluded the simplest models of nonsupersymmetric and supersymmetric grand unified models.
The committee notes that in a broad class of supersymmetric grand unified models (SUSY-GUT) the predicted lifetimes for p → e+π0 and p → vK+ are only a factor of 10 to 30 beyond the present limits, motivating the next generation of
detectors. Grand unified models aside, many theoretical arguments point to proton instability, as mentioned above. The search for proton decay is therefore compelling science.
For the proposed proton decay experiments, shielding from the bulk of cosmic rays is necessary, requiring a depth of about 2,000 mwe or greater. Detection of some modes of nucelon decay (e.g., proton decay to a lepton and two neutrinos, or even neutron decay into three neutrinos), may, however, require a much cleaner environment and, hence, much greater depth.
It is worth remarking that these proton decay detectors, thanks to their large masses, sensitive particle detection methods, and long lifetimes, made discoveries beyond their original purpose. Kamiokande and IMB detected neutrinos from a supernova in the Large Magellanic Cloud (SN1987A), and they confirmed the theory of Type II supernova as the death of a massive star forming a neutron star. These two proton decay experiments studied neutrinos produced in the atmosphere from the collision of cosmic rays and saw the first hint of neutrino oscillations and hence finite neutrino mass. This observation was later established by the bigger proton decay experiment Super-Kamiokande and corroborated by the Soudan-II and Monopole Astrophysics and Cosmic Ray Observatory (MACRO) experiments. Kamiokande and Super-Kamiokande have demonstrated that neutrinos come from the Sun, confirming the Sun’s power source as the nuclear fusion process. They have also shown that there is a deficit in the neutrino flux relative to the predictions by the standard solar model. And now Super-Kamiokande also serves as a target detector for accelerator-based, long-baseline neutrino oscillation experiments using the neutrino beam produced at KEK (the Japanese high-energy accelerator research laboratory) at a distance of 275 km.
The increasing cost and size of next-generation proton decay experiments make it important for such an experiment to serve multiple purposes. A proton decay experiment that can act as a target detector for an accelerator-based, long-baseline neutrino oscillation experiment seems particularly attractive. The technology for building a detector 20 to 40 times bigger than Super-Kamiokande (that is, megaton-scale) is in hand. New technologies are being proposed and studied actively. In a few years, we will know more about the feasibility of these new options.
NEUTRINOS, SOLAR ENERGY, AND THE FORMATION OF THE ELEMENTS
Apart from the interest in their properties, neutrinos can also be used to probe the nuclear processes that fuel our Sun and the processes that create the elements.
Because neutrinos can escape from an enormously dense material, they carry direct information about the interiors of stars and gas that cannot be studied by optical telescopes. A number of measurements can potentially be carried out in an underground laboratory.
The “burning” of hydrogen in the Sun (the conversion of protons into helium) is the source of energy that makes life possible on our Earth. The reactions believed to be responsible for most of this energy production cannot be observed directly because they occur in the interior of the Sun—what can be measured well is the total thermal energy radiated from the solar surface. Neutrinos provide a direct window on these processes (see Figure 4.4).
The solar neutrino detectors using water Cerenkov technology, such as Super-Kamiokande and SNO, are able to look at higher-energy neutrinos only. Direct experimental confirmation of the basic features of the solar neutrino spectrum is lacking. And despite our confidence in understanding the Sun’s operation, some questions do remain. In addition to the main cycle, in which hydrogen converts into helium, a second cycle of thermonuclear reactions can occur, in which the elements carbon, nitrogen, and oxygen (CNO) serve as nuclear catalysts for converting hydrogen to helium. Only about 2 percent of the Sun’s energy is believed to come from the CNO cycle (this percentage depends on the presolar abundances of these elements relative to hydrogen and helium). Additionally, the differences in relative abundances on the surface and in the solar interior may be revealed by the solar neutrino spectrum. Direct measurements of the associated neutrinos below 2 MeV in energy may resolve these issues.
There are smaller experiments in Europe and Russia that are sensitive to the much more abundant lower-energy neutrinos. But precision experiments to de-
termine the spectrum and absolute total neutrino flux in the key low-energy region will be difficult and certainly will require very carefully controlled background conditions, which a sufficiently deep underground facility would provide.
The processes in the Sun are typical of similar stars that are too far away for their neutrinos to be detected on Earth. A detectable signal from distant stars is the result of extreme conditions, and by far the most likely cause is a supernova.
When a star’s nuclear fuel has been sufficiently exhausted, it begins to contract, and if its mass is sufficiently high, it will collapse under gravitational forces. This collapse leads to a supernova: an enormous explosive release of energy during which the star can outshine a galaxy in visible light, yet 99 percent of the energy is released in the form of neutrinos. It is only in such very hot environments that elements heavier than oxygen can be released—the heavier elements in the solar system, such as iron, gold, and platinum, are the product of past supernova explosions. The details of this cataclysmic collapse and the location of the rapid neutron capture that must produce the elements from iron through uranium are still poorly understood. At the extremely high densities that are reached, neutrinos are momentarily trapped and escape over many seconds, cooling the star. The intense flux of neutrinos is believed to reenergize the explosive shock wave that is otherwise stalled by infalling matter and that ultimately flings the mantle of the star into space. The remaining matter will be captured and form a neutron star.
Supernova events have caught the attention of astronomers since ancient times. They are visible when sufficiently close and not hidden by dust. This happens roughly once every few hundred years in a typical galaxy. The light from most supernovae in our galaxy is obscured by galactic dust—it is believed that there may be a few such events per century—but the neutrinos are undeterred by dust. Indeed, some estimates suggest that there are more nearby, optically clouded supernovae (i.e., visible primarily through neutrinos) than there are optically visible ones. In 1987, light from a supernova in the Large Magellanic Cloud (a nearby dwarf galaxy) was seen by telescopes, and simultaneously, 17 neutrinos were detected in the large water volumes of two operating underground proton decay experiments. To understand the mechanism of supernovae better will require detecting many more (thousands) of neutrinos from a single supernova and measuring the flux of the emitted neutrinos, the energy spectrum, the time distribution (the pulse lasts a few seconds), and the distributions among the different flavors of neutrinos and antineutrinos.
The best signals will come from nearby supernovae, but the times of their occurrence are not predictable. Existing detectors such as the Large Volume Detector (LVD), Super-Kamiokande, SNO, and KamLAND, as well as planned future detectors, would primarily detect electron antineutrinos and provide a wealth
of information about the temperature evolution of a supernova as well as neutrino properties. The sensitive detectors that are likely to be built in an underground laboratory for proton decay and long-baseline neutrino oscillation experiments, as well as those for solar neutrinos or double beta decay, will certainly provide a signal (and information) when subjected to a supernova neutrino pulse. However, a novel type of detector may be required to obtain new information about the inner workings of the supernova explosion by studying the spectrum of the muon and tau neutrinos. New designs have been proposed for this purpose.
Coordinating the observation time of supernova neutrino signals with the observation of a visible light signal and possibly with a gravitational wave pulse would provide yet more information. Neutrinos are a unique source of information about supernovae and will provide a better window on how the elements heavier than oxygen that are essential to life came to exist on Earth.
OTHER SCIENCE AT AN UNDERGROUND LABORATORY
Other uses of a laboratory deep underground have been suggested. Some are studies of neutrinos from other sources of geological and astrophysical origins; the committee did not assess these. Others range over a wide variety of interesting possibilities that were beyond the charge and expertise of this committee, from geologic processes to subterranean life forms to other uses of such environments with ultralow backgrounds, including, possibly, applications related to national security.
UNDERGROUND SCIENCE IN AN INTERNATIONAL CONTEXT
Underground science is a burgeoning effort in most scientifically advanced countries outside the United States, with laboratories of various sizes and at various depths in operation or planned for operation. Historically, the United States has been the leader in underground physics because of the pioneering Homestake and IMB (the precursor to the Japanese Kamioka experiment, in fact) experiments, which gave it a tradition of excellence and discovery. Currently operating labs are summarized in Figure 4.5, which plots the depth of the laboratory against the cosmic-ray muon flux at that depth. The “depth” is not the actual depth, but rather the equivalent water depth in meters (mwe) that would reduce the muon rate by the same factor.
The Baksan Laboratory in Russia is the first deep facility (4,700 mwe) specifically excavated for physics. It played a major role in solar neutrino physics with the discovery in the SAGE (Soviet-American Gallium Experiment) experiment
that the flux of low-energy neutrinos from the Sun was also suppressed. The research in Baksan on solar neutrinos and cosmic rays continues, despite difficult conditions.
In Europe several small laboratories have been built contiguous to road tunnels (Canfranc Underground Astroparticle Laboratory (LSC), Fréjus-Modane Underground Laboratory (LSM), Gotthard, and Mont Blanc), as well as a large multipurpose installation, the Gran Sasso National Laboratory in Italy. While some are no longer in use, Fréjus (4,900 mwe) is in demand, and Canfranc (2,450 mwe) and Gran Sasso (3,800 mwe) are being expanded. The Gran Sasso Laboratory hosts the Gallium Neutruino Observatory (GNO) experiment; Borexino (a liquid scintillator detector for solar neutrinos); and two long-baseline experiments to detect neutrinos from CERN. A 1,000-ton supernova detector, LVD; two double beta decay detectors (Heidelberg-Moscow and Cuoricino); a dark matter detector, DAMA; and two accelerators for nuclear astrophysics round out the scientific program. In addition, R&D for new detectors such as the LENS low-energy solar neutrino detector, is proceeding in Gran Sasso. Although at present oversubscribed, Gran Sasso’s 18,000 m2 of laboratory space is being expanded, with two new halls and an independent access tunnel for safety. Fréjus, with 3,400 m2 of space at one of the deepest locations, is home to two double beta decay experiments (NEMO-3 and Telescope Germanium Vertical (TGV)); the Edelweiss bolometric-ionization hybrid dark matter detector; and a low-background counting facility. The laboratory is 130 km from CERN and is a possible site for a megaton-scale detector for neutrino oscillations, solar neutrinos, supernovae, and proton decay.
A recent and noteworthy addition has been SNO in Canada, situated in an operating nickel mine. At 2,092 m (6,000 mwe), SNO is currently the deepest operating laboratory. With new funding from the Canada Foundation for Innovation, the underground and surface experimental facilities at Sudbury are being expanded with the excavation of a new 15,000 m3 cavity at the 2,092-m level and a 2,000-m2 laboratory building, respectively. The first of the new experiments is PICASSO, a supercooled droplet detector for dark matter WIMPs. A cadmium-tellurium detector is under investigation to search for double beta decay in both 116Cd and 130Te simultaneously in a single device. Proposals from the international community for other physics experiments are being actively encouraged. The first new underground space will be available in 2003, and the facility will be complete in 2005. Eventually, when the present program in the SNO detector is completed, that 10,000 m3 cavity will also become available for new research.
In Finland, scientific work is being carried out at 90, 210, 400, 660, and 900 m in the Pyhäsalmi Mine in Pyhäjärvi (Center for Underground Physics in Pyhäsalmi
Mine—CUPP), and a maximum depth of 1,440 m (4,050 mwe) is available for new projects. There is access by both a vehicle ramp and a single-shaft hoist. Experiments currently under way are measuring cosmic-ray interactions and the fast neutron background from radioisotopes in the rock. Proposals exist for a search for multimuon events and for a large long-baseline experiment with a beam from CERN, 2,288 km away.
The Boulby Mine (3,350 mwe) in the United Kingdom is operated by the U.K. Dark Matter Collaboration and contains dark matter detectors using xenon and sodium iodide. It is being augmented with new surface facilities in preparation for the next generation of ton-scale detectors (DRIFT and ZEPLIN).
The most ambitious and successful program in underground science has been developed in Japan. There are two major centers, one at the Kamioka Mine (2,700 mwe) and the other in a disused rail tunnel near Oto (1,400 mwe). Kilogram-scale double beta decay and dark-matter searches are in progress at Oto, but the shallow depth will not permit significant future increases in detector sensitivity. The large water Cerenkov detectors Kamiokande and Super-Kamiokande, built principally to study proton decay, have recorded one milestone after another in neutrino physics. Kamiokande was the first active solar neutrino detector, demonstrating the solar origin and the 8B spectrum of the solar neutrinos. It recorded for the first time the burst of neutrino emission from a supernova. Super-K confirmed Kamiokande’s indications for oscillation of atmospheric neutrinos and is in the process of searching for the same phenomenon with a neutrino beam from the KEK accelerator 250 km to the east.
Kamiokande was dismantled and replaced by the KamLAND detector, a 1,000-ton liquid scintillator experiment that is well positioned to observe oscillations of reactor antineutrinos now that the mixing parameters are known well enough from solar neutrino data. Such a measurement, if successful, will precisely determine the two parameters that define two-flavor mixing. KamLAND may also be a detector of low-energy solar neutrinos (from 7Be), but at a depth of 2,700 mwe the cosmic-ray production of 11C is a serious background.
Other new experiments at Kamioka include a small lithium fluoride dark-matter experiment, a gravitational wave detector, and a 100-kg prototype liquid xenon detector for dark matter, low-energy solar neutrinos, and double beta decay. The xenon project (Xenon Massive Detector for Solar Neutrinos—XMASS) is to be scaled up eventually to reach 10 tons. Further in the future (about 2007 or later), to take advantage of the “superbeam” of neutrinos from the new Japan Hadron Facility being built in Tokai 295 km east of Kamioka, a megaton-scale water Cerenkov detector, “Hyper-K,” is under consideration both for a proton decay search and for long-baseline precision studies of neutrino oscillations and CP violation.
Within the United States, there are two operating underground laboratories, the Waste Isolation Pilot Plant (WIPP) in Carlsbad, New Mexico (1,600 mwe) and the Soudan Mine in Minnesota (2,080 mwe). Until recently, the pioneering chlorine solar neutrino experiment was active in the Homestake Mine in South Dakota. This mine is scheduled for closure, and there is a proposal that it be converted to a deep underground laboratory. Among its assets are its great depth (7,000 mwe or more) and existing infrastructure. Other sites in the United States are also being considered for such a national laboratory. One is a new site at San Jacinto Mountain in California, where the laboratory could be situated at a depth of 6,500 mwe and would be accessible by vehicle via a level tunnel. Greater depths would be possible with a sloping tunnel. The other is at WIPP, which is currently used as an active safe repository for low-level nuclear waste and could be expanded to accommodate an experimental physics program.
The scientific community in the United States has played a very significant role in the accomplishments in underground science in the past 30 years. For instance, the U.S. experiments at Homestake and IMB pioneered the field of underground physics. Started as a proton decay experiment, IMB also served as an atmospheric neutrino detector and was one of only two experiments to observe the neutrino flux from supernova 1987A (the other being the Japanese Kamioka project). U.S. researchers are now actively engaged in preparation for the next generation of experiments. As described elsewhere in this document, the experiments have important and fundamental objectives. Do they require a new facility, one that could be sited in the United States, and would such a facility then be unique?
In the case of the megaton-scale detectors, no underground cavity of this scale exists anywhere in the world. Such a detector serves many purposes, one of which is long-baseline neutrino oscillations, and for that purpose a relevant matter is the distance to the neutrino source. (Great depth is not required.) Possible baselines are limited in Japan, but the sizes of the North American and European continents offer a range of possibilities.
Double beta decay, solar-neutrino, and dark-matter detectors are more demanding with respect to depth. Each experiment has a different tolerance to cosmic-ray-induced backgrounds. For instance, at the depth of the Homestake chlorine-argon experiment, 4,200 mwe, cosmic-ray activation was a source of background with an attendant experimental uncertainty, whereas at the depth of the SNO experiment, 6,000 mwe, there is no significant contribution from cosmic-ray activation. Only the expanded Sudbury site appears to be both deep enough and large enough to meet the needs of some of these experiments. However, that site has only 25 percent of the excavated volume of the expanded Gran Sasso site.
With the intense activity in this field at laboratories elsewhere, will the science
move forward faster than a U.S. facility and its experimental program can be built? In general, that appears not to be the case, although important exceptions exist. Megaton-scale detectors are long-term projects at an early stage. Since long-baseline neutrino physics is an objective, it is desirable to know more about the neutrino mixing parameters before committing to a design. That may take several years. Of the five ton-scale double beta decay experiments proposed, one is committed to Gran Sasso, two are sufficiently advanced that underground sites will be needed soon, and the other two are in the R&D stage. The low-energy solar neutrino experiments that will follow Borexino and KamLAND are also still in the R&D stage. Large dark-matter detectors are under construction now and can be sited at a number of locations.
In principle any of the intended experiments can be carried out at an existing site somewhere. The added value of a dedicated U.S. deep underground laboratory derives from such factors as priority use for science rather than commercial mining, freedom of access, expandability, common use of infrastructure to support many experiments, and the opportunity for the United States to retain a position of equity and leadership in a major worldwide scientific endeavor.