The Neutrino in 1980

Home , Bruno Pontecorvo, Clyde Cowan, Frederick Reines, Inverse beta decay

I SCIENCE IDEAS

Neutrino experiments planned at the Los Alamos Meson Physics Facility and at the Nevada Test Site may hold the key to many advances in particle physics and cosmology.

by Thomas J. Bowles and Margaret L. Silbar

ifty years ago scientists postu- suits of a recent and controversial ex- laws, which have been so successful in lated the existence of the neu- periment, which indicate that neutrinos describing the phenomenology of weak F trino, a neutral, massless, spin- oscillate or spontaneously change their interactions, will no longer apply to all ning particle traveling at the speed of “flavor” as they travel through “empty” physical processes. light, passing unimpeded through the space—almost as if the apple falling Experiments are now being designed whole earth and filling the universe in from Newton’s tree transformed itself at Los Alamos and elsewhere to search copious numbers. After numerous ex- into an orange before it struck him on further for evidence of a finite neutrino periments and reformulations, the prop- the head. If this result is proved correct, mass. erties and behavior of the neutrino still it will mean that the neutrino, long The Los Alamos Meson Physics Fa- challenge our theories about the forces assumed to be a massless particle, does cility (LAMPF) provides the best testing and symmetries of nature. have a nonzero rest mass. It will also ground for certain classes of ideas about The present puzzle concerns the re- mean that some of the conservation neutrino oscillation, for it produces neu-

LOS ALAMOS SCIENCE 93 SCIENCE IDEAS trino fluxes of uniquely high intensity electrons have a wide spectrum of miens. He assumed that the neutrino, and special composition and time struc- energies, as shown in Fig. 1. Pauli sug- like the other fermions, has a particle ture. At the Nevada Test Site, the ex- gested that the missing energy was car- and an antiparticle form, but since the tremely high instantaneous neutrino flux- ried away by a highly penetrating (and neutrino has no charge, and perhaps no es available when nuclear devices are hence, undetected) particle with no magnetic moment, the distinction be- detonated provide other unequalled charge and little or no rest mass. He also tween neutrino and antineutrino was not capabilities to test concepts of weak noted that, to conserve angular momen- then known. interactions, tum and to preserve the “law of spin and However, the mathematical form of The Laboratory has played a central statistics,” this neutral particle must be a the interaction dictated that fermion role in the development of neutrino phys- fermion, a particle with an intrinsic spin number (the number of fermions minus ics, for it was a Los Alamos the number of antifermions) be con- team—Frederick Reines and the late In 1934, two years after the discovery served in the reaction. Since the neutron, Clyde Cowan, Jr.,—who first observed of the neutron, Enrico Fermi used the the proton, and the electron are particles, the neutrino in reactor experiments. This neutrino hypothesis to formulate a and by convention are assigned a fer- role will continue as experiments on theory of beta decay. This theory cor- mion number of + 1, the neutral, un- neutrino oscillation and neutrino masses rectly predicted the shape of the electron detected particle emitted in beta decay help test grand unified theories of fun- spectrum and provided the central ideas must be the antineutrino with a fermion damental particles. New data from these for what was to become the theory of all number of —1. Then the total fermion experiments may indicate the existence weak interactions. number is + 1 before and after the reac- of totally new interactions in nature. The basic process in nuclear beta tion. (As we will see, this and other They may even suggest that our expand- decay is the change of a neutron (n) into number-conservation laws have played ing universe will come to a halt and a proton (p) with the emission of an an important role in understanding and finally contract. predicting weak-interaction processes.) But the question of whether or not the Is the Neutrino Massless? neutrino has a mass remained open. That same year Hans Bethe and The possibility of a nonzero rest mass Fermi postulated that this process takes Rudolf Peierls pointed out that Fermi’s for the neutrino has been considered place through the direct interaction of theory allowed the neutrino to induce since 1930 when Wolfgang Pauli postu- these four particles, which are all fer- inverse beta decay: lated its existence to save the laws of energy and momentum conservation in beta decay of radioactive nuclei. Ori- ginally beta decay was thought to be a This reaction provided a means for veri- two-body process in which a nucleus, fying the existence of the neutrino by such as RaE, changes to a unique final observing the emitted neutron and state, in this case RaF, and emits an positron. However, the probability for electron (e-): inverse beta decay is so low (the cross section is about 10–44 cm2) that neutrino detection was not pursued until almost 20 years later when an intense source of Conservation of energy and momentum antineutrinos became available from fis- implies that all electrons in such a sion reactors. two-body decay reaction come out with Fig. 1. Typical energy spectrum of elec- In 1953. Frederick Reines and Clyde the same energy. Instead, the emitted trons emitted in beta decay. Cowan, Jr., reported results consistent

94 LOS ALAMOS SCIENCE SCIENCE IDEAS

Fig. 2. Neutrinos can be detected by observing the events nucleus with a high neutron-capture cross section, such as associated with inverse beta decay, the interaction of an cadmium, to produce capture gamma rays. The positrons antineutrino with a proton to produce a neutron and a and/or the gamma rays are observed by their interactions with positron. The positron annihilates with an electron to produce a detecting medium, such as a liquid scintillator. two gamma rays. The neutron eventually interacts with a

with the occurrence of inverse beta de- fermion that does not participate in the cay in an experiment at the Hanford reactor. Incontrovertible proof of the and v) is assigned a lepton number of+ 1 neutrinos existence came three years later when Reines and Cowan. together and the decay of the muon, lepton number of –1. With these as- with F. B. Harrison, Herald W. Kruse, signed numbers, the law of lep- and Austin D. McGuire, repeated the ton-number conservation says that total experiment at the Savannah River reac- lepton number remains constant in weak tor. Figure 2 shows how neu- Although the neutrinos were not ob- processes. This law, which emerges natu- trino-induced inverse beta decay can be served, their existence was assumed, this rally from the assumed mathematical observed. time to preserve lepton-number con- form of weak interactions, is consistent By this time other weak processes servation, a special case of fer- with all observed weak processes and were known, in particular the decay of mien-number conservation. A lepton is a explains the nonoccurrence of processes

LOS ALAMOS SCIENCE 95 SCIENCE IDEAS

that would violate it. Then, in 1957, the startling discovery that parity is not conserved in beta decay led to a major breakthrough in our conception of the neutrino and of weak interactions in general, At the suggestion of T. D. Lee and C. N. Yang, C. S. Wu and her collaborators measured the direction of electrons emitted in the beta decay of polarized Co60. They found that 40% more electrons were emitted along the Co60 spin axis than opposite it. This near-maximum vio- lation of parity, or right-left symmetry, could be explained if the antineutrinos emitted in beta decay exist only in a longitudinally polarized form, that is. with spin vector pointing either along or opposite the direction of motion (right- or left-polarized, respectively), but not both. The property of longitudinal polariza- tion led to a new and very appealing theory of the neutrino. Unlike other known fermions, whose wave functions have four components corresponding to particle and antiparticle in both right- and left-polarized states, the neutrino has only two components: the neutrino is Fig. 3. In the standard two-component neutrino theory, the neutrino is a massless particle with intrinsic spin, or angular momentum, of 1/ h. The neutrino is always always left-polarized, or more precisely. 2 left-handed, and the antineutrino is left-handed, that is, tile direction of its spin vector is opposite that of its momentum; always right-handed.* The other two the antineutrino is always right-handed with the direction of its spin vector the same components (right-handed neutrino and as that of its momentum. This distinction between particle and antiparticle impossible left-handed antineutrino) are missing. only if the neutrino is traveling at the speed of light (and is therefore massless). Such a two-component theory in which Otherwise, transformation to a reference frame moving faster than the neutrino would neutrino and antineutrino are distin- reverse the momentum and cause the neutrino to appear to be an antineutrino. guished by their handedness implies that The massless two-component neutrino weak interactions. Second, the the neutrino travels at the speed of light theory had two important consequences left-handed character of the neutrino and is therefore a massless particle (see for weak interactions. First, since the helped to establish a universal Fig. 3). neutrino and its antiparticle were not the left-handed mathematical form for all weak-interaction processes, thus explain- *Right- or left-handed is not exactly the same as lepton number assignments of + 1 and ing why even massive particles emerging right- or left-polarized except for massless particles, but for this discussion we will ignore the –1, respectively. Thus lepton-number from such processes are partially difference. conservation remained an exact law for polarized.

96 LOS ALAMOS SCIENCE SCIENCE IDEAS

We must emphasize that although the itself change into a member of the elec- tween members of different lepton fami- standard two-component theory implies tron family. For example, a negative lies. that the neutrino is massless, it does not muon cannot decay into an electron prove this as fact, It is possible to unless a muon neutrino is emitted in the Neutrino Oscillation— formulate a two-component theory for a A Recurring Idea massive neutrino consistent with the observed polarizations of participants in y does not. The first suggestion that neutrinos weak interactions. However, this alter- Lepton-number conservation was thus might oscillate—periodically change native formulation (in which the two separated into muon-number conserva- from one form to another—came from components correspond to the right- and tion and electron-number conservation. Bruno Pontecorvo in 1957. He noted left-handed neutrino) does not produce a Again, these conservation laws reflect that if the massless two-component neu- lepton-number conservation law because the observed occurrence or nonoccur- trino theory was wrong, that is, if neu- the neutrino and the antineutrino are the rence of various weak interactions. * trinos were massive and lepton-number Evidence for a third lepton family, the conservation was violated, then the neu- community favored the massless tau and the tau neutrino, has now pro- trino, like the neutral K meson, might two-component theory. It was simple. It duced a tau-number conservation law. oscillate between its particle and anti- was aesthetically appealing. And, most The existence of three neutrino flavors particle forms. This possibility was not important, it preserved lepton-number did not, however, change the notion that explored because at that time the physics conservation. each one was a massless, two- community was just beginning to accept This simple picture of the neutrino component particle. the massless two-component theory of became more complex when, in 1963, That brings us to the present status of the neutrino. experiments showed that neutrinos come theories and observations about the neu- In 1963, as a result of the discovery of in two “flavors.” The quantum number trino. The hypotheses of a massless the muon neutrino, the idea of oscillation flavor was postulated to explain the fact two-component neutrino and lepton- surfaced again—this time between the that antineutrinos from beta decay never number conservation appear correct for electron neutrino and the muon neu- produce positive muons. That is, if we all observed weak interactions and are trino—and in 1969 gained some respect- incorporated into the Glashow- ability as a possible explanation for the then Weinberg-Salam model that unifies the solar neutrino puzzle. (See Kolb’s com- electromagnetic and the weak forces. mentary, “Neutrinos in Cosmology and Although all experiments to date sup- Astrophysics.”) It nonetheless remained port this simple picture of the neutrino, on the fringes of theoretical physics but there is no fundamental reason why the because it violated the empirically estab- neutrino should be massless. In fact, lished conservation laws of muon modern views of fundamental forces number and electron number. which incorporate the Glashow- Only in the last few years has the Thus weak interactions separate the lep- Weinberg-Salam model into a larger possibility of neutrino oscillation been tons into two families, the muon family model suggest that neutrinos have a taken more seriously, Indeed, this consisting of the muon, the muon neu- small rest mass and that effects of this phenomenon is a natural result of radi- trino, and their antiparticles, and the rest mass can be observed through the cally new schemes that attempt a unified electron family consisting of the electron, phenomenon of neutrino oscillation be- description of all fundamental interac- the electron neutrino, and their anti- tions. (See Goldman’s commentary, particles. Further these two families are *Experiments to search for reactions that violate “Neutrinos and Grand Unified Theo- muon-number conservation are discussed by Cy not “mixed” by weak interactions: a Hoffman and Minh Duong-Van in Los Alamos ries.”) To achieve unification, these theo- member of the muon family cannot by Science, I, No. 1,62-67 (1980). ries must postulate that at least some of

LOS ALAMOS SCIENCE 97 SCIENCE IDEAS the empirical number-conservation laws cle. The experiment, performed at the measured with a magnetic spectrometer are not exact. The theories suggest that, Savannah River reactor by a group from and analyzed for evidence of a nonzero at very high energies (or, equivalently, at the University of California at Irvine, neutrino mass. Recent work by sol- very small distances), neutrinos and the provided evidence for neutrino oscilla- id-state physicists at the University of other leptons become indistinguishable tion, evidence that implies a nonzero rest Amsterdam and at the Massachusetts from the quarks, which are the consti- mass for at least one neutrino type. Institute of Technology promises atomic tuents of those particles (such as the Oscillation experiments determine not a beams in the near future of sufficient proton and the pion) that interact value for the mass of a neutrino, but intensity to permit verification of the through the strong force. In this Moscow result. framework neutrinos acquire a nonzero and m2 are the eigenmasses of the neu- Physicists at Los Alamos National rest mass just as do all the other fun- trino mass eigenstates defined by the Laboratory also plan to address the damental constituents of matter. Thus, lepton-mixing mass terms referred to question of neutrino mass by searching the major philosophical objections to above. This difference is said by the for evidence of neutrino oscillation both oscillation have been swept away. Irvine group to be about 1 (eV)2. at LAMPF and at the Nevada Test Site. In fact, the structure of unified theo- The implied existence of massive neu- ries suggests that neutrino oscillation trinos is supported by another recent The Mechanism of Oscillation does occur. The terms in these theories experiment. A group at the Institute of that could give the neutrino its mass are Theoretical and Experimental Physics in Oscillation of one neutrino flavor into not likely to have the same symmetry as Moscow measured the neutrino mass another and back again seems very those describing weak interactions; these directly from a careful analysis of the mysterious because it conjures up the mass terms might mix members of the spectrum of electrons emitted in the beta image of matter suddenly disappearing different lepton families and thereby pro- and just as suddenly reappearing. But duce oscillations between different neu- The deviation between their experimen- this quantum-mechanical phenomenon tally determined spectrum and that ex- becomes more intelligible when we re- produce oscillations between particle pected for a massless neutrino indicates alize that it is analogous in many ways a value for the rest mass of the electron to the classical phenomenon of energy originally suggested by Pontecorvo. antineutrino of at least 14 eV, with a transfer between the two (identical) bobs Further, if the neutrino is massive, it of a double pendulum system (see Fig. 4). might actually be a four-component fer- rest mass of the electron is 511 keV.) We know that if we start the pen- mion whose two unseen components Although no one has yet found fault with dulum by swinging only one bob, after take part in “superweak” interactions this experiment, there is concern that the some time the second bob will be swing- that have yet to be observed. Thus the observed deviation may be due to in- ing and the first will be stationary; still discovery of massive neutrinos through teractions of the electrons as they leave later the second bob will be stationary oscillation experiments or direct meas- the sample, which in this case was solid and the first bob will again be swinging. urements would open up a Pandora’s (tritium in an amino acid deposited on an The change of an electron neutrino into box of new possibilities for the elusive aluminum backing). a muon neutrino and back again is and still puzzling neutrino. To eliminate this uncertainty, a Michi- analogous to this periodic transfer of gan State University-Los Alamos group energy between the two bobs. New Results and New Plans is planning a similar experiment at Los Why does this transfer occur? In the Alamos, in which the source consists of case of the double pendulum it occurs We can now understand the excite- a cold (10 K), monatomic tritium beam. because the pendulums are coupled by a ment created by the announcement in As the beam passes through a l-m long spring between the bobs. In the case of 1980 of the controversial experiment decay region, the spectrum of electrons massive neutrinos it would occur if dif- mentioned at the beginning of this arti- from beta decay of tritium will be ferent neutrino flavors are coupled by

98 LOS ALAMOS SCIENCE SCIENCE IDEAS

The coupled double pendulum has two stationary states, or normal modes. The mode in which the two pendulums swing in phase has a lower than the mode in which the pendulums swing completely (180°) out of phase.

mass eigenstates causes the combination of mass eigenstates representing the electron neutrino to evolve with time into that representing the muon neutrino.

LOS ALAMOS SCIENCE 99 SCIENCE IDEAS the proposed lepton-mixing mass terms. at t = O propagates in time, the relative distance x from the source is given by The time it takes for energy to transfer from one pendulum bob to the other depends on the frequency difference will begin to appear. between the double pendulum’s This is exactly analogous to the time Eqs. (1) and (2), determines the ampli- two normal modes of oscillation; similar- evolution of the double pendulum. We tude of the oscillation. The oscillation ly, the time it takes for one neutrino length L in meters is given by flavor to oscillate into another depends describing the swing of one or the other on the difference between the squares of the two neutrino eigenmasses normal modes, or stationary states, of the pendulum corresponding to the two To understand this dependence we bobs swinging exactly in phase and must continue our analogy on a exactly out of phase, respectively. The mathematical level. Consider the possi- swing of one bob or the other corre- Designing Oscillation Experiments bility that the electron and muon neu- sponds to addition or subtraction of trinos produced in weak processes ac- equal parts of the two normal modes, The equation for Pee(x) says that os- quire mass as a result of new interactions cillation effects are maximum at that mix members of the muon and n/4]. But the two normal modes have half-integer multiples of L. Therefore, to electron families.* Then although these different frequencies. (The mode in detect oscillation the distance between neutrinos do not have definite masses, which the bobs swing in opposite direc- the neutrino source and the detector tions has a higher frequency because the must be at least the same order of described by linear combinations of two spring restricts the separating motion.) Consequently the relative phase between hence L are unknown. the two normal modes changes with time Theoretical considerations suggest that and m2, respectively. so that the additive combination describing the swing of one bob [Eq. (l)] be very large. Consequently, to be sensi- (1) eventually changes into the subtractive combination describing the swing of the should be placed at the farthest distance and second bob [Eq. (2)]. from the source consistent with obtain- Analysis of the electron neutrinos ing a measurable signal. In addition, to (2) time evolution is similar and is given in minimize the oscillation length as much the accompanying note “Derivation of as possible, experiments should exploit Neutrino Oscillation Length.” Here we sources of low-energy neutrinos. The extent to which ve, and vµ are mixed by simply state the results. As an electron most convincing experiments will be the new interaction. If ml does not equal neutrino produced at t = O travels those that demonstrate, by changes in mz, oscillation between ve and vµ will through space at almost the speed of the source-detector distance, the vari- light, c, it will periodically disappear and ation with distance of the number of reappear over a characteristic length neutrinos of a particular type. Neutrinos of a particular type are *In actuality, we may be dealing with the mixing of more than just two neutrino types. The for- trino oscillation. L is called the oscilla- detected by observing the reactions they malism, however, remains the same. Ultimately, tion length. Thus the probability Pee(x) induce in a detecting medium. For exam- the question of which neutrinos mix with which no ple, if protons are the medium and doubt depends on how many different neutrinos that an electron neutrino has not os- exist. cillated into another type of neutrino at a electron antineutrinos from beta decay

100 LOS ALAMOS SCIENCE . Derivation of Neutrino SCIENCE IDEAS

Oscillation Length are the source, a reaction to observe is

ome simple algebra can show how neutrino oscillation effects depend on the The signature of this reaction is the mass difference of the neutrino mass eigenstates. We express the quan- delayed coincidence of photons from s tum-mechanical wave function for an electron neutrino produced at t = O as a annihilation of the positron and from capture of the neutron. A decrease in the rate of this reaction from that predicted for massless neutrinos would be evidence for neutrino oscillation. (An experiment where 0, the mixing angle, characterizes the extent of mixing of the mass eigenstates in of this kind is known as a disappearance the weak-interaction eigenstate. At a later time t, the wave function is experiment.) To determine such a change in an already low reaction rate is a tremendous challenge. The experiments require extensive shielding to minimize back-

we can approximate El and E2 by ground events and sophisticated detectors to differentiate background from events of interest. Detectors are similar in concept to that used by Reines and Cowan, but they are more sensitive. Made up of multiple units monitored by modern electronics, these detectors obtain fine-grained information about the sources of the measured signals. speed of light, we can replace t by x/c, where x is the distance from the source of Despite the size of present detectors, electron neutrinos. Then we calculate Pee(x), the probability of finding an electron the expected count rates are still quite neutrino at x. low. For example, at LAMPF with a typical neutrino flux of 1010 neutrinos per second incident on a 50-ton detector (a not abnormally large size) located 100 m from the LAMPF beam stop, one can expect to observe an interaction of interest in the detector only once every three or four hours. obtain As diffiicult as these experiments may be, their implications for grand unified theories and cosmology provide a com- pelling motive to attempt them. Thus the probability oscillates with distance and the oscillation length L is given in meters by Oscillation Experiments So Far

The Irvine group, who have claimed positive results, used neutrinos from the Savannah River reactor to search for the

LOS ALAMOS SCIENCE 101 SCIENCE IDEAS

experiment performed in Grenoble at the check calculations of the neutrino spec- ured the cross sections for both a Laue-Langevin Institute’s reactor. A tra. Preliminary data from a direct meas- “charged current” and a “neutral cur- group from the California Institute of urement by the Grenoble group support rent” reaction* induced by an electron Technology, the Nuclear Sciences In- their calculated spectrum. This group antineutrino on a deuteron: stitute (Grenoble), and the Technical has since moved its experiment to a University of Munich also searched for reactor in Gösgen, Switzerland, where they plan to obtain data at termining the rate of inverse beta decay. source-detector distances of 35-80 m. The detecting medium, located 8.7 m The Irvine group, in turn, has begun from the reactor core, consisted of 375 direct spectrum measurements at the The detecting medium, located 11,2 m liters of proton-rich liquid scintillator for Savannah River reactor and is obtaining from the reactor core, consisted of 268 photon detection and neutron mod- further oscillation data at distances of kg of heavy water in which were placed eration and 3He-filled proportional coun- 11-38 m. 10 3He-filled proportional counters for ters for neutron detection. Using a neu- Other information on oscillation exists measuring coincident and single neutron trino energy spectrum calculated by from an experiment performed at counts. Liquid-scintillator anti- Davis and Vogel, the group found no LAMPF by a group primarily from Yale coincidence counters surrounded the de- deviation from the expected rate of in- University and Los Alamos. This experi- tector, which was enclosed in a lead and verse beta decay and therefore con- cadmium shield. cluded that oscillation at the level seen The cross section for the neutral cur- by the Irvine group does not occur. For rent reaction is independent of the type of neutrino that initiates the reaction. noble experiment set* an upper limit on by these experiments. The charged current reaction, in con- Whether the evidence for neutrino trast, can be induced only by the elec- oscillation is confirmed or eventually tron antineutrino. If neutrino oscillation refuted, these experiments have stimu- occurs, the count rate for two-neutron The largest uncertainty in the results lated tremendous interest in the scientific events will be reduced relative to the of both these experiments lies in the community and many experimental count rate for one-neutron events. energy spectra of neutrinos from the groups are planning even more sensitive From their data, the Irvine group reactors, which for the original analyses experiments to determine neutrino derived a “ratio of ratios” (the ratio of were calculated theoretically. The Irvine masses and mixing angles. the experimental cross sections for the group claims that its ratio of ratios is two reactions divided by the ratio of the insensitive to both experimental and the- Neutrinos at LAMPF theoretical cross sections). For massless oretical uncertainties, but both groups neutrinos this ratio of ratios should be are now performing experiments to As a source of neutrinos, LAMPF unity. The Irvine group’s currently re- offers two great advantages: neutrino *As reported (at the 90% confidence level) by F. fluxes much greater than those of any claim that neutrino oscillation does oc- Boehm et al. in “Neutrino Oscillation Experiment other accelerator facility, and low neu- cur. at the ILL Reactor at Grenoble,” presentation at trino energies (20-53 MeV) particularly the International Conference on Neutrino Physics The controversy associated with this and Astrophysics: Neutrino 80, Erice, Sicily, June suited to oscillation experiments. It also experiment is due to the fact that its 23-27, 1980 (to be published). offers a unique opportunity to study **As reported (at the 68% confidence level) by F. results are not confirmed by another Reines et al. in “Evidence for Neutrino Instabili- muon neutrino oscillations at low ty,” presentation at the Spring Meeting of the *Charged current or neutral current refers to the American Physical Society, Washington, D. C., charge of the hypothetical intermediate vector April 28-May 1, 1980 and submitted to Physical boson that mediates the reaction. Review Letters. are at the 90% confidence level.

102 LOS ALAMOS SCIENCE SCIENCE IDEAS

Fig. 5. Results of completed neutrino experiments appear to be contradictory. The Irvine group’s positive results for

(claimed at the 68% confidence level are shown by the unshaded region. Negative results from other experiments a function of 0, shown here at the 90% confidence level. According to these lat- upper limit curves.

LOS ALAMOS SCIENCE 103

SCIENCE IDEAS energies. Neutrinos at LAMPF are produced by the decay of pions and muons. Pro- tons that remain after passing through the upstream targets are brought to rest in the beam stop at the end of the accelerator. In the process, they collide with atoms in the beam stop to produce photons and the negative pions are almost completely absorbed in nuclear reactions. The positive pions come to

+ vµ. The positive muon in turn decays that these two decays yield no electron antineutrinos. The two oscillation experiments scheduled to begin at LAMPF’ in early 1981 take advantage of this fact. Both are designed to search for the transition by determining the rate of inverse beta decay. Occurrence of this reaction at a rate significantly higher than expected from background events will be clear evidence of neutrino oscillation. (Such experiments are called appearance experiments because electron Marilyn Barlett is shown assembling several of the 600 d@ tube modules to be used antineutrinos appear in a beam that in the anticoincidence counter of the Irvine-Los A lames neutrino oscillation experi- originally had none.) ment at LAMPF. One experiment, an Irvine-Los Alamos collaboration, is an outgrowth of a long-planned study of the elastic scattering of electron neutrinos by electrons. The heart of this experiment is a 14-ton “sandwich” detector to be located inside the neutrino facility, an “iron house” about 10 m from the beam stop. The detector, which contains 10 tons of plastic scintillator and 4 tons of polypropylene flash chambers, will measure the energy, position, and direction of motion of the recoil electrons

LOS ALAMOS SCIENCE 105

SCIENCE IDEAS

from elastic neutrino-electron scattering. simultaneously an appearance and a and the total number of counts expected In the search for neutrino oscillation, the disappearance experiment (and could from each test is quite small. It may be same quantities will be measured for the thus serve to refute or verify the Irvine necessary to use several detectors on positrons from inverse beta decay reac- group’s results). each of several tests. The return for such tions. The Irvine-Los Alamos group ex- Neither of these experiments in their extra effort, however, is the possibility of

2 currently approved forms constitutes 0.4 (eV) , assuming maximum neutrino 0.0005-0.005 (eV)2, values that may not what has been called the definitive test: mixing. be achieved in other ways. detection of the same transition at the The other scheduled experiment, a same neutrino energy but at varying Los Alamos effort, is in some sense a The Next Generation Experiments source-detector distances. However, preliminary experiment: its first objective when the 5-ton Los Alamos detector is is to test a 5-ton detector being as- The high interest in oscillation moved to the Nevada Test Site, it will, sembled at LAMPF but planned for phenomena and the need to explore all again if warranted by observation of eventual use in neutrino oscillation ex- channels of neutrino mixing have positive results, be positioned at various periments at the Nevada Test Site where focused continuing attention on source-detector distances. tests of fission weapons provide an ex- LAMPF, where an intense source of tremely brief but extremely intense pulse Three new proposals for experiments muon neutrinos as well as electron neu- of neutrinos. The detector, containing at LAMPF involve the search for oscilla- trinos will enable experimentalists to 4470 liters (-5 ton) of gadolinium- tion effects as a function of distance. We search for oscillation with increased sen- loaded liquid scintillator and surrounded will briefly discuss the Nevada Test Site sitivity. At this writing, three such pro- by an anticoincidence shield, will be experiment and three proposed LAMPF posals have been submitted to and relocated 33 m from the LAMPF beam experiments. viewed by the Program Advisory Com- stop in a hole drilled in tuff and will be mittee, and its recommendations are covered with about 8 m of sand and iron Table I summarizes the status and awaiting review by Louis Rosen, Direc- for shielding. If cosmic-ray backgrounds basic parameters of all the neutrino tor of LAMPF, are found to be sufficiently low, the oscillation experiments discussed. Two of these experiments are repre- sentative of neutrino experiments that Neutrinos at the Nevada Test Site can be done at the LAMPF beam stop. from inverse beta decay will be detected. One proposal is by a group from Rice During a 110-day run with one detector The Los Alamos experiment at the University, the University of Houston, (the possibility of more detectors is being Nevada Test Site, a disappearance ex- and Los Alamos; the other is by a group considered), the group hopes to set an periment like those performed at reac- from Ohio State University, Argonne tors, will be a search for a rate of the National Laboratory, Louisiana State warranted by observation of positive University, and the California Institute results, the detector can, with modest than expected in the absence of oscilla- of Technology. In both experiments, de- effort, be relocated at other tion. Here, sensitivity to small values of tectors will be placed at distances from source-detector distances. Background the source that are variable and greater information garnered by the group will because of the larger source-detector than those to be investigated by the also be used by later experiments. distances (200-1200 m). Test of a fission Irvine-Los Alamos group. Variable dis- The group also believes that it may be weapon produces a short but very in- tance helps to eliminate uncertainties due tense pulse of electron antineutrinos. to beam-associated backgrounds and anything by determining the rate of the Signal-to-noise ratios will therefore be neutrino flux magnitudes, high. On the other hand, ground motion The Rice-Houston-Los Alamos group e–. If so, then their experiment would be from the blast creates huge problems proposes an appearance experiment.

LOS ALAMOS SCIENCE 107 SCIENCE IDEAS

of 150 MeV. At such energies the reac- + P + e- and thus to search for the + µ- are possible. Observation of these than expected from background events. The Ohio State-Argonne-Louisiana- reactions at rates lower than expected For positron tracking and energy meas- Cal Tech group proposes both disap- will indicate the occurrence of the transi- urement, the group will use a 40-ton pearance and appearance experiments. array of liquid scintillator modules and The former will consist of a search for drift chambers. Detector-source distances will be in the range 50-75 m with the inverse beta decay reactions ve + n the detector placed in a deep tunnel for p + e-. Properties of the emitted elec- cosmic-ray shielding. The projected up- trons will be measured with a detector proposed 50-ton detector, to consist of consisting of drift chambers in heavy liquid scintillator and drift chambers, will positrons are detected, and 0.06 (eV)2 water. With source-detector distances of be positioned at source-detector dis- if, as is hoped, both neutrons and 25-60 m, the group expects to set an tances of 40-300 m. positrons are detected simultaneously. In addition, this experiment may pro- tion of ordinary water in the detector will various proposed experiments are shown vide information about the muon neu- permit the group to perform an ap- in Fig. 6. trino lifetime, which is of relevance to pearance experiment by searching for detecting the neutrino background as- Conclusion sumed to exist as a remnant of the Big the rate of inverse beta decay. In addi- Bang. If a muon neutrino decays to a tion, if the group’s proposal to build a The neutrino has been one of the most lower-mass neutrino plus a mono- more energetic muon neutrino source at successful inventions of theoretical phys- energetic photon, the decay could be the LAMPF beam dump is approved, ics. In the 50 years since its existence detected by observing the elec- the group will search for disappearance was first postulated, it has profoundly tron-positron pairs produced by the of muon neutrinos by determining the altered our theories of weak interactions photon. The group claims they can im- and of astrophysical phenomena. Now prove the limit on the muon neutrino p. we realize that if this particle has even a lifetime from its present value, (2.6 X The third proposed experiment was very small mass, it can influence our conceived by a Los Alamos group and understanding of all fundamental in- calls for the construction of a new beam teractions and the dynamics of the cos- mass. line at LAMPF. Here pions, created by mos as well. The new experiments to The Rice-Houston-Los Alamos group collision of part of the proton beam with measure neutrino masses and mixing may also perform a disappearance ex- a thin target, will decay in flight in a angles will just begin to explore the periment. Replacement of liquid scin- decay region to produce muon neutrinos many possibilities created by the ex- tillator with heavy water would permit and antineutrinos with average energies

108 LOS ALAMOS SCIENCE Fig. 6. The approved and proposed experiments at LAMPF and at the Nevada Test Site may not observe neutrino oscillation effects. Even so, they will establish upper limits on

angle 0. The curves here are the upper limits (at the 90% confidence level) projected for the various experiments.

SCIENCE IDEAS

LOS ALAMOS SCIENCE 109 SCIENCE IDEAS

Thomas J. Bowles, Robert L. Burman, Herbert H. Chen, Los Alamos National Laboratory Los Alamos National Laboratory University of California at Irvine

Scientists Working on Current Neutrino Experiments at Los Alamos.

Experimental Collaborations

University of California at Irvine: E. Pasierb, F. Reines (spokesman), and H. W. Sobel

California Institute of Technology: F. Boehm (spokesman), A. A. Hahn, H. E. Henrikson, H. Kwon, and J. L. Vuilleumier Nuclear Sciences Institute (Grenoble): J. F. Cavaignac, D. H. Koang, and B. Vignon Technical University of Munich: F. v. Feilitzsch, R. L. Mossbauer, and V. Zacek

Yale University: V. W. Hughes, P. Nemethy (spokesman), and S. E. Willis Los Alamos National Laboratory: R. L. Burman, D. R. F. Cochran, J. Frank, and R. P. Redwine Saclay Nuclear Research Center: J. Duclos National Research Council of Canada: C. K. Hargrove Swiss Institute for Nuclear Research: H. Kaspar University of Berne: U. Moser

University of California at Irvine: R. C. Allen, G. A. Brooks, H. H. Chen (co-spokesman), P. J. Doe, D. Hamilton, F. Reines (co-spokesman), A. N. Rushton, and K. C. Wang Los Alamos National Laboratory: T. J. Bowles, R. L. Burman, R. Carlini, D. R. F. Cochran, T. Dombeck,* M. Duong-Van, J. Frank, V. Sandberg, and R. Talaga

Los Alamos National Laboratory at LAMPF (approved): D. Bartram, A. Brown, H. W. Kruse (spokesman), J. Mack, C. Smith, and J. Toevs

Los Alamos National Laboratory at the Nevada Test Site: D. Bartram, A. Brown, H. W. Kruse (spokesman), R. Loncoski, J. Mack, C. Smith, and J. Toevs

Rice University: G. C. Phillips (co-spokesman), H. Miettinen, G. S. Mutchler, and J. B. Roberts University of Houston: A. D. Hancock, B. W. Mayes, and L. S. Pinsky Los Alamos National Laboratory: J. C. Allred, A. A. Browman, R. L. Burman, D. R. F. Cochran, J. B. Donahue, M. Duong-Van (co-spokesman), M. V. Hynes, and B. W. Noel

110 LOS ALAMOS SCIENCE SCIENCE IDEAS

Thomas W. Dombeck, Visiting Staff Member, Minh Duong-Van, Herald W. Kruse, Los Alamos National Laboratory Los Alamos National Laboratory Los Alamos National Laboratory

Ohio State University: T. Y. Ling (co-spokesman) and T. A. Romanowski (co-spokesman) Argonne National Laboratory: L. G. Hyman and B. Musgrave Louisiana State University: R Imlay and W. J. Metcalf California Institute of Technology: R. B. McKeown

Los Alamos National Laboratory at LAMPF (proposed): T. J. Bowles, R. L. Burman, and T. Dombeck* (spokesman)

Michigan State University: A. Ledebuhr and R. G. H. Robertson (co-spokesman) Los Alamos National Laboratory: T. J. Bowles (co-spokesman), J. Browne, R. Hardekopf, and R. Mills

*On leave from the University of Maryland.