Detecting Massive NeutrinosA giant detector in the heart of Mount Ikenoyama in Japanhas demonstrated that the neutrino metamorphoses in flight,strongly suggesting that these ghostly particles have massby Edward Kearns, Takaaki Kajita and Yoji TotsukaCopyright 1999 Scientific American, Inc.
One man’s trash is another man’s treasure. Fora physicist, the trash is “background”—someunwanted reaction, probably from a mundaneand well-understood process. The treasure is “signal”—areaction that we hope reveals new knowledge about theway the universe works. Case in point: over the past twodecades, several groups have been hunting for the radioactivedecay of the proton, an exceedingly rare signal (if it occursat all) buried in a background of reactions caused byelusive particles called neutrinos. The proton, one of themain constituents of atoms, seems to be immortal. Its decaywould be a strong indication of processes described byGrand Unified Theories that many believe lie beyond theextremely successful Standard Model of particle physics.Huge proton-decay detectors were placed deep underground,in mines or tunnels around the world, to escapethe constant rain of particles called cosmic rays. But nomatter how deep they went, these devices were still exposedto penetrating neutrinos produced by the cosmic rays.The first generation of proton-decay detectors, operatingfrom 1980 to 1995, saw no signal, no signs of protondecay—but along the way the researchers found that thesupposedly mundane neutrino background was not soeasy to understand. One such experiment, Kamiokande,was located in Kamioka, Japan, a mining town about250 kilometers (155 miles) from Tokyo (as the neutrinoSUPER-KAMIOKANDE DETECTOR residesin an active zinc mine inside Mount Ikenoyama.Its stainless-steel tank contains 50,000 tons ofultrapure water, so transparent that light canpass through nearly 70 meters before losing halfits intensity (for a typical swimming pool the figureis a few meters). The water is monitored by11,000 photomultiplier tubes that cover thewalls, floor and ceiling. Each tube is a handblown,evacuated glass bulb half a meter in diameter,coated on the inside with a thin layer ofalkali metal. The photomultiplier tubes registerconical flashes of Cherenkov light, which signaleach rare collision of a high-energy neutrinowith an atomic nucleus in the water. Techniciansin inflatable rafts clean the bulbs while the tankis being filled (inset).INSTITUTE FOR COSMIC RAY RESEARCH (ICRR), UNIVERSITY OF TOKYODAVID FIERSTEINDetecting Massive Neutrinos Scientific American August 1999 65Copyright 1999 Scientific American, Inc.
UON-NEUTRINOMUONLECTRON-NEUTRINOELECTRONSHOWERCHERENKOV LIGHTCONES OF CHERENKOV LIGHT are emitted when high-energy neutrinos hit anucleus and produce a charged particle. A muon-neutrino (top) creates a muon,which travels perhaps one meter and projects a sharp ring of light onto the detectors.An electron, produced by an electron-neutrino (bottom), generates a small showerof electrons and positrons, each with its own Cherenkov cone, resulting in a fuzzyring of light. Green dots indicate light detected in the same narrow time interval.flies). The name stood for “KamiokaNucleon Decay Experiment.” Scientiststhere and at the IMB experiment, locatedin a salt mine near Cleveland, Ohio,used sensitive detectors to peer into ultrapurewater, waiting for the telltaleflash of a proton decaying.Such an event would have been hidden,like a needle in a small haystack,among about 1,000 similar flashescaused by neutrinos interacting withthe water’s atomic nuclei. Although noproton decay was seen, the analysis ofthose 1,000 reactions uncovered a realtreasure—tantalizing evidence that theneutrinos were unexpectedly fickle,changing from one species to another inmidflight. If true, that phenomenonwas just as exciting and theory-bendingas proton decay.Neutrinos are amazing, ghostly particles.Every second, 60 billion of them,mostly from the sun, pass through eachsquare centimeter of your body (and ofeverything else). But because they seldominteract with other particles, generallyall 60 billion go through youwithout so much as nudging a singleatom. In fact, you could send a beam ofsuch neutrinos through a light-year oflead, and most of them would emergetotally unscathed at the far end. A detectoras large as Kamiokande catchesonly a tiny fraction of the neutrinosthat pass through it every year.Neutrinos come in three flavors, correspondingto their three charged partnersin the Standard Model: the electronand its heavier relatives, the muon andthe tau particle. An electron-neutrino interactingwith an atomic nucleus canproduce an electron; a muon-neutrinomakes a muon; a tau-neutrino, a tau. Formost of the seven decades since neutrinoswere first posited, physicists have assumedthat they are massless. But if theycan change from one flavor to another,quantum theory indicates that they mostICRR, UNIVERSITY OF TOKYOlikely have mass. And in that case, theseethereal particles could collectively outweighall the stars in the universe.Building a Bigger Neutrino TrapAs is so often the case in particlephysics, the way to make progressis to build a bigger machine. Super-Kamiokande, or Super-K for short,took the basic design of Kamiokandeand scaled it up by about a factor of 10[see illustration on page 64]. An arrayof light-sensitive detectors looks in towardthe center of 50,000 tons of waterwhose protons may decay or get struckby a neutrino. In either case, the reactioncreates particles that are spotted bymeans of a flash of blue light known asCherenkov light, an optical analogue ofa sonic boom, discovered by Pavel A.Cherenkov in 1934. Much as an aircraftflying faster than the speed of soundproduces a shock wave of sound, anelectrically charged particle (such as anelectron or muon) emits Cherenkovlight when it exceeds the speed of lightin the medium in which it is moving.This motion does not violate Einstein’stheory of relativity, for which the crucialvelocity is c, the speed of light in a vacuum.In water, light propagates 25 percentslower than c, but other highly energeticparticles can still travel almost asfast as c itself. Cherenkov light is emittedin a cone along the flight path ofsuch particles.In Super-K, the charged particle generallytravels just a few meters and theCherenkov cone projects a ring of lightonto the wall of photon detectors [see illustrationon this page]. The size, shapeand intensity of this ring reveal the propertiesof the charged particle, which inturn tell us about the neutrino that producedit. We can easily distinguish theCherenkov patterns of electrons fromthose of muons: the electrons generate ashower of particles, leading to a fuzzyring quite unlike the crisper circle from amuon. From the Cherenkov light we alsomeasure the energy and direction of theelectron or muon, which are decent approximationsto those of the neutrino.Super-K cannot easily identify the thirdtype of neutrino, the tau-neutrino. Such aneutrino can only interact with a nucleusand make a tau particle if it has enoughenergy. A muon is about 200 times asheavy as an electron; the tau about 3,500times. The muon mass is well within therange of atmospheric neutrinos, but onlya tiny fraction are at tau energies, so66 Scientific American August 1999 Detecting Massive NeutrinosCopyright 1999 Scientific American, Inc.
ZENITH(perpendicular toearth surface)most tau-neutrinos in the mix will passthrough Super-K undetected.One of the most basic questions experimentersask is, “How many?” Wehave built a beautiful detector to studyneutrinos, and the first task is simply tocount how many we see. Hand in handwith this measurement is the question,“How many did we expect?” To answerthat, we must analyze how theneutrinos are produced.Super-K monitors atmospheric neutrinos,which are born in the spray ofparticles when a cosmic ray strikes thetop of our atmosphere. The incomingprojectiles (called primary cosmic rays)are mostly protons, with a sprinkling ofheavier nuclei such as helium or iron.Each collision generates a shower ofsecondary particles, mostly pions andmuons, which decay during their shortflight through the air, creating neutrinos[see illustration at right]. We knowroughly how many cosmic rays hit theatmosphere each second and roughlyhow many pions and muons are madein each collision, so we can predict howmany neutrinos to expect.ATMOSPHEREZENITHθθOSCILLATING NEUTRINOθθSUPER-KINCOMINGCOSMIC RAYSTricks with RatiosUnfortunately, this estimate is onlyaccurate to 25 percent, so we takeadvantage of a common trick: often theratio of two quantities can be better determinedthan either quantity alone. ForSuper-K, the key is the sequential decayof a pion to a muon and a muon-neutrino,followed by the muon’s decay to anelectron, an electron-neutrino and anothermuon-neutrino. No matter how manycosmic rays are falling on the earth’s atmosphere,or how many pions they produce,there should be about two muonneutrinosfor every electron-neutrino.The calculation is more complicated thanthat and involves computer simulationsof the cosmic ray showers, but the finalpredicted ratio is accurate to 5 percent,providing a much better benchmark thanthe individual numbers of particles do.After counting neutrinos for almosttwo years, the Super-K team has foundthat the ratio of muon-neutrinos toelectron-neutrinos is about 1.3 to 1 insteadof the expected 2 to 1. Even if westretch our assumptions about the fluxof neutrinos, how they interact with thenuclei and how our detector respondsto these events, we cannot explain sucha low ratio—unless neutrinos are changingfrom one type into another.We can play the ratio trick again toDAVID FIERSTEINCOSMICRAYAIR NUCLEUSPIONSHIGH-ENERGY COSMIC RAY striking a nucleusin the atmosphere (below) generates a showerof particles, mostly pions. The pion’s sequence ofdecays produces two muon-neutrinos for everyelectron-neutrino. Equal neutrino rates should beseen from opposite directions (above) becauseboth result from cosmic rays hitting the atmosphereat the same zenith angle, θ. Both these ratiosare spoiled when muon-neutrinos travelinglong distances have time to change flavor.MUON2 MUON-NEUTRINOS1 ELECTRON-NEUTRINOELECTRONDetecting Massive Neutrinos Scientific American August 1999 67Copyright 1999 Scientific American, Inc.
MUON-NEUTRINO CREATED INTHE UPPER ATMOSPHEREPION(DECAYS)MUONHow Quantum Waves Make a Neutrino OscillateTWO WAVE PACKETS OF DIFFERENTMASS TRAVEL AT DIFFERENT VELOCITIESWHICH FLAVOR ISDETECTED DEPENDSON THE INTERFERENCEMUONORINTERFERENCE PATTERN OF WAVE PACKETS DETERMINES PROBABILITY OF THE NEUTRINO FLAVOR100% MUON-NEUTRINO0% MUON-NEUTRINO100% MUON-NEUTRINONOT ENOUGHENERGYTO MAKE TAU0% TAU-NEUTRINO100% TAU-NEUTRINO0% TAU-NEUTRINOWhen a pion decays (top left), it produces a neutrino. Describedquantum-mechanically, the neutrino is apparentlya superposition of two wave packets of different mass(purple and green; top middle).The wave packets propagate atdifferent speeds,with the lighter wave packet getting ahead ofthe heavier one.As this proceeds,the waves interfere,and theinterference pattern controls what flavor neutrino—muon(red) or tau (blue)—one is most likely to detect at any pointalong the flight path (bottom).Like all quantum effects,this is agame of chance, with the chances heavily favoring a muonneutrinoclose to where it was produced.But the probabilitiesoscillate back and forth, favoring the tau-neutrino at just theright distance and returning to favor the muon-neutrino fartheron. When the neutrino finally interacts in the detector (topright), the quantum dice are rolled. If the outcome is muonneutrino,a muon is produced. If chance favors the tau-neutrino,andthe neutrino does not have enough energy to create atau particle,Super-K detects nothing. —E.K.,T.K.and Y.T.LAURIE GRACEtest this surprising conclusion. The clueto our second ratio is to ask how manyneutrinos should arrive from each possibledirection. Primary cosmic rays fall onthe earth’s atmosphere almost equallyfrom all directions, with only two effectsspoiling the uniformity. First, the earth’smagnetic field deflects some cosmic rays,especially the low-energy ones, skewingthe pattern of arrival directions. Second,cosmic rays that skim the earth at a tangentmake showers that do not descenddeep into the atmosphere, and these candevelop differently from those thatplunge straight in from above.But geometry saves us: if we “look”up into the sky at some angle from thevertical and then down into the groundat the same angle, we should “see” thesame number of neutrinos coming fromeach direction. Both sets of neutrinosare produced by cosmic rays hitting theatmosphere at the same angle; it is justthat in one case the collisions happenoverhead and in the other they are partwayaround the world [see illustrationon preceding page]. To use this fact, weselect neutrino events of sufficientlyhigh energy (so their parent cosmic raywas not deflected by the earth’s magneticfield) and then divide the numberof neutrinos going up by the numbergoing down. This ratio should be exactly1 if no neutrinos are changing flavor.We saw essentially equal numbers ofhigh-energy electron-neutrinos going upand down, as expected, but only half asmany upward muon-neutrinos as downwardones. This finding is the second indicationthat neutrinos are changingidentity. Moreover, it provides a clue tothe nature of the metamorphosis. Theupward muon-neutrinos cannot be turninginto electron-neutrinos, because thereis no excess of upward electron-neutrinos.That leaves the tau-neutrino. Themuon-neutrinos that become tau-neutrinospass through Super-K without interaction,without detection.Fickle FlavorThe above two ratios are good evidencethat muon-neutrinos aretransforming into tau-neutrinos, but whyshould neutrinos switch flavor at all?Quantum physics describes a particlemoving through space by a wave: in additionto properties such as mass andcharge, the particle has a wavelength, it19301933195619621969Wolfgang Pauli rescuesconservation of energy byhypothesizing an unseen particlethat takes away energy missingEnrico Fermi formulatesthe theory of beta-decayincorporating Pauli's particle,now called the neutrinoFrederick Reines (center) andClyde Cowen first detect theneutrino using the SavannahRiver nuclear reactorAt Brookhaven, the firstaccelerator beam of neutrinosproves the distinctionbetween electron neutrinosRaymond Davis, Jr., firstmeasures neutrinos fromthe sun, using 600 tons ofcleaning fluid in a mine in68 Scientific American August 1999 Detecting Massive NeutrinosCopyright 1999 Scientific American, Inc.
can diffract, and so on. Furthermore, aparticle can be the superposition of twowaves. Now suppose that the two wavescorrespond to slightly different masses.Then, as the waves travel along, thelighter wave gets ahead of the heavierone, and the waves interfere in a waythat fluctuates along the particle’s trajectory[see box on opposite page]. Thisinterference has a musical analogue: thebeats that occur when two notes are almostbut not exactly the same.In music this effect makes the volumeoscillate; in quantum physics it is theprobability of detecting one type ofneutrino or another that oscillates. Atthe outset the neutrino appears as amuon-neutrino with a probability of100 percent. After traveling a certaindistance, it looks like a tau-neutrinowith 100 percent probability. At otherpositions, it could be either a muonneutrinoor a tau-neutrino, dependingon the roll of the dice.This oscillation sounds like bizarrebehavior for a particle, but another familiarparticle performs similar contortions:the photon, the particle of light.Light can occur in a variety of polarizations,including vertical, horizontal, leftcircular and right circular. These do nothave different masses (all photons aremassless), but in certain optically activematerials, light with left circular polarizationmoves faster than right circularlight. A photon with vertical polarizationis actually a superposition of thesetwo alternatives, and when it is traversingan optically active material its polarizationwill rotate (that is, oscillate)from vertical to horizontal and so on,as its two circular components go inand out of sync.For neutrino oscillations of the typewe see at Super-K, no “optically active”material is needed; a sufficient mass differencebetween the two neutrino componentswill cause flavor oscillationswhether the neutrino is passing throughair, solid rock or pure vacuum. When aNUMBER OF MUON-NEUTRINO EVENTS200150100500PREDICTION WITHOUT NEUTRINO OSCILLATIONPREDICTION WITH NEUTRINO OSCILLATIONSUPER-KAMIOKANDE MEASUREMENTARRIVAL ANGLE AND DISTANCE TRAVELED BY NEUTRINO12,800 km 6,400 km 500 km 30 km 15 kmNUMBER OF HIGH-ENERGY MUON-NEUTRINOS seen arriving on different trajectoriesat Super-K clearly matches a prediction incorporating neutrino oscillations(green) and does not match the no-oscillation prediction (blue). Upward-going neutrinos(plotted toward left of graph) have traveled far enough for half of them to changeflavor and escape detection.neutrino arrives at Super-K, the amountit has oscillated depends on its energyand the distance it has traveled since itwas created. For downward muon-neutrinos,which have traveled at most afew dozen kilometers, only a small fractionof an oscillation cycle has takenplace, so the neutrinos’ flavor is onlyslightly shifted, and we are nearly certainto detect their original muon-neutrinoflavor [see illustration on page67]. The upward muon-neutrinos, producedthousands of kilometers away,have gone through so many oscillationsthat on average only half of them canbe detected as muon-neutrinos. Theother half pass through Super-K as undetectabletau-neutrinos.This description is just a rough picture,but the arguments based on the ratio offlavors and the up/down event rate are socompelling that neutrino oscillation isnow widely accepted as the most likelyexplanation for our data. We also havedone more detailed studies of how thenumber of muon-neutrinos varies accordingto the neutrino energy and thearrival angle. We compare the measurednumber against what is expected for awide array of possible oscillation scenarios(including no oscillations). The datalook quite unlike the no-oscillation expectationbut match well with neutrinooscillation for certain values of the massdifference and other physical parameters[see illustration above].With about 5,000 events from ourfirst two years of running the experi-LAURIE GRACETIMELINE BY HEIDI NOLAND; AIP EMILIO SEGRÉ VISUAL ARCHIVES (Pauli and Reines); COURTESY OF THE INSTITUTE FORADVANCED STUDY, PRINCETON (Davis); DAVID MALIN Anglo-Australian Observatory (SN1987A); ICRR (Super-Kamiokande)1975–1977The tau lepton and b quarkare discovered, revealing athird generation of quarksand leptons1983W and Z bosons are discoveredat CERN: they are the carriersof the weak force, whichmediates neutrino reactions1987Neutrino astronomy: the IMBand Kamiokande proton decayexperiments detect 19 neutrinosfrom Supernova 1987A in the19890 0The Z decay rate isprecisely measured atSLAC and CERN, showingthere are only three active1998Super-K assemblesevidence of neutrinooscillation usingatmospheric neutrinosDetecting Massive Neutrinos Scientific American August 1999 69Copyright 1999 Scientific American, Inc.
Another implication is that the neutrinomass should now be considered inthe bookkeeping of the mass of the universe.For some time, astronomers havebeen trying to tabulate how much massis found in luminous matter, such asstars, and in ordinary matter that isdifficult to see, such as brown dwarfsor diffuse gas. The total mass can alsobe measured indirectly from the orbitalmotion of galaxies and the rate of expansionof the universe. The direct accountingfalls short of these indirectmeasures by about a factor of 20. Theneutrino mass suggested by our result istoo small to resolve this mystery by itself.Nevertheless, neutrinos createdduring the big bang permeate space andcould account for a mass nearly equalto the combined mass of all the stars.They could have affected the formationof large astronomical structures, suchas galaxy clusters.Finally, our data have an immediateimplication for two experiments thatare soon to commence. Based on theearlier hints from smaller detectors,many physicists have decided to stoprelying on the free but uncontrollableneutrinos from cosmic rays and insteadare creating them with high-energy accelerators.Even so, the neutrinos musttravel a long distance for the oscillationeffect to be observed. So the neutrinobeams are aimed at a detector hundredsof kilometers away. One such detectoris being built in a mine in Soudan,Minn., optimized to study neutrinossent from the Fermilab accelerator nearBatavia, Ill., 730 kilometers away onthe outskirts of Chicago.Of course, a good atmospheric neutrinodetector is also a good acceleratorneutrino detector, so in Japan we areusing Super-K to monitor a beam ofneutrinos created at the KEK acceleratorlaboratory 250 kilometers away.Unlike atmospheric neutrinos, thebeam can be turned on and off and hasOther Puzzles, Other PossibilitiesThere are other indications of neutrino mass that particle physicists are tryingto sort out. For more than 30 years, scientists have been capturingsome of the electron-neutrinos that are generated by nuclear fusion processesin the sun. These experiments have always counted fewer neutrinos thanthe best models of the sun predict [see“The Solar-Neutrino Problem,”by JohnN. Bahcall; SCIENTIFIC AMERICAN, May 1990].Super-K has also counted these solar neutrinos,finding only about 50 percentof what is expected.We are studying these data,hoping to find a clear signatureof neutrino oscillations. In May the Sudbury Neutrino Observatory in Ontariodetected its first neutrinos. It uses 1,000 tons of heavy water, which greatly enhancessolar neutrino detection.Other new detectors will start up soon.An experiment performed at Los Alamos National Laboratory provides a furtherhint of neutrino oscillation:it detects electron-neutrinos from a source thatshould produce only muon-neutrinos.The signal is mixed, however, with backgroundprocesses.The result has not yet been independently confirmed, butsome experiments will check it in the next few years.Mass-induced oscillations between muon- and tau-neutrinos seem themost natural explanation for the Super-K neutrino data, but there are otherpossibilities.First,the most general scenario has mixing between all three neutrinoflavors,and Super-K’s data can accommodate some oscillations betweenelectron- and muon-neutrinos at the energies it covers.Yet results from an experimentat the Chooz nuclear power station in Ardennes, France, greatly limithow much electron-muon oscillation could be occurring at Super-K.Another possibility is that the muon-neutrinos are oscillating to a previouslyunseen flavor of neutrino. Still, studies of the so-called Z 0 particle atCERN, the European laboratory for particle physics near Geneva, clearly showthat there are only three active flavors of neutrino. (“Active” means that theflavor participates in the weak nuclear interaction.) A new flavor would thereforehave to be “sterile,” a species of neutrino that interacts only throughgravity. Some physicists favor this idea, because current evidence for threedistinct effects (solar neutrinos, atmospheric neutrinos and the Los Alamosdata) cannot be accounted for by one consistent set of masses for the electron-,muon- and tau-neutrinos.Other oscillation mechanisms, relying on more esoteric effects than neutrinomass, have also been proposed.—E.K.,T.K. and Y.T.a well-defined energy and direction.Most important, we have placed a detectorsimilar to Super-K near the originof the beam to characterize the muonneutrinosbefore they oscillate. Effectively,we are using the ratio (again) ofthe counts near the source to those faraway to cancel uncertainty and verifythe effect. As this article is being printed,neutrinos in the first long-distanceartificial neutrino beam are passing underthe mountains of Japan, with50,000 tons of Super-K capturing asmall handful. Exactly how many itcaptures will be the next chapter in thisstory.The AuthorsFurther ReadingEDWARD KEARNS, TAKAAKI KAJITA and YOJITOTSUKA are members of the Super-KamiokandeCollaboration. Kearns, a professor of physics at BostonUniversity, and Kajita, a professor of physics at the Universityof Tokyo, lead the analysis team that studiesproton decay and atmospheric neutrinos in the Super-Kdata. Totsuka is spokesman for the Super-K Collaborationand is director of the Institute for Cosmic Ray Researchat the University of Tokyo, the host institutionfor the experiment.The Search for Proton Decay. J. M. LoSecco, Frederick Reines and DanielSinclair in Scientific American, Vol. 252, No. 6, pages 54–62; June 1985.The Elusive Neutrino: A Subatomic Detective Story. Nickolas Solomey.Scientific American Library, W. H. Freeman and Company, 1997.The Official Super-Kamiokande Web site is available at www-sk.icrr.u-tokyo.ac.jp/doc/sk/The K2K Long Baseline Neutrino Oscillation Experiment Web site is availableat neutrino.kek.jp/The Super-Kamiokande at Boston University Web site is available at hep.bu.edu/~superk/index.htmlDetecting Massive Neutrinos Scientific American August 1999 71Copyright 1999 Scientific American, Inc.