Japanese and American physicists from a collaboration called KamLAND have released analysis of data from their first six months of experimental operations, in which they claim for the first time to have detected a disappearance of electron anti-neutrinos flying to their detector from nuclear reactors at typical distances of 175 km in Japan. This experiment represents the culmination of nearly half a century of effort to detect oscillations with reactor neutrinos. The results appear to confirm that the long observed deficit of neutrinos from the sun (the "solar neutrino problem") is indeed due to neutrino oscillations, this accomplished in a totally independent and completely terrestrial experiment. That oscillationes were the root cause of the solar neutrino problem had long been suspected in a series of 7 solar neutrino experiments over the last 40 years. But there have always been questions about whether the deficit was due to the solar model or some peculiar form of neutrino disappearance, such as might be caused by magnetic fields in the sun, the decay of neutrinos, new types of neutrinos, and various hypothetical interactions.
These results make the case for neutrino mass and neutrino oscillations seemingly inescapable. Oscillations has to do with the peculiar ability of neutrinos to change from one type (electron, muon or tau) to another while flying along at close to the speed of light, a quantum mechanical oddity found nowhere else than with neutrinos. These results, totally unpredicted by theory are outside of the much acclaimed Standard Model of particle physics, which while clumsy seems to make correct predictions for most elementary particle interactions and decays. The Standard Model needs patching for neutrinos, but perhaps more importantly, such unpredicted observations may point the way towards a true unified theory of elementary particles, the holy grail of physics.
The mixing of these neutrinos is surprisingly strong, and the neutrino masses fall into a different regime than all the other particles which make up the standard zoo of constituents of our universe. These results move into uncharted territory. In fact, on the cosmic scale there is at present much talk in theoretical circles about neutrinos providing a possible route to explain the mysterious tiny excess of matter over anti-matter in the universe, which arose at the time of the Big Bang (and to which we owe our existence). Understanding the masses, mixing and other properties of neutrinos is crucial to this line of exploration.
Neutrino mass is not required for any known reason in the so-called Standard Model of particle physics, but can be easily incorporated. In the universe at large it appears to amount to roughly as much in total mass as all the visible stars in the sky. It seems that we live bathed in a constant ethereal breeze of ghostly neutrinos, those left over from the Big Bang, those streaming from the fusion burning sun and stars, those from all the radioactive decays throughout the earth, and nowadays quite a number of neutrinos from man-made power reactors around the world. (One need not worry about any danger from this flow of neutrinos, since the radiation damage they cause is miniscule compared to the natural radioactivity in our bodies.)
In fact observation of the total neutrino flux from the earth has been a desired measurement for many years, since there has been controversy about how much heating radioactivity contributes to the earth's interior (and which has implications for cooling of the earth, continental drift, and geology generally). These measurements from KamLAND give the first significant upper limit on that power, and in a few years should yield the first positive measurement of the total radioactivity of the earth. The KamLAND measurements also set the first significant limit on a hypothesized reactor in the earth's core (not a widely accepted idea in any case); again limits that will tighten with the accumulation of more KamLAND data over the next few years.
A further implication of the KamLAND results has to do with elimination of one proposed solution to a quandary posed by an experiment from Los Alamos, called LSND (Liquid Scintillation Neutrino Detector), which found as yet unconfirmed evidence for a tiny amount of neutrino mixing, but at only few meter distance. One escape from the seeming contradiction of the LSND results contrasted with the picture formed from all other experiments was for violation of one of the sacred conservation laws of physics, relating the properties of particles and anti-particles. This door seems to have been closed by the KamLAND results since the anti-neutrinos observed from Japanese nuclear reactors behave just as do the neutrinos from the sun. (Yet, as of 9 Dec 02, clever theorists claim to find another escape scenario to save this model.)
The KamLAND Experiment
The KamLAND detector is built inside a cavity in the Mozumi mine of the Kamioka Mining and Smelting Company, in the Kamioka Underground Laboratory complex operated by the University of Tokyo. The laboratory is located on the Western side of the Japanese alps near the city of Toyama. The group is lead by Professor Atsuto Suzuki, the Director of the Research Center for Neutrino Science at the University of Tohoku, in Sendai, Japan. The collaboration includes about 92 physicists from Tohoku University, U. Alabama, UC Berkeley, Cal Tech, Drexel U., U. Hawaii, U. New Mexico, Stanford, U. Tennessee, Triangle Universities Nuclear Lab, and IHEP in Beijing. The project support comes largely from the Ministry of Education, Culture, Sports, Science and Technology of Japan, a significant contribution from the US Department of Energy and a number of smaller contributions from collaborating institutions and agencies.
The construction was begun in the summer of 1998 when the previous distinguished experiment, Kamiokande, was dismantled (this experiment has resulted in a Nobel Prize this year for Professor Masatoshi Koshiba from the University of Tokyo). The old cavity was slightly enlarged and a new stainless steel tank of 18 meters diameter installed, with surrounding water bath and 225 photomultipliers to tag incoming background particles. The inner section consists of an 13 meter diameter balloon, which is filled with a scintillating material dissolved in paraffin oil. The scintillation molecules convert elementary particle motion and disturbance into light flashes whose intensity is proportional to the particle's energy. This light is viewed by a surrounding spherical array of 1879 photomultipliers, which light detectors provide electrical signals from which one may deduce the exact timing, location and energy of events occurring in the balloon. The sensitivity of the experiment reaches down to an impressively low level of about 4 keV (and about 50 times lower than the larger neighbor experiment, Super-Kamiokande). The price for lower threshold is however that no directional information can be extracted from the low energy neutrino products observed in the 1000 tons of scintillating detector. All information must be extracted by the type of event, the energy, time and location recorded (see below for more discussion of this).
This experiment has a long history of predecessors, none of which saw any sought for deviation from expectations without oscillations. In fact the very same reaction was used by Reines and Cowan in their original experiments in the mid-1950's in the first detections of neutrinos (then thought to be perhaps undetectable). Reines received a long overdue Nobel Prize for this (Cowan having died) in 1995. Nuclear reactors create many radioactive isotopes which decay with electron anti-neutrinos as one product. Other products remain in the reactor and heat it up, providing the energy sought for generating electrical power. But about 3% of the total energy escapes in neutrinos. Reines and company originally detected powerful military reactors from distances of about 10 meters. There have been a series of about a dozen experiments over the years (including some false hints at oscillations) moving progressively further from the reactors, with bigger and more sensitive instruments.
All of this previous study also means that the expected neutrino flux from such reactors is well understood to several percent uncertainty. KamLAND made a giant leap from the previous experiments. The most recent were conducted in the last decade in France (CHOOZ) and the US (Palo Verde) at about 1 km range. KamLAND detects reactors from all around Japan and even Korea (2.5%, all others <0.7%), but on average senses reactors at a mean range of 175 km. We now know that the oscillations which explain the solar neutrino problem require many kilometer distances to become manifest, and this is what KamLAND has observed.
The detection utilizes the interaction of electron anti-neutrinos with protons in the scintillator solution to make neutrons and positrons, the so-called "inverse beta decay" reaction. (It is somewhat the opposite to the "beta decay" of a free neutron, which yields a proton, electron and electron anti-neutrino. The inverse reaction requires 1.8 MeV to make a more massive neutron and positron from the incoming neutrino hitting the proton, and this sets a threshold on neutrino detection energy.) The positron, being an anti-matter electron, almost immediately annihilates with an electron in the scintillator, creating two gamma rays. These gammas fly about a meter and make splashes of secondary particles and hence light as they careen through the scintillator. Meanwhile the neutron, like a ball in a pinball machine, rattles around only getting trapped by another proton in a time of around 200 microseconds later. Upon capture the newly formed Deuterium releases about 2.2 MeV of energy, which is detected. This two part signature, with the pulses required to be sufficiently close in time and space, acts as a powerful tool to remove essentially all other background processes.
The experimental paper, submitted to the Physical Review letters on 6 December 2002, reports the detection of 54 neutrino events recorded from a period of 145 days of operations, in the energy range from 1 to 10 MeV, to be compared with an expected number of 86.8+/-5.5 events if there were no oscillations. The ratio of seen-to-expected is 0.611 +/-0.085 (statistical error) +/-0.041 (systematic error). The probability that this could be just a downward statistical fluctuation (the null hypothesis) is less than one in a thousand. This is reinforced by the energy distribution of events (which are less than expected under the null hypothesis at all energies). This demonstrates that neutrinos are disappearing between the reactors and the detector, but does not reveal the cause of the disappearance, which requires further analysis of hypothesis testing.
Analysis of the number of events and their energies allows the team to deduce that neutrino mixing is taking place (electron types being mixed with muon and tau types), and to set an approximate mass difference scale and mixing strength. The results agree entirely with those deduced from the solar neutrino experiments, and lie in a region known as the Large Mixing Angle solution. [The best fit numbers are sin^2(2 theta) = 1.0 and delta m^2 = 6.9 x 10^(-5) eV^2.] The results pin the neutrino mass difference down with a precision which will surprise most observers, and appears to be due to a fortuitous value for the mass difference relative to the reactor energies and distances. Moreover, quite unexpectedly, the mixing angle seems to be on the larger side of prior expectations, being close to the maximum allowed. This result will be the grist for many theoretical papers no doubt, but at the moment we have no understanding of why it is so.
This experimental triumph would seem to close (after about fifty years, with most progress in the last decade) the exploratory chapter of work to understand the gross nature of neutrino mixing and masses. Now it seems that further progress will involve improving the details of measured masses and mixings, and the search for some more rare types of mixing which have great theoretical importance.
It remains also that the actual mass of the neutrinos is not measured, since the oscillations observations only measure mass differences. It is possible that if all the masses are close to the present upper limit of total neutrino masses, about 2 eV, neutrinos could play a major role in galaxy formation and clustering. Indeed, indirect measurements of the total neutrino mass could come first from cosmological observations. This is a nice demonstration of how results on the opposite extreme scales of the universe from the most grand to tiniest subnuclear, can interact in surprising and exciting ways helping unlock nature's secrets.
The KamLAND project will continue to operate for some years, making refined measurements of reactor neutrinos, hopefully detecting solar neutrinos (after some further improvement in purification from radioactive materials) and looking for new physics, including supernovae in our galaxy.
0+th Draft of KamLAND backgrounder, 11/6/02, jgl, re-edited 11/7 jgl, sp comments incorporated 11/8; some updates from fs, as and jgl 11/15, 11/30, updating scince results 12/2 jgl, and 12/9.