A series of three workshops has considered a new generation of experiments at nuclear reactors, which could help to pin down the neutrino mixing matrix.
In late March in Japan only a few buds from the cherry blossom trees are beginning to show their shades of pink, but in Niigata this year new ideas for neutrino experiments at nuclear reactors were in full bloom. It was here that an international group of physicists met to discuss these ideas at a workshop hosted by the University of Niigata and the Tokyo Electric Power Company. This was the third in a series on future low-energy neutrino experiments that had begun in Alabama in April 2003 and proceeded to Japan via Munich, in October 2003.
The basic idea being considered is to use several detectors to search for anti-electron neutrino disappearance, as this can provide evidence for a non-zero value of the parameter θ13 in the Maki-Nakagawa-Sakata (MNS) mixing matrix, the analogue for neutrinos of the Cabbibo-Kobayashi-Maskawa matrix for quark mixing. In its simplest form the 3 x 3 neutrino MNS matrix can be parameterized with three angles and one phase. Experiments using atmospheric neutrinos have shown clear evidence for neutrino oscillations, with the mixing angle – the parameter θ23 – near its maximal value of 45°. The long-standing solar neutrino problem has also been solved by neutrino oscillations with a large value of the parameter θ12. This result has been confirmed by the reactor neutrino experiment KamLAND, which has an average distance to the reactors of 180 km.
The current best limit for θ13 comes from the reactor experiment CHOOZ. This was originally designed to look for a large signal from θ12 related to the atmospheric neutrino anomaly, and used only one detector. Now, however, it has been realized that an experiment with two (or more) detectors could greatly reduce the dominant systematic uncertainties from the reactor fuel cycle and detector efficiencies. This would allow a more sensitive search for θ13.
The meeting in Niigata began with a number of talks reviewing the theoretical situation. Hisakazu Minakata from Tokyo Metropolitan University described how a neutrino measurement of θ13 could be combined with measurements from future long-baseline accelerator experiments to measure the sign of Δm2 and the CP parameter δ. He also introduced a concept for a θ12 experiment that would use the Kashiwazaki-Kariwa nuclear-reactor complex near Niigata and a detector on Sado Island, about 70 km away. Morimitsu Tanimoto from Niigata University then addressed the issues of why θ13 is so small, why θ23 is near maximal and why θ12 is not maximal. He covered several theoretical frameworks: anarchy (in effect, random oscillations), radiative origins, grand unified theories and texture zeros (very small entries in the mass matrices). In one example of texture zeros the small value of θ13 could be related to the smallness of the neutrino masses. However, even in that case he concluded that only experiment could reveal the real size of θ13.
The recent increase in interest in a new reactor experiment goes hand-in-hand with ideas for large “off axis” long-baseline neutrino experiments at accelerators to measure νµ-νe oscillations. Unlike reactor experiments, accelerator experiments are also sensitive to CP-violating effects in the neutrino sector and to matter effects. This is both an advantage and a disadvantage. The advantage is that there is obviously a richer physics programme to investigate. The disadvantage is that a particular measurement is more difficult to interpret due to ambiguities and degeneracies. Takashi Kobayashi from KEK described the status of the T2K experiment, which was recently approved to send a beam from the new JPARC accelerator at Tokai to the Super-Kamiokande facility in the Japanese Alps, a distance of 295 km. The first beam is currently expected in 2009. Bob McKeown from Caltech then showed how reactor and accelerator measurements could be combined to provide greater precision and insights. As examples, he used both the Japanese experiment T2K and the proposal for an off-axis experiment at the NuMI beam in the US, now called NOνA.
The main parameters of a reactor neutrino disappearance experiment were outlined in the talk by Karsten Heeger from the Lawrence Berkeley Laboratory. This also served as an excellent summary for the three workshops on this subject as he addressed three main questions. Why do a new reactor experiment? How would such an experiment be configured? What are the experimental challenges that new multi-detector reactor experiments face? Heeger concluded that a disappearance measurement of θ13 with reactor neutrinos is a promising method to measure the true value of sin22θ13, but it is experimentally challenging. Combined with the result of a five-year experiment with a high-intensity neutrino “superbeam”, reactor measurements can provide significant new constraints and perhaps even decide the neutrino mass hierarchy and yield information on the CP-violating phase angle in the MNS matrix. The sensitivity to a normal mass hierarchy is better. An optimized baseline of 1.7 km helps to reduce the impact of systematics, and limits of the order of sin22θ13 < 0.014 are achievable. Smaller, quicker reactor experiments will yield sin22θ13 < 0.04.
Talks and discussion at the meeting also explored six projects that are under development in five countries and on three continents (see table “A comparison of the features of some of the proposed reactor sites”). In Japan, the Kashawazaki-Kariwa nuclear-power complex consists of seven nuclear reactors. Located about an hour from Niigata, it is the highest power nuclear site in the world. The Tokyo Electric Power Company arranged a tour for conference participants, who – after the appropriate security and radiation protection requirements – were able to stand on top of one of the seven cores, 20 m from the release of enough energy to power 3% of the Tokyo area. Fumihiko Suekane from Tohoku University showed the design for the KASKA (Kashawazaki-Kariwa) project, in which 8 tonne gadolinium-loaded liquid-scintillator detectors would be placed deep in shafts at two near-detector locations and one far location, with an average distance 1.3 km from the reactor cores. Osamu Yasuda from Tokyo Metropolitan University demonstrated that for multiple reactors and near-detectors the uncorrelated error is reduced and there is no loss of precision.
Turning to the US, Ed Blucher from the University of Chicago and Jonathan Link from Columbia University described progress on the proposal to use a site at Braidwood in Illinois, about 80 km from Chicago. One way to control systematic errors is to move the detectors between the near and far sites, and the relatively flat terrain in Illinois allows this to be done relatively inexpensively. A cost estimate has been made for the underground construction of two shafts, 300 and 1800 m from the centre of the two reactor cores at the Braidwood nuclear plant. The next step is to drill boreholes to full depth at the positions of both shafts, to provide information about geology, radioactivity and density. A site-specific estimate of isotope production by muons is being used to calculate the optimum depth for the detectors. Karsten Heeger described similar considerations for the Diablo Canyon site in California.
In Europe the original CHOOZ experiment, 1050 m from the reactor cores, is still available. Thierry Laserre from CEA/Saclay and Herve de Kerret of APC (AstroParticule et Cosmologie)/Collège de France presented the Double CHOOZ concept, with a larger detector and the possibility of placing a relatively shallow near-detector about 100-200 m from the reactors. A dense mound of shielding would probably be needed to reduce backgrounds. A letter of intent by a proto-collaboration from France, Germany, Italy and Russia is nearing completion, and early stages of approval have already been obtained.
For South America, a site at Angra in Brazil is a possibility, as described by Orlando Peres from the State University of Campinas (UNICAMP) and David Reyna from Argonne National Laboratory. Due to a favourable local geology, two 50 tonne detectors could be placed 350 and 1350 m from the reactor core. Back in Asia, the site at Daya Bay in China was described by Yifang Wang from the Institute for High Energy Physics in Beijing. Located near Hong Kong, the power plant has four reactor cores in two clusters, providing a total thermal power of 11.6 GW, with two further cores (6 GW) planned for 2011. A tunnel that would service two near-detector locations and one far-detector location is being considered, as well as a design for multiple 10 tonne detectors.
In addition to describing the site characteristics, most speakers also addressed a myriad of issues, including optimal distances, detector design, scintillator properties, backgrounds, calibration and systematic errors. Two talks focused on the progress in understanding gadolinium-loaded liquid scintillators. The neutron absorption cross-section on gadolinium is so high that it provides an attractive target for this kind of experiment. Both the high cross-section and large energy release (8 MeV) provide a high efficiency to look for the neutron in coincidence with the positron in inverse beta-decay. However, previous experiments have found large degradations as a function of time of the light attenuation length in gadolinium scintillators, which would make a precision experiment more difficult. Francis Hartmann from the Max Planck Institute in Heidelberg described the progress that is being made there in scintillator chemistry, including the promising metal beta-diketone structure that is being investigated. Dick Hahn from the Brookhaven National Laboratory reported on a series of tests undertaken at Brookhaven and elsewhere to understand the optical properties of gadolinium-loaded scintillators in various solvents and with a variety of concentrations.
The unit of luminosity for reactor experiments is gigawatt-tonne-years, a product of the reactor power, the detector size and the running time. Manfred Linder from the Technical University Munich had shown in earlier workshops that there were two limiting cases, low luminosity (below 400 GW-tonne-years) and high luminosity (above 8000 GW-tonne-years). The former allows a measurement of rates, while the latter allows the shape of the energy distribution to be studied. However, different systematic errors are important for the different ranges of luminosity. David Reyna from Argonne National Laboratory focused on the advantages of using larger detectors to get enough statistics to see the change in shape of the energy distribution due to electron-antineutrino disappearance.
Although the main goal of the experiments being discussed in Niigata is to discover and measure θ13, there are other physics goals that could be pursued. Valery Sinev from the Kurchatov Institute considered the sensitivity of such an experiment to sterile neutrinos. He also looked at the issues of “burn-up” (changes in fissile content based on changes in the antineutrino rate) and changes in the energy distribution at the near-detector as studies in reactor physics. Michael Shaevitz from Columbia University presented a study showing that the events from the near-detector, if it is deep enough, could be used to measure neutrino-electron elastic scattering with an accuracy good enough to make a measurement of the weak mixing angle. This could be valuable as the NuTeV neutrino experiment has a measurement of this angle that is somewhat in conflict with other ways to measure it. A measurement of the antineutrino flux from reactors could also prove useful for the International Atomic Energy Association in its monitoring of the fuel cycle of nuclear reactors, as Thierry Laserre described.
The three workshops in this series have been useful in providing motivation for the experiments and sharing strategies for how to go about them. While the theorists refused to give a firm prediction for θ13, the experimentalists in Niigata conducted a poll of their expectation of what θ13 might turn out to be. More than 80% of their values were within the sensitivity of the proposed new reactor experiments. Since no large civil construction is needed, the quickest opportunity is for the Double CHOOZ experiment, with a detector that could be taking data in 2008. Participants were also in agreement that another experiment beyond Double CHOOZ was necessary in order to cover the range of parameter space that is reasonably accessible. They left Niigata convinced that they needed to form the collaborations, get the experiments approved and find the value of θ13.
The international working group of 126 physicists from 40 institutions in nine countries has collaborated on writing a white paper entitled “A New Nuclear Reactor Neutrino Experiment to Measure θ13“, which was published in January 2004.
