Doda Khan Badini

Doda is double majoring in physics and philosophy at MIT. He is interested in philosophy and foundations of physics and hopes to pursue graduate study in history and philosophy of physics after graduation.

Plutonium is a prime material for building nuclear weapons; it exists only in trace amounts in nature, but uranium fission reactors produce plutonium in abundance—and clandestine illicit programs can divert it from spent nuclear fuel and use it to produce weapons. Adam Bernstein, a physicist at Lawrence Livermore National Laboratory, has been working on antineutrino detectors to track plutonium in reactors as a way to expose possible diversion attempts. Years ago, he and a colleague had investigated using antineutrino detectors to search out possible illegal nuclear bomb detonations. "The results were not hopeful," he says. "However, during my investigations I also happened to look into antineutrino output from nuclear reactors, and the idea just seemed right."

Nuclear reactors produce a very large number of antineutrinos, on the order of a billion trillion per second (to get a sense of how big that number is, all the beaches in the world together contain on the order of a billion trillion grains of sand). Antineutrinos are the antiparticle of neutrinos, and like neutrinos they are electrically neutral. They rarely interact with matter, and travel through it unimpeded. This makes them ideal for peering into reactor cores. "You cannot hide them, and given that they come from the core, antineutrinos are a potent method for analyzing the composition of the core," says Bernstein. Although their ethereal nature makes them very difficult to detect, it is possible to observe antineutrinos under certain circumstances. As a general rule, the nearer a detector lies to the source and the larger the number of antineutrinos the source produces, the greater are the chances of detection.

Nuclear reactors generate power through nuclear fission, the same process behind the explosive burst of energy released when an atomic bomb is detonated. During a fission reaction, a nucleus breaks into multiple components, releasing energy. For instance, the fission of a uranium nucleus can produce a barium nucleus, a krypton nucleus, and three neutrons. This releases energy via Einstein's famous E=mc²—since the total mass at the end of the process (the fission products) is less than the total mass in the beginning (the uranium nucleus), the difference shows up as a release of energy! But unlike the runaway atomic bomb explosions, nuclear reactors control the fission process, which is usually fueled by a mixture of plutonium and uranium. Bombarding these materials with neutrons starts the chain reaction—the nuclei absorb neutrons, become unstable, and break apart, producing smaller nuclei, heat, and additional neutrons to continue the chain reaction.

Reactor fuel contains isotopes of uranium and plutonium. An isotope is a particular variant of a chemical element—it has identical chemical properties but, importantly, has a different number of neutrons. Fresh nuclear fuel contains uranium-235 and uranium-238 (the number 238 indicates 3 more neutrons than 235); the former is the main fission isotope. During a reactor's operation cycle, typically 600 days, the supply of uranium-235 decreases as it is burned. At the same time, uranium-238 absorbs neutrons, becomes more unstable, and decays into neptunium-239, which finally decays into plutonium-239, which also burns. By the end of the reactor cycle, the quantity of plutonium may contribute as much as half of the fissions. So while uranium dominates the fission initially, plutonium is an important source at the end of the cycle.

As plutonium-239 builds up in the reactor core over the cycle, the antineutrino count rate drops by about 5 to 10 percent. The fission process itself does not produce antineutrinos. Rather, unstable fission products undergo a chain of rapid decays that release antineutrinos. On average, about 6 antineutrinos are produced per fission. The number and energy of those antineutrinos depend on the original nucleus that was fissioned—so the energy spectrum of antineutrinos from uranium-235 fission will look different from that of plutonium-239 fission. Thus, analyzing a reactor's antineutrino energy and count rate makes it possible to establish whether any plutonium has been diverted from the reactor core. And this is precisely what Bernstein's detectors exploit.

The International Atomic Energy Association (IAEA) has expressed interest in antineutrino detectors for reactor monitoring. The IAEA, established in 1957, works with member states and other partners to promote safe, secure and peaceful use of nuclear technology. One of the IAEA's major goals is to ensure nuclear non-proliferation. Monitoring of illegal activity at civil reactors can be done through direct methods such as reactor fuel analysis, video surveillance, and random visitation from IAEA agents. Unfortunately, this monitoring is not always feasible because it involves interference with reactor operation and other logistical hurdles.

Bernstein's goal is to develop detectors that are simple to operate and maintain, and have good long term stability requiring no maintenance for months at a time—perfect for the purposes of the IAEA. He has been leading the development team behind such detectors for about a decade now. It is a joint collaboration between researchers at the Lawrence Livermore National Lab and Sandia Laboratories. The Lawrence-Sandia detectors measure "inverse-beta" reactions, where an antineutrino collides with a proton to produce a positron (the antiparticle of the electron) and a neutron. This interaction creates two consecutive flashes of light in the detector which coincide very closely in time. This unique signature gets picked up by light detectors, called photomultiplier tubes, which convert the light into electric signals. It is then further analyzed to determine the fuel composition.

Bernstein's team has tested three prototype detectors to date. The first prototype, SONGS1, uses a proton-rich liquid scintillator doped with gadolinium to induce inverse beta-decay interactions. The gadolinium is added to increase efficiency for capturing neutrons. The detector has proved successful at a distance of 25 meters from the reactor core. However, there is one drawback. "Although the double flash signature of the inverse-beta reaction is [usually] easy to differentiate from those occurring in the background, the possibility of noise from stray neutrons still exists," says Bernstein. The dual signature is not foolproof because background flashes can sometimes occur so close together that they might get picked up as the dual flashes of a positron and neutron pair. To counter the noise, shielding is required, and this increases the size and weight of the detector.

The SONGS1 prototype measures 3 meters per side by itself, but shielding adds 20 tons of water to it. This kind of weight is a problem for transportation and deployment for monitoring purposes. As a result, Bernstein and his team have designed two other prototypes called SONGS2 and SONGS3. The smaller SONGS2 uses a more compact plastic scintillator instead of a liquid. However, like SONGS1, it still has to be installed underground to provide additional shielding from background noise. SONGS3 is different in that it uses water mixed with gadolinium and measures Cherenkov radiation. Cherenkov radiation produces blue light when charged particles travel faster than the speed of light in water. Though this type of anti-neutrino signature is fainter than that produced by liquid and plastic scintillators, it has two advantages. First, water is benign, whereas scintillator liquids are toxic, flammable and carcinogenic. Secondly, water-based detectors do not pick up false signals from background noise.

Earlier prototypes were tested only in underground settings. However, SONGS3-type detectors have also been tested on the surface and the results are promising. "The last detector we tested was the size of a refrigerator," Bernstein says. This small size is a great feat in itself, but more work needs to be done. "We want to get to the point where one could load a detector in the back of a truck, park at a nonsignificant distance from a reactor, and collect data … We are not quite there, but we are close," he adds. Bernstein thinks that further development will open new vistas, and may prove a practical means for non-intrusive, real-time surveillance of nuclear reactors.