Search for sterile neutrinos triples up

This summer, two 270 m³ steel containment vessels are making their way by land, sea and river from CERN in Europe to Fermilab in the US, a journey that will take five weeks. Each vessel houses one of the 27,000-channel precision wire chambers of the ICARUS detector, which uses advanced liquid-argon technology to detect neutrinos. Having already operated successfully in the CERN to Gran Sasso neutrino beam from 2010 to 2012, and spent the past two years being refurbished at CERN, ICARUS will team up with two similar detectors at Fermilab to deliver a new physics opportunity: the ability to resolve some intriguing experimental anomalies in neutrino physics and perform the most sensitive search to date for eV-scale sterile neutrinos. This new endeavour, comprised of three large liquid-argon detectors (SBND, MicroBooNE and ICARUS) sitting in a single intense neutrino beam at Fermilab, is known as the Short-Baseline Neutrino (SBN) programme.

The sterile neutrino is a hypothetical particle, originally introduced by Bruno Pontecorvo in 1967, which doesn’t experience any of the known forces of the Standard Model. Sterile-neutrino states, if they exist, are not directly observable since they don’t interact with ordinary matter, but the phenomenon of neutrino oscillations provides us with a powerful probe of physics beyond the Standard Model. Active–sterile mixing, just like standard three-neutrino mixing, could generate additional oscillations among the standard neutrino flavours but at wavelengths that are distinct from the now well-measured “solar” and “atmospheric” oscillation effects. Anomalies exist in the data of past neutrino experiments that present intriguing hints of possible new physics. We now require precise follow-up experiments to either confirm or rule out the existence of additional, sterile-neutrino states.    

On the scent of sterile states

The discovery nearly two decades ago of neutrino-flavour oscillations led to the realisation that each of the familiar flavours (ν_e, ν_μ, ν_τ ) is actually a linear superposition of states of distinct masses (ν₁, ν₂, ν₃ ). The wavelength of an oscillation is determined by the difference in the squared masses of the participating mass states, m²_i – m²_j. The discoveries that were awarded the 2015 Nobel Prize in Physics correspond to the atmospheric mass-splitting Δm²_ATM = |m²₃ – m²₂| = 2.5 × 10^–3 eV² and the solar mass-splitting Δm²_SOLAR = m²₂ – m²₁ = 7.5 × 10^–5 eV², so-named because of how they were first observed. Any additional and mostly sterile mass states, therefore, could generate a unique oscillation driven by a new mass scale in the neutrino sector: m²_{mostly sterile} – m²_{mostly active}.

The most significant experimental hint of new physics comes from the LSND experiment performed at the Los Alamos National Laboratory in the 1990s, which observed a 3.8σ excess of electron antineutrinos appearing in a mostly muon antineutrino beam in a region where standard mixing would predict no significant effect. Later, in the 2000s, the MiniBooNE experiment at Fermilab found excesses of both electron neutrinos and electron antineutrinos, although there is some tension with the original LSND observation. Other hints come from the apparent anomalous disappearance of electron antineutrinos over baselines less than a few hundred metres at nuclear-power reactors (the “reactor anomaly”), and the lower than expected rate in radioactive-source calibration data from the gallium-based solar-neutrino experiments GALLEX and SAGE (the “gallium anomaly”). Numerous other searches in appearance and disappearance channels have been conducted at various neutrino experiments with null results (including ICARUS when it operated in the CERN to Gran Sasso beam), and these have thus constrained the parameter space where light sterile neutrinos could still be hiding. A global analysis of the available data now limits the possible sterile–active mass-splitting, m²_{mostly sterile} – m²_{mostly active}, to a small region around 1–2 eV². 

Long-baseline accelerator-based neutrino experiments such as NOvA at Fermilab, T2K in Japan, and the future Deep Underground Neutrino Experiment (DUNE) in the US, which will involve detectors located 1300 km from the source, are tuned to observe oscillations related to the atmospheric mass-splitting, Δm²_ATM ~ 10^–3 eV². Since the mass-squared difference between the participating states and the length scale of the oscillation they generate are inversely proportional to one another, a short-baseline accelerator experiment such as SBN, with detector distances of the order 1 km, is most sensitive to an oscillation generated by a mass-squared difference of order 1 eV² – exactly the region we want to search.

Three detectors, one beam

The SBN programme has been designed to definitively address this question of short-baseline neutrino oscillations and test the existence of light sterile neutrinos with unprecedented sensitivity. The key to SBN’s reach is the deployment of multiple high-precision neutrino detectors, all of the same technology, at different distances along a single high-intensity neutrino beam. Use of an accelerator-based neutrino source has the bonus that both electron-neutrino appearance and muon-neutrino disappearance oscillation channels can be investigated simultaneously.

The neutrino source is Fermilab’s Booster Neutrino Beam (BNB), which has been operating at high rates since 2002 and providing beam to multiple experiments. The BNB is generated by impinging 8 GeV protons from the Booster onto a beryllium target and magnetically focusing the resulting hadrons, which decay to produce a broad-energy neutrino beam peaked around 700 MeV that is made up of roughly 99.5% muon neutrinos and 0.5% electron neutrinos.

The three SBN detectors are each liquid-argon time projection chambers (LArTPCs) located along the BNB neutrino path (see images above). MicroBooNE, an 87 tonne active-mass LArTPC, is located 470 m from the neutrino production target and has been collecting data since October 2015. The Short-Baseline Near Detector (SBND), a 112 tonne active-mass LArTPC to be sited 110 m from the target, is currently under construction and will provide the high-statistics characterisation of the un-oscillated BNB neutrino fluxes that is needed to control systematic uncertainties in searches for oscillations at the downstream locations. Finally, ICARUS, with 476 tonnes of active mass and located 600 m from the BNB target, will achieve a sufficient event rate at the downstream location where a potential oscillation signal may be present. Many of the upgrades to ICARUS implemented during its time at CERN over the past few years are in response to unique challenges presented by operating a LArTPC detector near the surface, as opposed to the underground Gran Sasso laboratory where it operated previously. The SBN programme is being realised by a large international collaboration of researchers with major detector contributions from CERN, the Italian INFN, Swiss NSF, UK STFC, and US DOE and NSF. At Fermilab, new experimental halls to house the ICARUS and SBND detectors were constructed in 2016 and are now awaiting the LArTPCs. ICARUS and SBND are expected to begin operation in 2018 and 2019, respectively, with approximately three years of ICARUS data needed to reach the programme’s design sensitivity.

A rich physics programme

In a combined analysis, the three SNB detectors allow for the cancellation of common systematics and can therefore test the ν_μ → ν_e oscillation hypothesis at a level of 5σ or better over the full range of parameter space originally allowed at 99% C.L. by the LSND data. Recent measurements, especially from the NEOS, IceCube and MINOS experiments, have constrained the possible sterile-neutrino parameters significantly and the sensitivity of the SBN programme is highest near the most favoured values of Δm². In addition to ν_e appearance, SBN also has the sensitivity to ν_μ disappearance needed to confirm an oscillation interpretation of any observed appearance signal, thus providing a more robust result on sterile-neutrino-induced oscillations (figure 1).

SBN was conceived to unravel the physics of light sterile neutrinos, but the scientific reach of the programme is broader than just the searches for short-baseline neutrino oscillations. The SBN detectors will record millions of neutrino interactions that can be used to make precise measurements of neutrino–argon interaction cross-sections and perform detailed studies of the rather complicated physics involved when neutrinos scatter off a large nucleus such as argon. The SBND detector, for example, will see of the order 100,000 muon-neutrino interactions and 1000 electron-neutrino interactions per month. For comparison, existing muon-neutrino measurements of these interactions are based on only a few thousand total events and there are no measurements at all with electron neutrinos. The position of the ICARUS detector also allows it to see interactions from two neutrino beams running concurrently at Fermilab (the Booster and Main Injector neutrino beams), allowing for a large-statistics measurement of muon and electron neutrinos in a higher-energy regime that is important for future experiments.

In fact, the science programme of SBN has several important connections to the future long-baseline neutrino experiment at Fermilab, DUNE. DUNE will deploy multiple 10 kt LArTPCs 1.5 km underground in South Dakota, 1300 km from Fermilab. The three detectors of SBN present an R&D platform for advancing this exciting technology and are providing direct experimental activity for the global DUNE community. In addition, the challenging multi-detector oscillation analyses at SBN will be an excellent proving ground for sophisticated event reconstruction and data-analysis techniques designed to maximally exploit the excellent tracking and calorimetric capabilities of the LArTPC. From the physics point of view, discovering or excluding sterile neutrinos plays an important role in the ability of DUNE to untangle the effects of charge-parity violation in neutrino oscillations, a primary physics goal of the experiment. Also, precise studies of neutrino–argon cross-sections at SBN will help control one of the largest sources of systematic uncertainties facing long-baseline oscillation measurements.    

Closing in on a resolution

The hunt for light sterile neutrinos has continued for several decades now, and global analyses are regularly updated with new results. The original LSND data still contain the most significant signal, but the resolution on Δm² was poor and so the range of values allowed at 99% C.L. spans more than three orders of magnitude. Today, only a small region of mass-squared values remain compatible with all of the available data, and a new generation of improved experiments, including the SBN programme, are under way or have been proposed that can rule on sterile-neutrino oscillations in exactly this region.

There is currently a lot of activity in the sterile-neutrino area. The nuPRISM and JSNS² proposals in Japan could also test for ν_μ → ν_e appearance, while new proposals like the KPipe experiment, also in Japan, can contribute to the search for ν_μ disappearance. The MINOS+ and IceCube detectors, both of which have already set strong limits on ν_μ disappearance, still have additional data to analyse. A suite of experiments is already currently under way (NEOS, DANSS, Neutrino-4) or in the planning stages (PROSPECT, SoLid, STEREO) to test for electron-antineutrino disappearance over short baselines at reactors, and others are being planned that will use powerful radioactive sources (CeSOX, BEST). These electron-neutrino and -antineutrino disappearance searches are highly complementary to the search modes being explored at SBN. 

The Fermilab SBN programme offers world-leading sensitivity to oscillations in two different search modes at the most relevant mass-splitting scale as indicated by previous data. We will soon have critical new information regarding the possible existence of eV-scale sterile neutrinos, resulting in either one of the most exciting discoveries across particle physics in recent years or the welcome resolution of a long-standing unresolved puzzle in neutrino physics.

LArTPCs rule the neutrino-oscillation waves

A schematic diagram of the ICARUS liquid-argon time projection chamber (LArTPC) detector, where electrons create signals on three rotated wire planes. The concept of the LArTPC for neutrino detection was first conceived by Carlo Rubbia in 1977, followed by many years of pioneering R&D activity and the successful operation of the ICARUS detector in the CNGS beam from 2010 to 2012, which demonstrated the effectiveness of single-phase LArTPC technology for neutrino physics. A LArTPC provides both precise calorimetric sampling and 3D tracking similar to the extraordinary imaging features of a bubble chamber, and is also fully electronic and therefore potentially scalable to large, several-kilotonne masses. Charged particles propagating in the liquid argon ionise argon atoms and free electrons drift under the influence of a strong, uniform electric field applied across the detector volume. The drifted ionisation electrons induce signals or are collected on planes of closely spaced sense wires located on one side of the detector boundary, with the wire signals proportional to the amount of energy deposited in a small cell. The very low electron drift speeds, in the range of 1.6 mm/μs, require a continuous read-out time of 1–2 milliseconds for a detector a few metres across. This creates a challenge when operating these detectors at the surface, as the SBN detectors will be at Fermilab, so photon-detection systems will be used to collect fast scintillation light and time each event.