On this page
The history of large neutrino detection runs along two parallel tracks. The water Cherenkov tradition — Kamiokande, IMB, Super-Kamiokande, Hyper-Kamiokande — builds enormous transparent water volumes and reads out the conical light pattern produced by relativistic charged particles. The liquid scintillator tradition — KamLAND, Borexino, JUNO — builds smaller volumes of organic scintillator with much higher photon yields and excellent energy resolution at low thresholds. Each tradition has its strengths. Water Cherenkov is cheap per ton and preserves event direction; scintillator delivers low threshold and good energy resolution but loses directional information because the light is isotropic.
THEIA is a proposed detector that refuses to choose. It is built around water-based liquid scintillator (WbLS), an engineered fluid in which a few per cent of organic scintillator is stably dispersed in ultra-pure water with surfactants. The mixture produces both prompt Cherenkov light with the directional pattern of a water Cherenkov detector, and delayed scintillation light with the photon yield of a (diluted) organic scintillator. With sufficiently fast photomultipliers and good time resolution, the two components can be separated in the readout, and the detector reads out simultaneously as a Cherenkov ring tracker and as a calorimetric scintillator.
A 25-to-100-kiloton THEIA could pursue several major physics targets in a single facility: the diffuse supernova neutrino background, solar CNO and ⁸B neutrinos, geoneutrinos, accelerator-beam oscillations from LBNF, atmospheric neutrinos, proton decay, and — with a doped inner region — neutrinoless double beta decay. This post is about how WbLS works, what THEIA aims to do, and where the technology stands.
What WbLS is
A standard organic liquid scintillator such as linear alkyl benzene (LAB) doped with PPO fluor produces about 11,000 scintillation photons per MeV of energy deposit by charged particles. Pure water produces no scintillation but emits about 300 Cherenkov photons per centimeter of relativistic charged-particle path, peaked in the ultraviolet. WbLS dilutes the organic scintillator into water — typically at 1 to 10 per cent by mass — using a surfactant such as Triton X-100 to keep the two phases stably mixed.
The resulting medium has approximately the optical density and transparency of pure water at the relevant wavelengths, instrumented with photomultipliers at meter scales just as Super-Kamiokande is. It produces scintillation light at a yield reduced by the dilution — about to photons per MeV depending on the scintillator fraction. It also still produces Cherenkov light from charged particles above threshold. The two light sources can be distinguished on multiple grounds:
- Time: Cherenkov light is emitted instantaneously by the passing particle (within a few hundred picoseconds); scintillation is emitted with a characteristic decay constant of order to nanoseconds, depending on the scintillator formulation.
- Angle: Cherenkov light forms a sharp cone around the particle direction; scintillation light is isotropic.
- Wavelength: Cherenkov light has a characteristic spectrum extending into the UV; scintillation light is dominated by the fluor’s emission spectrum, typically at 400-450 nm.
With sufficiently fast photomultipliers — 100 ps timing resolution or better — the two components can be separated event by event. Recent photomultiplier development, particularly silicon photomultipliers and the next generation of fast PMTs aimed at THEIA, has brought this within reach for kiloton-scale detectors.
What you get from combining the two
The combination delivers capabilities neither pure water Cherenkov nor pure scintillator can.
Direction reconstruction down to lower energies. In pure water Cherenkov, the threshold for directional reconstruction sits at the Cherenkov threshold of about 0.78 MeV for electrons and significantly higher for muons. Below that, no Cherenkov light is produced and the event is essentially invisible. In WbLS, the scintillation light still detects the event below Cherenkov threshold, and even above threshold, the additional scintillation photons reduce the energy at which directional information becomes useful. The result is directional reconstruction down to a few MeV — useful for sub-MeV solar neutrinos and for distinguishing fluxes from different directions in the diffuse supernova background.
Particle identification. The ratio of Cherenkov to scintillation light depends on the particle type. A relativistic electron produces both abundant Cherenkov and abundant scintillation; a non-relativistic proton produces only scintillation; a gamma ray produces primarily scintillation from secondary electron showers. The ratio therefore distinguishes the particle type, providing the same kind of event-by-event tagging that LiquidO achieves through topology and that SK-Gd achieves through delayed coincidences.
Sub-MeV energy resolution. The scintillation light yield is sufficient to give energy resolution better than 5 per cent at MeV-scale energies, comparable to pure scintillator and far better than pure water Cherenkov. This is essential for the spectral shape of CNO solar neutrinos and for separating individual electron-recoil populations.
Reduced Cherenkov-direction ambiguity. Pure scintillator detectors lose directional information entirely; WbLS retains it. For accelerator-beam oscillation experiments at THEIA’s proposed kt scale, the directional reconstruction matters because the neutrino-beam direction is known and the resulting muon or electron should point along it.
THEIA’s physics targets
The THEIA collaboration’s design proposal envisions a 25-to-100-kiloton fiducial volume hosted in an underground laboratory — proposed sites include LBNF/DUNE’s South Dakota cavern and a planned cavern at the SURF site. The detector would simultaneously address several physics goals.
Diffuse supernova neutrino background. The signal is approximately 1 to 10 events per kiloton per year at MeV energies, and SK-Gd is currently the leading effort. THEIA’s larger mass plus its ability to particle-identify the prompt-positron signature through Cherenkov/scintillation ratio would push the DSNB sensitivity well below current limits.
CNO solar neutrinos. Borexino’s recent CNO detection set the framework; a kiloton-scale WbLS detector with sub-MeV directional reconstruction would measure the CNO spectrum at much higher statistics, constraining the solar metallicity problem at percent-level precision.
⁸B solar neutrinos. With directional reconstruction and good energy resolution, THEIA would refine the existing Super-K and SNO ⁸B measurements and probe potential energy-dependent oscillation features in the MSW transition region.
Geoneutrinos. Antineutrinos from terrestrial U and Th decay produce inverse-beta-decay events at the rate of about 10 per kiloton per year, accumulating to thousands of events over THEIA’s exposure — substantially better statistics than KamLAND or Borexino on the same physics.
Long-baseline oscillation. With its proposed LBNF beam from Fermilab, THEIA could serve as a complementary or alternative far detector to DUNE’s LArTPC. The Cherenkov ring at GeV energies plus the scintillation light gives both directional and energy reconstruction at long-baseline scales.
Atmospheric neutrinos. Like Super-K, THEIA would measure atmospheric muon-neutrino oscillation parameters and tau-neutrino appearance.
Galactic supernova burst. A core-collapse supernova in the Milky Way would produce tens of thousands of events in THEIA, with the flavor decomposition cleaner than in any current detector thanks to the Cherenkov-scintillation separation.
Neutrinoless double beta decay. A smaller inner volume of WbLS doped with a 0νββ isotope — tellurium-130, molybdenum-100, or another candidate — would provide a kiloton-scale 0νββ source with low backgrounds, competing with or surpassing the next generation of dedicated 0νββ experiments.
The technology readiness
WbLS itself has been demonstrated at the tens-of-tonnes scale in dedicated test stands and at the kiloton scale in design studies. The key challenges have largely been controlled:
Long-term optical stability of the surfactant-mediated emulsion has been tested over multi-year timescales without significant phase separation. Aging studies show the scintillation yield is preserved at the few-per-cent level over years.
Radiopurity of WbLS components — the surfactant, the organic scintillator, and the water — has been brought to part-per-trillion levels comparable to what KamLAND and SNO+ have achieved. Continuous purification systems handle the maintenance of this purity.
Photomultiplier timing at 100 ps resolution is achievable with current PMTs, particularly the next generation developed for IceCube-Gen2 and JUNO. Silicon photomultipliers offer an alternative path at smaller pixel sizes but lower individual coverage. Whether to instrument THEIA with PMTs or SiPMs is one of the design choices still under evaluation.
Reconstruction algorithms that combine Cherenkov and scintillation pulses are under active development. Machine-learning approaches that take the full PMT waveform as input have shown promise for cleanly separating the two light components even when the timing overlap is significant.
The CHESS test facility at Lawrence Berkeley National Laboratory has demonstrated WbLS performance with prototype detectors and is the principal R&D testbed. ANNIE at Fermilab has run a one-kiloton WbLS sub-volume in an existing water-Cherenkov detector and validated the in-situ separation of Cherenkov and scintillation. The next step is a several-kiloton demonstrator, planned in the 2030 timeframe.
Why this matters
THEIA represents a deliberate move away from the single-purpose detector model. Most large neutrino experiments target one or a few specific physics goals — DUNE for long-baseline CP violation and supernova bursts, JUNO for mass ordering, SK-Gd for DSNB, KATRIN for neutrino mass. THEIA aims to be multi-purpose: a single detector that addresses a broad range of MeV-to-GeV neutrino physics through the flexibility of WbLS.
The economic argument is real. A 50-kiloton water-Cherenkov detector and a 5-kiloton scintillator detector together would cost significantly more than a single 50-kiloton THEIA, and the THEIA would do everything they would individually do — with the added benefit of cross-physics consistency from operating one calibrated detector. The science argument is also real: many of THEIA’s targets benefit from the combination of Cherenkov and scintillation, not just from one or the other.
Whether THEIA is ever built will depend on funding decisions and on the demonstrated maturity of WbLS at the kiloton-plus scale. The technical case is increasingly solid, and the physics case is broad enough that any major detector approval in the late 2020s — whether for DSNB, for 0νββ, for solar precision, or for long-baseline — could plausibly anchor a THEIA-class facility.
Summary
THEIA is a proposed 25-to-100-kiloton neutrino detector based on water-based liquid scintillator, an engineered medium in which a few per cent of organic scintillator is stably dispersed in ultra-pure water. The detector reads out both prompt directional Cherenkov light and delayed isotropic scintillation light, separated by photomultiplier timing at 100 ps. The combination enables directional reconstruction down to sub-MeV energies, event-by-event particle identification, and energy resolution approaching pure-scintillator performance — none of which a single existing detector technology delivers. THEIA’s physics program spans the diffuse supernova background, CNO and ⁸B solar neutrinos, geoneutrinos, long-baseline accelerator oscillations, atmospheric neutrinos, a galactic-supernova burst, and — with a doped sub-volume — neutrinoless double beta decay. WbLS technology has matured through dedicated test stands and one-kiloton in-situ demonstrations; a several-kiloton demonstrator is planned for the early 2030s, with full THEIA construction conditional on funding decisions and on the maturity of the underlying technology.