KATRIN: How to Weigh a Neutrino You Cannot Catch

Measuring What Is Missing

When tritium undergoes beta decay, it emits an electron and an antineutrino, and the available energy is shared between them. Usually the split is uneven. Very occasionally the electron takes almost everything, leaving the neutrino with almost nothing - and in those rare events, the neutrino's rest mass becomes the limiting factor, because even a stationary neutrino must be given the energy equivalent of its mass.

That sets a hard ceiling on the electron's energy, slightly below where it would sit if the neutrino were massless. The size of the shortfall is the neutrino mass. So the measurement reduces to determining the exact shape of the electron spectrum in its final fraction of an electronvolt.

The difficulty is that this is where almost no decays occur. Of the roughly 10^11 tritium decays per second in KATRIN's source, only about one in a trillion lands in the region that carries the information. Everything about the apparatus follows from needing to collect enough of those events and measure each one precisely.

The virtue of the method is that it assumes nothing. It uses energy conservation and the kinematics of a two-body decay, and nothing else. It does not care whether the neutrino is its own antiparticle, which nuclear-decay searches for neutrinoless double beta decay must assume, and it does not depend on cosmological modelling as limits from the early universe do.

A Seventy-Metre Instrument for One Number

The apparatus begins with a windowless gaseous tritium source, in which tritium gas circulates continuously at low temperature. Electrons from the decays are guided out magnetically along a 70-metre beamline while the tritium itself is pumped away, because letting it reach the spectrometer would flood the measurement with background.

The centrepiece is the main spectrometer: a vessel 23 metres long and 10 metres in diameter, evacuated to around 10^-11 millibar. It works as a magnetic adiabatic collimation with electrostatic filter, which in practice means it acts as a very sharp high-pass filter for energy. Only electrons above an adjustable threshold can climb the electrostatic potential and reach the detector at the far end; everything else is turned back.

By stepping that threshold across the last few electronvolts and counting what gets through at each setting, the experiment reconstructs the spectrum's shape point by point. The precision required is such that the electrostatic potential must be stable to a few parts per million over months.

The spectrometer itself became briefly famous for its delivery. Built in Deggendorf in Bavaria, roughly 400 kilometres from Karlsruhe, it was too large for any road, so it travelled by river and sea around half of Europe - about 8,600 kilometres, through the Danube, the Black Sea, the Mediterranean, the Atlantic and the Rhine - to cover that distance.

The Experiment in Numbers

KATRIN operated at the Karlsruhe Institute of Technology, with data taking beginning in 2019.

• Method: precision spectroscopy of the tritium beta-decay endpoint, model-independent
• Beamline: approximately 70 metres from source to detector
• Main spectrometer: 23.3 metres long, 10 metres in diameter, vacuum around 10^-11 millibar
• Source activity: on the order of 10^11 tritium decays per second
• Useful fraction: roughly one decay in 10^12 falls in the region carrying mass information
• Limits achieved: below 1.1 eV (2019), below 0.8 eV (2022), below 0.45 eV (2025)
• Design sensitivity: about 0.2 eV

Why the Number Matters Beyond Particle Physics

Neutrinos are the second most abundant particle in the universe after photons. Even a tiny rest mass, multiplied by that abundance, adds up to a gravitational contribution comparable to all the stars combined - which means the neutrino mass is a parameter in how cosmic structure formed.

Cosmological observations already constrain the sum of the neutrino masses, and often more tightly than KATRIN does. But those limits are only as good as the cosmological model they are derived within. A laboratory measurement that depends on nothing but energy conservation provides an independent anchor, and a disagreement between the two would be interesting rather than embarrassing.

The mass also bears on why neutrinos are so much lighter than every other massive particle - at least a million times lighter than the electron. The leading explanations tie that lightness to physics at energies far beyond any accelerator, so the measured value is a constraint on theories nobody can test directly.

For the broader context of what oscillation experiments can and cannot determine, see our pages on neutrino oscillation and the Standard Model, which treats the neutrino as massless in its original formulation.

What KATRIN Does Not Settle

It has not measured the neutrino mass. It has bounded it. Every result so far is an upper limit, and the possibility remains that the mass lies below the reach of the method entirely, in which case KATRIN's final answer will be a smaller number with no detection attached.

It also measures an effective mass - a weighted combination of the three mass states - rather than any individual one, and it says nothing about the ordering of those states, which is what DUNE and other long-baseline experiments are built to determine.

And it has no bearing on energy applications, which is worth stating because the experiment is sometimes cited loosely in that context. KATRIN does not capture neutrinos or extract anything from them; it watches electrons from a radioactive source and measures a subtle distortion in their energy distribution. The question in neutrinovoltaic research - whether ambient radiation and thermal fluctuations at a material surface can drive a measurable current - is unrelated to what KATRIN measures.

What the experiment does deliver is a number that constrains cosmology, particle theory and the structure of matter, obtained without trusting any of them. That independence is the whole point of the design.