PTOLEMY: How to Catch the Cosmic Neutrino Background

The cosmic microwave background — relic photons from roughly 300,000 years after the Big Bang — was first detected in 1965 by Penzias and Wilson. The discovery confirmed the hot-Big-Bang cosmology and inaugurated modern observational cosmology. The 1978 Nobel Prize followed.

The CMB has a less-discussed companion: the cosmic neutrino background (CνB), the relic neutrino population from approximately one second after the Big Bang. When the universe cooled to about 1 MeV temperature, neutrinos decoupled from the rest of the cosmic plasma — analogous to how photons decoupled at recombination several hundred thousand years later. Since decoupling, the relic neutrinos have propagated essentially freely through the universe.

The CνB exists. Its existence is confirmed indirectly: the number of relativistic species in the early universe (parameterised as $N_{\text{eff}}$, with measured value ≈ 3.04) is consistent with three thermalised neutrino flavors. The CMB and BBN both constrain the relic neutrino population at the few-percent level. There is no theoretical room for the CνB to be absent.

But the CνB has never been directly detected. No experiment has yet observed an interaction caused by a relic neutrino. The reason: the relic neutrinos have temperatures of approximately 1.95 K, corresponding to energies of about $10^{-4}$ eV — far below the threshold of every conventional neutrino detector.

A direct detection of the CνB would be the cleanest possible confirmation of the hot-Big-Bang cosmology in its full neutrino sector. It would also provide tests of neutrino properties (Dirac vs Majorana, mass scale) inaccessible by any other technique. The experiment that could do this is being proposed and developed: PTOLEMY, the Pontecorvo-Olbermann’s Tritium Locator with Enhanced Methods of Y-ield.

This post is about what PTOLEMY would do, why the CνB is so hard to catch, and what success would mean.

The cosmic neutrino background in detail

The neutrino population from the Big Bang is set by thermal equilibrium during the radiation-dominated era. Neutrinos coupled to the rest of the cosmic plasma through weak interactions at temperatures above about 1 MeV. As the universe cooled, the weak interaction rate dropped below the expansion rate, and neutrinos fell out of equilibrium.

After decoupling, the neutrinos continued to redshift along with the rest of the universe. Their present-day temperature is: $T_{\nu }={\left(\frac{4}{11}\right)}^{1/3}{T}_{\gamma }=0.7138\times T_{\gamma }$

For the CMB temperature of 2.725 K, this gives $T_{\nu }=1.945$ K.

The corresponding number density of all three flavors (counting neutrinos and antineutrinos separately) is approximately: $n_{\nu }=\frac{6\zeta (3)}{11{\pi }^2}{T}_{\gamma }^3\approx 336{\text{ cm}}^{-3}$

This is similar to the CMB photon density. Every cubic centimetre of the universe contains approximately 336 relic neutrinos.

The neutrino energies follow a Fermi-Dirac distribution at temperature $T_{\nu }=1.95$ K, with a typical energy of approximately $⟨E_{\nu }⟩\approx 5\times 10^{-4}$ eV. This is enormously lower than any other neutrino source: solar neutrinos are at MeV, atmospheric and beam at GeV, supernova at tens of MeV. The relic neutrinos are 7-9 orders of magnitude lower in energy.

Why standard detectors don’t work

Conventional neutrino detection methods all have kinematic thresholds:

• Inverse beta decay: $E_{\nu }>1.806$ MeV. CνB is below by a factor of $10^{10}$.
• Elastic scattering on electrons: cross-section drops as $E_{\nu }^2$ at low energies. At meV energies, the cross-section is negligible.
• Coherent elastic scattering on nuclei: cross-section is enhanced compared to incoherent, but still requires energy transfer above the nuclear binding energies (eV scale).

All practical detection schemes that involve a neutrino interacting and depositing energy in a detector require the neutrino to have at least enough energy to produce some final-state product. The CνB does not.

The way around this kinematic problem was identified by Steven Weinberg in 1962: neutrino capture on a beta-unstable isotope. The idea is that the captured neutrino provides additional energy that allows the otherwise-forbidden direction of beta decay to proceed.

Neutrino capture on tritium

Consider the spontaneous beta decay of tritium: ${}^3\text{H}\to {}^3\text{He}+e^{-}+{\bar{\nu }}_e$

The electron has a continuous spectrum from zero up to the endpoint energy $Q=18.574$ keV. The reverse process — neutrino capture — is: ${\nu }_e+{}^3\text{H}\to {}^3\text{He}+e^{-}$

This is forbidden for free ${}^3$He atoms in their ground state if the neutrino has no energy. But if the captured neutrino has energy $E_{\nu }$, the resulting electron is monoenergetic at: $E_{e^{-}}=Q+E_{\nu }$

That is, the captured neutrino’s energy adds to the beta-decay Q-value, producing an electron above the standard beta-decay endpoint. The electron is monoenergetic — at energy $Q+E_{\nu }$, not part of a continuous spectrum.

For the CνB, with $E_{\nu }\approx 5\times 10^{-4}$ eV, the resulting electron is approximately 0.5 meV above the tritium beta-decay endpoint. Such a small energy shift would be undetectable in a standard spectrometer with eV-scale resolution. But with a sufficiently large tritium target and an extremely high-resolution electron spectrometer, the monoenergetic feature can in principle be resolved from the continuous beta-decay spectrum.

This is the PTOLEMY concept.

The PTOLEMY architecture

The proposed PTOLEMY (currently led by Princeton with international collaborators) consists of several integrated components:

Tritium target. A monolayer of tritium chemisorbed on graphene. The graphene substrate provides:

• A mechanical platform for the tritium target
• Reduced final-state distribution uncertainty (compared to molecular tritium)
• Compatibility with subsequent ion-optics elements

A few-tonne tritium load is required to give a measurable capture rate. The target mass is much larger than KATRIN’s 100-gram-scale tritium source.

Spectrometer. A radio-frequency (RF) tracker-spectrometer combining the principles of:

• Cyclotron radiation emission spectroscopy (CRES) — measuring the cyclotron frequency of the electron in a magnetic field, with frequency precision corresponding to meV energy resolution
• Magnetic separation — separating capture-peak electrons from beta-decay-continuum electrons by their slightly different kinematics
• Electrostatic filter — providing additional energy discrimination

The combined system aims for an energy resolution of approximately 50 meV — close to the kinematic gap for CνB capture.

Background suppression. The dominant background is the tritium beta-decay continuum itself: a few percent of beta-decay events fall within 1 eV of the endpoint, producing continuum events at energies near (but not at) the capture peak. PTOLEMY’s design strategy is to combine extreme energy resolution, low-background backgrounds-of-backgrounds rejection, and high statistics.

What PTOLEMY would measure

If the CνB is detected at the predicted rate, PTOLEMY would:

Confirm the hot-Big-Bang cosmology in the neutrino sector. Currently confirmed only indirectly through $N_{\text{eff}}$ measurements at the CMB and BBN epochs. Direct CνB detection would provide an independent, in-situ verification at the present epoch.

Measure the absolute neutrino mass. The capture-peak position is at $E=Q+E_{\nu }$. For non-relativistic relic neutrinos (which they are: $T=2\times 10^{-4}$ eV is well below the neutrino mass for any reasonable mass-scale), the captured energy equals the neutrino mass: $E_{\nu }=m_{\nu }$. So the capture-peak position directly measures the neutrino mass.

Test Dirac vs Majorana nature. The capture rate for Majorana neutrinos is approximately twice that for Dirac neutrinos, because Majorana relic neutrinos can be captured in either helicity. The capture rate measurement, normalized appropriately, would discriminate.

Measure the local relic-neutrino velocity distribution. The peak shape provides information about the kinematic distribution of the CνB, including any local clustering due to galactic gravitational potentials. Local CνB density could potentially be enhanced by orders of magnitude over the cosmological average.

Why this is so hard

The required experimental specifications are extreme:

• Tritium mass scale of approximately 100 grams to 1 tonne (for sufficient capture rate)
• Energy resolution at the 50-100 meV level (compared to KATRIN’s ~1 eV)
• Background rejection at the level of $10^{14}$ relative to the spontaneous tritium beta-decay rate
• Operations underground or at depth to suppress cosmic-ray backgrounds

The combination of these requirements means PTOLEMY is far more ambitious than any single existing technology can deliver. The development effort spans:

• Graphene-tritium chemistry (Princeton, INFN-Trieste, others)
• CRES spectrometry (Project 8 and related at University of Washington)
• Cryogenic electronics for RF readout
• Vacuum and thermal management at unprecedented scales

Each component is being developed in parallel through the late 2020s. First-stage demonstrator results are expected in the 2030s. Full physics-run sensitivity for CνB detection is in the late 2030s or 2040s.

The complementary programmes

PTOLEMY is not the only proposed approach. Several alternative concepts have been suggested:

Coherent enhancement at nucleons. The CνB could in principle produce coherent forward scattering on nuclei in macroscopic targets, with the cross-section enhanced by the square of the nucleon number. However, the coherent enhancement is only useful if the scattered signal can be observed — which requires energy transfer above experimental thresholds. Current proposals are mostly speculative.

Resonant nuclear capture. Some isotopes (e.g., ${}^{106}$Pd) have resonance-like enhancements for relic-neutrino capture at specific energies. The required precision in the resonance position is extreme. No working prototype exists.

Direct detection via Stodolsky-Cabibbo effects. Coherent forward scattering of relic neutrinos on macroscopic targets produces a tiny momentum transfer that could in principle be observed in ultra-low-temperature systems. No working prototype exists.

Cosmological constraint refinement. Improved CMB and BAO measurements (CMB-S4, Stage-IV galaxy surveys) will tighten the $N_{\text{eff}}$ constraint to ~0.03 precision — substantial improvement but still not a direct detection.

PTOLEMY remains the most concrete and promising direct-detection path. Its development is therefore central to the broader programme.

What success would mean

If PTOLEMY detects the CνB:

The hot-Big-Bang cosmology will have been confirmed in its neutrino sector with the same kind of in-situ measurement that the CMB provided in 1965. The relic neutrino population from the first second of the universe will have been directly observed.

The absolute neutrino mass will be measured. If the mass is at the 0.1 eV scale predicted by some models, PTOLEMY’s expected ~50 meV resolution should allow a clean measurement. Combined with KATRIN, Project 8, and cosmological bounds, the absolute mass scale will be pinned down at sub-eV precision.

The Dirac vs Majorana question will be answered. The capture-rate factor of 2 between the two scenarios would be unambiguous given sufficient statistics.

The local CνB density will be measured. Standard cosmology predicts ~336 cm⁻³; gravitational clustering could enhance this near galactic structures by factors of 10-1000.

These would be transformative results for fundamental physics. The direct detection of the CνB would close the experimental verification of the hot-Big-Bang framework in the neutrino sector, while simultaneously providing data on the absolute neutrino mass, the Dirac/Majorana nature, and the present-day relic neutrino distribution.

The long arc

The CνB has been theoretically expected since the late 1940s, when George Gamow and Ralph Alpher first developed the hot-Big-Bang framework. Steven Weinberg proposed neutrino capture on tritium as a detection method in 1962. The PTOLEMY collaboration formed around 2015.

The arc from theoretical prediction to direct detection is, in modern physics, often measured in decades. For the CMB it was 17 years (Gamow 1948, Penzias-Wilson 1965). For gravitational waves it was 100 years (Einstein 1916, LIGO 2015). For the Higgs boson it was 48 years (Higgs 1964, ATLAS/CMS 2012).

For the cosmic neutrino background, the arc is now approaching 80 years. Detection, if it comes, will be in the 2030s or 2040s. By then the CνB will be the oldest predicted-but-undetected fundamental phenomenon in physics — a record that, given the difficulty of the measurement, is in some sense earned.

For now, PTOLEMY represents the most credible path forward. Its development continues at multiple international institutions. Its eventual success — or failure — will resolve one of the longest-standing experimental gaps in modern cosmology.

The neutrinos are out there. They have been propagating freely for 13.8 billion years. They fill every cubic centimetre of space, including every cubic centimetre of your body, at a density of about 336 per cm³. Eventually, with sufficient patience and the right experimental architecture, we will see one of them interact. PTOLEMY’s job is to be ready when that happens.

For more on the relic-neutrino prediction and the cosmological background, see cosmic neutrino background and N_eff: how cosmology counts neutrinos. For related tritium-based experiments at the laboratory mass scale, see KATRIN’s tritium endpoint and Project 8 cyclotron radiation.