Early Galaxies Put New Limits on Hidden Neutrino Interactions

A particle question written in galaxies

The earliest galaxies formed within clumps, or halos, of dark matter. Their abundance reflects both how matter accumulated under gravity and how efficiently gas inside halos formed luminous stars. Ranjini Mondol and colleagues use that connection to test a hypothetical interaction between neutrinos, the elusive particles present throughout the universe.

Their analysis combines observations already made by the James Webb Space Telescope (JWST) with measurements of the cosmic microwave background (CMB) from Planck. It reports upper limits, not a discovery of neutrino self-interactions. The work is an arXiv preprint, version 2611.07013v1; its landing page does not report journal acceptance.

From neutrinos to starlight

In the standard cosmological picture, early-universe neutrinos eventually begin to travel largely unimpeded, a behavior called free streaming. An additional interaction among them could delay that transition. This would change the evolution of gravitational potentials and, in turn, the growth of fluctuations in matter density.

The researchers describe these fluctuations with the matter power spectrum, a measure of how strongly matter clusters on different scales. In their model, neutrino interactions can enhance clustering at some scales and suppress it at smaller ones; they do not simply erase all structure below a sharp cutoff. The effects discussed in the paper reach roughly 15% in the power spectrum, though the resulting change in the abundance of rare halos can be larger.

JWST does not measure this spectrum directly. Instead, the authors use the ultraviolet luminosity function: the number density of galaxies in different ultraviolet brightness ranges. Because smaller structures form first in the hierarchical picture of cosmic growth, galaxies at high redshift can provide information about small-scale clustering. Redshift, denoted by z, indicates how much the universe has expanded since light was emitted; larger values generally correspond to earlier times.

How the comparison works

The team calculated power spectra for interacting-neutrino models and used a halo mass function to estimate how many dark matter halos of each mass should form. They then linked halo mass to galaxy brightness using a star-formation model calibrated against Thesan-Zoom simulations. The model allows for scatter in brightness associated with bursty star formation and includes a dust correction.

The JWST likelihood uses ultraviolet luminosity-function data at redshifts above approximately 9. According to the paper, the measurements and limits through about z = 16 are spectroscopic; the higher-redshift inputs, from about z = 17 onward, are photometric and include galaxy candidates. The analysis also uses Planck 2018 temperature and polarization data and a Pantheon Type Ia supernova likelihood. Its combined-data setup adds priors informed by nucleosynthesis and CMB measurements.

Using Bayesian sampling, the authors varied standard cosmological parameters, the sum of neutrino masses, the interaction strength and parameters describing uncertain galaxy astrophysics. They compared two cases: one interacting neutrino flavor with two free streaming, and all three flavors interacting. Their principal analysis allows the neutrino mass sum to vary above 0.06 electronvolts and uses a prior range of −6 to −2 for the base-10 logarithm of the effective coupling.

What the limits show

For the one-flavor case, the reported 95% upper limit on log10 of the coupling, measured in inverse megaelectronvolt squared, changes from −2.00 in the paper's Planck-only analysis to −2.75 with the combined data. The Planck-only bound sits at the upper edge of the chosen prior, reflecting weak sensitivity in that case. For three interacting flavors, the corresponding limit changes from −3.29 to −3.41: an improvement, but a comparatively modest one. These are model-dependent bounds on an effective coupling, not measurements of an interaction strength.

The authors attribute the additional sensitivity mainly to the highest-redshift luminosity functions. In their calculations, combinations near the upper limits can suppress the small-scale clustering and halo abundance needed to account for early galaxies. Neutrino mass also suppresses structure, so its assumed range matters: the authors note that a higher lower bound on the mass sum would strengthen the inferred interaction limits.

What remains uncertain

Galaxy brightness is not a direct census of dark matter halos. Star formation, its variability and dust must be modeled, and the high-redshift observations have substantial uncertainties. The authors assume their astrophysical model parameters do not evolve with redshift and note that simulations explicitly including the proposed neutrino interactions would improve the halo calculation. Switching the smoothing filter used in that calculation changes the reported upper limits by 3–4%, but that check does not test every modeling uncertainty.

For the three-flavor case, the paper says current constraints are weaker than those from some other probes, including the Lyman-alpha forest. More galaxy data and joint analyses could improve the test, but the paper's discussion of possible sensitivity down to a logarithmic coupling near −5 is an illustrative estimate, not a statistical forecast or an achieved limit.