Deep divefrom Episode 1Cosmology24 August 202616 min

Axion search in the diffuse supernova neutrino background at

A deep dive into one paper: what was known before, what the researchers did, what the numbers say and what they don't.

The paper: An axion constraint from the diffuse supernova neutrino background indicated by Super-Kamiokande (arXiv)

Chapters

Also linked

Reference: Search for Diffuse Supernova Neutrino Background with Super-Kamiokande Gadolinium Dataset (ApJ)

Transcript

The script the two AI voices read from. The times are worked out from the script, not heard, so a line can start a few seconds early or late.

Read the whole episode

Cold open

Host 1Welcome to the Hastrup Review of Science — where artificial intelligence reads real science, so you don't have to pretend you understood the paper. One paper today. A group of physicists has figured out how to hunt for an invisible particle using the collective whisper of every star that has ever exploded.

Host 2Mori, Takiwaki, Kohri and Mori have used the first hint of the diffuse supernova neutrino background, recently detected by Super-Kamiokande, to set a new constraint on axion properties. The paper is called An axion constraint from the diffuse supernova neutrino background indicated by Super-Kamiokande.

Core-collapse supernovae

Host 1That title packs in a lot. Let's unpack it from the bottom up. What happens when a massive star dies?

Host 2When a star with more than roughly eight times the mass of the sun exhausts its nuclear fuel, its iron core can no longer support itself against gravity. The core collapses in less than a second, compressing roughly one and a half solar masses of material into a sphere about thirty kilometres across. That is a proto-neutron star. The gravitational binding energy released is enormous — roughly three times ten to the fifty three ergs.

Host 1And almost all of that energy comes out as neutrinos.

Host 2About ninety nine percent. The spectacular optical display of a supernova, the explosion visible across galaxies, represents less than one percent of the total energy budget. The real action is invisible. A burst of neutrinos of all flavours streams out over roughly ten seconds as the proto-neutron star cools and contracts.

SN 1987A

Host 1We have detected that neutrino burst exactly once. Supernova 1987A.

Host 2In the Large Magellanic Cloud, February the twenty third nineteen eighty seven. The Kamiokande-II detector in Japan and the IMB detector in the United States recorded a combined twenty four neutrino events over about thirteen seconds. That single detection has been the foundation of supernova neutrino physics for nearly four decades. It confirmed the basic picture of core collapse and set constraints on a wide range of new physics.

The diffuse supernova neutrino background

Host 1Twenty four events from one supernova. Now, what is the diffuse supernova neutrino background?

Host 2Every core-collapse supernova that has ever occurred in the observable universe has emitted a burst of neutrinos. Most of those neutrinos are still travelling through space. They have been redshifted by cosmic expansion, losing energy as the universe stretches, but they are still there. The sum of all those neutrinos from all those supernovae across cosmic history forms a faint, isotropic background. It is called the DSNB, or sometimes the supernova relic neutrino background.

Host 1How faint are we talking?

Host 2A few electron antineutrinos per square centimetre per second in the detectable energy range. For context, the sun sends about sixty five billion neutrinos per square centimetre per second through your body. The DSNB is roughly ten orders of magnitude weaker than the solar neutrino flux.

The Super-K detection

Host 1So how did Super-Kamiokande detect it?

Host 2By changing its own chemistry. Super-Kamiokande is a fifty thousand tonne water Cherenkov detector located under a mountain in central Japan. The DSNB signal it looks for is inverse beta decay: an electron antineutrino hits a proton in the water, producing a positron and a neutron. The positron produces Cherenkov light and is easy to see. The problem was always the neutron.

Host 1Why is the neutron a problem?

Host 2In pure water, a free neutron takes about two hundred microseconds to capture on hydrogen, producing a single low-energy gamma ray that is very hard to distinguish from backgrounds. Starting in 2020, the collaboration dissolved gadolinium sulfate into the water. Gadolinium has an enormous neutron capture cross section. It grabs the neutron within about thirty microseconds and emits a cascade of gamma rays totalling roughly eight megaelectronvolts. That is a distinctive, unmistakable coincidence tag.

Host 1And with that tag, they saw the DSNB.

Host 2They saw a 2.6 sigma indication of it. Using 956 days of gadolinium-loaded data, the collaboration measured an electron antineutrino flux of 3.6 plus 1.6 minus 1.5 per square centimetre per second, in the energy range 13.3 to 81.3 megaelectronvolts. That result was published in the Astrophysical Journal earlier this year. It is the first indication of the DSNB from any detector.

Host 1Not a discovery. 2.6 sigma is a hint, not the five sigma threshold.

Host 2Correct. But Mori and colleagues argue that even a hint is enough to start doing particle physics with. And the particle physics they choose to explore is the axion.

The strong CP problem and axions

Host 1Which brings us to the particle. What is an axion?

Host 2To explain axions, you need to start with a puzzle called the strong CP problem. The theory of the strong nuclear force, quantum chromodynamics or QCD, allows a term in its equations that would violate a symmetry called CP — the combined operation of swapping particles for antiparticles and reflecting space. If that term existed at any significant strength, the neutron would have a measurable electric dipole moment.

Host 1And it does not.

Host 2Experiments constrain the neutron electric dipole moment to be vanishingly small, which implies that the CP-violating parameter, called theta, is less than about ten to the minus ten. The puzzle is why. Theta could naturally be anywhere between zero and two pi. Nothing in the standard model forces it to be that close to zero.

Host 1So there is a dial that could be set to any value, and nature has set it to essentially zero.

Host 2In 1977, Roberto Peccei and Helen Quinn proposed a solution. They introduced a new symmetry that dynamically drives theta to zero. When that symmetry is spontaneously broken, it produces a new, very light pseudoscalar particle. Steven Weinberg and Frank Wilczek independently realised this particle was a necessary prediction of the mechanism. Wilczek named it the axion, after a brand of laundry detergent, because it cleaned up a problem in physics.

Host 1Named after a cleaning product.

Host 2The essential property of axions is that their mass and coupling strength are inversely proportional to a single energy scale called the Peccei-Quinn breaking scale, f sub a. A very high f sub a means a very light, very weakly coupled axion. Current constraints suggest f sub a is at least around ten to the nine GeV, corresponding to axion masses below about ten millielectronvolts. Such a particle would interact so feebly with ordinary matter that direct detection is extraordinarily difficult.

Axions in supernovae

Host 1But a supernova core might produce them.

Host 2This is where astrophysics becomes a particle physics laboratory. The interior of a proto-neutron star reaches temperatures of roughly thirty to fifty megaelectronvolts and densities several times that of an atomic nucleus. Under those conditions, axions can be produced through nucleon bremsstrahlung: two nucleons scatter, and in addition to exchanging a pion, they radiate an axion.

Host 1And because axions interact so weakly, they escape immediately.

Host 2Unlike neutrinos, which are trapped in the dense core for seconds and must diffuse out, axions at the relevant coupling strengths stream freely out of the proto-neutron star the instant they are produced. Each axion that escapes carries away energy. That energy would otherwise have heated the star and ultimately been radiated as neutrinos.

Host 1So axions act as an additional cooling channel. More axion production, faster cooling, fewer neutrinos.

Host 2Precisely. And this is what the classic SN 1987A cooling argument is based on. The neutrino burst from 1987A lasted about thirteen seconds. If axions had been carrying away too much energy, the proto-neutron star would have cooled faster and the neutrino burst would have been shorter. The observed duration constrains the axion-proton coupling to roughly six times ten to the minus ten.

Host 1But that bound comes from twenty four neutrinos and one supernova.

Host 2Correct. And that is the limitation this new paper addresses.

A word from our sponsor

Host 1A moment of domestic science. Do you own wet laundry?

Host 2The Hastrup Review of Science is brought to you by clothes pegs. The original clamping technology.

Host 1Humanity split the atom. We sequenced the genome. We detected the neutrinos of dead stars. And we still hang our socks on a line with a spring-loaded wooden clip.

Host 2Good clothes pegs do not break in wind, do not leave marks on fabric, and do not require a physics degree to operate. Unlike most things we discuss on this programme.

Host 1Clothes pegs. Because gravity is not your laundry's friend. Our referral link is in the show notes.

The method

Host 1Back to the science. So the authors want to derive an axion bound from the DSNB instead of from a single event. Walk me through what they did.

Host 2They ran one-dimensional general-relativistic neutrino-radiation hydrodynamic simulations of a core-collapse supernova, following the proto-neutron star evolution for about fifteen seconds after core bounce. They added axion emission via nucleon bremsstrahlung using the one-pion-exchange approximation with spin fluctuation corrections. Then they ran seven models spanning a range of axion masses up to eleven millielectronvolts.

Host 1Which axion model did they use?

Host 2The KSVZ framework — Kim, Shifman, Vainshtein, Zakharov. In this model, the axion couples primarily to protons, with negligible coupling to electrons and a suppressed coupling to neutrons. So the dominant interaction in the supernova core is axion-proton. That makes the axion-proton coupling constant, g sub a p, the key parameter to constrain.

Host 1How do they go from a single simulated supernova to the DSNB?

Host 2They took the modified neutrino spectra from each simulation and integrated them over cosmic history, weighting by the empirical core-collapse supernova rate as a function of redshift. That gives a predicted DSNB spectrum for each value of the axion coupling. Then they compared those predictions to the Super-Kamiokande measurement.

The result

Host 1What did they find?

Host 2At an axion-proton coupling of 8.5 times ten to the minus ten, the electron antineutrino flux at ten megaelectronvolts drops by about thirty three percent compared to the no-axion case. The proto-neutron star cools measurably faster, and that faster cooling propagates directly into a reduced DSNB flux.

Host 1A third of the neutrino signal vanishes.

Host 2And comparing to the Super-Kamiokande measurement, they derive an upper limit on the axion-proton coupling of 1.3 times ten to the minus nine at one sigma. In terms of the Peccei-Quinn scale, that corresponds to f sub a above 3.5 times ten to the eight GeV, or equivalently an axion mass below about sixteen millielectronvolts.

Why the DSNB matters

Host 1How does that compare to the SN 1987A bound?

Host 2The conventional SN 1987A cooling bound on the same coupling is about six times ten to the minus ten. So the DSNB constraint is currently about a factor of two weaker in raw numbers.

Host 1Weaker. So why should we care?

Host 2Because the two constraints fail in completely different ways. The SN 1987A bound depends on everything being typical about that one event: the mass of the progenitor star, its metallicity, its rotation rate, whether it formed an accretion disk. If SN 1987A was unusual in any of those respects, the bound could shift significantly.

Host 1A sample size of one.

Host 2Exactly. And there are proposals in the literature that specific progenitor configurations — for example, a rapidly rotating core that forms an accretion disk — could produce a neutrino signal consistent with the 1987A data even in the presence of strong axion cooling. The SN 1987A bound might be circumventable if that particular supernova happened to have unusual properties.

Host 1And the DSNB bound cannot be circumvented that way.

Host 2That is the central point of the paper. The DSNB represents the cumulative neutrino emission from the entire cosmic population of core-collapse supernovae. Unusual individual events are averaged out. You cannot escape the constraint by arguing that any particular supernova was atypical, because the DSNB includes all of them. One unusual supernova hides in a sample of one. It cannot hide in a sample of billions.

Caveats

Host 1What are the main caveats?

Host 2Several important ones. First, the DSNB detection itself is 2.6 sigma. That is an indication, not a confirmed detection. If the measured flux shifts downward with additional data, the axion constraint weakens or could vanish entirely.

Host 1Second?

Host 2They used a single progenitor model — a 9.6 solar mass zero-metallicity star with very low compactness. Real supernovae come from a wide range of stellar masses, metallicities, and evolutionary histories. The proto-neutron star temperature and density profile varies across that population, and those properties directly determine axion production rates. A complete treatment would integrate over a distribution of progenitors.

Host 1Third?

Host 2Neutrino oscillations. As neutrinos travel from the proto-neutron star through the supernova envelope and across cosmological distances, they change flavour. The paper assumes no flavour conversion, which simplifies the analysis but is almost certainly wrong. Collective neutrino oscillations inside the supernova envelope, in particular, remain an unsolved problem in theoretical physics that could modify the predicted antineutrino spectrum at Earth.

Host 1Fourth?

Host 2The axion production rate itself carries theoretical uncertainty. The nuclear physics at supra-nuclear densities is not fully understood, and different treatments of the nucleon-nucleon interaction could shift the constraint in either direction.

What comes next

Host 1So the number should be read as a proof of concept rather than a precision measurement.

Host 2That is a fair summary. And the concept extends well beyond axions. The authors note that the same methodology applies to any feebly interacting particle that could cool a proto-neutron star. Axion-like particles, sterile neutrinos, dark photons — anything that provides an additional energy loss channel could be constrained using the DSNB.

Host 1What about future detectors?

Host 2Hyper-Kamiokande, currently under construction in Japan, will have roughly eight times the fiducial volume of Super-Kamiokande. It is expected to reach a five sigma discovery of the DSNB within about a decade of operation. The JUNO detector in China will add complementary sensitivity. As the DSNB measurement moves from a hint to a precision measurement, every particle physics constraint derived from it will tighten in proportion.

Host 1So the interesting thing here is not the specific number.

Host 2The interesting thing is the method. For the first time, a cosmological neutrino signal can constrain new physics in a way that does not depend on the accident of which particular supernova happened to explode close enough for us to detect. That is a qualitative shift, not just a quantitative one. It opens a new channel for particle physics that will only get stronger.

Sign-off

Host 1That is the deep dive. An axion constraint from the diffuse supernova neutrino background indicated by Super-Kamiokande, from Mori, Takiwaki, Kohri, and Mori. Links in the show notes.

Host 2This has been the Hastrup Review of Science. See you next week.