Title: Constraints on Axions from Patchy Screening of the Cosmic Microwave Background
Authors: Goldstein, S., McCarthy, F., Mondino C., Hill J. C., Huang, J., and Johnson, M. C.
First Author’s Institution: Department of Physics, Columbia University, New York, NY 10027 USA
Status: published in Physical Review Letters [closed access]
One of the main mysteries of modern astronomy is the nature of dark matter. Decades of observational evidence tell us that there must be a lot of unseen mass out in the universe. Since this matter doesn’t interact with light (or only interacts very weakly with light), we can’t directly observe it the way we would observe normal matter out in space. Instead, astronomers and physicists have to come up with novel theories that could explain dark matter, and design experiments that can shine a bit of light on the nature of these perplexing particles.
Many different kinds of particles have been proposed as dark matter candidates, each with their own implications for both particle physics and astrophysical environments. Axions are a kind of particle originally proposed to solve a problem in quantum physics called the CP problem and are a natural consequence of many beyond-standard-model theories, including string theory. They have since also become a popular dark matter candidate, with many Earth- and space-based experiments designed to detect axions.
A unique feature of axions is their ability to turn into photons (and vice versa) in the presence of a magnetic field. The mechanism behind axion-photon oscillations is similar to what allows neutrinos to oscillate between different flavours. The probability of a photon turning into an axion depends on the properties of the medium through which it is travelling. Luckily, the low-density, highly-ionised, and magnetised environment of galactic halos (regions around galaxies) are a great place to look for axion-photon oscillations.
But, in order to observe axion-photon oscillations in galaxy halos, like today’s authors, you need to understand your source of photons very well. Otherwise, it’s impossible to tell whether the light you’re receiving has been affected by axion-photon oscillations. Luckily, the cosmic microwave background (CMB) is an ideal background source to use.
The CMB is light left over from the Big Bang; we see it in every direction in the sky and it’s a very even source of light. It’s been mapped in great detail by telescopes like the Planck Satellite, so that even the very slight deviations in the CMB are also well-understood by astronomers. This makes it a great candidate as a background source in searches for axion-photon oscillations.
A schematic with the cosmic microwave background on the right, emitting photons shown by white wavy lines, which travel through galaxy halos. In the halos, two wavy lines meet and turn into a white dashed line, symbolising the production of an axion. On the left is an image of a space telescope observing the CMB and seeing blue patches in the regions associated with galaxy halos.
Figure 1: This schematic shows you how photons travel from the CMB, through galaxy halos where they might turn into axions, and how this then affects our view of the CMB. Axion screening creates distinctinctive patchiness in the CMB. Figure 1 in today’s paper.
Figure 1 shows you how the experiment set up by today’s authors works. CMB photons travel through space and eventually encounter the diffuse, magnetised material of a galaxy halo. Some of these CMB photons might be converted into an axion as indicated by the dashed line. As a result, you would expect the CMB to be ever so slightly “dimmer” in regions of the sky coincident with the locations of galaxy halos. Today’s authors looked for correlations between the location of patchiness in the CMB and the location of galaxies to look for evidence of axion-photon oscillations.
In order to model the expected signature of axion screening, the authors had to make a model of the galaxy halos the CMB photons would be travelling through. To map the distribution of halos in space, they use a map of galaxies from the unWISE catalogue, a catalogue constructed using data from the Wide-Field Infrared Survey Explorer. They then use the results of models like the IllustrisTNG simulations to model the distribution of electrons and magnetic fields in galactic halos.
A graph showing the strength of cross-correlation as a function of scale. Blue and red lines show that axion screening should create a negative cross correlation that remains fairly constant across scale. The observational results are shown by black points, which are all consistent with zero cross correlation.
Figure 2: Strength of cross-correlation between CMB patchiness and galaxy halo locations (y-axis) over a range of scales (x-axis). The expected signatures of axions of two different masses are shown in blue and red, and the results of today’s paper are shown by black squares with errorbars. Figure 2 in today’s paper.
The results of today’s paper are shown in Figure 2. In the simplest terms, you can think of the x-axis as being the scale on which you’re looking for a correlation between galaxy halo positions and axion-photon oscillation-induced patchiness, and the y-axis to be the strength of the correlation at that scale. The red and blue lines show what we would expect to see for axions of two different masses, which capture the range of axion masses that this experiment is most sensitive to. The black points with errorbars show the results of the analysis. You can see that the observed correlation strength is mostly consistent with 0 – or no correlation.
A graph showing the sensitivity of different experiments to axions of different coupling strengths (y-axis) and masses (x-axis). CAST is shown by a blue rectangle, showing it's sensitive to all masses, but only a small range of coupling strengths. The sensitivity of the experiment described in this article is shown by a red region which extends to weaker coupling strengths at intermediate masses.
Figure 3: Sensitivity of different experiments to axions of different mass (x-axis) and coupling strengths (y-axis). The region excluded by today’s paper is shown in red. Values excluded by CAST are shaded in blue, and constraints from other astronomical sources are shown in grey.
Does this mean that axions don’t exist? Not quite. The analysis in today’s paper is only sensitive to axions of a certain mass and coupling strength. Figure 3 shows the masses and coupling strengths that this particular analysis technique could detect. The results of today’s paper tell us that if axions do exist, then they can’t have a mass or coupling strength within the red area on the graph.
There are many other methods of searching for axions, some of which are shown in Figure 3. You can see that they’re all sensitive to different masses and coupling strengths, allowing astronomers and physicists to gradually whittle down the potential properties of axions. The results of today’s paper significantly improve on the constraints made by the CERN Axion Solar Telescope (CAST, the blue region in Figure 3). Future CMB measurement and galaxy surveys will allow this analysis to be done with greater sensitivity and open new avenues of research. Better, more detailed models of the magnetic fields and distribution of free electrons in galaxy halos might also improve future versions of this experiment. There are also many different ways we can look for axions in space, which you can read more about here and here. The hunt for axions continues!
Astrobite edited by Skylar Grayson
Featured image credit: Goldstein et al., (2025)
Author
Nathalie Korhonen Cuestas is a second-year PhD student at Northwestern University, where her research focuses on the chemical evolution of galaxies.
View all posts