Usually, we believe that dark "matter" is fermionic. My question is simple: is there any reason why a SM-inert massless unknown boson could NOT be the thing we know as dark matter if it only couples to gravity? For instance, is there any problem with massless-type axion-like particles?

  • $\begingroup$ Wouldn't massless particles be "hot" dark matter, and hence not fit halo formation models? $\endgroup$ May 2, 2022 at 11:23
  • $\begingroup$ @AndersSandberg Interesting point! Indeed, I understand hot dark matter as "bosonic", but there is something missing (with the exception of some possible massless neutrinos) in the Standard Model: massless fermions (I am not sure if they are considered or not hot dark matter!). We do know, as far as I know, that SM neutrinos, even if one of them is Weyl-like and massless, can not be DM, but sterile neutrinos or even weird fermionic fields (I thought also in ELKO, even if controversial) can NOT be hot dark matter as we know... $\endgroup$
    – riemannium
    May 2, 2022 at 12:03

1 Answer 1


Actually, below a few electronvolts in mass, the phase space density of dark matter becomes so large that dark matter has to be bosonic. Axions are the most famous bosonic dark matter candidate, and there are of course others, dark photons for example. So, yes, massive (though light) bosons can be dark matter.

Usually, we believe that dark "matter" is fermionic.

This is certainly not true for the community working in the field.

The allowed dark matter mass range goes down to $10^{-21}$ electronvolts. Below that, the de Broglie wavelength of dark matter particles would become larger than the size of (dwarf) galaxies, and since these do contain dark matter, that mass range is ruled out. It is good practice to think of dark matter lighter than a few electronvolts as bosonic, i.e. a wave. Above that mass, it could be fermionic (but might still be bosonic too)

Now, you mention massless particles coupling to gravity. For starters, the "charge" of gravity is mass, but as one commenter pointed out, photons couple to gravity too, so that's not necessarily a show stopper. However, massless particles travel at the speed of light. In jargon, that would classify them as "hot" dark matter. Studies of the structure formation of galaxies however show that dark matter is "cold", i.e. non-relativistic. Thus, massless particles can not be the dark matter. (Or at least not all of it... neutrinos are dark matter, but only a tiny fraction of it)

  • $\begingroup$ This answer is great (+1), except for the last sentence. Massless particles absolutely can couple to gravity -- photons are an example. $\endgroup$
    – Andrew
    May 2, 2022 at 12:15
  • $\begingroup$ Fair point; Editing. $\endgroup$
    – rfl
    May 2, 2022 at 12:22
  • $\begingroup$ 1st. Massless particle do couple to gravity (i am well aware of it, since gravity couples both to mass AND energy). 2nd. Dark photons (as also some forms of "bosonic" dark matter), couldn't they be massless at all if "confining"? I could imagine some type of confined nonabelian massless gauge field up to scale of dwarf galaxy, couldn't I? 3rd. Matter/radiation nomenclature is fuzzy sometimes. 4rd. I believe superstrings predict,e.g, a large big amount of moduli (scalars), and some of them could be in principle massless (or almost) if "confined", or is that impossible? $\endgroup$
    – riemannium
    May 2, 2022 at 13:32
  • $\begingroup$ Indeed, lot of fun when you told me about the zeV particle...I am going to make my students "suffer" with a wave-form that is essentially an oscillating toy model for that class of DM oscillation field, but not in the finals, or they will KILL ME hahaha... If they read that problem, they will be crushed for sure... $\endgroup$
    – riemannium
    May 2, 2022 at 13:44
  • $\begingroup$ I don't immediately see a reason why confinement shouldn't work. Depending on the actual model I would guess that yes, you could make something like that work. $\endgroup$
    – rfl
    May 2, 2022 at 14:43

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.