New paper!

Supernova Muons: New Constraints on Z' Bosons, Axions, and ALPs
https://arxiv.org/abs/2006.13942 

w/ Djuna Croon ( @QuantumMessage), @GillyElor + Sam McDermott

We use supernova muons to find some of the strongest existing limits on light new particles coupled to muons! Thread:

1/ Deep in the Large Magellanic Cloud, on the outskirts of the Tarantula Nebula, a blue supergiant named Sanduleak once shone with the brightness of over ten thousand Suns.

2/ That was, until one day, hydrostatic burning ended, no more nuclear energy could be released by fusion, and Sanduleak's iron core began to collapse under its own gravity.

3/ Eventually, its core became so massive that even electron degeneracy pressure could not stabilize it. As Sanduleak contracted,
photons began to dissociate the iron atoms of the inner core, which decreased the energy of the star even more and caused it to further contract.

4/ Electrons were absorbed onto protons, and converted into neutrons and neutrinos. The escape of the neutrinos further lowered the electron degeneracy pressure, until Sanduleak's core became unstable, and *collapsed*.

5/ Once Sanduleak's core reached nuclear densities, the collapse was abruptly *halted*. A shock wave formed between the outer and
inner core and moved outward from the core through the star.

6/ The implosion of the inner core ignited an explosion, and a core-collapse supernova was initiated. A mere 0.3 seconds after the collapse, the shock wave ejected the entire contents of Sanduleak's outer layers, leaving only a compact remnant (likely a neutron star) behind.

7/ Neutrinos screamed out of the stellar inferno, and were observed in 1987 by the Kamiokande II, IMB, and the Baksan collaborations, in an event known as Supernova 1987A (SN1987A).

8/ The transformation of Sanduleak into SN1987A has gifted us with a wealth of new insights for particle, nuclear, and astro physics.

9/ By comparing with simulations, it appears that the number of neutrinos that were observed in SN1987A is consistent with that expected from known physics.

10/ This provides an exciting avenue to probe other light new particles. That is, if new particles were also created during this explosion, they might leak energy out of the event, leaving less energy available to power the production of the neutrinos.

11/ This would be in conflict with the neutrinos observed in SN1987A. This means that any other light particles that live long enough to escape the system, can be constrained by the fact that they are not allowed to steal away too much energy.

12/ When producing these new particles, research in the past has mostly focused on production (and interactions) using the protons or neutrons in the supernova.

13/ We do something largely different. We consider the impact of interactions instead with the *muons* produced in 1987A.

14/ While these are less abundant than the protons or neutrons, they can provide a more sensitive probe of particles that would interact directly with muons, or leptons.

15/For the first time, we considered limits that arise from Z' bosons (force carriers) interacting with these muons. To get such limits,we exploited the most recent simulations of the muon number density profiles in the explosion, as well as temperature profiles of the explosion.

16/ The profiles give us a measure of these variables as a function of the radius from the center of the supernova. This allows us to determine how strongly these new interactions would occur, and how much energy loss they would lead to during Sanduleak's demise.

17/ When the Z' only interacts with muons and taus, in a particle model called gauged muon number minus tau number, we get the following limits, where the y-axis is the Z' coupling, and the x-axis is the Z' mass:

18/ The dip on the right hand side comes from neutrinos in the supernova pair coalescing to produce the Z', which would have then have run away stealing some of the supernova energy.

19/But it isn't allowed to do that in order to match observation, so it is shown as the excluded hot pink region.The bolder and lighter hot pink excluded regions come from taking a conservative version of the profiles, and a less conservative version of the profiles respectively.

20/ The flat line on the left hand side of the plot comes from something called a semi-Compton process, where a Z' could be produced after a photon interacts with a muon, and outputs a Z' and a muon. Again, as this would make the Z' a greedy energy thief, it is excluded.

21/ Compared to other processes, we find these limits are largely independent of the Z' mass, and can extend down to arbitrarily low Z' mass, even though they are only shown in a smaller mass window here.

22/ We also show the previous estimate for this bound (dashed), which had not used the charged muons in the supernova, along with other relevant constraints from Neff, neutrino tridents ("CCFR"), and g-2.

23/ As our greedy Z' exclusion covers some of the g-2 region, we exclude the g-2 explananation for this model for Z' masses less than about 10^-5 MeV.

24/ We also considered the case that kinetic mixing was present in the Lmu-Ltau model. In this case, the bounds don't change much, but instead, the competing constraints do. You can see this in our plots for different amounts of mixing:

25/ The resonance in the larger mixing case around 20 MeV comes from producing the Z' from a kinetic mixing loop, off protons rather than muons.

26/ We also extended these limits to a popular model class, called "B-L", which is gauged baryon minus lepton number. This just determines which particles the Z' will interact with. In that case, we also find new bounds:

27/ Here, the biggest change from previous estimates arises from the neutrino-pair coalescence process, which gives more sensitive constraints at higher Z' masses.

28/ Separate to the Z', there are other types of light particles that could be produced in this event, called axions, or axion-like particles. Probing these particles using supernova muons was studied in a nice recent paper by Bollig, deRocco, Graham, and Jenka (2020).

29/ This is also the paper that produced the awesome new simulations.

30/ That work focused on tree-level couplings of the axion. We were curious about how much these bounds might change when loop processes were also considered. We also extended the calculation to higher masses, to determine the full constraint across *all* axion masses.

31/ We found that the loops didn't change the bounds in Bollig et al for their mass range. In the higher, new mass range we were also considering, we found that loops were very important.

32/ We found the following exclusion on the axion-muon interaction parameter space, where the x-axis is the axion mass, and the y-axis is its muon coupling:

33/ The flat lines arise from a combination of processes from muon bremsstrahlung, and the semi-Compton process. On the right hand side, the larger couplings lead to the axion getting trapped (and therefore not stealing energy), and so the limits cover less of the couplings.

34/ We find we can constrain axions up to masses to nearly 800 MeV with one of the profiles. This is really high! The reason we have such good sensitivity here, is that the core temperature is really high, and the trapping is less efficient.

35/ In our work, we have pointed out and explicitly demonstrated for the first time the broad applicability of supernova muons to provide a sensitive probe of models of new physics. This certainly motivates further studies of how muons behave in supernovae!

Shout out to Sanduleak, whose ultimate sacrifice gifted us with so many interesting stories to tell, and so many things we can learn about new physics. :)

Also shout out to my awesome collaborators, Djuna, Gilly, and Sam! Stay tuned for another paper we have in the works. ;)