 Okay, I think we're live. So, hello everybody and welcome to this 85 first series of the Latin American webinars on physics. Today, I'm going to be the host of this very interesting webinar that is going to be given by Johannes Herms from the Technician University of Munich. And he's going to talk about anti-nuclear cosmic rays and the potentiality sources and the potential for discoveries. So, for all the people that is following the webinar, please, you can ask questions using our YouTube chat system in the YouTube channel. And please, you have to be aware that there is a small delay between the live transmission that we are having using Zoom here between all the people in the webinar and what is showing us in what is showing in YouTube. So, please write a little bit in the head your questions to Johannes to be able to ask them at the end of the webinar. So Johannes, whenever you want, you can start. Excellent. Thank you very much for the kind introduction and to all of you for the opportunity to talk to you. I'm very happy to take part in this project of bridging oceans and mountains with physics. So let's get right to it. I will talk to you about galactic cosmic rays and what we can learn from them for fundamental particle physics and how anti-nuclear fit into that picture. So galactic cosmic rays are energetic charged particles that are accelerated either by astrophysics in the vicinity of very violent events in our galaxy, or potentially by interesting particle physics. And once they're produced, they, their movement in the galaxy is governed by magnetic fields because space is mostly empty. And it's more astrophysics onto these cosmic rays. Now I'm a particle physicist and I'm interested in fundamental particle physics. So the fundamental question for me in doing astrophysical physics is how can we separate the astro from the particle. And in the case of cosmic rays, this goes towards rare cosmic rays. There's one big difference between astrophysics and particle physics. One big difference is the baryon asymmetry of the universe. Astrophysics accelerates lots of cosmic rays from the baryon dominated into the stellar gas. So the cosmic rays that we get from that are lots of protons, some helium, some metals, but it's going to be dominated by baryons. On the other hand, particle physics, if you have enough energy, you can do pair production. Anti-meta, meta pairs in one-to-one correspondence. So that if you look at anti-meta cosmic rays instead of baryon cosmic rays, you get a huge background suppression. So you can hope that you can look only maybe at the particle physics that you're interested in, not at the astrophysical background. So this plot here shows, for example, the flexes of protons and helium, and then of rare cosmic rays of electrons, positrons, also anti-meta, but lighter than anti-proton, so they're easier to produce. And then anti-proton, and you can see that just by looking at anti-proton, instead of protons, you put many orders of magnitude of background suppression. And one additional advantage compared to gamma rays is that charged cosmic rays accumulate in the galaxy. So the magnetic field of the galaxy confines them, and they take about 10 million years to escape. So if we're looking at rare events like dark metal annihilation, then there's also some signal enhancement from that. So in this talk, I will first talk a bit about cosmic rays to get you the basics and the context of what we're doing here. And then we are talking about sources of anti-meta in cosmic rays, of anti-nuclear in cosmic rays, specifically, which is the main motivation in this whole topic about dark matter and a bit about primordial back-hole evaporation. Before, in the end, we can then look into the different anti-nuclear, anti-proton, anti-dutron, and anti-hedium, and see what they will tell us. Anti-proton have been measured to astounding precision recently. There's much to learn from that. Anti-dutrons, there's a dedicated experimental program just for anti-dutrons underway. And then in the end, there's some very exciting hints about anti-hedium from the AMS experiment. So let's get right into it. We start with cosmic rays. The life of cosmic rays starts at birth. So for most of the galactic cosmic rays, that is an acceleration. And the greed upon source of galactic cosmic rays is that an exploding star in a supernova creates a shockwave. The shockwave leaves the star and now it propagates through the interstellar gas, which you can see, for example, here in an optical in the yellow shockwave. And at the shockwave in the interstellar plasma, what happens is what's called first-order Fermi acceleration. So it's acceleration by a plasma shockwave. It's acceleration not like on Earth with electric fields, but there is stochastic acceleration around the shock. And this is the source of the ubiquitous power laws in cosmic rays that we see. And this process is so efficient that 10th of percent of the kinetic energy of the supernova can be dumped into accelerated cosmic rays and B-field turbulence. Evidence for this or proof of this has only recently been determined by looking at this supernova remnant and others and seeing that there is a pi M decay spectrum coming from there, which is evidence for the fast protons that hit interstellar gas there. Once they're produced, they escape into the galaxy. Now this is a picture of a typical disk galaxy like ours. But if you look into that using radio waves, what you see is this. It's a big, a synchroton halo around the galaxy created by electrons that move in the magnetic field that are confined close to the galaxy and by being bent by the magnetic fields emit radio radiation. What we can infer from that is that there is a magnetic field that extends many kiloparsec above the disk of the galaxy and can diffusively confine cosmic rays like electrons there. And we can describe the movement of cosmic rays in this turbulent magnetic field by diffusion with the diffusion coefficient that mostly depends on the rigidity R of the particle. Rigidity is how easy a particle is to bend by a magnetic field. So it's momentum over charge. And this is the, so the interactions with the magnetic fields are the reason that we often use this quantity of rigidity to talk about cosmic ray spectra. There's other processes than just diffusion going on in the galaxy. So for example, the magnetic field is attached to the movement of the magnetized plasma. So if in the galaxy, there is a convective wind that blows the plasma out from the disk, then the cosmic rays will be connected along with the magnetized plasma. Also, there might be acceleration happening there by random waves in the magnetized plasma. So if you look at the upper picture, it's clear that there is also stuff, not just magnetic fields. So this interstellar gas can be in the way nuclear collisions can happen, cosmic rays can lose energy by ionization or Coulomb energy losses. And all of that can be described by this diffusion convection equation for the flux of cosmic rays. The time evolution of the flux density is source term from injection, for example, by the supernova remnants or from dark matter for that matter. Mine is a loss term by hitting interstellar gas and being lost. Then there's the diffusion term, the convection term, and we have a energy redistribution and reacceleration term and some discontinuous losses. Now that's the differential equation and we just need to model our galaxy or the geometry of our galaxy in which this takes place. And what is often used is a two-zone diffusion model, what we have in green here, the diffusive halo. Then it's the zone through which this magnetic field extends and in which the galactic cosmic rays are confined. And in the middle there's a thin galactic disk with all the stars and the supernova remnants and the gas and the sun. Now having defined all of these things we can move on and calculate what should happen. So we look at a source and think what kind of cosmic rays should this source accelerate. If it's a shock wave then it will accelerate all the elements that are abandoned in the abundant in the interstellar gas like protons and helium and carbon. There are some elements which are not abundant in the interstellar gas, so that's what this chart shows that the open circles show the abundance in the solar system or the interstellar gas. You can see that helium and hydrogen are very abundant and carbon, nitrogen, oxygen also, but lithium, beryllium, boron are not. Now the black circles show what we find in cosmic rays. And you can see that there's a large abundance of boron generated somewhere on the way. And we call these cosmic rays here secondary cosmic rays because they are not accelerated by the primary source like proton and helium. But they must be created as a secondary product somewhere else. They're created by carbon and oxygen and so on hitting interstellar gas and fragmenting into smaller nuclei. So then during all of the propagation of the carbon nucleus, this boron to carbon ratio will record how often this carbon nuclei hit interstellar gas particles. And this is a very sensitive probe of galactic cosmic ray propagation. And we can measure that with our instruments by just dividing the carbon flux by the boron flux by the carbon flux. And that's what this plot shows here. And you can see that we have really precise measurements of this. And we can now use this to tune our models of propagation and see whether models of propagation are any good. And different implementations of models of propagation, for example, are the very well known Galprop code. So the Dragon code, both are fully numerical. And then in this study what we, or in my work, what we use for semi analytical to zone diffusion model, which is, has recently also become available as a pre implemented code for you to use. Should you be interested. Of course, the galaxy is not the only thing that stands between a supernova remnant and us there's also. So for cosmic rays to reach detectors on Earth, there's a few more obstacles to overcome. So the next one will be the solar system, the solar wind. So what happens here is that the solar wind carries the magnetic field of the sun outwards. And any cosmic ray that needs that reaches at Earth in the center eventually needs to propagate against this conductive wind. And that shields us from some of the cosmic rays. For example, this can be seen here. This is cosmic rays recorded by Voyager as it leaves the solar system. This blue curve is low energy cosmic rays created by the sun. And they drop as Voyager leaves the termination job. Whereas higher energy cosmic rays that come from the galaxy, they suddenly rise as Voyager leaves the solar system, which is again evidence that the solar wind shields some of the galactic cosmic rays. In the end, if you have this geometry of propagation versus convection and you approximate that as one dimension of just the radial part, you take the stationary approximation, so then you can approximate this by a force field. So that you can effectively describe this as an energy loss so that the kinetic energy of cosmic rays that we observe at Earth is just the interstellar one minus some force field potential. And there is some attenuation that comes with propagating against this force field. So the flux that we observe is also lower than the flux that would be observed in the interstellar medium. Of course, there are some caveats to this very simplified pictures. The real problem is three dimensional, there's time evolution in the sun. This in the end needs to charge sign dependence so the solar modulation is not the same for protons as it is for anti protons. But for the moment, I'm looking at larger energy cosmic rays, this should suffice. Finally, the last barrier is the Earth's magnetic field. So the magnetic field of the earth thankfully shields us from many most of the low energy cosmic rays. For experiments, if we want to look at low energy cosmic rays, of course that's not so good. Also the Earth's magnetic field extends far up into the atmosphere so even if we have some low Earth orbit satellites, we're still influenced by this. What you can see here on the right is a map with the vertical rigidity cutoff below which cosmic rays cannot propagate to the height of 50 kilometers. So if you want to look at cosmic rays of only one G, the energy, you'd better build your experiment at the pole. So that's just this context for why many of the experiments that we're looking at are either on satellites which have an inclined orbit so that they also pass the areas with lower cutoffs or high altitude balloon flights over the poles. So much for the life of cosmic rays, standard astrophysical cosmic rays from birth to detection at Earth. Now we're going to look into sources of anti-nuclear cosmic rays, which is what we're actually interested in. The first one we already talked a bit about are these secondary cosmic rays, which come from some primary cosmic ray species, i.e. hitting an interstellar gas atom day, and producing in a high-energy particle physics process an anti-nuclear and some other things. In more maths that can be written as the flux of primary cosmic rays hitting a density of interstellar gas atoms, and then we have to integrate over the differential cross-section of producing anti-nuclear in here. And we know the average density of gas in the galaxy. We can measure fluxes of primary cosmic rays at least at our position of the Earth and extrapolate from there. And these production cross-sections, for example from proton hitting a proton going into an anti-proton, we can measure that on Earth in our colliders. And then to calculate this we can either use an analytical parameterization of the data, so some fit function to the data, or we can use a Monte Carlo generator, tune that to all of the data that we have, and then also use it here. This will be the main background towards all the interesting particle physics searches that we want to do, right? I'm interested in not just in supernova rems and electric cosmic ray propagation and what's going on in the galaxy out there, you know the main motivation is dark matter. As all of you are probably aware, there's a set of observations from the length scales of the CMB to large-scale structures, to galaxy clusters, to individual galaxies that can all be explained with one new thing, which would be a new particle. New particle that is collisionless and that is very slow and doesn't do a lot. And the simplest particle model for dark matter, the arguably simplest, is WIMP dark matter, a weakly interacting massive particle dark matter that is produced by a free sound in the Earth universe, which means for this discussion that the measured abundance of dark matter today predicts the annihilation cross-section of what's the cross-section for annihilation when two particles, two dark matter particles, hit each other. Now, we know that in the center of the galaxy and all through the galaxy, in fact, there should be, there is lots of dark matter. We can infer that from the rotation curves and the movement of stars in our own galaxy. So we expect, in this model, production of high-energy particles, possibly including anti-nuclear in our galaxy today. If we want to calculate the flux that comes from that, we need to get a bit more concrete. So for example, this was in the dark matter annihilation picture. Of course, dark matter, there could also be decaying particles producing fast particles that also produce antiprotons and whatever fancy things. But in this wind picture, what's most interesting is annihilation into some high-energy final state, like 2B quarks or W bosons. And these particles, of course, we're talking about astrophysics, right? They are not stable. They decay very fast into other things. And if there is a hadronic channel here, then in the end, what we get are lots of protons, pions, gamma rays, antiprotons, possibly anti-neutrons. And the source term for that then looks like this. So we have the density or number density of dark matter squared times this annihilation cross-section, for which we have the theory expectation, times then the spectrum of anti-nuclear that are produced for annihilation into a channel F, like BB bar. The halo profile is reasonably fixed by stellar movement or by simulations. We take very often in studies of charge cosmic rays of anti-protons, BB bar or WW as representative annihilation channels. And then for the analysis in the end, we take the annihilation cross-section or the decay radius free parameters. And the spectrum, of course, we cannot observe at Earth. We have unfortunately never seen dark matter annihilate on Earth, so we don't know what the spectrum is going to look like. Instead, what we do is we use all that we know from collider experiments and tune them on to Carlo, like Pithya to that. And that then tells us the spectrum of anti-nuclear that comes from this. And these look like this. So what is shown here in blue are the spectra, source spectra for BB bar, final states, and in red is for WW, final states. And for our purposes, what is important here is that all of these spectra are peaked at low energies. So in terms of rarity of observing these, it would be most promising to place a detector to build something that is sensitive to low energies. And on the other hand, the astrophysical backgrounds are these steep power laws that drop as energy to the two, to the minus. And so they drop very steeply. So that actually in anti protons, they tend to stand out against the astrophysical background, which goes like this at higher energies. Now, this is the biggest motivation for looking for anti protons in terms of particle physics. And there's one more interesting source that I'm going to talk about now, which are primordial black holes. So that goes like this in the early 70s, Stephen Hawking discovered that there's a possibility of forming tiny or about black holes of all kinds of sizes in the early universe by just statistical large density fluctuations or in phase transitions in the early universe. Their masses could be very small, which, so for example, for a permanent black hole of mass of 10 to 13 gram, that's 1.5 femtometer size. Now, of course, as you were saying, that's pretty tiny right so let's look into quantum effects. This prompted him to look into quantum effects of black holes, leading to the celebrated theoretical result. And black holes should emit particles as if they were hot bodies with a temperature that can reach DB if the masses only small enough 10 to 13 gram is pretty small for a astrophysical masses. So the sun is 10 to the 33 gram. So this is a very, very tiny black hole. And this evaporation and the emitting particles leads to mass loss so this black hole should in the end evaporate and bent. And this result also predicts that if in the early universe black holes were formed with mass smaller than five times 10 to the 14 gram, they should have evaporated by now. So if you have a black hole of mass 10 to the 13 grams, then it's already one GB hot. And it will only last for 10,000 years until it's completely gone. So as I said, the, it's the emission is expected to look as if from a body. So the elementary particle emission rates from one promote black hole or from one black hole looks like this. So just thermal here depending on the spin of the particle that you're emitting, the energy of the particle and the temperature of black hole. So there's some gray body factor. So it's not, not a perfect black body, but there's this additional factor that comes from the fact that the black hole is in size comparable to the part of the wavelength of the particle you're emitting. So then, then in this picture, the black hole should admit if it's very massive, only massless particles. So if it's very massive, the temperature is small. So the energy of the particle you're emitting also needs to be small. So only massless particles like gravitons, photons, maybe neutrinos should be a minute. Then as the back hole evaporates and the mass drops, the temperature grows and as the temperature becomes large enough also electrons and positrons and pions should be emitting. And as the temperature goes above lambda QCD, you expect emission to be described by an element, emission of elementary quarks and nuance, that then later hadronize into stable particles like maybe anti protons. So we want to calculate what the spectrum look like of anti protons from black holes. For that, of course, well, we have to, so for the spectrum of anti nuclei from permanent black holes. We have to integrate here over the emission rate of elementary particles quarks and blue ones that later hadronize into the anti nuclear nuclear that we want to look at. So we have to multiply that by a fragmentation function of some of the initial particle radiated like an up quark, fragmenting how many anti protons do we get from that. Of course, we're not interested just in a single back hole, but we want to look at the spectrum that's emitted from all of the black holes in the galaxy. So we need some information about the black hole mass distribution that we have in the galaxy. So here on the left, if we take this, these two examples of what might be the initial mass distribution after formation of black holes. And they look like this. If we then take into account that these black holes lose mass and the smaller the mass, the faster they lose their mass. This means that the spectrum evolves and it evolves fastest for the lightest black holes. So for black holes that are smaller than this mass of black holes that has evaporated by now, there is a universal spectrum that we obtain just from this vibration law, which is the dndm should rise as m squared. Quite independently of what complicated shapes the spectrum might look like as long as it is kind of smooth around here. So there's a universal spectrum, there's a universal mass distribution mass function for black holes that can emit nuclear or anti nuclear so that have gv temperatures. That means that we can calculate the primary black hole source term without many assumptions on what the mass distribution of black holes actually looks like. So we're taking what we had here, the single black hole spectrum and folding that with this dndm, which is just m squared. And then again, what we get from that is a one parameter spectrum. The only parameter here is the normalization. It is very predictive in the sense that the shape is completely fixed by a primary black hole evaporation plus this universal mass spectrum and the low mass primary black hole range. So why maybe it's not as exciting as dark matter it is a very interesting benchmark to look into and to compare data to. And actually look into the data. So now we're going to talk about experiments and observations and how to explain them. And the very big role here in place the instrument that has observed most and most cosmic ray anti proton so far, which is the AMS or two experiments experiment that you can see here, which is on board of the International Space Station. This is orbiting the earth and 100 kilometers high. And the idea is that cosmic rays come in from here and then hit all of the detector parts inside here. And then we can analyze that this detector is built like that. So it's a quite similar to the particle detectors that we used to in our collider experiments on earth. So it has a tracker system, it has a magnet for deflection. And it's built through this from from the incoming cosmic ray direction. So let's say there's a cosmic ray coming in here it hits the tracker here. Here there's a position radiation detector that is built to separate electrons from nuclear. Then it hits the inner part of the detector where there's a strong magnetic field and it's deflected. So then it leaves a track in the in the tracker, from which we can reconstruct the particles rigidity. It leaves the time of light system, which is used for velocity determination, if the particles not too fast. And then here it hits the ring imaging to rank of detector, which is used for velocity determination if the particles a bit faster. And in the bottom here there's an electron calorimeter which we're not interested in because this is a talk about anti nuclear. And the interesting thing here and why it really changes all of these collected cosmic ray discussions is that they for the first time have percent level precision for galactic cosmic rays over large range in energies. So this for example is all results from 2015, but you can see that you cannot even see the error bars in the anti proton to proton ratio here. It's really a very precise measurement. So it's interesting to ask, is there anything exciting in there? And if you ask the ams people if there's anything exciting, they will come up with something so they always have something exciting. So if you compare the spectrum it's the same as before just your net linear scale instead of the box scale and compare that to the secondary expectation from some old paper. Then you would conclude that, well, there's a huge gap here, what is happening? And of course they also offer the explanation of what might be happening, right? There could be dark matter annihilating, which would be amazing. Now it turns out that if you really take into account also the other measurements that ams did on the proton flux and the helium flux. This secondary production line here is no longer tenable, but still with these data you can make very careful analysis, which these people here have done a few years ago. So what you can see here again is the, now this is the anti proton flux multiplied with some rigidity scaling here so that you can, it's easier to see the shape. And it's very hard because the error bars to see anything in here because the error bars are so small. But what they find when they fit secondary production, plus a possible dark matter component to this, is the spot here on the right where depending on the dark matter mass, you get limits on the dark matter annihilation cross-section, which here this range here is for different propagation models, so it kind of shows the uncertainty in propagation, which are much better than the gamma ray limits from that time. And excitingly there is about a three sigma excess around here, which is very hard to see in this spot. But it's very clear that there is a three sigma excess, at least in this paper here, which lies right at the thermal annihilation cross-section, which we expected all along, right, if we believed in the WIMP. So that's very exciting. Now what's the status today? It's unfortunately not as simple, so because if you look at them, then you see that the signal is just 10-20% of the overall flux. So we are looking really into a precision measurement with cosmic rays, so that it's not that easy. And the picture today looks more like this. I took this slide from a recent workshop last week actually by Michael Cosmeier, so one of the authors of the previous slide, where he summarizes the recent developments on this anti-proton excess, if you like. So there's been a few studies. The main thing is that AMS has not yet published any covariance matrices, any correlations between their errors. So theorists have done, have tried hard to come up with their best guesses for what the correlation of the errors might be. That's one thing. Then you need to be really careful in accounting for the propagation errors. So in the previous study, they just used a few example realizations of the galaxy, for example, of propagation models. One of these guys here, they really scanned over propagation models and seeing, in many cases, a joint fit of the boron to carbon ratio to constrain the propagation model and then the anti-proton to proton ratio. Another big thing is, and that's maybe best seen here in this lower right hand plot, so in black that's the, these are residuals to some best fit. But what you see is that the errors of the measurement are smaller than this green band by a factor of two, something like that. And green is the cross-section uncertainty. The fact that we know too little about anti-proton production to be able to predict the anti-proton flux to the same accuracy as the experiment can measure. And somehow, the Michael Cosme and Phil Anseldonato and other people have, people from this community here have initiated a push to measure more anti-proton production cross-sections to be able to take more out of this measurement. And until then, wow. So at the moment it looks a bit like there's 1.1 sigma, very nice limits, but no excess to other people claiming there's maybe up to 4.7 sigma and a robust excess, which is also consistent with the Galactic Center DEV excess. So there is some excitement still around. And in the end it will get settled by the experiment publishing their own error correlations would be one thing. And then better knowledge about anti-proton production would really help us out. Okay, there is one other exciting option, which is to look at a channel that has less background, which is where anti-dutrons end of the day. So anti-dutrons are one anti-barrier in more. Now, if you look at the secondary production, what happens is mostly one proton hits another proton produces, let's say an anti-dutron and barrier number in here must be conserved. That means that in addition to the anti-dutron, which has barrier number minus two, if we start up with barrier number two, we need to produce four more nucleons. So that the threshold energy in this process is six times the nuclear mass. And the threshold energy for an incoming cosmic ray proton needs to be 16 GB, which is quite a lot already. So three times more than for anti-proton production. That suppresses the background, of course, but there is this other effect than anti-dutrons or a larger anti-nuclear like anti-helium will always be produced boosted because you need so much energy in this fixed target in natural state. There cannot be, there is some energy below which there cannot be anti-dutrons produced. And that gives a huge signal to background boost for low energy anti-nuclear because secondary production of anti-dutrons drops precipitously at low energies. So if you now look for dark matter, which we saw was peaked at low energies, here at these low energies, there's going to be almost no background. If you look in anti-dutrons. Okay, so let's calculate how many anti-dutrons do we have. The problem is that anti-dutrons are quite rare and the experimental data is not sufficient to get a data driven differential anti-dutron multiplicity. For example, proton-proton collisions. So we have to do some phenomenology, some modeling. The simplest model you can do is the factorized coalescence model where you just say, well, okay, I look at my process. Look at the anti-proton multiplicity and then let's see what's the probability, so to say for an anti-proton produce with this momentum and with this momentum. Okay, and then there's this coalescence factor some probability that these two anti-proton and anti-nuclear neutron will merge into an anti-dutron. And this is works and so on so empirically this works well for high multiplicity processes so for heavy ion collisions at the Large Hadron Collider, this can describe the data quite well. But it's quite clear that this will not work for example close to the anti-dutron production threshold, where the anti-proton multiplicity doesn't know anything about the fact that we are really interested in only processes that have two anti-nuclear ions produced. So the next best thing that you can do is to do event by event coalescence which is kind of state of the art of how this is done or used to be done. Which is that we generate lots of Monte Carlo events and look into those that have two anti-nuclear ions. So we run our Monte Carlo and let it do all of the heteron section and stuff and then we look at all of the events that have two anti-nuclear ions. Now if their momenta are aligned enough that so the physical picture that the binding energy of the anti-dutron can overcome the momentum difference. So that their momenta are close enough, they're close enough in momentum space that the binding energy can overcome the difference then they will merge into an anti-dutron. This is obviously physically not a nice model because we know there's quantum mechanics and this sharp cutoff in momentum space has nothing quantum mechanical to it. So it fares on formal grounds already. So what in practice we do is we take this as a logical model as a one parameter model for the differential anti-dutron multiplicity. And we determine this coalescence momentum not from this theoretical hand-baby argument with binding energies but from data. And even then it turns out then we have to allow for different coalescence momenta P0 for different processes. Which kind of makes that so. Proton-Proton collision says lots of, at 7 TV has a lot of patterns, large multiplicity, many things going on, far less clean than electron-electron collisions at 10 TV. So doing this and then using the coalescence momentum that we determined from proton-Proton collisions for the secondary background which is mostly proton-Proton collisions. And from cleaner, higher energy particle physics processes like clean processes like electron-positron annihilation for the dark matter process or here ZDK for example dark matter going into WWW it makes sense it's kind of similar to ZDK. That's how we calculate our spectrum. Yeah, so recently there's been some some nice developments on building a more quantum mechanical and maybe a physically more correct picture to this but I don't have time to go into it. And in the end, this is what we get. The future prospects given the anti-proton levels. So this top row of clouds shows anti-proton spectrum for different secondary types. Then here black is always secondaries. So the astrophysical background. And these other lines are chosen large enough so that they're just compatible with the data plus secondaries. So you can see that all of the interesting contributing the things that we interested in dark matter and permanent vehicles can only make up about 10% of the flux. So in anti-proton we really need to do a precision study if we want to learn anything about anti-neutrons, sorry, anything about dark matter or permanent vehicles. Whereas if you look at anti-neutrons you can see that the secondary flux drops a lot towards lower energies. Whereas the dark matter and permanent vehicle flux is not affected by that. So if we build a detector that is sensitive to these kinds of fluxes, that would be and then found anti-neutrons. That would be a very strong indication that there's something interesting going on here. Okay, so let's do that. There's actually an experiment planned and set to fly in the next year over Antarctica, which is the gaps experiment. It's an Antarctic balloon mission and is specifically purpose built for low energy anti nuclei. So it's built to have very good separation between anti-neutrons and anti-propons. And that works like this. So let's have an anti-neutron coming in. Speed is measured in the time of flight detector here and then there is a very thick target detector system in which the anti-neutron has actually stopped and then forms an exotic atom, so the negative anti-neutron with the silicon nucleus. As this exotic atom de-excites, it emits characteristic x-rays that are different between anti-propons and anti-neutrons. And then in the end, of course, we have the annihilation signal of the anti-neutron actually eating up part of the silicon nucleus, from which we will get lots of finance. Many more than from an anti-proton annihilator. So this is very exciting and it might find something and we will know into years. Until then, we have some other exciting things to talk about if we want, although they are a bit more speculative, which brings me to anti-heal. There have been claims, but no published articles about anti-heal events being observed at the AMS or two experiments. So this is an event display of an anti-healium candidate event where they measure that it's a particle with charge minus 2 and mass 3.8 plus minus something. So actually the mass is even more consistent with anti-healium 4 than anti-healium 3, which is very exciting and hard to believe. And the reason for that is like this. If we want to calculate, make predictions for anti-healium, we are in even more trouble than for anti-heal neutrons because there is even less data for anti-healium than for anti-neutrons. But the general trend is that this shows something like a normalized multiplicity for anti-nuclear neurons of some mass number A. So this would be anti protons and then in the same type of collision observed at Alice, this would be anti-neutrons. This would be the anti-healium 3 multiplicity and this is the anti-healium 4 multiplicity. And there is a factor of a few hundred for each anti-nuclear that you had. Now we have measured anti protons. So we know how many anti protons are out there. And for all of the components for a secondary for dark matter for permanent black holes, all the particle things. If you want to produce some anti-healium, you're going to produce lots more anti protons in the process and those we can compare to what we've seen. Yeah, so we can do the same exercise that we did for anti neutrons. So taking the observed anti proton spectrum, looking how much anti neutrons can there be, and we can also look at how much anti-healium can there be. And we did that for anti-healium 3 and this is the result. So here is different secondary production of anti-healium. These are the maximum possible fluxes from dark matter without overshooting anti protons. If dark matter annihilates into W bosons or B quarks, or here for permanent black holes. And what you see is that the projected AMSO2 18 year limit for the detection of anti-healium is above all of these. Maybe if so, this uncertainty is related to the uncertainty and coalescence of anti-healium, which is huge because we have so little data. It's really hard to make it work out for anti-healium 3. Most of the time the observed anti-healium flux is an order of magnitude. At least above the secondary expectation. And even with dark matter, it's hard to make it work. Some other people who were made some much, much more conservative assumptions about anti protons, they find that the anti-healium flux, if these events by AMS are actually anti-healium, then they would be marginally maybe compatible with a dark matter origin. But this event that we looked at two slides ago looked like anti-healium 4 and 2 sigma. So how is that going to work? If particle production with particle physics processes with coalescence drop by a factor of few hundred for every anti-nuclear that you have. So that's what these people here looked into. So here this is the secondary prediction of anti-healium 3 and anti-healium 4 compared to the AMS 5 year sensitivity. So you can see that maybe if you're really, really good, if there's something that we really misunderstood about anti-healium coalescence or a galactic cosmic rays, then maybe you can make this intersect here. Maybe. But there's no way that there's a secondary anti-healium flux that can be detected at these points. And that holds true for all of the particle physics processes where you produce the anti-nucleus by coalescence. Of course, if you ask, well, if you put it to a dark matter phenomenologist, it's impossible that dark matter can do this. Then a good phenomenologist will come up with a model that does it, even if it looks very strange. Now if we took these anti-healium 4 events for granted, this would mean either some extremely involved complex, really on dark matter model, or it would mean that there is anti-healium somewhere in our galaxy, that there are anti-matter regions in our galaxy, either as clouds or as compact anti-matter objects. That would be a very extreme conclusion and the kind of evidence that AMS has until now is in no way sufficient to support them. But it's not unheard of in theory. So there's been, for example, this paper of Baryogenesis, which predicts a small fraction of the mass density of the universe to be in the form of compact anti-baryonic regions. So though extreme, it's not impossible. Yeah, this brings me to my conclusions and summary. We talked about anti-nuclei in cosmic rays. Why they are very promising detection channel for exotic particle physics in the galaxy, first and foremost dark matter annihilation. We looked into the state of the arc of what's going on in antiprotons. Antiprotons place competitive limits on web annihilation, but the secondary background limits the discovery potential. So it's really hard to convince the seasoned particle physicists that what you find there is actually particle physics and not just an astro. But for inter-neutrons, there's the promise that the secondary background would be orders of magnitude lower, so that even with nothing visible in antiprotons, low energy inter-neutrons may have a signal from dark matter annihilation orders of magnitude above the secondary background. And in the end, there was a fun excursion about anti-heating events that MSO do and what they would if true imply amount the universe we live in. Thank you very much for your attention and I look forward to questions. Thank you very much Johannes. It was very interesting webinar, your talk. In fact, it was very interesting listening because it's been my area. So let's start with the for the people that isn't in YouTube. Please start to write a question because you know there is a small delay between what we are living here inside the Zoom session and what you see in YouTube. So for the moment, we're going to start with the question from the same people that are here in the Zoom session. I don't know if some of the other coordinators want to start with the first question. No, anyway, I have a question. Nicholas. So thanks first for the super nice talk. If you go to, I think it was just like number 14 or 15, I'm not sure. You showed a spectrum for the anti-deuterance. No, I think it was previous one or after. Where? Dark matter. That one. Yes, exactly from dark matter. So for instance in the 10GV in the right hand side, so you have like this peak at exactly that. So what's that? Yeah, so in the B quark, there's always, in the B quark channel, what you can form is there is a, you can form a B meson. So all of the low energy tail here is just from, for example, the strength fragmentation of your quark gel. But for the specific case of the B quarks, there's always some additional stuff from the leading quark, so to say. And that's what gives you this peak and this peak here and also this peak here. That's D. Yeah. What are D? So, okay. I see. Yeah, so we're looking into light quarks back to them that should be absent. Okay. Thank you. So another question. Johannes, I have a question about the, about the coalescent factor. There are other models than the one that you presented because since it's quite, it's rather seem to trying to, because there are no data as you said. Yeah. So the biggest problem is that if you don't have lots of data, then the whatever model you make must not be very complex, because you cannot determine so many parameters from the data. Nevertheless, it should capture the relevant physics so that you trust it where you don't have data, which is the whole point of the model point. And for this factorized coalescence model, the simplest thing you can do. It's been investigated where it fails and that's the reason why people today mostly use event by event coalescence. So as you say this classical description is not very trustworthy. Recently, there's been some developments. So this one here can be made totally fine. There's been some, some developments on starting with heavy iron, iron collisions, like the room and others and determining the coalescence factor there and then using some arguments to extend the model from here from the case where it works to also the other ones. And then the most exciting one is this are these recent developments where they actually do a quantum mechanical description of coalescence. So what they do is they, they use Monte Carlo and to determine the distribution of protons and neutrons and protons and neutrons. And they calculate the two particle wave functions from there. And then they project them onto an anti neutron bounce day, which is what you would do right so that sounds like a sensible quantum mechanical description of what should be going on. And then they came up with a nice week, a nice week now function based approach in which you can use Monte Carlo to take care of the of the correlations between anti protons and anti neutrons that you have and characterize them and then analytically do the quantum mechanical part of projecting that onto the pond state. And that looks very promising, though so far it hasn't been as far as been tested on the energy data. So, to me it's not clear yet whether it also works close to threshold, but I hope it would because in. So it looks like a much more sensible model than what came before. Yeah, kind of ahead of the state of the art at the moment. So, but in that sense, another, I mean one question related with this. Because since we know so few about these coalescence factor coalescent momentum could be it could be also a possibility that there are kind of resonance in this coalescent factor that make more visible to in some coalescent processes, you coalescence more than in other point that what we can see here in the same table that you're presenting that for low energy, you have much smaller coalescent momentum, and for high energy using just PTA tune coalescent momentum because also it's very big the spread between Monte Carlo. That's, that's, that's possible so what that effectively would correspond to is that now at the moment one so for example in this event by event simple event by men coalescence model. There's this sharp cut off. So there's 100% probability to merge at very low momentum separation and then drops to zero abruptly. What you would suggest is that there might be something where maybe at low relative momentum there's a large probability to merge and maybe then there's a resonance later. Maybe. Yeah, it happens with the with the P mess on production like when there is the resonance you produce more and so on. Oh, anti protons because this be messing is everywhere especially at low energy. So the coalescence modeling is definitely is the weak point of anything involving anti neutrons. So anti neutrons you cannot really use them to put constraints on dark matter, because we don't really know how much we expect of anti neutrons. The other way around it works that we say if we find anti neutrons, then they cannot come from secondaries, or if we find anti neutrons at some, and at this level for example, then we can be quite sure because there's orders of magnitude in between here. Yeah, that it's not so that way around it works. It's more of a detection channel not so much constraints channel. So it's more curiousity. How large is that our 70. I mean the impact of the coalescent momentum in the flags. Like, where. Yeah, so if you think about three of the Okay, it goes as a factor. So so to say about like like the volume in momentum space of the ball within which all the momentum coalesce. And here there's about depending on the process maybe you can take there's a factor of two or even more of uncertainties. So that gives you a factor of 10. Yeah. So, yeah, yeah. That is very important for the, yeah, to establish some constraints. Yeah. I mean, I have more questions but I'm going to let the other people to ask. I don't know if the other guys here in the, and also in YouTube, you'll still have time to make some questions. I don't know. Okay, so, because I'm very interested in this coalescence function because it's the key part of the, of the production of anti-antideuter. But for instance, have you, I mean, have you tried or other people have since you read more about this. I mean, you're working on that to try to use the same coalescent factor or coalescent model for the function of the ethereal. I mean, since it should be more abundant in, I mean, it's more messy the signal, but or is it expected that it's symmetrical. You can form the ethereal in the same way that you're from, you're from, I mean, I started from a proton and proton neutron, proton, antiproton neutron. Yeah. Yeah. Um, for high energy collisions, like at Alice, the result, the experimental result is that it's pretty much equal. So the entire neutron to neutron ratio if you have 70 the proton proton collisions is pretty much one. So for that you then also infer the same coalescence parameter. And the problem is done. So for using this with low energy processes where it's not clear that the, so for proton-proton collisions at smaller energies where it's not clear then should be symmetric. It's unclear how much we can learn from that because we anyway have this really big process dependence here. So what it's done at the moment is to calculate the secondary production, you use the measurement of the P zero that you infer from proton-proton collisions. And for a production from dark matter you use the one that you infer from Z decay or from, from leptonic processes. Yeah, because they're more cleaner. Yeah, not production is not clear on that. But of course, I mean, I'm pretty sure that the people here for the plus and minus where there's no variant number involved in the initial state. I'm sure that for the determination of the P zero, they took into account neutrons and neutrons. Okay. One more question. But not all coalescence because it's a bit too much question. I mean, I don't know if you didn't present, or maybe you didn't, you're exploring also the case of decay that matter because in that sense you don't have the restriction because of the annihilator cross-section of the wave particle. Yeah. It's able to observe these kind of processes in, in AMS for instance. So, the only thing that really changes between annihilating dark matter and decaying dark matter is the spatial distribution of sources. So annihilating dark matter is proportional to the dark matter density square. So it's more peak towards the center of the galaxy. Decaying dark matter has a gentler slope because it's only in proportion to the number density. But the spectrum is going to look the same. So in analysis, we treat the annihilation cross-section and the decay rate as three parameters anyway. The normalization of this is basically limited or fixed by anti protons. And the results for what you could get in anti-duperons are basically the same. And also, that also implies that all of these fancy, these limits here on dark matter annihilation can be pretty easily recast into limits on dark matter decay. So in terms of this diffusive process in galactic cosmic ray propagation, the different density, the different source profiles for annihilation and decay don't matter much. They can not influence the spectrum a lot. It can practically directly translate this into a limit on dark matter decay. Okay. So I'm going to take YouTube for question for Johannes. It seems there are no questions here in YouTube. So for the people that is following us following this transmission, please, you still have time. I don't know people here from the Zoom session. You have questions, you can unmute yourself and ask the question to Johannes. Yes, I did. I did sorry. So I'm going to go for it. I was, I was hesitating because I think a previous speaker mentioned the excess that you were mentioning before and discussed the dark matter explanation for this. I think I asked the same question at that time, but I'm going to do it again because I remember the answer. So the question was, if dark matter direct detection, the bounds have an effect in the possible explanation of dark matter as a source of the excess, in case the excess is there. Of course, if it's a decaying dark matter, of course, there's no wish, but in the case of it's a coenilating, I was wondering if there's anything coming from the direct search bounds. Yeah, so the, as soon as you specify model, yes. So if you have a model of dark matter, say it's a, I don't know, some real scale is saying the dark matter by the Higgs portal, then you can calculate correlations. But if you don't have a model, then it's possible that the particle physics, where it depends on the, on your more detailed model. So if there is, you can have dark matter direct detection suppressed by some velocity dependent operators by different couplings to the quarks. Whereas the annihilation is always kind of fixed by by production line, as long as you have wind dark matter, there is a quite clear prediction except for the final state of the annihilation of what the annihilation rate should be. Now for the dark matter direct detection, there is no such clear prediction depending on the model. If it's a simple contact interaction, which again would be a model. Then there would be a prediction, and then typically direct detection bonds are very strong. Right, so if we have like an ultralino dark matter or something like this. I mean, would that be able to avoid balance and explain it simultaneously? I don't know. I really don't know much about neutral dark matter. I just came in. I think you can suppress. So now we're talking about dark matter direct detection, right? How can you suppress dark matter direct detection to comply with bounds? Because the annihilation cross section is going to be close to the thermal one. Okay, unless you do co annihilation. I would really out of my field here, but yes, maybe you have dark matter annihilation on the Higgs resonance, then the section to be and today would be the same as in the universe. But the dark matter direct detection could be suppressed because you have tiny cuttings. Sure. I'm sure there can be examples. Okay. Great. Thank you very much. So yeah, I was me. Sorry. So I guess for, I don't know for the people here and also in YouTube, there are no more questions. So I guess it's a good time to finish the webinar. So first of all, I want to thank you, Johannes, because of your talk that was very interesting. Indeed, I like it a lot. Thank you very much. Take it for the, especially for all the reference. There are many reference that I was looking in the past, but I couldn't find when they were in the first part in the first part of the introduction. And for all the work that you are done also for your that you're doing your PhD, so on. So thank you Johannes, thank you all the people that participated of the webinar and all the viewers in YouTube. And in two weeks, we're going to have Carlos, Oscar, sorry, from the University of Valencia, that he will give also a webinar. So for all the rest, have a good weekend and so on. See you on the next time. Thank you very much.