 Okay, let's start. Hi, everyone. I'm Nicolás Bernal from the University of Antioenariño in Bogotá, Colombia. And today we are super happy to have Josef Prada from Vienna, from the Austrian Academy of Science, the Osterreiche Akademie der Wissenschaftler, or something like this. He will tell us about direct detection and how to break this no-go theory in direct detection. So, please, Josef. First, let me remind you that you guys can ask questions via the Q&A system of YouTube to Josef, so at the end of the talk we'll take all the questions. So, please, Josef. Hi, Nicolás. Hi, everyone. Thanks very much for tuning in. So, I have to go to full screen now, right? Right. One second. Ah, I have to share my screen. Sorry. Okay, thanks very much, Nicolás, for inviting me to give this talk here. So, the title of the talk is How to Break a No-Go Theorem in Dark Metadirect Detection. And it's work which I've done in collaboration with Chris Covarris from CP3 Origins in Odense, Denmark. And the archive address you can find right here, and we have submitted it to PRL. So, I presume this is a very educated audience, but nevertheless let me sort of start out with a few basic facts, which you probably know all very well. So, how does direct detection work? So, what we are trying to do is we have a dark matter particle from the galaxy coming into our detector and it scatters on the nucleus here with mass mn. The dark matter goes out with a certain scattering angle and then includes a recurrence. And so, what we are after in direct detection is the so-called recurrence spectrum. So, that's the rate of recurrence as a function of energy, of recoil energy. And this is sort of the simple master formula right here. So, mt is the number of target nuclei per kilogram of detector material. Rho over m dark matter is the local number density of dark matter. So, that sets the overall rate. And then we have, we fold the recoil cross section, which is given to us by particle physics with the galactic velocity distribution of the wind boosted into the frame of the detector. And this factor V here is just to make the number density times V becomes the incoming flux of dark matter. And in this business, we measure events in counts per day and kilogram in KAD. So, this gives you a sense that this is a rare event search because we do this per day and now currently the backgrounds are at the 10 to minus 3 level. So, these are really super rare event searches. So, experimentally, what do we do once the nucleus is kicked? We try to pick it up in various channels and here I show you three representative or three pictures or cartoons of how this is done experimentally. Either we might directly observe the heat upon the recoil so the nucleus is kicked and we can operate at the moment which is superconducting right at the transition and so that a small increase in temperature leads a large increase in resistance and this is something which can be read out. There are also detectors which yield more striking signatures such as bubble nucleation here in the liquid. The other possibility is to use ventilation so you can think of this as sort of a wavelength shifter. You are kicking the nucleus and usually you are deposited in KV energy and what comes out is energy in the optical regime in form of scintillation light. Here you see this blue light or we can pick up the ionization that is the nucleus is kicked and because of the kicked some electrons fly off which can then be drifted in an electric field. This leads to an amplification of the signal and then there is a delayed ionization signal that can be read out and very often these detectors use two technologies in order to discriminate electron versus a nuclear recourse. This field started out with high hopes more than two decades ago when people realized that if we think of WIMP dark matter really weakly interacting massive particles as being the dark matter massive with masses in the ballpark of the electric scales of 100 GVs and if we interpret weakly as being c-mediated then the cross-sections for scattering on the nucleon of the dark matter particle on the nucleon on the order of 10-39 cm2 or 10-3 pico barn and that is this error up here. Now this is the typical plot where we are working in this talk so where you see the dark matter mass on the x-axis and the WIMP scattering cross-section the elastic scattering cross-section on the nucleons of the C-minus here and so C-mediated cross-section is somewhere up here and these lines here show upper limits on what is the allowed cross-section there has been a lot of commotion about the various anomalies but pretty much the status today is that we have only null results and no positive detection with the exception of Dharma. But let's not go into this. So today we are pretty much probing Higgs-mediated interactions which are sort of at the 10-45 level. Now this is not the most updated plot. This one neither but this is a summary plot from the SNOMAS exercise which basically shows you a summary of two decades of experimental effort of detecting dark matter and the detection and all these lines here are either upper limits and the dashed lines are projections. Some of them actually have already come into being like Lux and Condex have now the strongest limits. Now if you want to make progress in this plot, we can do this in two directions. Either we probe small and small across sections in the y-direction here and that basically requires bigger experiments because we need more exposure, more kilogram days. If you want to make progress in the x-direction here to probe lower dark matter masses, then what we really need are lower detector thresholds and I will elaborate on this a little bit more. Okay, so let's look at the simple kinematics of direct detection. This is a very simple picture and you know this very well. Dark matter comes into your detector. It strikes a nucleus and now you ask what is the recoil energy of the nucleus? What is the kinetic energy with which the outgoing nucleus flies off? That depends on the momentum transfer Q here. The recoil energy is nothing else than Q squared over 2 mm. If we have a back-to-back scattering in the center of mass frame, then we have a maximum recoil and we can now turn this little equation around and say suppose our experiment has a minimal and acceptance of it. We require a certain recoil in order to detect the signal. Then this certain recoil demands a minimum relative velocity and so that's just given here. Let's say for example our detector requires a 500EV recoil in order to detect the scattering then for a 1GV particle what you need is if you want to schedule an oxygen then you need to have a relative velocity of 600 km per second and if you want to deposit 500EV on a maximum nucleus then already for a 1GV particle you need 1700 km per second. And so that immediately tells you that we might be out of luck there because we believe that the velocity distribution first of all the velocity distribution of dark matter in the galactic halo is to good approximation a Maxwellian. So this here is a Maxwellian shape. But most importantly it's not a Maxwellian with a long wing tail but it's a Maxwellian with a cutoff and the cutoff is just given by the fact that the galaxy has a finite gravitational potential and so there's a finite escape velocity and so it cannot host particles above a certain speed and this escape velocity is somewhere in the range of 650 km per second. And so if you go back to this equation now we realize that say dark matter has really mass of 1GV then there is no particle that is bound to the galaxy that could induce say a 500EV nuclear recall on a xenon atom. And so in a sense that's a kinematical no-go theorem we will not be able to detect dark matter particles with a mass smaller than 1GV in wind nucleus scattering if the detector doesn't have a lower threshold than 500EV. And so that's the no-go theorem I was alluding to in my title. Now you can ask okay well given the situation how can we make progress in this plot of dark matter mass versus wind-nuclear cross-section because there's a lot of parameter space vastly open and these are perfectly viable dark matter candidates. Dark matter doesn't have to be under GV it can also be below the GV scheme that's a perfectly fine possibility. So of course people discuss experimental alternatives to this for example you may not consider the scattering on the nucleus but instead you may consider dark matter scattering on the electron and that's a fair game to play and people use this scattering to put constraints for dark matter below their GV scale. Another possibility is to actually directly produce dark matter in the lab this goes under the framework of the so-called intensity frontier so for example you can shoot electron or proton beams on fixed targets and produce dark matter directly there. Or there are also a whole number of ideas out there you know superconductors using color centers in crystals and what not. So there is a plentitude of ideas how we can actually gain access to sub GV dark matter and these methods it will the time will tell what a prospective proposal said. But for my talk I would like to propose a different mechanism and that is simply leaving the elastic scattering and going instead to the inelastic channel. And so what we have in mind is now dark matter scatters again on the nucleus and now the nucleus emits a continuum photon. So basically it's prem straddle at this point and we can just work through the exercise we can just derive the cross section for this and see if there is any prospect that this gives us something better. All right so let's do this. It's very well known that the soft photon emission here can be described by a so-called by a factorized matrix element where the matrix element for the whole process is just the elastic scattering matrix element times basically a gauge invariant factor epsilon is the polarization vector and dotted into the to the momentum k is the momentum of the photon here. It's important to note that this sort of factorization holds for any nucleus and that's because basically this is a semi-classical process. The nucleus moves on a classical trajectory and the emissionist quantum. So this is a really universal statement here. And so given that the matrix element factorizes also the cross section does. And so we immediately obtain the cross section for this inelastic process here. So it's very simple at this point. All right so let's compute the cross section and here it is. And now the important part is so again the cross section here now you can see it factorizes this is just the elastic piece and there's sort of a cutoff of the maximum photon energy omega and times this factor here. Now importantly if you look at the maximum energy of the photon we realize that it is basically the kinetic energy of the wind. And so if we compare the maximum photon energy with the maximum recall energy we find that for light dark matter the maximum photon energy wins. And that's the first key why this method actually gives an improvement. So we have a possibility to reach higher energies. The second key is that it's really hard to detect a soft nuclear recall experimentally. However if we have a soft photon then it's basically never missed. And so these are the two key facts which help us to exploit this inelastic channel. Of course we should make sure that this sort of factorization which we use here in deriving this cross section holds and indeed this does so that the change of momentum transfer delta q with the additional photon except for the photon endpoint energy is always much smaller than momentum transfer itself and so we can really use this simple picture here. Alright of course if you do that we pay a price going to the inelastic channel costs. First of all it costs a factor of alpha so find structure constant but actually this factor of alpha is even overcompensated for having nuclei by a coherence factor charge squared and so this is not the big punishment. The big punishment is a factor recall energy over nuclear mass. So basically here there's a q squared which translates into recall energy over mn. And so if you look at xenon we find this suppression factor is actually enormous. Eight or seven orders of magnitude. So now the question is well can we overcome this suppression? And the answer is yes because the recall spectrum is actually an exponentially rising function with smaller recall energies and so we can tap this exponential to overcome the suppression. Maybe I just show you a plot that's the easiest thing to understand. So here I show you a recall spectrum now of either nuclear recoil so the usual thing one is after in direct detection or now a photon energy and the dotted line here that shows the usual elastic the usual elastic scattering spectrum and it's cut off here for these values for a 1gb particle. It's cut off at basically 100 eb recall energy. It's cut off because there's escape velocity. And now you see the benefit of going to the inelastic channel because even though there's a big suppression in the rates energetically the brainstorming sort of seeps out of the elastic spectrum and so this part of the spectrum you can now use to constrain these light particles because the energies go up to a here 3kV for example. So that's very nice. But we should be a little bit worried because maybe we have been a little bit too naive about the way of how this photon emission really works. Namely we have treated the nucleus as being a bare nucleus fully ionized and we have derived a cross-section for Bremstrahl that actually scales as 1 of omega. So let me just go back. You can see here there's a 1 of omega factor. This is a well-known infrared divergence for soft photon emission. And so if we go to softer and softer photon energies the whole picture becomes modified by the fact that the nucleus is not bare but it's actually sitting in a balance state with electrons. And so we should really work in this picture where the WIMP scatters on the nucleus inside this electron cloud and what happens is this is how to understand the process. The WIMP scatters on the nucleus. It displaces the nucleus from the electron cloud and it creates a dipole here a dipole between plus and minus and from that dipole you can emit a photon. In order for this picture to hold there are some certain conditions which have to be fulfilled. For example the interaction time of the WIMP with the nucleus has to be rather small so that this displacement really is not an adiabatic move of the whole atom but we really instantaneously shift the nucleus and then we can actually emit the photon from this polarized atom. But it turns out that this picture holds rather well for our purposes. All right so this is a quantum mechanical calculation which we then did so we compute now this process of nucleus scattering in this balance state of electrons and I here is the initial electron configuration F is the final electron configuration and quantum mechanically all intermediate states N are available and doing this in the end of the day what we find is a cross-section where we now get an omega cube factor in front which has to be there because omega cube is basically the energy scaling of dipole emission so the dipole of this polarized atom so that's very nice. This alpha of omega is not the fine structure constant but this alpha is something called what is called the polarizability of the atom. So it's a certain part of this scattering tensor of an atom and which is there if we consider that the final state equals the initial state so there will be further contributions to the process if the final state is not the initial state but if we want to put conservative bounds we can just take i plus f and in that case this becomes here in the cross-section of polarizability and then we get these usual factors here. Most importantly and I think this was the nicest part in this whole project is when we did this quantum mechanical calculation if you let the energy, the photon energy go to its end point we actually recover our naive calculation. So the polarizability, the properties of it are such that we just get this c squared alpha factor back it has alpha has a one over omega squared dependence and so in the end of the day we couple precisely into the naive result so that's very nice and that gives us sort of a confidence that we're doing things right here. Okay so let me show you this picture again. Now the dotted line was the naive calculation of BAMS trial and now in addition here we have this blue solid line and that's given through the polarizability and most importantly this polarizability is something which is tabulated so we can actually go into the atomic data tables and just read out the values and we don't have to do any more calculation but we can make connection to experimentally inferred values and so now we are sort of really slashed down all the theoretical uncertainties of this process. Okay so now we have done the important part of calculating that process in detail now we need to understand how we can put it to use in the reflection experiments. The first thing to note is that if we emit a photon from the nucleus it's a so-called electron recoil event and it's not going to show up in the signal band where one usually expects dark matter scattering to happen in the so-called nuclear recoil band. Instead it happens in the electron recoil band where we have a lot of conch and gammas and beta radiative activity and all these sort of things show up. And so this band is pretty crowded coming because of radiative activity and it's crowded here. This is the example of a crested vector so it's a solid state vector, it's a crystal and also the crystals have typically KAB thresholds and so this makes a solid state detector currently less suited for this type of search where we look for this soft photon. Instead the way to go is to use liquid simulators and the reason why it's because even a very soft photon can induce a signal the ionization energy of xenon is only 12 eV and so that's basically all unique 12 eV and you can already get a signal. And so if you say inject 100 eV photon it already produces multiple electrons and so in principle this is easily picked up, the electrons are ionized and then they are drifted in the detector and it gives an already strong signal. And so it turns out one of the best data are actually also one of the oldest data on this. So this here is the Xenotene experiment and in 2011 they put out ionization only analysis so they throw away the scintillation signal which is usually much weaker and they only look at ionization and here you see their observed spectrum so this is here the number of one electron events here the number of events with two electrons, four electrons, five and so on and so forth. And so we can use this plot to put a super conservative limit on our process of self-parametric emission because if you just take the whole plot here and we just say we don't care how many electrons are produced in the end which is in everything together we can get a very robust limit by just demanding that the photon emission rate is smaller than the ionization rate observed in this plot and doing that we get this limit here. So now these gray areas right here those are the existing dark metal limits coming from the elastic scattering and usually all limits end above a GV so now here this is MAV energy they end above GV, only crest 2 they are the only experiment which put a very aggressive limit down to 500 MAV on dark metal nuclear scattering but below that there is no existing limit and so now we were able to actually put the first limit from that detection in this mass range here. Actually we can do a little bit better I outlined to you sort of the conservative procedure of taking this whole plot in one pin we can also do a differential limit and that's something the referees for example requested for our paper now that's not online yet on the archive but the revised version will have a limit which makes use of the differential information of how many electrons are produced and so in our revised version we consider in more detail the process the microscopic process how many electrons are produced upon emission of the photon and details are not so important here what is important here is that the limit can actually be improved and now we go from the 10 to minus 30 level you improve it by another order of magnitude and so this here is now the Xenon 10 limit all right we should still be a little bit worried because of these cross-sections right here are not too small 10 to minus 30 versus 10 to minus 36 or smaller and so it turns out that dark metal may actually start to scatter before it reaches the detector on the ground and so if you compute the mean free path we find that for 10 to minus 30 cross-section we actually get into the order of the overburden of the detector and so if we really want to put a precise limit we should start doing a Monte Carlo analysis of how the incoming flux gets modified and how the incoming energies get modified through this elastic scattering we did this a little bit let me skip over the details here what you can see is if you take a light particle and you let it scatter in the earth you see how its velocity gradually degrades so the spectrum becomes softer but it doesn't do it very quickly and that's because the energy loss is pretty much suppressed by dark matter mass over nuclear mass and so the lighter dark matter the longer it takes for actually the particle will ping around but it will not lose energy very quickly and so it takes really a whole number of scatterings before it is too soft to induce a real signal here you see how the flux goes down in any case for this paper here for this first paper we have refrained from separating this Monte Carlo analysis into the limits and we merely say that above 10 to the minus 30 cm² the limits eventually will disappear and how quickly they disappear this depends now on the details of this degradation and I want you to do this Monte Carlo I just highlighted before but nevertheless there is a fraction of parameter space which survives okay there is more data Xenon 100 this year put out finally their S2 only analysis it was a little bit disappointing in the sense that they sort of had a rather high few electron rate in their detector which this is background which is sort of somewhat poorly understood but we can still put it to use and we get a somewhat better limit but not much better in any case what we have derived is now the first limit below 500 mV now if you want to go forward larger detectors might not be necessarily better detectors so here is for example Xenon 100 and the reason is the background rate really but a low background rate is only achieved in the fiducial volume so that means only in a sort of in a core region of the detector which is far away from photomultiplies and all that now you can ask how do you actually fiducialize and the way to fiducialize is by observing the scintillation signal so you see two photons flying off and that gives you the ability to construct the position of the event and that allows you to say if it's in the fiducial or outside the fiducial volume so these small rates require actually the scintillation signal and so it very much depends how good your detector is in picking up the weaker scintillation signal in order to make use of the fiducialized rates let me skip over the details here and just show you our projections for sort of an experiment a liquid xenon experiment where we require scintillation and where the detection efficiency of these photons of each of the photons vary between 40% and 100% 40% is what the best detectors do right now 100% is very futuristic and sort of this gives you I would say the most optimistic reach of this technology when we want to exploit the inelastic channel on the other hand actually turns out that Crest so the solid state detectors might do better so Crest they produce now their own crystals so they are homegrown and they promise us very low radioactive rates and their crystals and so using their own numbers from the paper for projections we actually get this line there so this is something interesting to do for them I should note that even if Crest lowers the threshold here they might get better in their limit with the elastic scattering but the elastic the recoil energy and the elastic signal it decouples with M chi squared whereas the kinetic energy of the wind goes linearly with M chi so there's always a region that Bramstrahlung will win over the elastic channel so even though this picture might look different in the future there will still be a region where the Bramstrahlung will be the best probe for low dark metal masses okay I think I should talk half an hour so let me I'm a little bit over time already so I'll be a little bit quicker over the rest one important thing is we should actually compare our proposal to dark matter electron scattering because constraints have already been put on sub-GV dark matter from dark matter electron scattering and if you look at the numbers here what are the cross sections which are constrained we find that basically we can only get an improvement over the dark matter electron scattering limits if dark matter electron scattering is suppressed at least by six orders of magnitude relative to dark matter nuclear if dark matter nuclear scattering wins over dark matter electron scattering by at least six orders of magnitude and so naturally we are drawn into so-called leptophobic dark matter models where dark matter only scatters on the nucleus but doesn't scatter on the electron and so let me just show you the most vanilla simple model of a leptophobic dark matter model which you can build you can take U1B gauged barrier number V would be the vector of this new U1 group and you can have a fermionic or scalar dark matter particle which then has gauging directions of this new U1 so dark matter interacts with the barrier and current essentially so Q bar Q is the Q bar gamma mu Q is the barrier and current and yeah so this model these sort of leptophobic models have been entertained a lot in the literature and so we can just put them to use for our purposes however what is important even if the kinetic mixing of the kinetic mixing with the ordinary photon is zero at tree level it might be induced radiatively and so radiatively we can actually generate dark matter electron scattering even if we had it turned off on the tree level and so just think of this process here we have the ordinary photon and here we have our gauge barrier number quantum run in the loop and they induce radiative radiative copper here and so if you look at sort of the parametric estimates we start out with scattering on the on the nucleon we have this value here and then there will be scattering on electrons it has an additional factor of copper squared alpha times alpha b versus alpha b squared so alpha b squared is sort of the barrier in coupling with a copper that is radiatively induced so it's on this order and if you compare this cross-section now we actually find that indeed we can achieve this hierarchy between between electron and maybe just show you a plot instead here in orange this line here is the Bremstrahlung emission this is now the theory rate so this is the rate in Xenon it's a function of gamma energy it's given here and in blue smaller is the rate of electrons scattering with a radiatively induced process and most importantly the gamma energy is sort of higher so it has a higher endpoint energy and so when you start to fold in the electron multiplicity is a settler so if you work out the detector signal this year will be a much stronger signal than this blue one here and so indeed we can cover new territory here finally there can also be constraints from the collider here from monogets this has been worked out in the literature and we can use this limit to put them on our plot as well in the end of the day the upshot is we have nice new limits and we break new ground with these things and so let me conclude and provide some outlook I've shown you that that already existing as well as upcoming energy batching experiments they can constrain a region in dark metamask which has been thought to be inaccessible namely the sub 500 MAB mass region and we break this kinematical no-go theorem from elastics scattering by going to the inelastic channel of photon emission which has the benefit of having a higher endpoint energy and I've shown you I've proven the existence of models that we can actually constrain better with this method I should also say there's a whole number of ways how we can and will actually improve our limits on this first proposal so for example we should look at not only the photon emission channel but we should also look at the direct shake-off contribution of electrons so you kick the nucleus some electrons fly off and so this is the soft signal and we should also include higher atomic final states we should work out the analog signal in a semiconductor versus which has a very different electron band structure and so the signal formation might be different and finally we should also re-evaluate the neutrino floor given this new signal here thank you very much I should switch back now to my camera over here hello hello hello can you hear me yes I can hear you host has a problem so now I'm gonna be host for now so thank you very much Joseph was very nice I like it a lot so maybe just to remember to the people following now the webinar that they can make questions to chat if you're following this live you can make questions for Joseph directly to him but now we start with a question from the public that is here in the session of the Hangout so if there is someone with a question we're gonna start now so I have a very hi so your issue with the cross section was it was involved the rescatterings with the shield of your detector right so would you suggest in that case removing this shield decreasing the shield how would that affect your backgrounds for those searches I think maybe I was not entirely clear on this point so if you look at if you look at the projections say from Xenon-1 ton so Xenon-1 ton they are trying to reduce their electromagnetic background by a lot and so they give us rates between one or even two orders of magnitude lower than in their current detector so they say they will so this is a reduction of electromagnetic background by a large factor and you can ask how do I go to the small rates and you can only do it by taking by by utilising your detector volume because on the out there exterior you know just the usually components of the detector sit they are dirty in a sense so they have a lot of contribution to the background and so as soon as it can start localising where your event was you can cut you can cut out all the crap but that localisation requires scintillation and scintillation is a weak signal and it requires sufficient energy deposition and so that's why just making a bigger detector and having smaller visualised rates does not necessarily gain you a lot in our proposal and so it depends a lot on experimental details and we have tried to quantify them to some degree and wow yeah okay I see thanks so there is another question I have a question when people start to prepare questions so when you were explaining in the model the idea of using the photo I was wondering if it is do you expect to have also effect like spin dependent cross sections because at the moment you just compare with the spin independent with the coherent scattering of the from the mission from the photo but I was wondering if there is some effect or different observable depending on which part I mean which part of the nucleus is emitting the photo in principle it should be only the proton if it is I know you don't have to worry about any nuclear form factor of any sort so we are also not considering the emission of a photon because of an excitation of the nucleus so there I mean I should maybe more generally answer that there are various ways of emitting a photon from a nucleus one way is the WIMP scatters pretty strongly and you go into an excited state of the nucleus then it de-excites and it emits a photon these things have been looked at but it requires quite sizable momentum transfer and you can only go into excited state using heavier dark matter so you need WIMP dark matter and then there are certain xenon isotopes which have you know 30 kV low-lying nuclear levels which you might hope to excite this is one possibility but this is not what we look at because we look at very light dark matter and so you never go into an excited state and now there you never need to worry about things like flipping a nuclear or any of these sort because the photon is relatively soft and so you never probe in the structure of the nucleus so this is basically a coherent emission with a factor of C squared the photon energies go at most a kV and you know Fermi scale is 200 MeV so you have a very soft photon with respect to the size of the nucleus you never resolve any of that but since you mentioned spin dependent what is interesting in the spin dependent case is that there are certain operators where you can couple to quarks and you never generate a coupling to electrons on no loop level so for example this is the that's an axial vector, axial vector coupling but it's also velocity suppressed but in principle I think what we should do is take our proposal and also look for the spin dependent case indeed there are there are really cases where you don't induce any scattering on electrons and then I'm sure it will be the only way of going to low masses did this answer your question? Yeah, in fact it was very enlightening so just the other extreme when you were showing lower limit of if I remember it was kind of 50 MeV something like that 10 MeV I can turn on the you want me to turn on the Yeah, if it is possible Yes, sure My question was was in the point of which is the would be your no-go theorem for this method in the sense because at some point your no-go was the minimal recall energy that can be detected in the I mean what would be your that's a great question so that actually depends on what's your method of detection say for example we look at these liquid scintillators and if you look at these liquid scintillators what you require is ionization so the minimum energy of the photon is 12 MeV that's the binding energy of the most shallow bound electron in a xenonucleus so that's the n equals 5 shell there's an electron with 12 MeV so that's the minimum energy you need and that energy so 12 MeV would be the minimum kinetic energy of the dark matter particle you need to have and so now you take M chi V square at half equals 12 MeV and you find out what's the dark matter mass and so it's I don't know it's somewhere in the MeV I don't know the exact value but that would be the answer but it's also possible that you are going lower and you say oh actually I don't require I don't require the ionization signal but you can directly try to detect this photon it's close to 7 MeV liquid cleaner becomes transparent and so it's like liquid cleaner is transparent to scintillation so it's transparent to its own scintillation that's why you can detect it and so in principle you can also try to be more ambitious and look really for the super soft part of the spectrum directly not easy I mean we have some ideas I'm not sure what what is going to work in the end we are thinking of this okay so and just the last question for the moment is the what about the experiment with another modulation do you think that they could also try to use this method yeah so I don't have my bonus material I made them so that is you can ask what you can do is you can ask for example take Dharma and you can speculate that maybe the modulating signal is not due to wait you know what maybe let me pull up a slide and show you I have a I have a different talk where I have this if you can just wait okay so oops I need to do it again and then I can show you the plot do you see my screen yes we can see it okay very good so oops okay it seems that the host I mean the speaker went out maybe he went to look for the slides to the next office but for the moment that he's coming back you can also remember to the people that you can make questions to him you can start to write a question to him in the YouTube chat live for the moment that he's coming back you can also remember to subscribe to our Facebook page and if you enter to the to the WordPress page that we have there is one of the menu that is to subscribe to the mailing list in case that you want to receive all the announcement of the webinars and just a reminder if you have any of it so you can write there so for the moment that he's connecting back also we can I don't know if you guys you were planning to make questions to them because it seems that the also he was gone you have a question for him yes but for him not for me let's wait for a moment and wait a couple of minutes because the most probably if you notice that when he he turned down his laptop it seems that maybe he disconnected the cable of his computer now it's coming back great I'm sorry did you hear any, sorry I think I hit the quick button on the wrong did you hear anything from what I said no it's remaining a secret I think maybe that was engineered no ok I'll show you I will make it very quick sorry ok wait ok so here it is so you can hypothesize that dark matter that the annual modulation is not induced by elastic scattering but by from the bramble straddle part of the elastic scattering but that happens at a lower energy and so we assume now that it's not deducted only the harder part of the spectrum which is the bramble straddle makes it into the deductible region and that's what indeed is observed and so here I went through the exercise of fitting the Dharma signal through this bramble straddle and that gives a very nice fit good chi-square and all of this if you look in the bigger context where this island shows up you see it right here so this here is now a third Dharma island there are these two islands down here somewhere at higher dark matter masses and now there's this here is the island coming from bramble straddle had we proposed this a number of years back this would probably be very exciting when the other limits were not in place but now it's actually this possibility is excluded so yeah it was a fun exercise but I don't think it has anything to do with Dharma okay so there were now we have the you're back here can you hear us? yes I can hear okay sorry so I have a question about the directionality can you use this to reconstruct the direction of coming dark matter? so the direction and something we are thinking about and we think it might be possible it's not easy so there's a direction of dependence in the in the photon emission and that is you basically dot q so the momentum transfer into the polarization vector that's the information you have okay and so there's a certain there's a preference radiation so yeah orthogonal to this to the direction of the if you want and so there's some information there you may try to correlate it with the recoil energy and we are currently looking tossing various possibilities around of doing this I don't think it's easy but if you go to low photon energies again what I said before the genome becomes transparent and you may actually you may actually try to do this but maybe not so we have the cross sorry maybe not photon by photon but statistically you can try to of course yeah this is statistically yeah but yeah so so we have all the formulas we have to put the things together now so do we have other questions for Joseph I guess the time is we already are right to the hour of transmission so I guess it was a very I'm going to take it but we have questions in the in the chat it's no we don't have questions there but anyway for the all the followers of the transmission you can you can watch again this webinar in our YouTube channel so I guess for the moment we are going to thank Joseph for this very nice seminar and the idea that was very good to explain and I guess just to talk to the rest of the people following us is just to join us again next week we are going to have another webinar from Pablo Roy from Simbestar, Mexico and and I hope you'll see you again next week so goodbye thank you very much