 Hi everybody, welcome back to this last episode of this season six of the Latin American webinars on physics law physics so today is a very Nice webinar because it's gonna be about the one of the topic that I am most Enthusiastic is that matter so before to go directly to the to the webinar first remember that if you are watching this be a youtuber on You get news about face over there You can give a like to the to our Facebook page to follow in follow us in Twitter and also to subscribe to YouTube so now The speaker of this webinar is gonna be Daniel Coder from the University of Freiburg He's postdoc there and he did his PhD in the room university bohem in Germany I guess so Daniel if you are ready you can You can start with your and welcome to the to this webinar and thank you for for you to Presentable the results of the Senon experiment. All right. Thanks for the introduction Go ahead and share slide deck. I have here All right, and if for some reason something stops coming through just let me So here we see already a very nice picture of the subject of the talk This is about the Xenon one-ton dark matter search. This is our beautiful detector and Support systems sitting underground the ground sassow in Italy So I'm gonna talk about this running system mostly and our plans for upgrading it Our collaboration to get an idea of the size. I'm representing about 20 institutes for more now And now we've grown to about 160 people. So it's for the Xenon end-ton stage We're about a 130 for Xenon one-ton. So it's a medium to large collaboration and Made up of institutes all around the world. So it's getting harder and harder to find a good time a meeting So this is a an experimental talk. So you only get one slide on dark matter But I've seen your webinar series. I think if people are following they have a little more familiar with the subject So the point of this so we're looking for dark matter, there's evidence the dark matter exists It's pretty convincing. I think every physicist would say there at least More or less certain dark matter is a component of the universe And we see this on a lot of different distance scales via missing mass So on the galactic scale we have galaxy rotation curves. We need dark matter explain the extra galactic scale We have moving dynamics of galaxy clusters Which we also need dark matter to explain the galactic collisions And on the universal scale we have these very famous point measurements where we see through the CMV that We need quite some mass in order to make the observations match our theory without the universe developed Xenon one kind is searching for not just wins several other models of dark matter to but the most The main search we do is for links. This is a very general form of dark matter Comes out of a lot of theories and we're basically looking for is weak weekly interacting massive particles So weak scale interaction no electromagnetic interactions. It's a pressed interaction to normal matter Masses you have to have mass. So when it's going to go MAV to TV for Xenon We go from a few TV So tens 10 GB or so to TV for our main search We also have other searches that try to go lower masses an important thing about winches. They have an energy density So put point three TV per meter square. This is rough But that means in our sector of the Milky Way every square meter of space has some whims in there We're flying through them or seen from our perspective. They're flying through us. So you can make a Earth-based detector make your detector out of some material and there's always a flux of whims going through it This is a principle of Earth-based direction detection and you're thinking about What material to make your detector out of? We picked liquid xenon liquid xenon is actually a great material for making a wood detector For a lot of reasons so first of all When you're looking for a wind coupling to matter it can either have a spin independent or spin dependent coupling It's been independent coupling experimentally easier Because it's and it's proportional to the atomic mass target and xenon isn't big at all. So it's 131 atomic mass And this means that we're looking for a relatively strong coupling This makes also a very nice So what we're looking for is coupling to the nucleus we get a nice recoil spectrum from xenon So this is shown here so you might think of using some other Elements like argon germanium other detectors do Xenon has a very nice recoil spectrum is in enhancement here at low energies for this for a hundred GB one So if you can measure energies around 20 kV or so you get a high statistics in nuclear recalls relatively high statistics. This is of course a very suppressed or nonexistent interaction Another thing is when you're looking for dark matter The only real way to do it is to close out every other known interaction So you have nothing left and whatever you have left you have to understand So we need a low background environment and Xenon can be purified a lot. It's a noble gas So we hit very high curies and except for Xenon 136, which is interesting its own right There's no radioactive isotope in Xenon Its density also means that it shields Radiation so the inner layers of Xenon are even more clean than the outer layers Background standpoint and a big thing when you're thinking of building a detector You have to actually build it so as to be possible to make so liquid Xenon is nice because it's heavy And you can also put it in a bucket right so it's liquid at temperature higher than with the nitrogen temperature So this opens up all opportunities in the lab You don't need some healing cooling or something and it's a simulator So it's actually simultaneous to the target and the particle detector because it simulates life and talk about that in a minute And for a lot of these reasons so liquid Xenon TPCs have kind of taken off in sensitivity Towards for between Winston matter that shown here So this is a plot from our first paper from Xenon one time our first result paper from last May And now we're going to talk about the new one the update so Xenon is a stage project because when you want to build a sensitive dark matter detector, you don't get around building a bigger one. There's all kinds of other Considerations you have to always make it more and more radioperials have to Solve a lot of practical problems when scaling up but in the end you need the mass So I kind of drew a cartoon here where you see the target mass It's a mess inside the TPC scaling up from Xenon 10, which was like a thermos Xenon 100 trash can size maybe like a short one Xenon one time is already like a big barrel So it's one meter diameter one meter drift length This is already a large thing and two tons of Xenon inside and I'm going to mention here I guess for the last time until the end our upgrade, which is already under construction So Xenon one time the pictures here. It's running right now Xenon n-ton. So this is You know n times we didn't know how many tons you're going to make it when we started We just know it was going to be bigger That will have a total of seven and a half tons of liquid Xenon which we already have stockpile Six tons inside the detector and the active volume and that'll be about a year and a half From now we'll start Operating it and you can see here the outer cryo sat here is actually the same outer cryo sat as you know one ton We made it extra big so it will accommodate now this bigger inner cryo set You can just upgrade in place so we can be really quick we can in the next stage out Which is important because this is a competitive field right now So here's a nice shot of our bottom photo multipliers during Installation I'll talk about how we actually detect dark matter with the CPC with liquid Xenon TPCs in general There's a few of them out there all kinds competing for the top spot But we're gonna have it Good so this diagram here shows how liquid Xenon detect TPC works. So It's one detector but two stages basically On the left. It's just a scintillator, right? So you have a particle coming in and it makes scintillation light as we said 178 nanometers in the liquid bulk and this is seen by two arrays of photosensors here on the top and on the bottom And that's it. So very normal detector familiar with if you do any type of experimental physics scintillation light collected on each end and then you get here on the bottom. We have a kind of Wave form so you get then this pulse right when the particle hits The second part is kind of the tricky part for the good Xenon detectors And this is where we get our second signal which is ionization. So the second energy loss channel for Particles hitting the good Xenon that's ionization They create free electrons if you don't do anything they just kind of sit there and we combine eventually But we put the entire volume under electric field so down here is a cathode grid on the bottom Up at the top is a gate. So this is we call it the gate. It's another electron It's at zero volts So we have then an electric field between the two of them causes the electrons to drift up towards the gate here's something a little tricky happens above the gate we have liquid stops and We start with gaseous in on filling the rest of the volume The liquid gas interface is right in between the gate and the anode Anode is by a stick four kilovolts are quite close together, and there's a really strong local electric field And this causes the electrons to be extracted into the gas where they make proportional simulation lights are amplified Sorry proportional implicated amplification so that they're amplified then in the gas phase What you get here is a localized signal in XY so you see kind of in the cartoon here You know where the interaction happened in XY so when they drifted straight up the time difference gets you to depth And so you get 3d reconstruction and you have the energy loss from both channels and one more trick we have up our sleeve We want to know when a wing gets a nucleus so Wink shouldn't in the electron as we said the coupling is they're electrically don't see the electromagnetic force They shouldn't couple of electrons in any measure of the way So most of our background though is electric so we're looking for nuclear recoils, and we want to filter out Recoils you can see electrons of electronic recalls that show here So electronic recoils are shown in blue where we plot the s2 signal versus the s1 signal And then in red isn't clear recalls when you see that we can separate them This is becoming very important later for the analysis Here's what it looks like in real life. So this is a waveform where you have the s1 signal nice tight See 0.4 microseconds here. So microsecond time sub microsecond time light Here you have the s2 signal blown up here Where you see several microseconds wider and there's this localized signal this is the top PNT rates Just the area in each channel see what the interaction happened already by eye and the drift tells you about how deep it was This was about a third of the way down So if you put the detector on the ground You measure kind of nothing except your backgrounds and if we want to understand our low-energy background or low-energy signal We calibrate in basically three ways So the first way you need to look at our low-energy electronic recoil backgrounds We need electronic recoil events and this is a little counterintuitive We have this nice detector we spent a long long time building a long time making a radio pure so that there's no Radioactive background in it and then we pump a radioactive source inside But right on 220 is kind of nice right on so everything here is very short-lived the longest This is 10 hour half-life. You can always wait with three days four days after pumping it in and you're back to background conditions So we do that the pump is right on 220 in and we get these data decays We've led to 12 that give us low-energy electronic recoils very nice So we measure every month or two the response of our detector to that make sure it's stable and we use this in later Training our analysis to know what the background looks like We use also krypton 83m. This is another internal source. We pump into the detector. It's also short-lived two hours We pump it in about every two to three weeks and this gives very nice two lines a nine and a 32 kv line About a hundred fifty nanoseconds apart. So sometimes you see them as one reaction sometimes it's two depending on how long they live And this is homogeneous and XYC So we wait a little bit after pumping it in and get some mixed around by the flow inside and you have just this homogeneous source of known energy we use this for a lot of calibration we'll talk about later and The last thing is neutrons for signal response. So neutrons look like a wind Looking for nuclear recoils and neutrons make nuclear recoils We've done a couple runs of emeralds and beryllium source. So we can see a diagram here This is a TPC and outside with this little box if you move up and down very simple mechanical thing and you move it into place and it creates Signals in your detector right here. We also have a DD neutron generator. We also lower in place and can power and The two days of that are about as good as several weeks of emeralds and beryllium. So we're happy to have this online We'll show the spectrum in these later So When designing the detector, so there's a few you have to kind of make it good at doing two things so that that's one signal And the s2 signal for the s1 signal You were worried about the light collection efficiency So you want to collect as much of the light from the original interaction as possible So the particle comes in light goes in every direction and you want to collect it The best would be if your whole detector was just made out of particle detectors Practically this doesn't impossible yet So the next best thing is we pick something really shiny that reflects a lot of light and we line our whole detector with that We use Teflon PTFP for this you can see here looking from the bottom up This is the top PMT array during construction. So this will be filled with liquid xenon later and these are Teflon panels That one's very good because it's You get it pretty clean. It's not very radioactive And you can polish it so that it's 95 or more percent reflective So it's very high chance of reflecting an incident Another thing is you need photo multipliers that have a good chance of detecting a single photon if one hits it and For that we use specially designed homo-mots with PMTs we have 248 about half on top half on the bottle They're designed to run in a liquid xenon with a high quantum efficiency so 35% that are assimilation light Pretty good high gain. So five million vector five million. They're operating voltage And we get electron or acceptances of single photo electrons about 94 percent on average So some better or some worse than this depends on the electronics in the particular PMT of course, but the average is about 84 percent, which is quite high And for most of our channels the gains are stable within one to two percent This is a plot of the whole science run Since we switched over to a new database, so it doesn't quite evolve it in it But it shows that these are really quite stable And the design is for low backgrounds. So it's minimizing material budget In every part of the PMT And another thing is the electrodes right so we have these electrodes have to span the whole TPC So the light has to go through them. We designed them to be as transparent as possible I think the worst transparency is the anode and that's already 90 percent all the other electrodes are better So you can see here if your monitor is good you can see the Meshes here your monitor is not so good then it might look like there's nothing there and that's kind of the point that you can see right through them Okay, so You see later in thoughts if you're astute corrected s1 signals We do one correction R and s1 signals and that's for light collection depending on where the s1 happens in the bulk so because of Interactions at the liquid gas interface if a reaction happens in the bulk We're more likely to see it in the bottom right in general than the top right So most of the light of these collected on the bottom right and just because of solid angle You have a higher efficiency if you're closer to the PMTs and closer to the middle you can see that in this plot here This is a radius of an interaction. This is krypton 83m fixed energy interaction versus the depth You see right in the middle bottom. We have the highest parent signal and at the top corner of the lowest This is well understood. We simulated Monte Carlo and we measure it and we correct the signals for this then we can monitor over time the Stability of the light yield using radioactive sources so radar 222 terrible background But we also useful in some ways we use it for monitoring Activated xenon so neutron calibration to make short-lived isotopes in on you see that here and krypton 83m And we see it's flat though science one. So you say our light yield is high and stable For the charge so different There's a few more things you have to worry about So first of all if you have very perfectly pure xenon which nobody has your electrons will drift to the top You'll get all of them out at the top that started In reality you have electronegative impurities in your xenon all the time that we're constantly filtering out and they can absorb these Electrons and so we have to correct for the absorption. This is a function of the depth of the interaction So how much xenon I had to travel through drift time actually and Period you can see that this plot here called electron lifetime because if you plot it It looks like an exponential falling and you can fit a slope and get the lifetime from this So it's the electrons are stable, but it's the lifetime in the detector of them means or this a drift We see this rising throughout a science run basically there are jumps the jumps are so known disruptions either an operational thing Here there's an operation where we intentionally change liquid level to make dust so far I know it's not a few impurities in but then we recover right away. We see it slowly rising and plateauing Sort of hitting about the maximum purity for a current setup. We monitor this with radon alpha decays Radon daughter alpha decays and a crypto again We have a second correction for this for amplification. So our analog and gauge remember We use the strong field between the two of them to extract the electrons into the gas phase And this is where you get this for fortune proportional amplification These are not perfectly the same distance apart all the time and they're not perfectly the same distance between the anode and the liquid level Because I have to spend one meter and they're very small But we can correct for this and so we see here the s2 correction You can see there's a bit of sagging in the middle of it higher on the outside But this is well understood so we see a corrected s2 signal later We correct for these for these things these are things we constantly have to think about when operating the detector If we look at our charge yield over time You see this is also quite stable with maybe a slight rise which we're attributing to our increasing purity over time And we monitor this also with the same sources that we use for the like you You get the energy now from this right so the s1 and s2 you've corrected them you get the actual s1 and s2 Energy at the interaction point and then you can convert this energy just very simply because you know that the particle I don't use energy either through light or charge so the two are anti-correlated so you can say the s1 signal is proportional to the Light quanta compared to total times some constant you fit to the data the charge signal is the same thing times another constant only with the Portional charge quantum of the electrons and if you exploit this relationship you can make a plot here No radioactive sources of fixed energies that are inside the detector And you see they lie around along the line and this is a good proof of this anti-correlation The light there that the energy as we lost you one of the two channels is called a dope plot But you can also do so you take these two equations You can solve them for energy you get a function of energy as a function of s1 s2 in these constants you fit And then you can make a nice energy spectrum We show here where we see The gray boxes are blinded for various rare vent searches. It's a radioactive Peaks here from these are activated xenon. These are material background peaks a big broad distribution from the two-nitry node double K of xenon 136 and we see this spectrum goes from zero to three I mean, but actually our search matter for winces most vanilla winces down here It's about the width of one of these ticks on the on the axis So we're going to zoom into that later for the rest of the talk and we get a very competitive resolution If we just fit the resolution we get from these peaks So here's our resolution curve versus energy versus some of the other experiments You see we're quite up there with the other TPCs and with ourselves from earlier. We've actually improved on our previous device So one last thing is the s2 position It's been too much time on this so the s2 position is reconstructed with a neural network that will train the simulation and Then we correct various effects in the detector like you know homogenetics in the drift field Inactive or misbehaving TMT is to give local problems We correct those using an empirical correction from crypt on data We get here is background data looks like this. You see more activity the edges This is XY RZ Here's a detector wall and then clean interior and then external source will have a different distribution Of course, it looks like this to see neutrons come again. You see that they're interacting mostly in the edge here We still get quite a few in the internal volume So Very important point. This is backgrounds. So For rare event search, you have to have your backgrounds down as low as low as possible Electronic recoil backgrounds. So we said before electronic recoils can be filtered out a bit later in the analysis. They're a bit less Dangerous the nuclear recoils. So Talk about those first and keep in mind we care a bit less about electronic recoil the nuclear recoil But it's still important. So we can see in this chart here is that most of the electronic recoil background comes in the radon 222 and This isn't too wrong right on 222 We just did a really good job at getting the other contributions down. So the material background is quite low Get this low by screening every screw and piece of steel and piece of plastic that goes into the TPC with Jirangian detectors beforehand And this is an example spectrum of this here We screen some high voltage connectors to make sure that they didn't accidentally and Put some radioactive impurities in your detector And we have a very aggressive screening campaign where we really screen everything Before building the detector out of that material Krypton 85 is also low and it might make you think it's not a big problem But it's a really big problem for liquid xenon TPCs. So when you buy ultra pure xenon from The distributor it comes with Krypton. It's not that much Krypton right so it's about So natural Krypton levels are suppressed by a few orders of magnitude already But Krypton 85 the way you activate the TPC is suppressed now by a 10 to minus 12 compared to natural Krypton But we still care about it because it will create electronic recoils in our detector So what we use is a dedicated distillation system. So we distill the Krypton out of the of the xenon Even during operation we did this and we see this here. So this is our background rate versus time and Here we see it Falling we're distilling the Krypton out. That's these gray periods here. It kind of flattens here We distill again it falls and the red points are measurements using rare gas max spectrometer Obviously taken out of the system We see that it's a Krypton drops. Eventually the background rate will flatten out here. We're radon dominated. So the xenon end tone Radon is going to be the big the big challenge And we're going to try to distill this in a similar way out so Backwards but using a similar system and radon. You also have to just try to minimize by minimizing leaks and measuring emanation of your components It should be components that ultimately make so much radon into the system. So that will be the challenge for any time Nuclear recoil backgrounds. So there aren't so many So the first thing is muon induced neutrons. We suppress that by putting the detector under a mountain So you get a factor of a million plus muons And then we suppress it further by putting everything. It's at water tank. So this attenuates all forms of radiation by becoming towards the detector And then we suppress it again by instrumenting that water tank is an active tranquil muon veto Which has a high efficiency of tagging muons crossing it. So we expect very few events from this 0.1 to events per ton here What we're worried about here is the the neutrons that can come from a muon shower Not the muon itself hits our detector. It doesn't look like dark matter. It looks like a looks like saturating every channel The most dangerous background for this stage is where the genetic neutrons these are is a material background There's not really anything you can do once they're in there You know the purities of the material it's alpha n reactions that cause the neutrons In your in your volume. This is a spectrum of what this might look like You see the penetrations a lot more than the electronic recoils But luckily this is this is scaled way up if you scale it down to the Level that we expect we expected out one event per ton here and we minimize this by minimizing the material budget and screening and selecting materials Good and the last part Neutrino scattering so this we don't see in xenon one time xenon n-tons will become an issue With solar neutrinos at lowest energies. So xenon n-tons start to get into this into a region of the parameters phase where We might see solar neutrons And this is irreducible. So they come It's a background that comes and you can't tell it from a whim. So you just have to accept that that background will be there Good, so there's a couple other sources of background here. So I'll go through a little quicker one So every time you run a xenon experiment, you find like a surprise one at the end So I think this was the one from xenon 100 Accidental coincidences where you get low in s1 and s2 signals Low in s1 and s2 versus time that can accidentally combine and then you get Fake events that look like signal and we build an empirical model to handle these expect us to want to venture next search region from it But we expect some more events outside the search region So it's a little disconcerting if you don't know what it is and you see the spectrum Some events creeping in and this this shows our search region and the heat map of where these events fall This wasn't our thing we found But it was luxus so they're surprised when they with their first result was uh rate on contamination both of the teflon surfaces and We see this as well. So we have rate on Daughters became on the teflon surface the s2 will lose charge And so you get an electronic recoil event. It's miss reconstructed and can pollute your signal region We see these very clearly if we select from s1 only cloning 210 We see our her teflon walls just lighting up. That's the spot here What we're really worried about are these led 210 beta decays they can Leak into our signal region and we build a similar type of empirical model there Selecting from data and we select as a function of radio error We expect as a function of radius decreasing events in the inner part of the detector and their inner most clean volume. It's zero Good. So just a few slides in the upcoming result. So Everything I've said now these are all plots from our upcoming results our new analysis We released a result on our first 32 days of data here um So this is just the time the calendar date since last october until february this year our first result was based on this exposure here whose Blue lines live days And our next result will be based on the full thing. So that's now 270 days of exposure The box is here at calibration time. So you see the line goes up flat and starting at calibrations because we're on the counting dark Matter here it flattens because we had a pretty strong and Rather tragic earthquake at grand sussle interrupted us for a couple weeks in here We recovered and started taking more and more and more science data. We've just been on a roll recently collecting good quality exposure Um, so it'll be about a ton your exposure Spanning more than a year of operation and the detector is still operational today um So selection conditions you don't take every we take about 20 000 triggers per hour All right, 365 days a year and most of them are high energy background around them Accidentally triggers things like this and we need a way to filter out the bad events So we use the series of selection conditions and here i'm going to laze over the work of about 15 hard working students and postdocs and one slack And just say that uh, we want to remove basically known backgrounds So some examples are double scatter events whims only scatter once So we can remove anything that scatters twice um The s2 will get wider in time versus depth through the diffusion you can see this here because of drift time versus s2 width Um, there's a nice curve here. So, you know things in the outlier. This might be mismatched s1 to s2 You might have your depth wrong so you can you can uh suppress those um The signal in the top and first bottom PMT is versus where your interaction happened So for s2, this should be constant interaction always happens at the gate. Yes one it depends on depth um And and x y also So we we we put conditions on all of these and in the end we cut out about 7 percent of events in our energy region that we say are are not so good The biggest condition is actually the fiducial volume So as I said before xenon absorbs a lot of radiation And we can see here r squared versus that in the detector This is the detector wall f1. This is the top eight bottom cathode and for science one one, we took this kind of blue square We've improved our analysis enough that for this Science one for this data set that we're doing we'll take this even larger volume and we added a few hundred kilograms So this is a good a good result of a lot of people working hard for a year on it Good So this looks like in the end Uh, so we have remember this plot with the blue and red at the beginning where we showed the nuclear and electronic Recoils, this is the 2016 2017 data at the top of the electronic records right on 220 Celebration you see the the bands. This is actually a fit to our the bands are a fit to our model the blackest of data um Here we have the new nuclear recoil. This is emeralds in beryllium and here's a neutron generator And emeralds in beryllium. It's a lot cleaner see because it's just quicker to collect And then here we have our background data. So for the first 32 days and for the second 247 days What you might notice here is there's a big blue box. This is blinding region. Uh, this is because we do a blind analysis and Actually, if you do a rare events or if you have to do a blind analysis because as I said we take 20 000 triggers an hour If you want to find dark matter, you'll you'll find it in there if you really want to And if you don't want to find dark matter, you'll cut every event away that you want to cut away So you have to do the analysis in such a way that you don't allow yourself that kind of bias And that's what we do. We're not allowed to look in this region until we fix the entire analysis and after that we Don't change it. Um, and we have enough data to fix the analysis We have our background data outside the blinded region all our calibration data We can study we have monocular data. We study a lot of the detector Without looking in here and in the end we look in and see If the detector actually found anything we would call compatible dark matter We have a second line of defense. We also assaulted the data That means that we have fake signal events added to the signal region. Maybe we don't know As an analyst, I don't know only a couple people in the collaboration now And after we unblind we will of course investigate these events if they're obvious background We missed we we would cut them away But having fake signal events in there prevents bias in this too that we can't just cut everything away We would know we biased ourselves if we cut these signal events Um, and the data is being unblinded Order of days, so we can expect a result out of this data set in order of weeks Good, um So just the last couple slides now how we parameterize these these band plots again So our signal and background, uh, we parameterize in the end in s1 s2 and radial space um for the WIMP signal electronic recoil background neutral clear recoil background we use simulations, uh, where The physical parameters are our fixed simulations that aren't fixed are fit to calibration data And this works quite well So we get a model that models really our actual detector response with all our biases for reconstruction all our efficiencies And what what should actually come out in the end what we see here is our model in s1 space Uh, which is the blue line is the total model the black points are data And this is fit to america brilliant data And then we can break it up into a single scatter of red multiple scatter components electronic recoil component It becomes a little more apparent if you take slices in s1 and look at the s2 You see the shape follows very well and we even describe this Um electronic recoil background that's in our in our neutron calibration our amber it seems really calibration naturally A couple of the backgrounds the surface background accidental coincidence I said a empirical so we don't have a physical model for those we fit them from data And they're all in the three-dimensional space. This is new for this result We used to just do the two energy dimensions So we don't do an actual box analysis like many years ago, but we interpret our results using profile likelihood method in this space And what we get for our sensitivity and this is uh preliminary projection Actually, I think it got slightly better recently, but it's a log class probably you can't tell our new result should Have a sensitivity for these three mass points Like this we only computed these three mass points so far We'll do the full sensitivity projection right before I'm blinding with the very final models that puts us Below this in one ton sensitivity looks the pandex is also here So we should be able to improve on our current result and it's worth saying that Uh We do have some discovery potential at three sigma. So if a wink at 50 gv had The zeno one ton limit as its cross section. We'd have a 50 chance of detecting it at three sigma. So this is Actually exciting to unwind And zeno n tons so the projection is down here. This would be another order of magnitude sensitivity Great, so that's the last slide already. So please stay tuned to us If we tweet On twitter, so we'll tweet our result that comes out We've got a blog. I think we tweet our blog entries probably too. So you can read about the results and There's been some like pr stuff to build up to it there. Um, so interesting stuff there You can also contact us, of course, or be a twitter in the blog or retweet or Not whatever they do on twitter And I should also say that this exposure will not just allow this one result Of course, everyone's interested in this one result first order, but we can also search for a lot of other channels low mass winx electronic recall searches annual modulation of the background rate And more so you should stay tuned to this this data set. There'll be a lot of nice papers coming out of it Hopefully not just limits, but we'll see and our next upgrade to non end time is moving forward very aggressively And we should see it in one and a half years So thanks Thank you very much. Daniel is was a very very very nice Webinar we're very nice talk so We can start to make Maybe I mean for the people that are are watching this streaming Please you can in the youtube page You can see that there is a chat box and you can write your question for daniel So for the moment, we are going to start first with the question for the people that is here in the hangout and Okay, so who wants to start to Basking about dark matter and and signal Please the people that is here that don't meet us. Yeah. Camilo, please Now you meet yourself. Camilo. Now you're immune. Okay. We can't hear you Camilo. I don't know if your microphone is working No, we can't hear you Maybe you can maybe you can try to type your question and I can ask it to to daniel So for the moment, I have a question for the moment because it was very interesting the I mean one of the questions that one of the Stuff that you were commenting in about sinon is this how you use neural network, you know this technique to try to To pinpoint from where the electron is coming in the electron when they are drifting in the in the vertical axis Yeah, that was I mean my question is related. Do you suspect that that this spread is energy dependent? I mean lower energy electron spread more in the vertical axis. I mean with respect with the With the vent in the where is the production event of these electrons? So How far are these electric? I mean this spread depends of how far is how deep is the the event happens and the energy of the electrons and how you how you manage to to To really trust in the in the in this training that you do to the neural network to to pinpoint the I mean it's You're asking about the energy So you expect that high energy electrons are more narrowing the spread and the lower in the drift Right. It's how many electrons You have more electrons with a larger electron cloud The spread doesn't matter too much because they all arrive in x y at about the same spot So if we spread more and see it then I know that's why with them So at high energy you get a lot of photons We get a lot of information in general. We have a very good resolution. So our position resolution is About centimeters a bit more Over a broad range of higher energies, right and our pmt size is Seven and a half centimeter diameter. So we achieve a much better resolution in our pmt sets at lower Just a few electrons. You have less information. So you have to build your influence based on just less data and Of course, the resolution gets worse there, you know, as in publicly at zero we'll get to zero but it's still A couple centimeters I think at the the lower energies we're looking at and we verify this by looking at for example, krypton 83 m As I said, they have these two decades. You have this nine and 31 kb um You have the in a lot of times they'll combine the two s2s because they happen too close together in time We can't separate them But there's other times where we can't separate them. So we see there's this 125 nanosecond Lifetime holds on long enough and we actually get a few lifetimes until the the second reaction comes We separate the two s2 peaks and then we can see Basically, we have two s2s that we know come from the same spot and data And we can measure what our position resolution is between the two And so we use this to quantify our errors and lower energy. So this is how we study it This is why we trust the Monte Carlo simulations, right? We because we don't trust them because we verify the data So we know that the Monte Carlo works because we checked out some of the data that it's working well Okay, thank you. So now Camilo, he wrote the question. I'm gonna tell to you this he's asking If you can comment about the impact of xenon one-ton measurement on the self interacting dark matter Or if you have a because he's commented about the Because last week there was a paper about I'll discuss in this type of impact, but in the case of panda x. I don't know if you I didn't actually can't say so off the top of my head Yeah, and in the sense he asked if you can comment or you have any information about the case of self interacting dark matter In the context of xenon one-ton Because there were last last week. There was a paper about the say this type of dark matter models In the case of panda x Okay, I mean it depends if you if you refuse I didn't actually read it. I actually can't say at the moment. Sorry. Um, hope Yeah, yeah, don't worry. Don't worry. Um, so we have I don't know if somebody has a question, but here we have a A question from in the youtube chat from arnie van das He's asking that xenon have any analysis for relatively strong interacting dark matter Where a very large cross-section cannot be ruled out I mean and it is beyond the wing the standard wing vanilla scenario Yeah, strongly so By strong, I guess you mean like a high cross-section. Yeah I'm not sure how we would actually Do that because if I had a high cross-section It probably wouldn't get all the way to our detector, right? It depends on how high so If there was something going on in our detector with a high cross-section, we would see it, right? We would see a high range and we would We would immediately immediately see it. Um, if it had too high a cross-section I guess it would be absorbed on the way in so we we wouldn't see it So, uh, I don't know. I don't know if that answers the question So we don't do a specific analysis for it But if there was something going on the detector that was causing a lot of events We try to characterize everything you see so I think we would we wouldn't notice it Yeah Like our background rates are really Very low, right? So in this search region, it's it's less than an event. It's maybe an event per day or two, so Okay, yes, so I guess anyone can comment in the chat if he want to more details or something like that But for the moment we can pass to another question. I don't know Somebody else here in the hangout Want to ask To unmute himself Anyway, yeah, I have a also have some question But this more if you can go to your slides So I'm gonna ask you to go to share screen again Okay Which slide? Yeah, in the first is in the slide 17 That's one Yeah in the the plot that you have in the in the right in the upper right Yep Which is the line that correspond to the case of xenon one tone What they extended I mean is around the the case of panda or So yours, I guess, uh, that's left up to inference. So it would be I think if you see my mouse, it would be somewhere here I didn't know what and and uh, okay So what I can say is that uh, so here it starts to flatten out our event rate and it remains flat throughout the science run We don't like to show the plot yet because we're going to use that data for annual modulation studies So you could infer from how flat it is Something like a limit on annual modulation So that's why we don't show the plot over the whole science run, but I can tell you at least it remains it remains Oh, okay. It's around there. Okay. Oh, yeah. I mean If there are kind of ongoing Research in that part better not to Yeah, so yeah, and then in the slide 21 You you you show the progress of xenon that is impressive how it how it's taking data with the time So you said that in one in the next year expected to make the upgrade to To end time. Yeah, so this is kind of a rolling process, right? So we're already started designing screening ordering components for end time even constructing parts We want to run xenon one time as long as possible And depending on what this result does it will kind of depend so if we find really a signal for dark matter that we Think is significant at all. We'll probably extend the science run of xenon one time But the plan now is to keep running xenon one time at some point its duties will kind of switch towards Prototyping and tests for xenon end time Because there's no better place to test new things and I'm a perfectly running detector And once the xenon end time has reached a certain stage that we we actually need to take this cross out of art and put it in Then we'll do that So we're going to run xenon one time Of course as long as we can collect all this very nice data We're doing and then kind of at the last minute when xenon end time is staged and ready and Getting to be built in the grand sassos and we will stop the operation and make the switch so the time was I don't know. I can't say for sure when xenon one time will stop But the xenon end time should be running in one and a half years I think you can assume xenon one time has to stop this year sometime If we can take that schedule So in principle, we could extrapolate this this straight line to 1.5 years more or less Yeah, but uh, I mean the exposure may or may not get so much better, right? So at some point you need a bigger detector because uh, you become background limited and your sensitivity doesn't scale Linearly anymore of your exposure. So, um I guess after the science run one is when we really are sorry I keep using eternal collaboration eternal slang after the science runs when we will really uh, kind of Do some discussions on how long we're going to run this thing and uh, Probably it doesn't make sense to run it for many many years. Just The extra six months or a year is enough Okay. Yeah, I understand. So in a little bit Regarding with this in the same so I mean seven one ton is suspected to point out to have new results in the in in this summer Like to to upgrade what was presented during this week last week or something like that Oh Or were some in in a conference if I I mean I was following the how was happening in twitter. So Yeah Some movement about xenon talks Yeah I mean, I think so the last week was this UCLA that might have come from I think they basically showed us This is just our projected sensitivity. It's not a not a result. It's not a limit even That could still be dark like we saw the chance of finding dark matter with this But this will be the result that comes out in in I guess march is our target right now So I end of march with a paper Don't hold me to that though, but it could be a plus of money But that's it order of time. So early spring and then uh, I guess you can expect over the next year Maybe the earliest ones in summer or a bit later Kind of some of these alternate models like axiom searches I know modulations and things like this papers to be coming out based on this one really high quality data set we've taken Oh, cool. So, um, I don't know if somebody has a question Please unmute and ask because I could continue so but in that plot that you you you you show With I with this Current level of you know one ton is not possible yet to see the to start to observe the neutrino floor No, I mean still too far So, uh, at least we we assume we're just missing it. Uh, xenon and ton this becomes then then a thing Mm-hmm sensitivity of all the world going to go through this this neutrino limits of expectancy a signal from these, uh, more neat neutrinos and then Okay, so Yeah, I Yeah, okay. We have uh, Arne van das the guy that asked before he Commented a little bit on his question was It wasn't the case when dark matter can scattered with the shielding this type of models So it's more than interact much more than the That the typical wind But not inside the the future for it's in the shielding itself So, I guess in this case the dark. Oh, I don't know. I've never really heard of a model like this But I guess in this case the dark matter would have to be somehow weekly interacting enough to make it through the mountain or And strong enough to scatter with uh With the shield. Yeah So I Don't know. I mean if it made it If I made a nuclear recoil, I guess we'd still see it in the detector. So this um This way would for sure still see So, uh, I don't know people the people here in the hangout Do you have questions to for Daniel? So I guess no, I mean For me, I I mean I could continue making question, but It's better if somewhere else couldn't could ask but if not, I guess for Has been a very very nice webinar. In fact, this is very very nice, especially all the information that you during the the whole webinar is very interesting especially for For people Interesting in dark matter and direct detection and also for the theoretician that they will bring many models to try to To see what their discretion of their models in in some So First of all, we have to thank Daniel for this webinar was very nice And second for all the people that is following the the webinars or if it is your first time in this Watching this type of research seminars In youtube, so please you can consider to subscribe to the to the youtube channel It's a very nice Parents for us and also for all the people that cannot attend to conference to follow in what is happening in In a specific research topic given by the same researchers and If you if you have phd students or people interested in this, please Spread the word and let's try to to make this More affordable for all the people So Daniel, thank you very much for for your webinar again And for all the rest of the people See you in the next time. This is the was the end of the season six. We finished with a very nice talk Next in two weeks. We're going to have another webinar. We're going to start the season seven that is going to be if I remember well all siri sumer and sari from enf and paoba and Also next in the next season in the law physics just to announce it Also, we're going to introduce other type of webinars, but this time is going to be we're going to start to make colloquium So we're going to Enlarge our topics and we're going to try to make it more affordable for all the fields in physics So thank you everybody and see you in the next time