 I would like to thank the links for inviting me here to give this talk. I hope it's going to be useful. So I think it's more in a lecture like a conversation of whatever what's what's happening in this new exciting field. And, you know, the few I if I can tell the few things that I've learned or something that may be useful for others to actually work in this in something that I believe is really like a revolution in science that happened in the last 10 years. So, without further ado, I will go to to the to the talk and I just want to say it's going to be structured this way. Let me first tell you a bit about x-ray FELs. So what they are how they work. I, I wasn't sure about the background of this, or this crowd I mean I know that some of you may work with extras other with neutrons. If, if anything, even if you know this very well from x-rays I think the the history is of how things happen that it could be useful to tell when you're giving on lectures and presentation. And then I will show you some examples about science that can be done there. I mean I will take it at a very high level I mean the point of the seminar is not to go into detail of any of all the science be done also because I could tell you in detail about my own field as a metaphysics, I will not be able to go too much about other fields. And then as I said, I mean I will maybe tell you a few things that I learned through the years and seeing how this thing works because how to get been tend to do experiments here. You see, it's as difficult as getting time to this facility almost as getting one of the big brands, but I wouldn't say that tricks that is way of thinking and way of preparing yourself I think preparation and thinking and doing hard work is the key to hopefully help and you know you can maybe start using these, these machines in your own research. Alright, so this is a very basic slide for anyone that does x-ray science, but I just wanted to put it there, just to tell you like you know the very basic of what has been using x-rays for the past 30 years, which is the synchrotron light. And the way the synchrotron works, it has, it accelerates electrons almost to the speed of light into a storage ring. Wait a second, sorry, yes. In the storage ring, in the storage ring the electrons emit the x-rays every time they actually get turned by a magnet. The ring is actually a polygon, so it's not round, it's a polygon on each corner as a called insertion device. The bending of the electrons done with the magnetic fields and the more magnets you put put it very simple, it's the more direction the x-ray emission is. And the big, so if you think about the generation of x-rays is basically, of course it's not only that, but the changing in how the light was produced by the advancing the technique of this first bending magnet and wiggler and then on the later, and then you can also use them to control the polarization of light. At best, so if you do the things very well, and Max4 and Lund is the best so far, you can actually get a pretty high degree of coherence, which is nevertheless not full because these are not lasers. And at best you get basically that's what Max4 has 10% of coherence in the soft x-ray range. And this is simply scales with the, well, with a lot of things Max4 has a particular design also to do this, but this is what basically you get. If you want to go to the next step, and have a fully co-event beam, you need to bring a laser. And this is what the FEL, the free electron laser, have allowed to open up for. And I will show you, I think it's one of the best videos so far, it's a bit old now but still explains the principle and FEL. I don't know if any of you have seen it, but this is the way they explain it. They put the video to explain how an FEL works at Stanford. Let me go and then I'll stop it at some point and go on. The LCLS is the first laser in the world to produce hard x-rays, which can be used to see down to the level of atoms and molecules, adding almost half a mile onto the original two mile long accelerator facility. The LCLS uses the final one-third of the accelerator to produce powerful pulses of x-ray laser light. Scientists at Slack and around the world will use these powerful beams to create movies about atoms and molecules move and behave on some of the shortest timescales imaginable. The LCLS starts with the drive laser, which generates a precise pulse of ultraviolet light, seen here in red. The drive laser pulse travels down to the injector gun, where it strikes the surface of a copper plate inside the gun. The copper capo plate responds with a burst of electrons, seen here in blue, which are guided into the linear accelerator. The electron bunch encounters the first of two magnetic chicanes, or bunch compressors. These chicanes help even out the arrangement of electrons of different energies in each pulse by sending the pulses along the side S curve. The compressed pulse emerges from the chicanes and is accelerated further, gaining energy as it travels. The electron bunch then encounters the second bunch compressor. The second bunch compressor is longer than the first, because the electrons in the pulse now have even greater energy. The electron pulse continues to the end of the accelerator at nearly the speed of light. Finishing the boost phase of its ride at an energy over 12 billion electron volts, the electrons enter the beam transport hall, along which they travel through a series of diagnostic monitors and focusing magnets that help keep the beam precisely shaped and on course. Here, into the undulator hall, the electron pulse enters the heart of the LCLS, where the X-ray laser light is generated. The undulator hall houses a long array of special magnets, which comprise thousands of alternating north-south magnetic poles, spaced only a few millimeters apart. These alternating poles cause the electron bunch to swerve back and forth in an undulating motion that forces the electrons to give off X-rays. As the electron bunch and X-rays proceed together, they start to interact with each other. The electrons arrange themselves in parallel sheets, causing the X-rays to become in tune with each other, or coherent, with an enormous boost in X-ray power. Once the X-ray laser light is generated, the electrons must be safely discarded before the X-rays can be used for experiments. The beam dump uses a powerful electromagnet to divert the electrons down to a special chamber that absorbs the electrons and dissipates their energy. The X-ray pulse, unaffected by the pull of the magnets, continues on in a straight line. When fully operational, this entire process will happen up to 120 times per second. The X-ray laser pulse is now ready for scientists to use in one of the six LCLS experimental stations. Okay, I think I can interrupt it here. So to the point where we see how these X-ray pulses are generated. And so I think this video made it very nicely explained it also with animations. But basically, there were two big ingredients to actually make this work. First, you needed a leaner accelerator, a leaner to accelerate the electrons at the speed of light. But the difference, or I would say the uniqueness when they do this here, of course, you have leaner given in the secret one, but then in a booster and you put it in a ring. Here you get to higher energy and you go into a straight line. And the reason is that you want to go to higher energy and then you want to have femtosecond bursts of this electron. So you start for a femtosecond laser and you can accelerate them and you can actually compress them. The reason why you can compress this bunch is so much the reason why you can make femtosecond electron buckets is because you actually bring this to such high energy that you're really in a relativistic regime. And you, it's a way to see actually how relativity works in this case. Typically you have space charge problems, you cannot put electrons too close to each other. But if you accelerate it to close the speed of light, there is a renormalization of the frame or where they live. So you can actually, in our reference frame, they're pretty more compressed. So that's the idea. So you need femtosecond lasers and you need an aline that is able to bring these electrons to such high energies. And the second part that was built in, it has the same name as the object given in the synchrone which is the undulator. But it's a special kind. This is much longer. So typically, I don't know, but it's a few meters in undulator as a synchrone. This is, it can be 100 meters. And this has to be perfectly aligned down to a micrometer precision to allow for this coherent alignment of the electrons. And this is basically the, it's called micromancing. So the first electrons start to emit photons. And since they're traveling basically at the same speed, the photons that emit the start to order the electrons. So if you get this wave, they get some getting amplified and more and more coherent. So, you start from from a mess and you end up with a ordered bunch of electrons. And this process is called self amplified spontaneous emission. If you go to FEL, you will always, you will often order what size or size the size of processes this one. You take electrons that are in coin and put in an undulator, and you actually make them coherent. And independent from three person is John Mayday, Evgeny Saldin and Claudio Pellegrini. Just to give you an idea how what an FEL can do, what was the jump given by the FEL in terms of brilliance, which is usually a quantity that we like to calculate for synchrotrons or for light sources. So, let's look at the, at the point of one kilo electron volt photon or 100 meter wavelength so x rays, soft x rays, let's put it this way. So the best synchrotron as this number, forget about the unit, I mean it's of course units that tells you how many photons per second what is the angle, what is the error what is the bandwidth okay it's it's the standard way to define brilliance. So this component is important is, is 10 to the 23. Well, when the FEL were introduced, the first FEL, it was jump of nine words of magnitude. And I forgot who did this job but the chances of getting Nobel Prize correlates directly with a jump in a logarithmic plot, and this what, what an FEL, the advent of the FEL as brought so it's not an example to think that this invention is creation will actually to a Nobel Prize. And the, what I will do now is, is just to show you some historical picture, just because I think it's good to do it on. It's not meant to give you a lot of information, technical information about what I'm going to show but just to have an idea what's what's happening because from the first. FEL, 10 years ago, LCLS, and now we already have, well, several we I think soon we're going to get close to 10 of those so it went very fast. And so this was the sets of the parameter just to get the slide from 01 but this is basically the what you see in the plot there is the first light in 26 of April 2009 that the lean the LCLS so. Yeah, just to show you so this was the first light a picture of the undulator hole at Stanford. And this is actually this saucer process that was mentioned. What you get is actually it's really gives an idea that it's it's nice you get coherent emission of the light, but you see if you look at the energy so this. And of course you have the energy that you want to have but you do have some noise around this so this is often called the size spikes. So this process we generate very intense first still have a typical spectral feature which is not the one that you use usually in the laser which is nice and and but this is still despite this, this is still such such a jump in in in technology that that this has been even in with this quality of light is not yet a full laser light is still being extremely useful and I will show you this was LCLS in the beautiful Portola Stanford Valley. And soon after that, there was the Japanese FL started to operate and the technical advancement that they did there was to use in vacuum on the latest is a technicality but basically what you can remember it allowed to build shorter machine and produce harder extra radiation meaning you can get to even shorter way. This was a big as a technology to develop and it sounds very simple but technical is very difficult to realize. And this is at West spring eighties. Then if you go to the other end so at spring eight they only have hard x-rays at LCLS they can generate both soft and x-rays. Then, if you go to the other side, the very soft x-rays that is Fermi at the letter in Italy, and this is the only machine still so far, which doesn't use a SAS a process to generate x-rays by uses a cold external seating. There's a laser which actually modulates your electron bunch and before it gets into the later so the quality of the laser is imprinted in then on the quality of the extra get it out, which is a sort of a harmonic generation. So this allows to give longitudinal cohenous. So the full cohenous that the FLF is to be the transverse cohenous, but with the seeding you also get the longitudinal cohenous. And if you look at the spectral shape of this, it's a perfect line in energy so this noisy sizes spikes are gone. The equation here is of course that this process is not easy to scale up I mean they're working on this but so far we limit to 250 300 electron volts so you cannot do are the x-ray diffraction or alleged the cross copy that are usually more common in the in the in the soft and hard x-ray community. And since then, I mean now there is also operating Swiss fell in a PSI Switzerland and parlor in South Korea. I've worked I've done through a postdoc the experiment at Swiss fell it's it's a very stable machine it operates very well and I heard also pal. So, the lessons learned from the first machine has been transmitted very well to the to this to these places. This is a logical jump that has been done recently because now is operating my group has been wonder first to use this is the, the, in this case the European expanse in Hamburg, and also else less to the great well so less stand for will see this that we go from the warm copper which, which has this cold price from stores of the object that accelerate the electrons to the speed of light, instead of making a copper that you need to cool down and you cannot run very fast, you make them out of a superconductor. And in this way, you can run up the repetition rate from hundreds typically that's what they operate the normal Linux to up to megas. And this is again a jumping technology because now what I show you there was this nine or the magnitude, if you now jumping in brilliance average brilliant. If you go and increase the factor of 1000 the, the repetition rate you go up again, even three years and so this is another, another jump which is important of course, you can imagine making a kilometer long accelerator made of superconductors not trivial you need to cool this down all at a very high cryogenic temperature so it is also technologically very difficult and very expensive. And then I will, I want us to mention that we are also making an effort to build a next an FL in Sweden, using the existing Linux at max for in this case, we would be in the soft x-ray regime only so the missing. Kind of the missing FL in the world, which is would do this only so Sackler does only hard x-ray we would do the only soft x-ray which is very in line with the traditional spectroscopy that is written as, and the unique, one of the uniqueness of this, this is that thanks also to the synergy with the Lund laser lab is that we want to be able to pump with basically any, any wavelength so I will tell you a bit more in a second but at these machines, you typically use the x-ray to probe your, your sample, but then you can drive them with different radiation and typically this is done with the visible light. The only way to extend this to from the terrorist range to the XB range. So an idea that you can control any sort of excitation you want in, in your research. And we're almost done so we have the drafts last week we made the first draft of the conceptual design report, and we'll submit it soon so we hope then we will manage to be able to find to find the some funding agency wanted to bring to the next step. So, I think I've been talking quite a lot already, I will try to go into this, a bit an overview of the science as we made the tool at this, at these machines. I'll try to be very high level as said just to want to get an example in each of these three fields which are the ones that be mostly represented into the, into the FL community and show you what, what is the, the quantum leap. The FL can do, you can still do the research that you were doing before the synchronous and I can do way more, way more things now. So, the biology is one of the first ones I am not biology so I hope I will say something that that makes sense to the people that is biology but there were, there was, there were two problems at doing protein crystallography as synchrotrons. And if, and one was that you need to prepare macroscopic crystals, because you, you needed a lot of them to get enough single, and the other is the radiation damage. So you put them in a, I've seen this beam as you put this in an environment and so those are not in people measurement with a free FL you solve both problems at once. And the radiation damage is actually, it's a bit controversial so it's actually way worse, but because the process so short you avoid it, you know, we'll tell you this is a principle diffraction before destruction. So this is the way that some of the experiment are done usually you inject whatever protein or whatever crystal you want to send through a jet, and then you synchronize the jet with the extras. And when they hit each other and then there is a detector behind the extras and then you get the collector scattering pattern. And so basically, every, so since every crystal and every protein comes in a random order so every image will be different but if you do enough statistics so you will be able to reconstruct it. This is a major computational work so I mean there's been consortia stuff this is this is difficult stuff to do it sounds easy but then we need to build 3D images down to from such a complex data is very complicated. And one of the problems that we easily figure out or discover that experts is that every time you do an experiment you go you go home with terabytes of data. And then you need to analyze them so this is one of the things that that the complexity experiments carries with them. And so there is the idea of this of the fraction before destruction comes because you basically had to think how something get destroyed in real in real life so the x-rays that we we use they move with the speed of light of course of course the material they go there is a reflection so on but basically that's the speed of light. What destroys something physically is not the, the light of the electrons is when the atoms fall apart, and this is given by the speed of sounds. So if you see from this number that I put on, these are orders make a difference so you can flash something with an extra deposit a lot of energy but when you take the picture, the thing is still is still intact. And then eventually will explode so that's why you have a continuous jet of new of new crystal so you have to refresh those. And what you see is still the crystal before it explodes you can, you can learn something from it. So that's that's the big jump that FEL allowed in biology. If we move to another field, which is closer to biology than physics so but still quite quite different is chemistry. Before I explain the, the experiments have been done there, I, I need to make an explanation of what, what does it mean to see something goes very fast so in the experiment of the biology we don't care we have one beam. We send it is very fast beam we send it through the sample, and it gets diffracted. But imagine that you want to see a very fast motion, and you're very fast means femtoseconds. How can you do this. This cannot be done because of course there are no mechanical shutter that works at femtosecond time skills. And there are no even electronics shutter the working femtosecond the best you can do is a nanosecond a bit better than that. So the idea to do this ultra fast experiment is to use the same trick they used to take the picture of the hummingbirds wings. And how do they do it so if you ever tried to in sunlight, we take your camera and take a picture of the hummingbird you will see the wings are fuzzy, because the wings are much faster than the mechanical shutter of the camera. So what they do when they do this, maybe they don't tell it but it's basically they use a flash. So, the flash the poor hummingbird, and, and then they have of course a camera shutter, which, which is still slow, but must most of the light if not all the light if they put in a dark room comes from the moment where the flash was on the flashlight can be much faster than mechanical shutter and much faster than the wings motion of the hummingbird. So here we do the same thing. So we have the textures which are slow, but we, we try to look at our sample only with very fast flashes. So there is a time difference with the hummingbird the hummingbird is moving his wings by his wings by itself. Here what we usually want to also what we need is a trigger. So we take our sample, whatever it is, we pump it and our pump is basically the start of the clock the trigger. And then we go in. So the pump doesn't go through the tour to our camera. And then we go in, we come with our probe. In this case could be the extra probe, and this is our flash, and this is what we look at. We look at this at a sample, you know, only the duration of our flash, but we look at the sample. After a certain time we give it a kick and so basically we can control this is very easy to do if you want. I will tell you how this is done but you can easily control the different time for the delay between the pump and the probe. If you do this stroboscopic you can actually build up a movie with a different frames taking a different delta T in this, in this slide. So that's the idea. So, and this is what we call the pump probe experiment, which is more than half of the experiment that you do an LFEL. So the example that I showed before was not a pump probe was just a probe with a sampling, you could do it a pump probe is very complicated but this first only the fraction only needed a probe. One of the key experiments is also one of a lighter the else less that was done is the try to probe a chemical reaction. And this is. So, I apologize for some chemistry or chemical physicists in the audience but I'll try to explain this but basically, this is the, the, this one or simplest way to picture a chemical reaction you have. There are molecules in some state that are that exist and you want to make them react and create something else. Okay, and the two energy states so the final state may well have a lower energy state but it's a better between good from the two states. And one of the trick that also was awarded a Nobel Prize was use catalysts so surfaces where would actually you lower this energy better and you favor, typically by providing them thermal energies you favor the transition from one side to the other. Okay, so you want to go from from from the left here to the right. The point is that in between. So there is what's called the transition states, which are very interesting in principle to know because this basically shows how the reaction is happening. And the problem is that they exist for only very short time. The question is, can we ever probably can we probe this, this, this transition state. And the conversion wisdom said no, because they're too fast if you do extra to the top of the single turn you just see before and after. And Anderson was one of my colleague and one of my mentors at Stockholm University, he was their professor at Stanford. It was one of their drivers of this of this. Yeah, I said, well, let's try to do this. We didn't have to tools. I mean this is something new we never tried. Let's, let's give you a try. People with skeptics and so on that is good work, but actually made it work and so basically what they did I mean now I will flash this slide through and so don't worry about this but basically, they did what is shown here is an X ray absorbance petroscopy at carbon edge, sorry, at oxygen edge, and a different time delays between these pump probe that I told you so there is a different delay between when they trigger the reaction in this case with a femtosecond pulse, and the arrival of the extras that probe the, the structure. So basically, they did some modeling, we skip it but basically what they could see is really how, in this case, CEO becomes here to on routine, and this is fantastic they made beautiful videos on on this and it's it's very pedagogical to see basically what you see is that you prepare a surface where you have a CEO molecule and oxygen around, and you start to drive it with lasers and this. And then molecules of these atoms start to move around, and then the motion start to get increased until the, the CEO, get quite close to the oxygen. And then what you can do really by doing this at different times it is you can see this happening and then you can calculate and you can figure out there's a probability for actually CO2 to form and then leave the subsurge or eventually the CO2 couple to another but this is basically the first time a chemical reaction was seen kind of live. So you could make models stuff but you actually here experimentally, you have a fingerprint so what's going on and this was fascinating this been like a revolution on how you look at this, this phenomenon. Okay, so, and this is was the first two experiments and I will go in a field that is a bit closer to mine. I don't have experiments but I think I will give you only one, because I don't want to load it too much. And so I'll skip the first one which was with a hard X rays, and, and peculiar pump which is there as but, but it's not important for this talk. I will talk something which is closer to my traditional background which is magnetism. And I would like to convince you that these machines which sounds so, you know, so so fancy so complicated and so very close to fundamental science actually can have an impact also in our society because they allow us to tackle problems that we where we actually have a roadblock. And in my case, it's something I started even doing my PhD, and it's was something that maybe some of you knows but the, but vast majority of the information the word is still stored in the form of magnetic drives. Okay, here it was, this is an annual picture, even magnetic tapes is coming back and I know the average able to audience here but when I was a kid, there were tapes around where you had your music store. There was before the CD even, and the VHS the video also. But this is I mean it's getting very fast and for long term storage tape can can stand 3040 years without losing the information so this is very very much useful for storing something that you need very fast. Nevertheless, even now, the Netflix servers so when all the videos are still magnetic. So the SSD which are on our laptops are not used they're too expensive. And not so reliable as compared to magnetic hard disk drives that maybe most of you have seen. So when you need a lot of data, and, and cheap then you go for magnetic disk. So this is something that is still, you know, used in the Netflix times and in lockdown times. And there is a bottleneck there is a technological bottleneck because these these hard drives are actually what was lost on the entire thing. So that's why I mean people trying to work on this and trying to improve this. Another issue, which is really crucial and people I started to tell this around when I go to conferences and we started to mix and calculations and there is people doing this. We are starting to use a substantial still not major but a substantial amount of electricity to heat and cool this data center. So, if we are not careful so of course I mean every exponential at some point flattens out but if the exponential slope that we see now the consumption where it continues, the predictions that by 2040 all the electricity in the world would be used to actually for data. That means we either live a life on Netflix, or we have to do something about it and of course it's not sustained. So, yeah, how can we make this more efficient. Okay, and one thing is of course I said data centers, more than half of the energy goes into heat. So can we actually try to get this magnetic storage to be more efficient so not to waste so much into heat. So the idea, this is very fundamental but in a field where very fundamental ideas like the giant when it resistance was actually done to to to commercial technology very fast, and it's still the way that is used nowadays. And one idea was to use ultra fast lasers to control the magnetization state, and the of a magnet of the way the data storage and it was discovered almost independently. And by the group of Johnny big more than 20 years ago now, when they noticed that if you take a ferromagnetic material nickel, and you shoot femtosecond laser minute, then the the magnetization quenches so it goes down. So you like a magnet of a certain strength, it becomes half the strength in a few in less than a picosecond. So it was against all the wisdom about magnetism mind is much slower, there is nothing it was thought to be in the textbook. So no one understood this, but somehow I mean it was of course, super exciting for people in in my case try to understand this. But until there was okay but what can we do about this, until it was this experiment 10 years later where they show that not all you can quench his magnetism but if you actually do it properly you can actually write with a femtosecond laser the state of a magnetic bit. So what you see here in the picture in black and white, this is basically the whitest north pole pointing towards you, and the, and the black is the south pole of a magnet pointing away from you. And, and so what you see the dots is where they basically they should single single laser femtosecond laser pulsing in completely reverse the magnetization and very reproducible. So of course it was a year, and they also made some calculation this is way more efficient that the way they we write my netization nowadays. So, the ultimate thing we want to do, and this is a big job I don't expect to go into detail to understand all all this but it is that what is the most stable and smallest magnetic object and can write information and make, you know, in every disk that rotates a consume some energy we have the largest amount of dots, and also stable. And one idea was to see this topological object which are called skirmish. And people have been trying to imagine those but imaging and try to see. So if we want to use them in the hardest we want to we want to see them understand how they react when we write them so that we can optimize the materials and try to make them usable. So what we want to do is an experiment which is both ultra fast and the looks at the ultra small if you wanted to learn. And the idea was to do this experiment that they newly built a European expel, which allows better statistics so that the image works even the other facilities but for some of this experiment was not enough. So let's try a different machine and try to do movies of how this skirmish structures look like and how the reverse upon illumination with a laser pulse. And the way you do log of fee is in. We have a technique called a magnetic holographic imaging. In this case of magnetism, and it is relatively simple. If you have a queen laser so you have your, you need to have a queen source, which is our x-rays with small enough wavelength so that you can actually resolve the thing. You need to have a reference hole so you have your beam going through the reference and the, and your object. This is the principle of a log of fee, because what what you do after is that after this, this thing you have two beams one from simple reference which will interfere. And from the interference pattern, if you take the Fourier transform this can be shown mathematically, you recover the original real space image. And this is the key holography that you try that you modulate the phase you encode the phase information in the intensity information by having to two references. And this is what the way it works. So we did this experiment. I just want to show you the results because this is new so we were actually one of the, I think now the first users with a new detector that can measure at this, this facility so you get this scattering pattern. On the left is the one that you get by one or the city of the, of the other photo so when you do magnetism you always want to do the right minus left electricity so that you get a magnetic signal you take the difference you get the image on the right. Because a lot of time is all playing in the data, but we were finally able not not too long ago to see the first magnetic images out of this and where are the magnetic images. They are here. And those are the, what you see here so called magnetic domains I zoom the on top of this and it was the, the first step now we have. These are the same samples where we expect to see the skirmish these small objects, but the first we wanted to make sure that this thing works, and I know it's noisy we're still working on this. But this we did with a femtosecond pulses and we also have time pump probe data to show you how a synchrotron can do these things in a way in, in, in, in a better way in a way in a way because it's a much more stable source but it doesn't have the temperature this is what you get. So this is we take the similar sample that's so late, you do this holograph reconstruction you see this magnetic domains here, done with a 50 because a second pulses. What you cannot do there is of course is to see or anything to see what happens when you go. When you want to see this ultra fast writing of magnetic bit. Okay, so this is we were really happy works now we still have to optimize it but I think it was it was an effort to work while doing and then hopefully we'll be able to make the first femtosecond movie at the nano scale and we really with a lot of frames and try to see how this evolves. So just to, this is just a summary of what I showed you so biology chemistry and physics. What you what you can do I think the summary of the summary is that if you go to an FL what you can do uniquely is the combination of both. The, the short pulses and the short wavelength. If you want only short pulses you have a table top light if you want short wavelength, you have synchrotrons. If you want both you need an FL. I think I don't have that much time left so I just want to tell you that maybe I'll, I'll just go quickly through these slides about how you set up a pump or experiment in FL. You need to think a lot about a lot of things. But I would say the most important things is that is the synchronization. So, when you do an experiment there you need to find a way rather quickly to, to align spatially and temporally your, your, your laser beam, I think we're getting better at this and the facilities are good at this. One thing that that that is very clear at the, the FL is that each experiment is different so you really need to think a lot through and if you do something strange you need to talk with the people doing this pen before to see if it's physical. Okay, I can actually. Well, okay I can show you this basically is a similar to a synchrotron in the sense you need to, sorry, no it's similar to a laser tabletop experiment. You need to find spatial overlap, temporal overlap on reference samples and then you go to the, to your actual sample. And this is sorry this is not something that you don't do in a synchrotron, most of the time because typically you only use one one beam. You need to align them in space and in time. This is, we're getting better at this but this is sometimes what takes a lot of time in a new experiment. Okay, I'll skip this because I think, I think somehow I will tell you this in the, in the, in how you write the proposal. Another thing that I wanted to tell you, if you decide to or interesting going to an FL is that this complicated thing of finding the sample and finding the timing and finding the overlap and so on is only the beginning the difficult part is actually start to actually get the data and analyze in it. And the, the reason why this is so difficult is because for most FL as I showed you before with disaster is very noisy process, each shot is different. So for each shot of the FL you need to, you need to analyze it independently. So that's very different from synchrotron where you just you know you assume a reasonably average flux, but here you need to go shot to shot. It's difficult to make cameras that allows to do this. We have them, but you need to do this you need to think you need to prepare your codes to be able to do this quickly deficient. And you need to have a good ideas because, as I was going to show you before I mean the data when you go home from FL, the first of us we had 10s of terabytes of data. In the estimate we did it at x well we had 500 terabytes, and we're going up even more. So you cannot just rerun your code on on on such data, you know such big files. So somehow you need to have an idea just before you start. And one thing that is also important. I think this I need to tell is that when you go to such high fluency is when you go to use this powerful machines. So what I want to communicate I think this will be online so people can maybe look at it afterwards is that you should always work in a linear dream. If you take to the textures usually there is always you always try to think to normalize your data. You look for a single and usually have an I zero monitor. So rather than the idea if you can in your experiment use the same I zero you should, otherwise you have to go and look for the regions, you plot always some sort of the signal of your, that you on your main detector against the reference shot to shop, and you try to see where the the linear region is, because otherwise normalization is impossible and you will not see anything. Okay, so that's as a promise to be to run over time but I would like to to go through this. The reason why I told you these technical things is because I think they're fundamental when you when you write a proposal that these machines. So the first thing is that, well, talk to the be my sense well out of time to discuss the issues that I briefly mentioned, is it feasible what we're going to look at, and so on. And do your homework. I think one thing that I heard from people that is usually just the day before the deadline they get hundreds of emails, you know, if you can do this or that. Just, just do it before I had of time and read what can be done so you can actually ask a specific questions. And so as he lasts and I think at the facility they decided to run two type of experiment one is called standard configuration and less standard. If you stand the configuration you're a bit more safe. But if not, you have to think that you will have to contribute with equipment with people to actually try to make it work and you need to, when you write your proposal right explicitly about this, because otherwise, it's a good way to reject a beam time proposal. It's another crucial thing. I mean, sometimes you cannot, but if you have experiment preliminary data that you actually can see a signal. That's crucial. And if not some sort of calculation so it's an expensive beam time that you have very little time to make it work so everything, except the new things you want to do as to be several. And the other thing that I think is important is how to think for this, this, when you write such a proposal. So, as I said, the beam to me is precious. So there was the estimate for an LCS one beam time cost the American taxpayers, about a week so five shifts, a million dollar. The US Department of Energy is giving any scientist there basically resources worth a million dollar to do this experiment. You want to, you know, be responsible and try to do the best you can to think to answer about under the big scientific questions in your field. You don't want to do incremental science there. And the people review the proposal top scientists in the different fields, more than one per field. So, you need to be scientific correct I mean what importance the science is sound. Okay, so that's, and that's people that would judge you I mean of course, as human factors you may have like a competitor independent but he's not alone and stuff. I don't think anyone will block you for having a great idea, but you need to convince extremely good scientists about this. And you need to do an essential your more about feasibility checks. Otherwise, that's a way to rejection. And this is what I say to my students and postdocs and in general people I work with so think deep but think simple so answer the questions that you can formulate simply. That's usually the fundamental questions. And finally, these, I mean some suggestions that I, I mean, I don't know you can do whatever you want with them but I think it can work. And this is the way I mean as a status a poster. Join a team that's been successful so you learn how you think and I prepare for this in time and offer to help because this is is a major effort for everyone involved so if you join a team. You don't want to be a burden you want to be a someone that actually brings new energy. So you can help with a lot of things with sample precalization with simulation. The first time I did a beam time I was keeping the logbook I was just listening and making sure I was recording what what people were telling. You know, the, the worst shift that there are, and you can work on the analysis afterwards and learn. And I think we all did it so you get some on experiments and then as a supporter and then before you lead, you did an investigation I think that's that's just common sense, but there's something so complex you really want to get a bit, you know, warmed up by by how it works in And this is a slide that I always use is. I mean, the amount of people is involved in some experiment is something that you cannot imagine. I mean I have one paper with I think 45 authors. It, it's crazy and this was not even one of the first I think now the first with a repeat expert before it close get close to 100 is close to a particle physics experiment. It's very stressful time. There's a lot of people different backgrounds from very technical very scientific is very easy to create misunderstanding frustration, you need to really make sure whatever your position is to try to make the communication clear the language is different But if you do it. That's usually a good way to succeed. I think one of the way that worked in the case of win on is that we always like we also recognize the expertise of different people and we ask for help and give them credit. So I think that's, that's the thing. And I think just to conclude sorry I forgot the conclusion slides but if you really want to get start to get your hands dirty but the thing is to join the, the today's at 5pm. I'm sorry, and the European expel is giving it is a town hall meeting so everyone is welcome so if you just go in the expel.eu and then you scroll down to announcement events, there is a zoom link. And each of these six beam lines will tell you the update what you can do what you're going to do and this is a way to start to learn what you can do because there's going to be a call. I think by the end of the next month, that's going to be in just before Christmas I forgot the deadline but they will tell you more. So, yes, I think with this I forgot the thank you slide so but thank you in person and if you have any questions I'm happy to take them. Okay, thank you so much, Stefano. And I want to invite the audience to watch really thanks, Stefano for his great talk. And it's time for questions. So I think participants can can unmute themselves. Yeah. So, donors. Do you hear me? Yes, yes. Well, that was really great talk. I have never been at Exfel but I've been, well, let's say supporting the teams from home. Very cool. By the way, what is the stages with our own Lund Exfel? What is the optimistic timeline? Yes, it's an excellent question. So I think we are yet a small delay due to COVID like everything but our report is ready. And I think the big difficult thing now is going to be I mean difficult I mean the one that we really need to find a way and to plan together is how to get the funding for eventually the technical machine. So the, so one thing is of course it will take I will say the best. We're talking about four years from now to get it done. This I think is reasonable but also optimistic. So let's say 2025. The good news is is I think, and this is I think a good thing that the community has done. We are not building identical machine. So each of us has his own strength and I have to say so it's pretty clear that a high rep rate machine can do something extremely well but maybe for other experiment. This hour will be a, you know, using the Linux 100 Earths machine. But if you give like in a very stable machine with a lot of pump possibilities. So let's be fantastic. I think the community will receive it very well. So, I will say 2025 and I think something was a bit question is our design. Interesting for the community by then the answer is getting clear and clear that is yes. So we will get. And I think together with Max for and the Swedish expertise in soft express petroscopy I think this is going to be a winning a winning card if we do it properly. Okay. Well, I'm looking forward. We are moving to a science village at about that time so there's will be fantastic. I think it will be a great thing and if we can. I think the cool thing is that if we have like I saw this there's been done with a neutron case the school, having young students young positives wanted to take this I think is a is is fantastic if there is not only like the machine but all the infrastructure around. Okay. Thank you. I've seen a chat. Any other question. If I can ask. Yeah. Yes, who is talking. I'm talking. I'm Chinta Janini from Italy. Hi. I couldn't see now I know you popped up. Okay. Sorry. Just a curiosity. I have been studied a lot. I mean, not a lot. A bit. Quantum imaging in the visible field. Let's see. And I wonder if there is any hope in, in our x-ray field because I work with x-ray but with normal standard techniques. And for example, in the, with this, what you showed for some courier experiments to be done to prove that without the brilliance of the pre electron laser we have any chance to reach. I wouldn't say the same, but something in the direction of the application in the quantum field. It is an excellent question. And I can tell you this is actually what a lot of people want to go and I think there's a lot of potential so the expert. I know I was fascinated by a talk that I think I hope I'm not misspelling his name. If I misspelling then you will cut out for the video. I think it's Joachim Sander in Germany. They did some preliminary fantastic water flash. I think the answer is such a question is yes, it's complicated, but I think it's also complicated because I mean I'm not an expert I'm getting fascinated by this stuff but I think these two communities could actually talk more because it could be that actually we are closer than we think to this. So, I mean a grain of salt that maybe it's not this as you said it's not this as as easy as this high quality as maybe in the with a normal laser. But it could be not that far. I know that some results has been done and I think it was a PRL on this, not too long ago. And I think it would be like, you know, quantum imaging goes x-rays at a physics thing so but we're going to the direction and I think it's an, it's a blue sea. So with the with the x-ray, I mean, not free later on laser I have seen more than one paper. No, sorry, I meant with the with the FL also. Okay, okay. Yeah, I will look for that. Yes. If you send an email I think I have them in my bookmark because this I think this is a frontier that if we open, I think people want to go there. So to do quantum macro science. It's it's where I think oh it's it's an infinite vast sea there I think. Okay.