 So today I'll spend some time on extreme imaging. So it's not so really extreme. So we tried really to push the limits of the standard or the conventional imaging. And it was fun actually. We had quite good results. And at the end, we could say that we enlarged the capabilities of the neutron imaging and also attracted more users coming for or coming for neutron imaging or using neutron imaging for their scientific questions. So it's only my name here, Nikolai Kachilov, but we are a small group at Helmholtz-Zentrum, Berlin, which is dealing with imaging in general and I'll present our activities also in relation to other applications which we have in the group. So I'm coming from Helmholtz-Zentrum, Berlin. It is really very nice area in the southwest part of Berlin. So we are almost the last building of Berlin. So a few hundred meters from our institute is, let's say the border to Brandenburg and Potsdam. So you see that our campus is really nicely located close to Wanzi. This is the Lake Wanzi here around. So we have a golf place here. So you see that when you are visiting us, you should take your golf equipment and then we can use the golf place. So actually Helmholtz society in Germany is has the duty or how to say, is related to large-scale facilities or take care of large scientific-scale facilities and this is also the goal or the purpose of our institute. So we have two large-scale facilities, the scientific or research reactor BR2 and the Synchrotron Ring BC2, which is around 20 kilometers away from our campus in Wanzi. So I'm speaking about two large-scale facilities, but situation has changed just one and a half years ago when actually the reactor was shut down on 11 of December, 2019. So now we have only one facility, but we continue still to be active in the field as users at other neutron sources like the research reactor in Greenville, Eilert Institute, or having experiments at Ephraim II in Munich or some other facilities. So this date, 11 of December, it was quite, let's say important for us because our situation has changed rapidly as you can see in this graph. So this is the power, the chart of the power of our research reactor, which you see has changed at 14 o'clock in the afternoon, so at 2 p.m., so the reactor was shut down. And with this, actually we just changed our activities from operating a facility to be user at other facilities. So you should just keep this also in mind that sometimes the situation can change really like this. I mean, from 100 to zero, I mean, having really some activity in such fields then at some point it can change. So for us, it was not really, really let's say big problem because we started several years ago to discuss what we will do after that and for everybody of us was some solution. So at the end, you can see that the staff or the scientists were not really very depressed from this event. And everybody from us had some further solution for continuing of his career. For us, it went in direction of Grenovo, ILN. So we were able to establish a joint research unit which is focused in neutron imaging. So together with partners from ILN and University Alps of Grenovo, we established this research unit and the aim of this research unit is to operate the existing imaging facility and also to plan a new imaging facility with much better performance for the future. So if we look at our situation before shutting down the reactor, we operated actually three instruments at large-scale facilities. And we, this is the Institute of Applied Materials at the Helmholtz Centrum Berlin where the imaging is actually the main experimental method used for investigation. So we used neutrons from our research reactor before, but in, how to say, in relation to this, we had also, we are having activities of using X-rays for tomography and also we are participating at the tomography beam line at the synchrotron source in Berlin. So here you can see some pictures from the three facilities. And here today I'll concentrate on in the activities which we had at the Neutron Tomography Facility at the research reactor from 2012 to 2019. So here I would like to show some statistics which we had at the facility just to get some impression. What is the, what are the applications which usually we have at such facilities? So you see that the main stream or the main part of experiments were related to material science and energy science. We had some geo-applications, life science, cultural heritage, and some fundamental science related to magnetism. So actually in the, in the, in this seven years of operation, we were quite successful and a lot of number of papers were prepared from our users together with help from us. So you can see that every year we had quite a good number of publications with good impact factors. And the reason for this good performance of the facilities, actually the method development or our, our contribution was to push the limits and trying to extract as much as possible from the experimental side and provide these features or these, let's say, these good capabilities to the, to the users. So here are some highlights from the recent years, so which are related to investigation of batteries, to plant science, magnetism, and here some material research and some research on the use of these materials. This is related to energy application where we investigated electric steels and all these good highlights and papers were able to produce by developing of new techniques which was actually quite important to implement and provide to the users. So you see here that we went in high resolution, in high speed in order to push the limits, we developed new methods using contrast related to a magnetic interaction and we were able to reconstruct the magnetic field as a vector and with new methods where the phase, let's say the phase nature of the neutron was related or was approached, we were able to use even interferometry. So you see here that we're speaking about high resolution, high speed. So we are trying really to push the limits and to get to some better conditions for the experiments of the users. And this was our main contribution. So being so extreme, it is something which we can relate to the Olympic motto. So this is just faster, higher and stronger and if we transfer this to the neutron imaging field, then it should be faster, higher resolution and larger objects. So this is what we tried to do and today I'll focus in all these aspects so starting with the higher resolution applications. So this is one result which we achieved quite a goal. So it's about nine years ago where we were able to perform imaging with resolution of about eight micrometers. And this is quite a good achievement for neutrons and you see why we are very proud with this. So finally, you can see here that we can resolve the resolution pattern in quite good way. So here the distance between the stripes is of about, I think they are 10 micrometer. So the resolution of eight micrometer was quite good achievement. So when we speak about resolution, we should have in mind that it's not only the, let's say property of the detector, but we should also have good experimental conditions for achieving of good resolution. And in case of neutrons, we have the problem that the source is isotropic source. In case of X-rays, it is possible to have really very small spot of X-ray radiation or in case of synchrotron beam lines, you can focus also the beam using refractive lenses. In case of neutrons, it's very difficult to influence the trajectory of the neutrons because they don't have a net charge and using electromagnetic or magnetic fields, it's very difficult to focus the neutrons. So in this case, in order to obtain really sharp image or image with good resolution, we need to use some coulimate or geometry or just point source geometry where actually to collimator, we can reduce the size of the source and using a large distance, we can reduce the geometrical blurring on our detector. So you can see here that this geometrical blurring, which we are obtaining on the detector surface on the detector plane, depends on the distance between the object and the detector and also on the collimator properties, which are defined by the distance of the from the pinhole to the sample L and also the diameter of the pinhole D. So this is the pinhole geometry, which is described here. And we can reduce this geometrical blurring or increase our resolution when we keep very close distance between the object and the detector or improve the properties of our collimator using large distance or small pinhole. Of course, reducing the pinhole or using large distance will reflect, this will reflect to the intensity, which we have at the sample position or the number of the neutrons, which have this solid angle for reaching the sample. And therefore this will reflect to increased exposure time. So finally, using certain pinhole is optimization question, where actually the exposure time should be, let's say feasible for the experiment what we are going to perform. So how to measure the spatial resolution or what is the, let's say, the methods, how we can determine the resolution. So usually if you have a perfect resolution, this means that the beam is really very good collimated or we have a parallel beam. And the detector has infinity capability of taking signals with, let's say, infinity, small pixel size. Then if we are visualizing, for example, if we are taking picture of one edge, then the response will be really very sharp or just this step function. In the reality, it's not the case. So at least we have some pixelization of the detector. So, and also if the beam is not really well collimated, this means that this step function will be smeared. And this D is actually here as shown here is representing first geometrical blurring from the beam collimation. And also the, let's say, the property of the detector or resolution property of the detector where the pixelization is also giving its contribution. So you can see here two different cases. If we have line spread function or edge spread function, actually the measure for the resolution is the edge response here between 10 and 90% of intensity what we are getting here. And this is a measure actually for the resolution. If we are taking the first derivative of this one, we'll get also the line spread function where the measure for the resolution will be just the full width at the half maximum. So this is just one, let's say, impression how the resolution can be measured. So this means that if you have a sharp edge, you can put it in the beam. And after analyzing the profile through this edge, we can get some first preliminary estimation for the resolution. So how we can, or what are the features or what are the experimental conditions which are influencing the resolution? So you can see here if we have such pattern with black and white stripes or with black stripes, then the imaging system will contribute to some, let's say, smearing of this pattern. And there are different contributions which we have. So it can be the lens, it can be the pixelization or the, let's say, the sensor. And you see here if we have a hood or a sufficient distance between the line pairs or these stripes, then finally through the lens and also through the pixelization of detector, we will be able to resolve these two pairs. But if they're very close to each other, then actually we'll have some overlap. And here in this region where we have overlap, this means that we have smearing of the contrast. The contrast is defined just by measuring of the intensity for maximum and minimum when we cross through this picture of these line pairs. And for the very tiny distance between the line pairs, then we have really reduced contrast. And here this is a new method or another method to measure resolution is just to see where or how to say frequency dependent reduce contrast. So this is shown here. If we have 100% contrast, this means that increasing the frequency or the number of the line pairs per millimeter, the contrast level is staying the same. So this means that the levels between high, between maximum and minimum are staying the same. But by reducing the contrast and smearing this image, then we see that with increasing the number of the line pairs, we lose contrast. And this can be seen here. Actually this is the original pattern and after contribution from the detector or let's say the response function from the detector is leading to a smearing of this pattern. And with the increase of the line pairs, we see that we lose contrast or the difference between the maximum and minimum intensity is reduced. And this is the definition of modulate transfer function where actually the reduced contrast with the increased frequency is coming down to 10%. So how this works in the reality, you see here, this is one step H or one H which we visualized with neutrons. So this is tin foil of, no, this was a tin foil of gadolinium which we had in the beam. And we measured this with a pixel size of 15 micrometers. So this is the signal which we obtained just by measuring the intensity. And you see that it's quite sharp edge, but nevertheless, there is a small broadening here. And after taking the empty F or the modulate transfer function, you see that with increase with the increase of the line pairs per millimeter we have reduced contrast at 10%. We have 20 line pairs per millimeter or this is corresponding to 25 micrometer. So here I would like to show you how this works. We actually not produced but programmed one plug-in which we used for the estimation of the resolution at our facility. And usually when you're trying to adjust, for example, your lens and see at what position you have the sharpest image this means that at the end you have a stack of images which are taken different, let's say, adjustment steps of your lens. That's why we developed this tool or this plug-in in order to estimate the modulate transfer function for stack of images. And you can see this here. So it is also given in the it is also the link to this plug-in is provided in the on the Indigo website. So you can download it. And you see here, this is this thin gadolinium foil which we measured at different rotation steps of our lens. And what we can see here is that in the beginning we have quite sharp image and with the rotation steps, the image is blurred. So this means that there is some optimum at some position but it's very difficult to estimate this by eye. Therefore, we can see that we can just, let's say, take one line profile here. You see, we can take this line and then we can use the MTF plug-in. Sorry, it is because I have two screens and it's always displaced on the second screen and I should move it here. So it's very simple. You should just define the pixel size which we have at our detector and it was seven micrometer. And so we want to have the contrast at 10% and the expected line pairs per millimeter is, let's say 25. So this means if you click here, then yeah, then we have the response here. So you see for the different pictures, we have here the line pairs per millimeter and also the calculated resolution. So finally, you see that the second picture has the best resolution and with increased, let's say, when we increase the number of the images, then the resolution is decreased. So we have also the option to see the derivation and also MTF and you see here that it's quite noisy. Actually, the result what we are getting here is based on quite noisy profile. Therefore, in order to increase or to improve the performance and they prove the statistics of our measurement, we can actually use just line profile which is taking more data points. So in image J by double click on the line profile, we can extend the width of this line and this means that we are taking all the points along the line width. So we are increasing or improving the statistics in our line profile. And it is also very important when we are taking this line profile that our line is really, let's say, normal or perpendicular to the surface of the edge. So here it is the case and we can calculate the MTF again. So here, then, actually I should start actually, yeah, let me see. Here, I should define again the pixel size and the parameters. So we have again the seven micrometer and they expected like there's millimeter is about 25. So here you see that we have much smoother profile, the resolution is, let's say, is improved or the calculation of the resolution is improved and obtained MTF is much more reliable. So this means that if you are going really to estimate the resolution, it is good to take more points for the statistics and also to have good orientation of your line profile. So for further training, I'll provide you also the experimental data so you can try this by on your own. And if you have questions, then you can ask me during the days at the school. So there are also some other chance or some other ways to estimate the resolution. And for this, there are certain patterns which are used. So this is the Siemens star, which I showed in the beginning. And you can use this pattern actually for very fast and let's say estimation and qualitative estimation of the resolution. So here you can claim that you have a certain resolution when you can distinguish between the line pairs far away from the center of the pattern. And in the case here, we can say that we have resolution between 200 and 300 micrometer and this was our standard setup in the year 2006. So since this date, we invested a lot in our detector systems to improve the resolution and to provide better conditions for the users. So the main things or the main components which are responsible for the resolution in such detector system as shown here are the scintillator and the lens of the camera which we're using for recording of the picture. I guess that my colleagues has shown yesterday the design of neutron imaging detector system. If not, we are just converting the neutrons to visible white through a scintillator and then we are observing the image on the scintillator through a mirror just by focusing the camera or focusing the picture through the lens. And by just using appropriate lenses then we can reduce the pixel size and improve the resolution. So the cameras which we are using, so this is not advertisement for the CCD camera which we used before because it's no more available commercially. So it's an old piece, but anyway, it's just for your information to have some feeling how the detector looks like. So it is quite bulky device and it is very sensitive for light. Here you can count even single photons. And we need this because the efficiency of the scintillators and the neutron plugs which we have are not really very high. So finally, you are not going to be able to see high, so finally you are fighting with quite a lot of noise, statistical noise and also reduce the efficiency of the scintillator. So finally you need really very sensitive detectors which are unfortunately very expensive. So such kind of cameras costs more than let's say 20,000 euro. So it's very delicate equipment. So for this digital cameras, we are using some commercial valuable lenses where we can change the effective pixel size. So you can see here that the pixel size which is provided from the chip of the camera is about 13.5 micrometers. So there are different types of cameras but by using different lenses, so you can project or you can zoom in or zoom out on the scintillator and then you have effective pixel sizes between 30, 50 and more than this micrometer. And actually this is connected to the field of view or to the area which you're recording. And especially when you're going to high resolution then the field of view is becoming quite small. So if we are going to one-to-one imaging then the field of view is just the projection of the chip over the scintillator area and then the size of the chip is the limited, let's say, so this is the limitation for the area which we can observe with this system. So for the, there are some, let's say, special specialties of the neutron imaging or neutron detection for imaging and the problem what we have is that the neutrons cannot be recorded or cannot be detected in a direct way. The reason for this is that usually the scintillators which we're using, especially in the area in the X-ray field so they're based on scintillating effect. So this means that we have interaction of the radiation with the electrons and due to the change or due to just changing of the energy level of the electrons then we have just emission of the electron visible white. So this energy transfer is just seen as emission of visible white. So due to the fact that the neutrons are not able to interact with the electrons, they're interacting with the nuclear of the atom. We are not able to have a scintillation using the electron type of some material. Therefore we have a two-step process or two-stage process and for this, the first stage is just to capture the neutron and use this nuclear reaction in order to produce secondary radiation like let's say alpha particles or electrons which actually on the second stage can produce scintillating effect in scintillating screen. So these two-stage effect is actually the reason why the neutron screens or the scintillating screens are two components. The first component is related to high absorbing material which is absorbing the neutrons and the produced secondary radiation is interacting with the second component which is scintillating material or material which has high probability for scintillation and then actually the secondary radiation is producing the visible white which we can detect with our camera. The problem here is that these two-stage process is reducing the resolution. So the light is not produced at the position where the neutron is captured but at the position where the secondary radiation is produced and also the light is produced that the secondary radiation is producing the light and usually this diffusion or this let's say diffusion path for the secondary radiation is the limiting factor for the resolution in case of neutron imaging. So here you can see some nuclear reactions which we are using for producing this secondary radiation for the scintillation and the main reactions which we are implementing in our scintillating screens are the lithium, the absorption of neutron by lithium-6 where we have production of alpha particle and tritone and this tritone is actually used for the scintillation. The alpha particles are also producing scintillating effect but the mean free path of the tritone is much bigger and this is actually the limitation for the resolution. So the second reaction which is quite important or the second element which we are using is gadolinium-based scintillators where converse electrons are produced and in this case actually we have not very big mean free path for the converse electrons and these types of scintillators are used for high resolution imaging. So you can see here the process of the detection for neutrons in the scintillating screen. So this is typical mixture of six lithium fluoride for absorber and things will fit for the scintillating material. So you can see here that the neutron is first absorbed in the six lithium fluoride and then we have emission of tritone and alpha particles. And this secondary radiation, the tritone and the alpha particles are interacting with the scintillating material and then visible light is produced. So the photon which is produced here we can detect by our camera. So here are some facts related to the limitation of the resolutions. Actually the mean free path of this tritone is of about 50 to 80 micrometer and this is actually let's say the limitation. So if we use very thin scintillators then we can go to resolution of 20 to 30 micrometer and this is actually the best resolution which we can achieve with this type of scintillators. What I mean with thin scintillators? So you can see here the effect of the scintillator thickness on the resolution just by imaging of the Siemens star and here you see actually the performance of the detector system for different thickness of scintillators and it is obvious that at thicker scintillators we lose the resolution. So also the effect of the geometrical blurring is shown here. So here for the second row here, we use the same scintillator thickness but different distance between the Siemens star and the detector. So at very short distances, we see that we have the best resolution and increasing the distance between the Siemens star and the scintillating screen. The resolution is quite worse. So here actually I can show you some method development related to improvement of our detector systems, which steps we used in order to improve the resolution. So in the beginning we started with scintillator of thickness of 200 micrometer and pixel size of 100 micrometer. Therefore it's not a surprise that we could detect or we could achieve resolution between 200 and 300 micrometer. Of course here the exposure times were quite reasonable or quite good. So in 20 seconds we had already a picture. So when we used a different type of lens where the effective pixel size went down to 30 micrometer, then we see you can see that the resolution was considerably improved. And we were able to go down to already about 100 micrometer. And for the next steps, we tried other type of scintillator which was based on gadolinium material, where the thickness was only five micrometer. And in this case you can see that we could achieve resolution of about 60 micrometer, which is not a surprise because usually the resolution is two times the pixel size because we need two pixels in order to distinguish between let's say between two stripes or line pairs. And for this, this is the maximum resolution which we can achieve with this system. In order to go further, so we should do something in order to change the pixel size to reduce the pixel size. And for this we tried optical magnification. So in case of standard imaging detector for neutron applications, we used just one lens and just we projected the scintillator over the chip by using a lens. And in order to increase the resolution, we tried a different setup where two lenses were used for optical magnification. So if we have two lenses which are facing each other and they are set to infinity, then you have the projection actually from one point through the two lenses on the chip of the camera where optical magnification is possible. Actually, at the end the optical magnification is the ratio between the focal lengths of the two lenses. And if the first lens has a focal length of 105 millimeters and the second lens of 15 millimeter then we have optical magnification of two. So this means with a pixel size of 13.5 micrometer, then we will end up with 6.4 micrometer effective pixel size but with reduced filter view. So here you see that we are ending up with about 13.2 millimeters. So it's about one by one centimeter. In addition to this, we used very thin gadolinium based scintillator and this helped us to achieve much better resolution of about let's say 15 micrometer. So we use this for different applications and one interesting application is the field of fuel cells. Probably you know that the fuel cells are quite, let's say new way or they're really very important for the development in the automotive industry where cars can be driven by hydrogen but it is not burning the hydrogen but using the electrochemical power which the hydrogen has where actually there is splitting of the hydrogen in proton and electron and then this electrochemical reaction is going in a way that the proton is going through proton exchange membrane and the electrons are used for the electricity production. From the other side, we have recombination where the protons together with the electrons and some oxygen from the air produce water. So this is a very clean way of producing electricity and using hydrogen for this. So the output product is water but the problem of these devices first if the proton exchange membrane is becoming dry then the proton conductivity goes down but in addition to this actually if we have too much production of water then, okay let me go to the next slide then this channel will be blocked and we have not enough oxygen for the reaction so the performance of the fuel cell drops down. Therefore, there was an idea from some industrial partner to make some good organization of the gas diffusion layer here where we have two flows. In one side we have water flow through the porous media in order to just get the water out or away from the area of reaction and the other flow is just gas flow of oxygen which we need for the reaction. So we can try some, this kind of geometry some drainage where if we use some hydrophobic material for the porous structure then the water will be forced to follow the channels and then we have free area for the gas flow for the reaction. So this was just a try to organize a bit better the flows in this gas diffusion layer. So for this we performed some neutron tomography imaging in order to observe where is the water at the operation of the fuel cell and for this we needed quite high resolution because the holes or the diameter of the holes was about 80 micrometer. So the high resolution tomography which we performed has shown us that the water is not in the hole but around the hole which was not expected because the material we used for this porous media was hydrophobic and we didn't expect that the water will be accumulated in the material itself. So this result was quite disturbing and we performed some additional study where it was found out that the drilling process with the laser is leading to some overheating of the material and this the heat affected zone changes its property from hydrophobic to hydrophilic. And finally we can have the water actually here in the walls of these channels. So at the end we were able to perform tomography experiment with high resolution synchrotron radiation where the contrast for water is very weak but anyway it was possible to see that actually the water or the distribution of the water for two types of production of these holes with laser perforation and mechanical perforation. And you can see definitely that in case of laser perforation the water is here around the channel and here you can see very well with the mechanical perforation that even we have some droplet and the water is clearly staying in the channel. And this helped actually to increase the performance of the fuel cells by around 40%. So finally the main result or the very important result here was provided by the neutron tomography. So here actually having high resolution is not only observing just using it for a neutron investigations but it is also a big advantage of combining X-ray with the neutron imaging because then we can match the same type of resolution using the two probes. The X-rays can be detected with really very high resolution and in this way increasing the resolution for the neutron investigation then we can have complementary information or complementary 3D data which can be merged together. And in this way we can perform our dual mode imaging with much better capability or much better quality. So here you see, for example, the X-ray and neutron image of sand with water or this is different sand which was compacted and it was interesting to see how the water distribution looks like in different stages of compactation. So the pixel size was similar for the X-rays and the neutrons and performing tomography we can see the complementary contrast that X-rays are seeing the silicon dioxide or the sand corns and the neutrons see very good the water actually between the sand. So combining the X-rays and neutrons, it is possible to perform some statistical study and see, for example, the coordination number for the neighbors for each sand corn, we can see the water or the areas of contacts and also to estimate the amount of water or where the water is presented. So the single tomography using neutrons or X-ray will be not able really to get this system in such a data. So we can go even to some energy application or energy applications examples where we want to see the performance of lithium batteries. So this is lithium or battery which is based on manganese oxide electrode and what is very interesting is the lithium intercalation on the lithium diffusion in the electrode in the manganese oxide electrode. And here we can combine the high resolution of the X-rays with the complementary contrast provided by the neutrons where we can see the distribution of the lithium. Here we used also some innovation where this curved or let's say rolled structure can be unrolled and in this way we can investigate just planar structures which is much convenient, much more convenient for the investigation. So what we can see here is how the lithium diffusion in the electrode. So you can see here with the time here the lithium is the dark area. So this means we have quite low attenuation for X-rays and here with neutrons it's the complementary contrast. We have high attenuation for the neutrons and low attenuation for the electrode, manganese based electrode. And with the time we can see how the lithium is decreasing, the amount of lithium is decreasing, the electrode is expanding and we have some intercalation of lithium in the electrode. This can be studied also with X-rays and neutrons. So you see here how the thickness of the electrode changed with the capacity. And here this is the contrast related to the lithium consumption with X-rays and neutrons. So we have complementary contrast. Unrolling the electrode, we can see also the lithium intercalation with the time. So you can see starting from the top to the bottom, you see that we have brighter areas and this is just the areas where we have more lithium diffusion in the electrode. So this means that combining the X-rays and neutrons, we can investigate such complex systems and even going, pushing the limit of temporal resolution, then it is possible in 3G to investigate the dynamics in such kind of systems. So this is a discharge of this type of battery. And what we see here is how the electrolyte is consumed, how here the lithium is the lithium intercalating in the electrodes. So in case of time-dependent tomography, we can investigate in 3D the process. We can see in 3D the lithium consumption in the battery. So all the changes which are related to the electrochemical process can be investigated or visualized in 3D. Here the time resolution was about 7.5 minutes per tomography with a pixel size of eight micrometers. So here this experiment was performed at ILL with really very high intensity, which is available there. I have a question about the in situ consuming of the electrolyte. How possible is it to let the battery run while it's rotating in the X-ray and neutron beam? Because the battery has to be rotated. Exactly. So there are some devices known as slip ring. So this means you have two cylinders where the upper cylinder can rotate and the contacts, for example, for the battery are guided through some rings which are connecting radially. So in this case, the upper part of the cylinder can rotate and the down part can stay. And then the contacts are connected to the battery with the upper part. And on the bottom part, you have the contacts which are staying static. So in this way, you can have really, let's say, transfer of the contacts to the battery without having the problem of twisting the cables or whatever. Okay, thank you. Yeah, you're welcome. Good, so pushing the limit of the temporal resolution here, the main component is the camera. So we should have really high speed. So this is 30 frames per second or we can even push it to 100 frames per second or 10 milliseconds. So this is very important for such kind of investigation. This is just two component flow of air in water. And you see here, the dark is water and the bright is air because it's just radiograph in transmission. And this study was performed by our colleagues at PSI so they use high speed camera and here with 800 frames per second, you can see the two flow pattern of air in water. So going to 3D, so what I have shown, we can investigate the water uptake in plants. For this, we can use just contrast agent D2O where the absorption is much less for neutrons. And here with 10 seconds per tomography, we can see how the D2O, which is transparent is pushing the light water to the top and we can observe this in 3D. And after this, this tracer, the heavy water can be used for estimation of the activity of the roots. So you see that some roots are losing contrast because the heavy water is uptaken by the roots and then the absorption property of these roots is decreasing. I also have a question here, sorry. So look at these four pictures, the plant wrapped by the water, heavy water. Are these waters, the modeling of the water molecule or are these actually the 3D images we get from the... These are the 3D images which are getting from the neutrons. So this means that every 10 seconds, you have a tomography, so these are 300 images over 10 seconds. So the exposure time was just, I think 30 milliseconds or something like this. And so that's why I wanted to show you how this looks in the reality. So last year, we were able to perform 1.5 second tomography at ILL with their high flux and this is how this looks in the reality. So this is the speed of, let's say, 1.5 seconds over 180 degrees. So this is just for your feeling, how is the speed and that we don't have really a problem with centrifugal forces for the water. So this is quite moderate speed. And then we can investigate the water update in the plant with 3D. So large samples, of course, we can investigate samples which fit in the filter view, but if not, then we can scan through the filter view. This is one sculpture from India. So, and these are the first examples in the history for producing such big casted figures. They were produced somehow in the year 1000 after Christus. And also some long samples like this Japanese bowl. Japanese bowl. So the example with this was, so the bowl was 2.5 meters. So we should put it in some holder and then scan just step-by-step over its, along its length. And finally, we can stitch all these volumes together so that we have a full picture of the electric bowl, sorry, of the Japanese bowl in 3D. And we can see that it's manufactured by different pieces of wood. So one, two, three, four, five, six, seven pieces of wood with different mechanical properties or elastic properties in order to have really very high performance of the bowl. So the next application is to investigate some lead ingots which are coming from Roman time, from some shipwrecks which are sunk around Syracuse. And it was interesting to find out the origin or where these ingots are coming from. This was the way of transporting lead material in the Roman time, which ships from the areas of just where this was gained. And the good thing is that all these ingots were stamped with the name of the manufacturer so that we can backtrace the area where this is coming from. But after let's say thousands of years in the salt water, these ingots are completely corroded. So we can see that there is some stamp here or area of stamping, but it's not possible really to read anything because there is very thick corrosion layer. So 10 centimeter of lead is really no way to transmit with X-rays, but the neutrons can transmit lead quite easily. So this sample is very heavy. It's about 50 kilogram to put it on the beam line and to investigate it was quite demanding work. But finally, when we remove the corrosion layer, then it's possible to read really on below what was the name of the manufacturer and in this case, this was Rossini. And also we were able here in the middle to reconstruct it, his trademark, it's a dolphin. So finally, together with the archeologists, we could find out that it's coming from Cartagena. So at that time, the ship was just sunk after some fire on the ship. And after thousands of years, it was able, we were able to reconstruct where it's coming from. So what is why this figure is so interesting because it's that due to the some primitive casting, it was, it had a lot of porosity. We found out that it has a lot of porosity and even the curators of the museum was we're quite disturbed that it can break after that. So they were really very surprised to see this, but observing this image, you can even make some, let's say, suggestion how the figure was in which position the figure cooled down after the casting because here you have much bigger bubbles than here. So this means that the face was on the floor and with the time, actually it starts to solidificate where the large bubbles went to the back. So here there were speculations that the old Indian manufacturer used this on a purpose in order to make it much more light for transportation because they use these figures in some street, how to say, manifestation or whatever in some religious days for religious purposes and then having this porosity, it makes the figure much lighter. But now the data is still under processing and we are in contact with the colleagues from the Eek Museum in Amsterdam. So I'm over the time and with this, I want to thank you for your attention.