 Okay, thank you, Dina. Thank you very much for the nice introduction. I would just like to mention that we have been developing Bragg Tyco also together with the group of Stefan Roskevich. So we are working mostly hand in hand since the very beginning, most of the time together and very good relationship. And of course, I would like to thank you, Dina, for inviting me to this seminar series and for insisting in having answers from me. I was really, really lazy in answering and I'm very pleased to be here and I will try my best to follow the next seminars. Yes, I will. I would like now maybe to share my screen. It should be here and there. No, sorry. Yes. So I guess you are, you see my screen, right? Do you? Okay. So I think I can start. It's a bit bizarre because here nobody's answering, but I trust that you are all online. So good afternoon, everybody. I'm very, I'm very pleased to see you. I've recognized many names and I'm very happy to show you some new slides. I try to make something a little bit different than usual, mostly based on questions that I had during these talks about Bragg Tyco Graphy. You will see that I'm assuming here that you are aware of what is Tyco Graphy and Coherent Infraction Imaging. I will not start from scratch, but I will insist on some very specific points related to the Bragg Tyco Graphy approach. So as Dina said, first bug. Yes, okay. So as Dina said, we are a group from Marseille. The team group is the comics team. And you can follow us on Twitter if you want to know what our daily science is made of. So these are the comics member and former members that you can then you can discover here. So a very interesting point about the comics group is that we are a very, we are a very small team actually, but we are gathering very complementary expertise in Coherent X-rays, in inversion problem with Mark Allen, and also in optical microscopy. And actually, when you look at the members of the group and you consider who is doing Tyco Graphy almost on a daily basis, you can see that almost everyone in the team is involved in the Tyco Graphy either for the development of the method or for the use of the method. So we are working in this field for quite some time and on a regular basis. So as expected, I will talk about Bragg Tyco Graphy with X-rays and why we are using this to investigate crystalline materials. But before I start to focus on X-rays, I would like just to mention that we are not only developing X-ray approaches, but we are also developing optical approaches and optical approaches in Tyco Graphy. And in particular, optical Tyco Graphy but meant developed to investigate materials which present an anisotropy, an optical anisotropy. So these are typically the refrigerant materials such as the prisms that you can see here on the left, which are a part of an oyster shell. And these prisms, they are B-refrangents as exhibited here by this cross polarized microscopy image. And with this optical vectorial Tyco Graphy that we are developing, we can extract a different series of optical parameters which can tell us a lot about the structure of the crystal which is produced by these animals. So I will not talk about optical Tyco Graphy here, but you can certainly go to these references here, Baroni 2020, and have a look at the references here, if you are interested by taking into account the vectorial aspects of the diffracted field to say it a little bit technically. Let's of course now start with X-ray bright Tyco Graphy. And I would like just to mention or to start as an introduction by defining why you may want to bother or to use Bragg Tyco Graphy. So Bragg Tyco Graphy is based on Bragg diffraction, so it is using the Bragg Law. So it means that you will use the diffraction onto your crystal lattice based on the application of the Bragg Law. And once, so it defines already a typical class of material, crystalline material, but however the crystalline material that you can and that you want to image, you really need to define why you need to image this material, but once you are sure you want to image it, you may have to make sure that the mosaicity of this material of this crystal is typically in the 2-3 degrees range. What you will get is a three-dimensional image with a spatial resolution typically in the 10-15 nanometer queue. You may expect to have retrieved a field of view which is rather large. Typically you can expect to go up to 10 micrometer large field of view. However, the sample thickness is still limited and it starts to be very difficult to probe sample or to image samples which present a thickness larger than 1 to 2 micrometer, let's say. So here on this series of images you can see samples which are well suited for this Bragg Tyco Graphic approach. On the right here you have samples which intrinsically present the correct geometrical properties to be imaged with Bragg Tyco Graphic. So typically they can be either intrinsically thin or intrinsically long. This other sample that is presented here has a specific shape that you do not want to disturb because the shape is directly related to the structural and to the physical properties of the sample. So this is also a kind of sample that you may like to image. This is a crystalline bridge, a germanium bridge, which has been patterned on purpose. And finally another typical class of material or geometry of material that you may like to image are samples which are specifically prepared for transmission electron microscopy such as this Sin lamella that you can see here and which is quite well suited for Bragg Tyco as well. So just a few words about the content of this talk. So as I mentioned already I will present some principle of typography assuming that you already know a lot about current diffraction imaging and forward typography. And here I have a warning there will be no typography inversion algorithm in this talk. I guess you have heard a lot about that already. And then we will see how to go from typography meaning forward typography to Bragg typography which are the important things to keep in mind to consider. And then we will see some applications in material science. So let's start with the principle of typography and as a school reminder I would say let's remind us what's happening when sample is humiliated by a by a beam. So if you have if you if you eliminate if you shine a beam onto your sample and you are lucky enough that there is a transmitted or diffracted beam going through your sample your sample is not totally opaque. The interaction between the material and the beam will create a perturbation in the field and this perturbed field this disturbed field will propagate until it reach if you want to do a microscopy if you want to consider a microscopy approach. This field will enter or will be catch or will be processed by a set of lenses by a set of by an objective for instance a set of lenses and these lenses typically for instance in optics will allow you to get back to to produce an image a magnified image of your sample and in optics the resolution of this optical setup is directly defined by the numerical aperture of your set of lenses and in optics this numerical aperture is very often close to unity meaning that you are very close to reach the the resolution limit of your instrument or of what is allowed by the wavelength. So this lens is actually as I said processing the diffracted field with x-ray the situation is quite different because you have to deal with optical setups which have a much less focusing properties power and typically you have to consider numerical aperture in the 10 to the minus 2 to 10 to the minus 3 so as a consequence your image the resolution of your image is strongly degraded. So in order to overcome this limit imposed by the numerical aperture of the x-ray optical of the x-ray lenses or an x-ray optical setup we have to consider what is happening here in the diffraction plane. So to understand what is the information contained in the diffraction plane let's start again from the object and to this object is associated to a diffracted field that you may calculate or anticipate by applying some propagation law to the field which is executing the object. So you have your representation of this diffracted field it is composed of amplitude in a white black scale and phase on a color scale and if you are able to apply back propagation laws on to this diffracted field you will retrieve a faithful image of your sample. During an experiment we are actually measuring the intensity of this field and the intensity is rated to the field through a square root relationship and if you apply this back propagation law to this quantity unfortunately you do not learn a lot about your object you have some information hidden here but it's not the image of your object this you will you will you will agree and understand. So all comes from the fact that you have lost the phase and this is exactly what it's all about retrieving the phase of the diffracted field. So ptycography has been proposed as a mean to retrieve this phase like many other approaches like holography or current diffraction imaging and in order to understand why ptycography allows you to encode the phase information I would like to start with I would like to show you what is happening during a ptycography experiment. So during a ptycography experiment you know that you have a probe which has a finite support a finite extent and here in this example my object is defined as a two well defined scattering particles and I will scan my probe across the object and you will see here on this little movie the probe scanned across the object and here the intensity pattern corresponding to each probe to object position and here the series of intensity patterns. Let's have a look of on how it evolves. So you see that the intensity is changing a lot when the probe scans through the object and there are there are different positions which are interesting. So for instance here at this position you see that the intensity pattern is decorated with very high frequency fringes which are just related to the distance between the scatterers these are this information here and when you are here for instance you are only sensitive to one of this particle and to its shape factor and when the probe is here you are sensitive to the other particles so you can see that from the set of intensity pattern which are all here you can really anticipate or you can realize how much information you have. So I'm used to present this slide at talks and often at the end of the talk I still have the question yes but how it works and I realized that it was a very difficult question to answer and maybe a simple way to explain how ticography works is to have a very very simple system and to try to give you a simple description of the of the ticography. So you know again we we will use a typical very typical system this time the object is composed of two scatterers but these are now pure Dirac functions so point like source and each of these scatterers are characterized by their scattering powers f and n prime f prime and they are located in plus of minus r zero being the the mid distance in between these particles and my probe is shown here it has some specific characteristic like again this finite support characteristic and the distance the size of the probe the width the extent of the probe is much larger than the distance between the scatterers and I will probe I will scan my probe along three at four three different position the position minus r one which is shown here zero the origin and plus r one so there is a typo here so at position r one minus r one I can express the exit field as the product of my probe times the scattering power of the particles so this is this expression here f times the probe located in minus r one times the Dirac function corresponding to the minus r position and so this is the exit field this is the field which is just produced by the interaction of the probe onto the sample and now this field is propagating and if I'm assuming that my detection is in the far field regime my far field so the field which is diffracted very far away from the sample can be expressed by the Fourier transform of the exit field and it has a very simple value in this case which is shown here and during my experiment I'm only sensitive to the intensity so I take the modulus square of this quantity so we have here the scattering power of these particles times the probe in r one minus r there is a symmetric position here in when the probe is located in r one and I have very very similar expression here for the exit field the far field and the intensity and when the probe is located at the origin so my two particles are illuminated together so I have to consider in my exit field the scattering from the two particles and this is this expression here which is slightly more complicated again I'm calculating the far field from the Fourier transform of this expression and I have also a more complicated expression and the intensity which is also a little bit more complicated but this intensity expression is very very interesting because it shows you something important it shows you that you have an oscillatory part here which is this cosine is two q r and you have a pre-factor which is a function of the scattering power of the particles and the position of the particles and this oscillatory part is indeed very important because from the frequency or from the period of this oscillatory part you will be able to extract the distance between the particles and knowing the distance between the particles you can inject in the two other equation that you have here and extract the scattering power of the two individual particles so you can see that by scanning this probe onto the sample you have you produce some intensity pattern and this intensity pattern they have an oscillatory part which is directly sensitive or directly related to the distance between the scatterers and the amplitude of this oscillation is related to the scattering power of the of the particles so in this case we understand how this simple problem can be solved and I hope it gives you a little bit of flavor of how tachography works so now I'm done with this introduction let's see how we can go from for what tachography to bright tachography so as a starting point here I'm just reminding you a very famous and well-known paper published by the Swiss the C-Saxx people at SLS so you have your sample here which is placed on the on the stage you have your detector here with which you are measuring the the diffraction pattern in the far field regime your probe here is smaller than the sample your diffraction pattern and all your setup make sure that you are in current diffraction regime and in order to get or to access or to build a set of of diffraction pattern you will move your sample across the beam along this and that direction across the beam and for and you will do this for any orientation of the sample theta along this orientation axis here and now you want to know or you will go from the detector plane to the sample so the detector plane is characterized by a two-dimensional reciprocal space parameter which is called q and which is defined as the exit wave vector minus the incident wave vector vector and so the intensity in the detector plane corresponds to the modulus of the Fourier transform of the exit field these are known properties and this exit field is recorded for all orientation of all theta orientation the exit field is given by the probe located in the rg's position times the project times the object which is a parameter also of of theta theta being these tomographic angles and so from to to get uh so you have all this projection and uh the object is uh expression is given here so for a given orientation and theta you are integrating all uh delta and beta along the beam pass so in order to retrieve your three-dimensional object what you will do is that you will take all these projected objects as a function of theta for all this tomographic angle and at the last step of this typography approach is to perform a tomographic reconstruction with all the objects for all these theta views so for in a typography in a forward typography your sample is not is not a crystal is not a crystal and therefore the intensity is distributed around the origin of the reciprocal space when you want to do crystalline microscopy with typography you have to consider that your material has some internal periodicity which is the crystalline which is the crystal periods or the crystal lattice in the in the material and this internal periodicity result in the series of Bragg peaks that correspond to secondary maxima located at very specific position in the in the reciprocal space and when you want to do Bragg typography you have of course to be close to one of these Bragg peaks so in order to be close to one to this Bragg peak you have to use what is called a Bragg geometry which is slightly different from the forward typography so we are now slowly entering into the specificity of Bragg typography so the Bragg geometry is defined as follows you can see here that in order to catch one of these Bragg peaks you have to make sure that you are in Bragg condition so it implies that your Bragg vector the modulus of your Bragg vector corresponds to the reverse of the lattice interdistance you want to probe and you have also to make sure that your Bragg vector is perpendicular to the lattice plane you want to measure so it's it's impose a certain number of geometrical constraint on to your your measurement and typically if you are in a reflection geometry you have your wave your incident wave vector here your diffracted wave vector or exit wave vector here and when you plot everything in the reciprocal space you can see here your detector frame which is crossing the or intercepting the intensity distribution along this plane shown here in gray and again this detector plane can be described by two parameters called q1 and q2 here which are the reciprocal space coordinates so if you want to do a Bragg typography microscopy approach you need to measure all three-dimensional free component in this geometry so a straightforward way to to imagine how to do that in in in the Bragg geometry would be of course to transpose exactly what has been done in forward typography to Bragg geometry meaning that you want to do a tomographic approach you want to perform a tomographic approach of your problem so you will try to measure for all a theta orientation of your sample where your sample is rotated around an axis parallel to the to the Bragg vector for instance so you want to record all the intensity distribution so it's this would be the straightforward application of forward typography and you will be able to retrieve all the projection of your object and using all this projection you could do this tomographic reconstruction we have just introduced so there are there are some weaknesses here or some difficulties here of course this is not an easy experiment to do you have to make sure that you always stay at the same at the same Bragg vector position and although when you are doing that another problem is that you have some missing information there is a part of the diffracted field that you will not be able to measure because your detector is inclined with respect to your rotation axis so another way to get access to the three-dimensional Fourier component of the of the of the object is actually to exploit or to take advantage of the specific geometry of the Bragg geometry the specific properties of the Bragg geometry so so this is what is usually called a rocking curve an example is shown here so by just making a small rotation around this axis here which is defined by this omega angle I am able to translate my detector plane across the diffracted field so with just only a very small angular range I can spend a very large reciprocal space intensity distribution because this little rotation transforms into a translation there is a difficulty here which is related to the fact that this is not describing an orthogonal space but I hide it somehow and I and I will not go further into this description which which which can be sold and which is not that difficult but please consider the following now we have catch how we have all these series of intensity pattern measured or acquired along the rocking curve how can I get from this set of two-dimensional diffraction pattern how can I extract my three-dimensional information so this is where things are a little bit more tricky than in the forward typography first you can use a free synthesis approach where you can write that actually all this intensity distribution can be gathered into a three-dimensional matrix which corresponds to this different plane here and this is when and where the q3 parameter is given or the q3 coordinate is given directly by is directly related to the rocking curve angle or to the rocking curve tilt so this is very very powerful because the whole typography problem can be treated as a three-dimensional problem directly however the difficulties that you have to make sure that your illuminated volume is constant along the rocking curve it is it is quite true if you do not suffer a lot a lot from drift and so on but it can become tricky if you if your setup is not that stable another approach is actually to use the back projection approach so you will consider in this approach that all diffracted intensity pattern can be associated to a projection of the object exactly like in the case of the tomographic approach but this projection of the object is somehow weighted by a phase expression which is given here this exponential iq3r which is given here which corresponds to a modulation of the sample by this modulated part and by summing up all these projected objects and their corresponding modulation you can retrieve the full three-dimensional object and this time you do not need anymore to consider that the project that the object is a whole three-dimensional object treated as a whole all projection can be treated individually and if needed they can be considered to be taken a different position for instance so everything is described in the in this paper here that we have proposed a few years ago with with Stefan Roskevic and this group it's a very powerful approach we are keep on we keep on working on it and developing it and it's this is something which is now becoming a much more standard in our development in our application so let's continue about bright tachography again very quickly some information about how to introduce the strain the strain field so here you have the full expression of the tachography problem where the exit field is related the sorry the diffracted field is related to the Fourier transform of the probe times the object and if I consider that the crystal the atoms of my crystals are not ideally placed but on the on the contrary they are misplaced by a small displacement vector u introducing this result in this expression with which is shown here which contains everything so you have the probe you have the crystal like electron density or the crystalline electron density and you have here this phase terms which contain the displacement displacement field within the crystals so this expression is really the the expression that we are we are using all the time to describe the relationship between the intensity and the object together with these crystalline properties so just before finishing on this part regarding how to go from forward tachography to bright tachography I guess although I didn't mention it clearly I guess what you have to maybe something we can have a look here we can here see when you look at this expression you have to you realize that actually the object is a three-dimensional object is rated to the intensity really has a has a whole three-dimensional object so it means that you cannot really do tachography in a two-dimensional manner it's it's kind of complicates not complicated but it's kind of dangerous and there are some exceptions to this problem so why that when you are looking at the quantity that you are measuring in the detector plane this quantity the exit the diffracted field in the detector plane is related through the Fourier slice theorem to the projection of the object times my probe but this quantity which is a 3d quantity is projected along the exit direction okay and what you would like to do if you want to do a fast or quick two-dimensional approach is that you want to use an expression which is something like that a function which is only depending on the probe and another one which is only depending on the object and you want these two functions being separate however this this expression here is not equal to that expression which is shown here unless you are allowed to make some approximation which are valid in two specific cases so you can do that if your exit direction kf is very close to your incident direction meaning that you have a very small brag angle here or that kf is almost collinear to ki this is a way to solve this problem or you can also do that do this use this kind of expression if your film thickness is much much smaller than the beam size there has been example in the literature of such use or success of two-dimensional brag typography and i want to show you two of them here so in this example the takahashi group is using a two-dimensional brag typography to probe dislocation strain field in a silicon film they are working in for in transmission in a transmission geometry and they have a very small two two theta two times the brag angle the brag angle is small and therefore they are allowed to use this this expression and to make this use of the 2d approach another articles or result where this has been also successfully employed or exploited is this work here from the the group of stefan where they were investigating polarization domains in ferroelectrics and they were recovering both amplitude and phase of these domains and they could use this two-dimensional approach because the film thickness was extremely small typically in the 25 it was a 25 nanometer film thickness very very small and the this approximation is is quite okay so i'm done now with the with my with the part regarding how to go to brag typography so i'm assuming now that you are a little bit more familiar with what you can how you can go from forward typography to brag typography and just before we finish this part i would like to to give some number so special resolution 10 to 30 maybe 15 nanometer the resolution is very often an isotropic you have to pay attention to this strain sensitivity that you can catch or that you can measure they are typically in the 10 to the minus 4 to 10 to the minus 2 and the lattice rotation sensitivity so the tilt that you can measure are typically between 5 but few 10 to the minus 3 degrees to a 2 1 degree these are also numbers that you have to take into account the field of view and the thickness we already mentioned the acquisition time so there has been a lot of progress from 10 hours a few 10 years ago we are now able to measure a full brag typo 3d brag typo scan in a couple of 10 minutes or so the probe size that we have been using typically a range between 18 nanometer to to the micrometer range the translation step size has to be shaped accordingly angular step size they are they must it must be designed so that it fulfill coherence condition so it creates a lot of steps when you are when you need to use such small angular step size like a few milli degrees and of course the the size of the data is very very big and the inversion time is something on which we haven't progressed a lot recently so it's still a few hours to a few days so I'm showing here some example of bright typography reconstruction which has which I've been obtained recently at different secretron and sls 2 id 1 beam line id 13 beam line at esrf and also some results obtained at nano max at max 4 um let's uh let's see now what we can do with typography um I think I'm going to be here a bit late Dina is it how much time do you I think it's fine go ahead for another 10 minutes how are you planning I don't know I will I think 10 is too long for for the first system I will skip this one and I will just go to this second system which is which is a result that I've shown already a few times but which is a kind of interesting and as you mentioned it in the introduction denied I I I rather present this one so we have used a bright typography different system this one that I want to show you now is a biomineral system it's a pearl the paradigmatic pearl oyster shell so why so so this biomineral as by definition is a heart tissue which is produced by an animal uh it's very intriguing it has a single crystalline like properties uh at the same time the animal is able to produce different crystalline phases for instance here calcite and also the necruse part is aragonite uh and you can see here these prisms which compose the external part of the of the shell and these prisms they are they are they present a single crystalline like behavior as exhibited here with a cross polarized microscopy but also even more visible here on this x-ray diffraction pattern where you can see the diffraction here the origin of the reciprocal space and all these diffraction peaks which are extremely isolated so these diffraction peaks correspond to the integration of the the diffraction pattern onto the hole onto a complete prism and you can see that they they stay very very strongly well defined so these these systems as I mentioned they are very intriguing because they they are able from environmental room temperature water and so on they are able to produce crystalline phases which in geology geology are not expected to happen it's a far for equilibrium growth of calcium carbonate crystalline polymorphs it presents various microscopic shape but what is very interesting is that at the very local scale at the submicrometric scale um you observe systematically in most of the most of the biominerals you observe systematically a granular structure shown here which is an organo mineral structure with typical granular granular granular granular size in uh below this micrometer range between 50 to 200 nanometers and the presence or the observation of these systematic granular structures indicates that biominerization presents some kind of generic mechanism so there is some generic mechanism in biominerization and we are interested in understanding this generic mechanism it also shows us the landscapes at which to focus the landscape to probe in order to understand this mechanism so this granular scale so in order to progress in the understanding of biomineration people are either doing post-mortem studies looking at the structure of these biominerals with chemical or crystalline approaches or another approach is of course to try to reproduce these biominerals from the lab by doing in vitro synthesis and the advantage of in vitro synthesis is that you can play a lot with the with the component and with the parameters the chemical and physical parameters and from this in vitro synthesis there is a general consensus which is actually the fact that these biominerals or the bio crystallization or the crystallization under organic mediation is likely resulting from a particle attachment concept there are different pathways which have been proposed in the literature and all these pathways involve the presence of particles so the nature of these particles depending on these pathways can be very very different from one case to the other they can be either liquid droplets amorphous particles polycrystalline particles or single crystalline particles and they can they also follow very different pathways in terms of crystalline or phase transformation some of them corresponding to a very smooth crystalline transformation where the whole amorphous solid which transform into a into a bulk crystals or some other transformation where your particles already in a crystalline state attach to each other on specific crystalline phases with more or less defect in between and looking at these different pathways at this different proposition that has been done in the literature you understand that it's very important to have very detailed view of the crystalline properties of these granules for instance if granules are strongly misoriented this is the sign of the single crystalline properties of the individual particles as well if you observe a lot of stacking fault in between two granules this corresponds likely to an attachment of the single particles on these crystalline phases if you have a lot of strain developed into these particles you can likely conclude or interpret that you have a lot of organics molecules so clued into the material the crystals and another important part to consider is the coherence the crystalline coherence so if the crystalline coherence propagates through several granules it corresponds to the existence of a continuous crystalline gross front which has propagated through different granules so we have investigated this biomineral this pearl oyster shell which is called pintada margaritifera in its juvenile form shape so you can see here the shell the oyster shell and a zoom in onto this shell show you its structures it is composed of these little tablets which are called prized prisms these are these single crystalline units and lucky us they have naturally a thickness which is suitable or which is compatible with a bright dichography so we have done this experiment at id 13 just by taking a small piece of the of the shell putting it on a needle and shining the beam on it so what I will show you is the reconstruction we obtain on a very small area of the of this prism so the the data at that time were very long to acquire it so we could only measure the very small area of the sample it's a one micrometer square area but we are going through the whole thickness and what you get is is shown here so in this small probed volume which is shown here in in yellow and here in gray we could identify a series of core oriented of iso oriented crystalline domain each color here encoding the specific orientation of these domains so each each each domains in in 3d in volume is defined by a specific orientation angle and you can see that they are uh miss oriented one with respect to to the other uh within a range of a few 1.5 1.8 degrees so this already shows that uh this single crystalline behavior that we were used to to to read or to to consider is actually not that true the prism is composed of several miss oriented crystalline domain these crystalline domains being however very uh very closely oriented one with respect to the other another important information we could extract is related to the crystalline coherence so with bright typography you can have a look on the properties of the lattice of the of the crystalline lattice and you can see how the crystalline lattice self-replicate when along a certain direction and if the crystalline lattice is unperturbed is homogeneously replicated the crystalline coherence is is very long so this is what we have we are seeing here in each of these iso oriented domains they contain several granules but the coherence the crystalline coherence is much larger than than this than one granule size one granule width so taking into account this information we could already identify some some specific pathways based on the in vitro fantasies approaches or proposition that have been made so either the crystal is produced uh by an association or an agglomeration of single crystalline with single crystals with different orientation which are partially fusing to co-orientate to co-orient or the the the crystal is produced by something which is more like a liquid amorphous droplet precursor which transform into a crystals so i i think i'm done with my all the the the content of my talks and just before i finish here i would like to give a few words to say a few words concluding words and perspectives so as a conclusion i would say that bright tachograph is challenging yes but doable and definitely worth the pain you will learn a lot of information of on samples that you cannot imagine by any other mean and i i think it's it makes it it it it makes sense to to do it if you have a very specific scientific question that you cannot address with other uh a crystalline oriented um microscopy approaches um something else uh the future of bright tachograph is uh kind of bright we have uh we are very pleased to see a lot of uh new or upgraded secretion sources and in most of these sources there are uh beam lines which are able to welcome bright tachograph experiments and we have been uh able to do a bright tachograph experiment at all these places here that you can see so max four uh es r f old and the new uh the the the new ebs soley uh ap s and nsl s2 i'm aware that there are other places where you can perform bright tachography but i'm i'm just considering here the the the one which i know better and please forgive me if i if i miss some other other places that you are most more familiar with uh regarding some future direction uh some things that i didn't mention a lot but uh which is inherent to bright tachography so far is that you need to know very well your probe this is a prerequisite to tachography why that is because with this tachography uh scan you are only scanning in 2d you are not able to scan in 3d or your probe is very elongated in along the third direction so it would not make sense to scan along the third direction but still you want to retrieve a three-dimensional information so getting simultaneously both the probe and the object is very very challenging so we have worked on this on this difficult point over the last years and this is a result which has been obtained together with the University of Oxford with Felix Hoffman and Nick Phillips and this is a result obtained by Penguin who has a postdoc with us since a few months so you can see here the reconstruction of the object the probe here and the object itself and what is i guess a little striking here is that the quality of this image is very very uh uh nice in the sense that it's very smooth there are not a lot of artifacts and you can distinguish both a large distance train field and also a very localized displacement field which are related here to this uh dislocation another difficulty of tachography is that it relies a lot on the stability of the setup uh these are experiments which are which requires to move to move a lot the samples so we are very very uh we have a strong interplay between all this motion of the samples and the need of stability for stability it degrades the resolution it adds inconsistency in the data set uh so it's it's it's it's very difficult uh however this is also something on which we are working and we are now able to retrieve the beam position uh corresponding to uncertainties into the into the jury obtained during the experiment and i'm showing here a result which is not yet published it's a work in preparation also obtained by a pangli together with a with dina carbon collaboration with the nano max beam line and you can you can probably appreciate the uh the quality of this reconstruction and just to give you a flavor of what it would be if we would not have recovered or retrieved the the beam position uh you can see here how degraded the image would have been without this this new new approach and finally uh something else on to which i think we should uh we should work specifically now we are using force generations including sources uh is the fact that samples are supposed to evolve uh during an experiment doing statics investigating static sample is is is kind of fun but being able to follow a sample which is evolving uh either because the crystal is growing or because you have some phase transition or because you are doing some misery to the sample i think it's uh it's a very very nice uh perspectives uh you may also face the fact that your sample is damaged under the beam from radiation and actually we have already started to work on this uh on this along this direction and uh we are more or less convinced we are convinced that there is probably no redundancy in the problem as long as the problem is well shaped or well defined and i'm showing here you some numerical results also obtained during the the postdoc of of pangli uh in uh in marseille where you can see uh that we could retrieve different states of an object which is evolving uh during an experiment so these are preliminary results and we are probably we are we want to continue along this direction so i think these are my final remarks i would like to thank all my colleagues who are working for so many years with uh with us on this uh on this project on this development methodological development and also application of bright technology so of course the people at comics uh a lot of people also at esrf who has been uh very welcome welcome us very often and also dina who is uh with us since the the very beginning of this uh bright technology uh journey so i thank you very much for your attention and i would be very pleased to answer any question if uh if you have some thank you thank you so much virgini that was really great uh we do already we have already a couple of questions some very specific and a bit more generic so i start with a question from angel i don't see the question dina shall i i can read it to you okay so why can you can you specify the limit of the thickness you were talking about one to two micron thickness for brag typography can you can you specify this bit more so uh i think this is a limit which is not uh which can which can be somehow pushed further although we have not made a lot of progress along this direction so i would stay i would say as a safe answer you have to make sure that uh your optical path lens difference uh is smaller than your coherence lens so basically you want to make sure that you are still working in coherence condition and uh this coherence condition uh the coherence lens condition applies very very strongly along the the the thickness of the sample so it depends on your brag angle it depends on your energy of course and and this one to two micrometer is just a typical range of my typical number but it's not it has to be calculated specifically for for your system but just to fully answer this thickness condition is exactly the same as in brag cdi so here this constrain is exactly the same as brag cdi so you can refer to brag cdi paper to have the exact exact relationship i believe in the future that we can get rid of this somehow of these limits by taking into account partial coherence effect but this is not something we have uh we have made a a lot of progress recently so this is still something to do uh yeah so i would like to add that anhel is working xx fel and he just sent a message to everyone the beam line mid is open for a ticography experiment so there is one on the list i'm sorry i didn't show the the result that we are obtained with the with the under smarts and and and and and shen so uh then there is another question from manuel from psi and he asks are brag ticography online reconstruction offered to general inexperienced users of beam lines that you have visited yes please believe me so manual thank you for these very nice questions i have to say that having been at your virtually at your at your at your instrument i really envy the fast the the the way you do ticography now or that the way you do ticography so we have so this was the aim of two experiments we have done recently at esrf with the new ebs with the new ebs facility so we have done one experiment at id one and one at id 13 where we try to retrieve the online so online meaning uh basically a few couple of hours after the experiment is finished or let's say just after the acquisition is finished and we have been partially partially successful at id one on at id 13 we we didn't have enough time to to decipher everything and we were not that that happy with how much we could achieve so this is really something that i would like to to to push further as much as possible i can see that this is not totally in my hand and i'm pretty sure that there are some there will be some additional efforts made by esrf and by Vincent Favon Nicolas to push further the possibility to do a bright ticography reconstruction online but at least we have shown that within a very reasonable amount of time like one hour we could get already some images not as nice as uh or as fluent or as easy the way you do it at c-sax but being inspired by this actually yeah i would like to add also when uh moving uh when the joining max four we had this long-term project of implementing actually themes that were available for users and then somehow the plans got a little bit changed but maybe nano max still has this actually i see alex bierling online if he wants to comment on this is welcome but meanwhile i see there is another question from carlos satudias from campinas hi carlos and he says when you make the bridge between bragg cdi to tico for retrieving the 3d diffraction can you also scan the energy of the beam is that yeah yeah it's very nice why you know if you cannot scan the energy of what is yes it's very it's very nice question as well um i think i should really consider having three hours talk now uh so we have done this uh so at max four together with dina we have acquired the energy dependent to ptycography scan so um yes scanning the the third direction of the reciprocal space with the by changing the energies uh and we have already very very beautiful reconstruction uh on this star i have silicon star i have just shown uh unfortunately we still there is still in the data set something we do not understand and we are kind of stuck here so so yes it is definitely possible we have shown we we did it and we have very good data but there is still something in the data that we don't understand uh which makes the image some thing um in the image something that we we don't understand so so far we don't show the the results so yeah thank you we have another question from ash three party yeah he says it's probably too technical but for difficulties in probe and sample recovering in bragg tico graph have you tried additional regularization or constraints to yes that's that's the trick actually exactly this is how we we solve the problem we have to add regularization so the the trick so this paper is is uh is submitted i don't know if it's available online or not i i don't remember i guess yes but i'm not sure uh so it should be available soon i hope uh so to retrieve both probe and samples simultaneously we have to use two uh two two constraints first we assume that the probe is not changing along uh when it's it it goes through the sample along the the sample intersection so basically the probe is self identical along the the propagation within the sample and this is pretty pretty reasonable as an assumption because our samples are always very very thin or thin with respect to the to the depth of focus of the probe and the second constraint that we are using is that uh we are regularly regularizing the the reconstruction onto the thickness of the sample so the probe cannot be everywhere it has to be contained into the sample yeah um so we have another question comment uh coming from dda could be yeah the broad issues that you have in the reconstruction with the energy scan could they come from higher order contamination i mean uh i i don't know it's uh maybe we should maybe we should consider again the data with this uh with this hypothesis i i i do not see anything in this in which goes in that direction so just to i i can explain briefly what we have as a as a problem so typically we did this energy scan during you see my screen right yes yes so we did this energy scan on a sample which is very similar to that one and what we retrieve is uh exactly this sample but the top parts and the bottom parts are shifted one with respect to to the other so both parts are extremely well retrieved but there is a shift in between these two parts and we don't understand whether it comes from uh the scanning stage that we form for which we are retrieving the position or from something more fundamental and we have not been able to to to go to go further as there has been a lot of efforts and pain and tears on this problem especially especially from from pain i think that this is also true for cd i isn't it there is a just single single inversion non-necessary dichography or so it's not an inversion it's just that the top part here is shifted by maybe uh i don't know 100 or 200 nanometer with respect to the bottom part in its own plane so like so like it's retrieving very well one part of the sample another part of the sample but he does not know that these two samples belong to the same data set so i don't know questions from the audience either you write them or you can unmute yourself and ask them in person or you can raise your hand and i can see it in the list um i have a question virgini i mean you have uh uh this special position you and your collaborative team uh to visit many facilities you do very specific experiment you have a very special expertise and you know we are now talking about information sources no no don't worry i'm not going to ask you what's best this is for private conversations but i'm asking you as a user of this very also very difficult technique first of all how do you see that dichography can really benefit from these brilliant sources i mean let's talk you know in a real in a practical way also to get it yes what manuel was saying yeah users can not be only for yes hyper specialists will for generation synchrotron help and if yes how and what do you see the potential so for a very long time bright dichography was was not appealing because it was extremely long to to to get a decent data set because you need stability because you need the specific samples because at the end the field of view is very small and so on and so forth and actually with the last results we have obtained we can show one step after the other that all these difficulties can be circumvented or can be solved so we don't need to know the probe anymore we can retrieve it we can see we can obtain here i didn't show the the field of view but this field of view here is a 10 or let's say 9 micrometer times 9 micrometer so this sample is huge and still we can see some dislocation we can we can handle problems in stabilities we can do so there are a lot a lot of problems that we have been able to solve and i think really what is needed now and which is unfortunately not so much on my power in my power is to have friendly user interfaces with designed by people who are not only computer scientists but also physicists uh or experimentally set some points we we urgently need this kind of of of tools of to to to be able to promote uh bright tachography so something i didn't mention no only only once uh now we are really able to break tachography in uh in an hour so the full dataset is obtained in uh an hour so it's extremely extremely fast and you you you don't have all the uh constraint that you have in break cdi is meaning that you need to have these isolated particles and so on so in in a lot of scientific cases you need to to you you can't do break cdi so i i yes i i think it's really really something the the synchrotron have to to to consider if they want uh if they have the volunteer to to to to propose these approaches to uh to to a larger audience to a larger material science audience for sure so we have a comment from from our good room who says maybe a research software engineer could do such a job we're talking about a physicist but what about uh software engineer do you think it could be enough i don't know i'm not sure or at least somebody who is extremely curious about the physics behind and yeah you mean that there there must be a strong collaboration anyway that people who are actually yeah because on the contrary to for tachography there are there are still a lot of i mean for tachography is not that easy it's uh there is a lot of efforts in order to make it suitable for for large audience because people at the beam line were willing to do so it's not that much the case for bright tachography and i would say that bright tachography has maybe a little bit more button to tweak you know like fine arrangement and so on and it has to be made as automatic as possible without losing uh the the physics behind yeah so there was the there is need of a big involvement of being scientists basically in this okay any other urgent question it's 16 i see that out of the 48 still 34 are here so it must be very interesting conversation i'm very happy about that any other urgent question otherwise i will close and thank enormously virgini for this intervention this discussions and everybody was participated in this discussion this is very important to keep it interesting and alive yes and uh yes and i invite everyone back for next week for more tachography uh in drug and in port thank you so much virgini thank you very much jina thank you again for for this very nice invitation and making a it's a a life during this lockdown cool thank you