 Okay, so I will start if that's okay. I'm Mariana Berezhak and today I would like to present you with a new method for visualizing defects and strain with high resolution that we call X-ray Tecographic Photography. And as Anna mentioned, I work at TSACS and this is where the method has been developed. And first I would like to thank all these people that have been involved in different stages of this project. So yeah, you probably know many faces from here. So it is known that not only a structure of material but also the deviation from this perfect structure would affect its properties and therefore performance. So strain can be defined in the simplest way as change in the volume of the sample compared to its original volume and it can be caused by different types of defects present in the structure. So the aim of our project was to develop a new method to visualize strain with high resolution in 3D but most importantly for extended samples and extended and complex strain fields. So during my talk I will introduce several methods that you probably know but just to remind you and to compare different advantages of this method to visualize strain starting with X-ray topography because here we will use the same concept of contrast formation and then following with coherent methods such as BRAC-CDI and different types of BRAC-Tecography. And then I will introduce the concepts of Bella Tecography again and a new method, Tec-Graphic Topography. And this was validated on ingeomentumonide micropealers and excuse me and we will discuss a possibility of implementing Tecographic topography in two different modalities in forward and diffraction direction. And then we will discuss what we observe in terms of strain fields in ingeomentumonide and possible dynamical diffraction effects present. So in X-ray topography we are sending a monochromatic beam into the sample and then we record a topogram and the base that the contrast is formed at the topogram is based on diffraction contrast. So in a simple in a simple case we will have a sample with the interatomic planes being almost at the BRAC condition but not quite and then we will have all the signal transmitted and we will have really high intensity in the detector in forward direction. But then if we have some defect that would bring a small fraction of the volume of the crystal exactly in a BRAC condition then some light will be diffracted and we will have a deep of intensity in a forward detector. And this is how the topograms are formed and we have in bright here the perfect structure and in dark we have defects and this is an example of a silica-carbide wafer topogram. Now the special resolution with X-ray topography can reach down to half the micron and if we would like to have higher resolution there are some interesting developments from ESRF in full field field X-ray microscopy. So here they would put place a particular grain of the crystal in a BRAC condition and then use X-ray lenses to magnify the topogram and achieve higher resolution down to 300 nanometers for sure I've seen and perhaps they can achieve better resolution but of course they will be always limited by lens and for hard X-rays it's quite complicated to make a nice lens. Now if we want to go in a further resolution we will probably use coherence and we are probably all know coherent diffraction imaging where we're sending monochromatic coherence X-rays into the isolated sample and we record number of rotations and far field diffraction patterns and then by phase retrieval we obtain the image reconstruction. Now if we place our sample in a BRAC condition then the resulting reconstruction will become sensitive to strain. The drawback here is that we will be limited to isolated samples and we would like to lift this limitation we would probably go for dichography where we scan the sample with a focused coherent beam in a set of overlapping position to increase the redundancy of the data and then we will reconstruct an object with phase retrieval and in a similar way while placing it in a BRAC condition we will be sensitive to strain. There are developments of the 3D BRAC dichography methods X-ray BRAC dichography and 3D BRAC projection dichography that already allows us visualizing strain field with high resolution but of course there are also some limitations of those techniques. What we would like to offer is an alternative method that would combine advantages from dichographies such as high resolution 3D information and topography with with its strain sensitivity and direct visualization of defects and also the fact that the setup would be much more stable and robust in topography case and one can look at really extended and complex strain fields. So we call the method dichographic topography and it features flat illumination which is improving the stability of setup requirement significantly. It of course will have strain sensitivity and ability to visualize defects and quite a robust reconstruction compared to some other BRAC methods and a possibility to look at extended samples. But before going into it we will first I will first refresh you on teletycography. So in normal dichography we will scan the sample and in teletycography instead we will place the pinhole few millimeters after the sample and scan exit way front and we will reconstruct it with space retrieval at the pinhole position and then we will back propagate it numerically to the sample position and get this nice and sharp reconstruction. Now in teletycographic topography we will use a parallel beam and we will place a crystal in BRAC condition similar to BRAC approaches but then we will place a pinhole few millimeters after the sample and we will scan a pinhole to obtain the dichographic images and the main advantage of this method is the fact that we are reconstructing exit way front after the sample so we have no assumption about interactions between object illumination and illumination and we have a constant probe so the reconstructions are quite easy and robust really but what is even perhaps more important is that we can study different interactions such as multiple scattering and dynamical diffraction effects because we are really looking at the exit way front after all interaction had been occurring. So the way we acquire the data would be we have a sample that we want to place in a BRAC condition so we will use a focus beam configuration and we will locate some place in the crystal where we have no defects and we will align a BRAC peak at the BRAC detector and perform number of rocking curves to align the BRAC peak well and then we will switch to parallel beam configuration and here we will rock the sample and at each angular position we will perform a 2D dichography with a pinhole and this is a basic idea of this method. So this is an example of one of few setups that I've built at CSUX we have here an incoming beam and the sample is located here on the rotation stage and several translation stages to find the sample we have a BRAC detector and a forward detector and we will need focusing optics that would come in and out and a pinhole for scanning after. So we decided to use indium and timonite object and those are the micropealers that had been feed out from the single crystalline edge and they've uniaxially compressed to the beginning of a plastic deformation as you can see here for displacement curve with the micro compression device prior to measurement and the reason why we thought this would be a good test sample for this method and many other strain visualization methods is because if this object is thin enough which in our case is two microns and that's that's thin enough and we compress it along a given crystallographic orientation which in this case is two one three we should activate a single slip deformation system and we thought okay this will be an easy first test object because we know exactly what is supposed to happen. I'm going to show you here an implementation of tachographic topography in two modalities. So the first one as I showed previously in forward direction so we will scan pinhole in forward after the sample and that this will offer us relatively easy reconstructions and the ability to reconstruct strain field but one should keep in mind that we will be sort of limited to strong Bragg reflections because the type of contrast we have here will be a superposition of transmitted signal and then some diffracted intensity will be retracted from it so a signal to noise will matter a lot if we have a weak reflection. This is something we noticed. On the other hand we try to perform this experiment in Bragg direction so basically we're moving pinhole here to the Bragg direction and we'll scan in Bragg and we will be here sensitive and I will show you in a couple of minutes two dynamical diffraction effects which is very interesting and of course here we become very sensitive to the slightest deviation from perfect periodicity but the interpretation of the images will become very challenging as you will see in a moment. So let's start with the forward direction and this is an example of an image that we obtain so we have we're reconstructing a pillar resolution based on Fourier-Shar correlation is below 13 nanometers which could be improved. We didn't really push for resolution here that was not an aim but basically at the bottom of the pillar you can see in the dark contrast features appearing that come from the strain field and defects present in this region. So I'm going to show you this movie perhaps it could help. The idea is that we will first use a focused beam and we will go at the position where we expect to have no defects so in this case at the pedestal of this system and we will perform a rocking curve so that we can decide on an angular range for rocking and angular step and then we will switch to parallel beam configuration and do dichography at each rocking point and here we will soon start seeing a dark feature appearing at the bottom here of the sample and as we go through further away from the bracket peak we have the sole shifting and in addition I just wanted to show you that we are reconstructing simultaneously a pinhole and in this case we use the three micrometer pinhole here so perhaps static images will be a bit easier to see when we are very far from the bracket so this is three degrees away from the bracket we have no dark features inside we have no information from strain and as we are reaching closer to the bracket peak we start seeing dark features appearing at the bottom half of the sample so perhaps you are asking yourself for me why don't we see anything at the top of the pillar because from the compression experiment this is where you probably expect to have most deformation occurring and I will show you why here so basically this is four rocking curves that were acquired with a focused beam at four different positions of the height of the pillar and at the pedestal we have this green rocking curve which looks very elastic seen bracket peak completely expected and then at the bottom part this curve in black that is a little shifted and widened which is also already presenting some strain but then at the middle and top part of the pillar we cannot really recover a bracket peak as such anymore because we have a lot of diffuse scattering and we are basically out of bragg and we cannot define a bragg peak as such and here is also a line scan so basically an evolution of this bragg peak as soon as we are reaching the half of the pillar we start being out of bragg basically we also wanted to compare our results with scanning transmission and diffraction microscopy maps so this is a mapping transmission we don't see much and this is the diffraction map so we can see that this part of the crystal at this angle is in bragg while the rest is out of bragg and I made this dark field maps here based on the the pedestal position so this kind of diffraction peak that looks very elastic and very symmetric and then in position two and three we already see a symmetry of in the bracket peak and a lot of diffuse scattering and position four and six is very very broad starting to split and at position five we cannot even catch a whole bracket peak anymore and also this is a map using only this region of interest so a single position of the bracket peak where we expect it to be from the perfect crystal and we see that this will be our area of interest what we will study so here we will be able to observe defects and this is out of bragg right so now we looked at forward direction and what we can get and let's move to the bragg direction so this is an example of an amplitude reconstruction in bragg direction and first of all it's an interesting shape and let's take a look a little closer so while going through the bragg peak this is the type of images that we obtain now we are closer to the bragg and then we go further from bragg we have this fringes of bright and dark contrast present everywhere but also we have this central feature that is present on all reconstructions sorry I go back one thing I wanted also to mention is that we might have a slight problem here with longitudinal coherence and basically I was able to reconstruct those images using single mode for probe but including a second incoherent mode was offering much better reconstructions just that you know okay so when we are very close to the bragg condition we have this sharp reconstruction that has this interesting shape and we have something that resemble a pillar shape and we have those fringes in the middle while we're going out of bragg we start seeing this other fringes going all over the field of view and basically let's try to decode those different contributions and what they might mean first of all is the shape of image why why the image has this shape I realize that as we have in this very difficult configuration I've calculated the projected thicknesses along the beam path in forward direction it looks like this and in bragg direction as we have 50 degrees from the transmitted beam in fact the pillar will the pillar shape will become something like that so this recover quite well what we observe in in a bragg direction so if we look at this more intense region this recovers it very well in addition we see that the pillar is present here it's of course shrunk in height and we see that the top half of the pillar would be sort of sticking out but we know that in our case it is out of bragg so we don't see that so we know also that the bottom part of the pillar will be present in this location so this is exactly what we see here okay now I wanted to also try to explain or understand those fringes that are present as we're going further away from the bragg peak and why they change with the bragg angle so they reminded us a pendulum effect that is present in single crystals of varian thickness based on dynamical diffraction effects and I did some simulations where I used the the maps of varian thickness in indium and so there's those maps I calculated the projected maps and then I calculated the diffraction based on the pencil beam approximation of course this does not recover our situation fully because we don't have a pencil beam experiment we have full we have full beam but already here we can see that the fringes that we would expect from just differences in thickness within our field of view recover quite well the effect that we observe and also they would change with the rocking angle and they recover they look quite similar let's say so that I would like to summarize a little bit so first of all in the forward echographic topography we observed defects and strain present in the bottom part of the pillar here that is probably caused by compression and in the bragg direction we have pendulum effect all over the place based on dynamical diffraction and then we also have this area within the bottom part of the pillar where we see additional fringes but the fact that those fringes are asymmetric is probably a signature of strain but it would be really really nice to have some modeling to understand this better so in addition we can extend this method to 3d and we will not need to do proper 3d but only rocking scan but if we would like to recover all components of strain field then we will need to study several reflections this is clear and we can make this method quantitative but it's very important to have some sample with really easy strain that we could simulate model and understand better what we see because as you can see there is a lot of different effects present and well one thing we try to do is to make direct comparison with existing methods so for example we had an experiment at id1 where we tried to do bragg cdi and like graphic topography on gold nanoparticles but we still we still need to manage we didn't manage yet to reconstruct data so yes if you have ideas about an easy sample that we are capable of simulating strain field in that i would be very interested and in addition this method can be in principle perhaps used in situ because we are not scanning the sample here and we will have a more relaxed requirement for the sample surrounding and we also try to use this method on different samples so one example we wanted to really look at extended sample and that was a silica wafer and the indent in a silica wafer so here we reconstructed a really really really large strain field and this is an example of copper micro pillars where we were able to see a single dislocation loops somewhere here or maybe not yes so this this is a single dislocation loop that we observe but that the quality here is much worse than what we can currently achieve and yeah in addition teletecography and tachographic topography kick-started more ideas of how to utilize this place in a sample in bragg and using teletecography and Angelou is going to talk about an interesting method studying dynamical diffraction effects in strain silica just in a few minutes and I would like to thank you for your attention and all people involved in this project and I'm happy to hear any questions thank you very much and for the nice presentation I'm of course familiar with your work but I really like the way that you have to do so I'm looking here at the chat in order to see whether there are some questions and I have a question now from Oleksandr, hi Oleksandr so the question says why couldn't you find bragg diffraction at the top of the pillar did you lose the crystallinity somehow yes so discussion on crystallinity is very I find it very difficult because I would say yes the crystallinity is not good enough at the top part of the crystal because we basically the diffuse scattering from the top part of the pillar was extending beyond our detector capabilities even if it would shift it etc it was just located in many in many places so it looked like we even had domains on the top part so one possible explanation could be that there were some cracks present on the surface and then we had really domains so we never managed to to have the to have an adequate crack peak from the top part it was just very very deformed so our expectations that we will only very gently push and have only single sleep deformation was of course wrong because there are other experimental constraints for example perhaps misalignment between the top part of the pillar and the compression deep or perhaps misalignment of a whole single crystalline wedge on the sample holder etc would of course violate this requirement all right so I think I have another question from Pablo getting the check but I think Ian has raised his hand before Ian would you like to ask a question um yeah can you hear me okay yes we can yeah yes excuse me excuse me for not showing my face I'm I'm actually on vacation and I don't have internet so I'm using a telephone which amazingly seems to work thank thank you for the very nice presentation I'm I'm really happy to see all of these exciting pendulousing effects starting to show up in in in the experiments you asked for what simpler sample you could look at and I think you probably know know this already but you shouldn't even deform the pillar you should just make a simple pillar with FIB and not touch it and that of course would be a simpler sample also indimantiminide is is quite strong in uh in dynamical effects and so you should go to gallium arsenide or silicon or something a little bit simpler which could be close to perfect but in terms of what to study I would suggest look at the damage due to the FIB because everybody wants to know what the FIB does in terms of whether it forms dislocations or whether it just degrades the surface of the of the material I think all of those would be very interesting questions yes thank you so much um it's it's it's really valuable I've been thinking about it one thing is that depending of how those pillars could be produced there are different ways we can um basically damage because one thing is if we will keep from top but now there are very exciting methods where the pillars is rotated and we have a focus being this way so I'm just curious how how how accurately can we simulate those effects and there are quite a few questions but well also there's the question of chemical etching that that is frequently used and nobody really knows how on the sort of 10 nanometer scale what what level of damage chemical etching does to these kind of materials we know that the some of the etching agents are very directional so they would etch certain facets more than others and if you take a pillar you would change the shape of the pillar but you would also introduce strain due to the etching so those would all be interesting experiments thank you so much it's it's very valuable thank you yeah I see the hand raised from butler also yes thank you for very nice talk I enjoyed very much I have one comment and one question concerning the forward beam that what you see is probably also due to dynamical effect because from definition kinematical diffraction a kinematical diffraction does not affect the transmitted beam so everything what happens in the transmitted beam is of dynamical origin so even if you are in the in the forward direction if you see some contrast it must be if this contrast is beyond the the usual absorption this contrast is due to multiple scattering to my to my opinion yeah so but in this direction no also what what what's in actually is that there is a lack of a dynamical diffraction in these areas no but if it is yes but if the crystal they're called purely kinematical then you would see nothing in in transmitted beam so there must be some something dynamical happen in the transmitted beam so this is one point then the other question can I comment on this point one thing so interestingly we did observe some dynamical effect in forward direction with the focused beam we can see it very well yes so even when we are measuring the normally you have either a black peak or that or you can have like an anti-black right yeah so in in case of having in case of having a focused beam we observe this very my question is what was the extinction length so this is the criteria if the diffraction is dynamical or kinematical so if the sample is larger than the extinction length it must be dynamic otherwise it's kinematic so what was the extinction length it depends on the wavelength on on the material of course but i'm not in i don't believe my heart it's how much it is i think it was about 4.5 microns 4.5 and just so this is just okay and concerning the diffracted beam do do did you consider something like Borman Fenn because then of course if there is a multiple diffraction then the the the contrast teachers do not originate from one point in in the sample but they are affected by the Borman Fenn therefore we better to use it we've been thinking about this but we had a difficulty to simulate this effect in in our situation it is not easy just take yeah you must take takagi equations and numerically integrated takagi equations so there are not many people they know it of course how to integrate the takagi equations but i've done it many years ago but i don't remember it anymore yes i've done it in 80s you know i am one of the very few people in this auditorium which is senior and i remember in 70s and maybe beginning of 80s was a big boom of dynamical theory everything everyone calculated the calculated takagi equations and this completely forgot maybe Jan understands it remembers it it seems we need to team up with some senior collaborator senior collaborator who remembers what takagi equations are yes takagi is a new thing also in the world of the ephioles and there are a lot of efforts due to the effect on the temporal effects of dynamical diffraction yes yeah yeah yeah and concerning this pendulum fringes that you showed i think this is a combination of pendulum fringes with simple thickness fringes this is just a kinematical origin or simple optical origin i'm not sure if it is really pendulum effect what you are seeing this is just a thickness fringes from interference not not interference from the shape not interference of the intrinsic wave field the dynamical wave field in the in the material itself so for the shape of the p like you mean yes exactly uh but i don't know of course it must be simulated it must be simulated yes yeah yeah but otherwise it was very nice talk thank you thank you thank you so much thank you so i mean i should now come back to pablo because i push i put him awake here on the chat he has a question for you mariana do you also observe the depth of focus effects do you need also this slicing but i didn't hear that so the question of pablo is if you observe depth of focus effects so if you need you would need more this slicing right like all of you would move this slicing in order to oh sorry ana yeah that's what i mean i was thinking about this pendulum effect that i found it really cool and really nice presentation so i was thinking about other thickness effects like uh the fact that uh when you have a really big sample and you have the depth of focus i mean you have to see the different planes with different propagators so if you when you are showing this to the image in the diffracted beam if it was affected by that as well so that was pablo this is not a focus beam that is used for it's it's parallel being so yes yes but still you are looking in diffraction you have the pinhole and you are resolving a diffraction pattern in the ecography so maybe your resolution of that diffraction pattern in 30 nanometers then it corresponds to a certain depth of focus and that it goes through your projected crystal so you will have the different slices of the projected crystal on that direction so if it would be forward experiment if it would be forward experiment 30 nanometers would be okay for that 10 micron even for the pedestal it would be okay it's about 20 no it's sorry it's 10 micron only even the pedestal the thickest part is on the but in the in the angle what was there the the fraction angle so i mean then yeah 50 degrees so let me see what is the maximum thickness i can tell you so in a normal forward experiment you could do 30 nanometer on an almost 30 micro sample at this end so i don't think that's how but just a moment i can tell you so the very maximum thickness that goes through will be close to 19 19 microns yeah so it's basically one over the cosine of the angle of you have and i would give you basically what is the thing that you have there and if that is above that maybe you have to have it those effects i don't know i mean i was just wondering because it was really nice to see this pendulum effect but i was wondering if something else could contribute that thank you so much i was waiting no problem so thank you mariana for this really cool talk i have a question very specific and to be honest i think i i would not i will not get an answer but i keep i i take the possibility to ask you anyway so my question is very specific i wanted to ask you whether uh when you are in the black geometry so not in the forward not looking at the forward beam but in the in the black geometry did you by chance at the look have a look at the three-dimensional speckled pattern and did you notice any specific change of the speckled pattern with respect to kinematical effect well yes we looked at the diffraction patterns for sure in 3d yeah in 3d did you did you try to no no no in 3d my question is is is i can explain of course why i'm asking these questions and i'm still interested by the answer so if you have the chance to have a look i would be very very curious uh we are we are measuring in bright taiko some samples which which are a bit thick and we are observing speckled patterns with uh speckled uh which are not inclined the way they should be uh they are parallel to the detector plane where they should be uh like say perpendicular to the incoming beam and we are we are we do not understand why we are observing that and i was wondering whether it could be due to dynamical effects so if by chance you you could have a look at your three-dimensional speckled pattern really in 3d yeah not in 2d i would be very grateful if you could let me know how they look in 3d yeah we didn't have a look at the in 3d for sure yeah i can imagine yes it's painful but uh yeah i mean if you if you if you are if you have a bit of time to to waste on these questions i would be very very grateful if you could let me know sure we should we should discuss this uh some other time together okay cool thanks okay i see here also a comment from olexander and i'm sorry i didn't see this before because it has to do with the previous discussion bucklatt is right in kinematical approximation direct beam is considered to be constant maybe a ten-weighted so everything you see is due to dynamical effects so i'm sorry for not saying this earlier olexander but i think i mean he's just been discussed already so okay then is bucklatt having his hand may i have another question to maybe to virgini hi virgini or to dina or to ana hi how is it with a phase retrieval in dynamical theory in kinematical theory the amplitude this Fourier transformation of electron density which is not true in dynamical theory is there any chance to to make phase retrieval to get the electron density in the case of dynamical reflection or how to do it so the the way i see it i think it's very close to what mariana and ana are doing i would retrieve the field because the field is a field right the so there is no problem with that and then you need to solve the inverse problem taking into account dynamical theory but first i would just retrieve the fees the field the phase and amplitude and then introduce the the the secondary effect of dynamical theory but to be honest i didn't tackle this this this point so far so there is no easy way how to make the inversion problem from the field from from from scratch including everything no and i i don't see really the i don't see really the interest except that you gain time but you know if you can get the first the the phase field the complex the complete phase field then for for almost no the price right yeah i i mean right now i would i would this is the way i would tackle the problem but maybe i'm missing the point but you are right that always you can switch off the dynamic effect using big beam or using yeah change the wavelength that everything is kinematic or go away from the breakpoint then you are more less kinematic yes yeah but to add something to this comment i think that the the way the this problem is tackled now is is to retrieve the field after the interaction as virgini said exactly here and then is is a is model approach again so there is no i'm not aware of any inversion but i know that a few i think in optics they are doing it in optics they are they are inverting the the the field taking i mean the field the complex value the field they start from this and then they retrieve the sample geometry or whatever refractive index and so on by solving the dynamical problem but they start from the complex valued field i have also seen some very interesting the black and i think i think i saw that presented in this tms session that we organized the march where they do incorporate in the iterations also some dynamical diffraction model in a very thin crystal right so obviously there must be some situation where this is easy and so but i am not familiar at all how this works it would probably be worth inviting tau joe from the nano probe to give a talk he's done a at argon he's done a ton of like automatic differentiation work with a dynamical diffraction as the forward model and he's using it to solve the you know the inverse problem exactly that i think that's what you saw exactly but this was like a very thin crystal right yeah he's been and trying to you know the bigger samples and it seems to be working yeah very interesting all right so i think we had a an interesting discussion thank you very much mariana i think you even had the some of your points that addressed for discussion some very good comments as well would you like to still bring out bring up legitudinal coherence and spin-off story and yeah we're a little bit worried that you mentioned that already right we're a little bit worried about partial coherence effects and we have been trying to do a little bit of geometry to see how this affects our problem so mariana has a couple of slides here to show us this and Steven should be quite familiar with this because this is from a very old paper from him so basically we're talking about here different optical path length differences between scatterers from points like A and C right which in black geometry they become quite large and this basically depends on the height of the sample along the direction of the of the black planes or the diffraction plane so in the case of the inhuman dimonite pillar because we have a very long pillar right there's a difference between the top and the bottom beams becomes huge certainly much larger than the 1.5 microcovidex that we have in our experiment but what we were thinking is that perhaps because we have a pinhole over there one does not need a fully coherent waypoint so so to speak it's important the coherence only within the size of the pinhole and then this is because this is two microns then this becomes really on the order of our coherence method perhaps our problems are not so severe because of that but I don't know if any of you has any insight kind of makes sense right that you only need coherence across the pinhole but well it has to be a it has to be a little bigger that maybe has to be twice the size of the pinhole because you you kind of need there to be some kind of phase relationship between two positions of the pinhole right between the beams at two positions yes the cups twice yeah I see I see your point yeah yeah for us in the group we have discussed this a little bit it's a little bit with the fact of reconstructing a phase over a very extended area so teletypography can reconstruct areas of 20 microns or more and how can we speak of a phase when we don't have coherence I can make some comment just on that as well sorry can can you hear me okay yes yeah I'm a little bit distant from the conversation but I'm following it um we found some quite interesting um improvements of the of the visibility of speckles that we get in very asymmetric geometries like this and we don't fully understand the book mark Sutton had something to say about them which was that there's a coupling between the spatial coherence and the longitudinal coherence which sometimes works in your favor so quite often you get a better visibility of diffraction um than than you expect by this argument okay so that that can be right that we can relax this condition a little bit perhaps yeah and it's all in his big paper which is that book chapter but it's quite a bit of heavy reading if um if you would like to follow it through okay thank you thank you all right well in that case I think we have had a very nice discussion so thank you very much Mariana for the great presentation and I will give this back to Dina so that she can introduce thank you thank you everyone uh now it's my pleasure to introduce the next speaker is Ankel and Ankel has done his PhD in the University of Oviedo and then he had an exciting experience at Cornell High Energy Synchrotron Source where he got in touch with um dynamical diffraction so I used to joke with him that he is one of the few people who understand dynamical diffraction so Vaslav you're in a good company today so Ankel came to max four and as a postdoc and we started to gather this project he had already been interested in optics for x-fell with a project with a Big Pedrini at a Swiss fellow project so he brought this interest at Nanomax and and then we went ahead with the measurement using tele-tycography so Ankel please show us the results of this research yeah so thank you very much Dina for the presentation and also for the opportunity to talk I really need to see that you are listening to me I don't know if my mic is on yes okay perfect sorry so yeah and I would like to actually show a little bit of a study say that they do a dynamical diffraction or a diffraction of the history by theme perfect crystals no so first of all I would like to introduce one guy that we are not talking usually about but this is Francesco Maria Grimaldi that was the first person ever talking about diffraction he also introduced for first time the idea of a light to be a wave company to water that was in the 17th century and actually if we see what he said in his book or he brought in his book he said that the light propagates or scatter not only directly by a refraction or by a refraction but also by a fourth wave the diffraction with that is to say that scattering and diffraction is not the same and scattering is scattering englobes many other effects no diffraction is something particular and the word diffraction actually this is interesting it's coming from the work Latin diffracted to break into pieces with this I jump 500 years almost into the future and we are in the world where we are actually in the new high coherence spatial sources with higher repetition rates no so we have the two facilities where I have worked in the last years max four and expel or european expel where we have this kind of world where we are meeting the high coherence crystals with the high coherence spatial sources and with the case of exfellows and self-sealing also a high longitudinal coherent sources in the world of fields but everything in the case of this project started actually in another project in a small bunker in the middle of a forest next to the german border at the polser institute called swiss weld where I was actually a invited to do a postdoc in self-sealing with the idea of improving the the longitudinal coherence of the salsa pulses to have an increase of the brilliance, the stabilization of the central wavelength and also the smoothing of the temporal pulse shape the idea is easy instead of having just one section of on the laters between the salsa pulse what you can do actually is try to use a deviate the electrons through a chicane and monochromatize your first salsa pulse to generate a seed that after is recombined with the electrons to lay in a second set of modulators the main proposed one was the one given by yaluca geloni that is the forward diffraction these are actually works that we have been doing in the last weeks in this case we see the cell seed in salsa 2 that is actually here we have the salsa that is around 30 v's and we can see that we have really a wonderful one eb with pulse coming at the end of salsa 2 and actually even yesterday and we brought the first time 1000 pulses to our spectrometer and actually we saw this kind of fancy angel appeared in the spectrometer as I said to and with that I would like to go back to diffraction all we are used to this definition in our textbooks about diffraction we have a incoming k vector so an incoming wave with a k vector that when we are set to diffraction condition what we have is a diffraction beam going in the direction defined by the two theta angle as the equation of plaque but also we have like a forward beam that is supposed to be just the transmitted beam so whatever it was not diffracted the same happened in the case of laue diffraction or this is what we think but this is actually not true what we actually have is something that looks much more like this we have this goes too fast we have an area that is a generated waves that are traveling and it's like a labyrinth of waves of photons traveling together and they are like matching themselves and when they have a proper face they are matching and generating different beams that are delaying time and also this place in the transverse direction to diffraction and we also have that the same in the diffraction direction for both the black geometry and the laue geometry we can see we go to a pulse so a focus beam how they look a little bit more we can see how here is the main effect of the incoming surface and what we have a huge effect of absorption and then we have a second beam that is coming actually from the back diffraction of the back surface or rear surface and we have the same or similar in the case of the forward where we have also like an increase of the absorption but these beams are all monochromatic in between each other like this guy as much monochromatic and in the case of laue we have actually something really nice that we are exciting the both surface of diffraction and actually what we have is a much more symmetric signal where we have the same intensity in both sides of the signal if we go to thicker crystal then we go to the approximation to the Borman effect where we will have here only like one big peak and also it will appear a big peak located somewhere here while the rest will be heavily or really strongly absorbed by the crystal and so if i continue these are some simulations i did with the code that i developed at swisswell in the case of a thin crystal 50 microns at 9.4 kiloelector volts for the diamond for zero zero reflection and we observe that we observe in the back direction a huge peak and then we observe these simulated echoes because they are displaced and delay in time so they are similar been assimil to what is the sound waves and what we see is the back or rare surface diffracting back in the direction and also we have the same in the forward direction now where we have also like the effects of the different surfaces appearing in between the different uh dynamical diffraction uh beam that are propagating parallel to each other in the laue direction we have the same but actually because we are exciting the crystal uh equally in all the areas we have actually a signal that looks much more symmetric to each in this area when we go to the thicker crystals already set is where we're going to have the Borman effect appearing in the forward direction we still have the symmetry but we can see a be much bigger effect of the absorption sorry again so if we translate that to time that is actually what I'm really interested in we can see that actually the uh transverse displacement uh as a question that was introduced by swiss called Lindbergh is related to the temporal delay multiplied by the speed of light and the cotangent of the diffraction angle or black angle and what we can see is the same now we see that we have pulses in this case that are like uh like 50 femtoseconds from each other and we have a effect of the back surface and in the case of the laue case we have some more evenly distributed intensities and again it is 250 that it will be the equivalent of the beam or the time that the beam needs to travel along the full crystal and if I continue then we oh too much if we went or we tried to do several experiments of this this is quite complex to do so we actually fail in three experiments two of them done at CES one of them done in material science also at the swiss light source and in the fourth not the third but the fourth was the right one and in this case we were able to observe the echoes of a diamond a single crystal that was located somewhere in between the focus of the kb's and the detector node so the in this case what we are trying to do is defeat divergence by focusing to a jack stream and with it be able uh like to not have the uh um can only like record amplitude uh if we look these kind of uh studies they depend a lot of the energy resolution our rotation states was not as good as we required to be able to record the 400 of the diamonds or and then what we were decided is to use energy scans we were scanning with a monochromator that it was a high resolution is 311 is not the highest but it's higher than a 111 that it was actually the same man with that we were using or we wanted to start our crystals no if I continue then we can see the first results for the bracket direction we see that when we are far away from the condition or diffraction condition we uh don't have the any sign of the echoes and we were closer we start to see this oscillation that actually when we are in the diffraction condition we can see in the logarithmic scale how nice they look like for our simulations starting with the code that I talked about before uh we can see also the same effect and if we compare the position of the experiment maxima and the position of the simulations they match as they do also the intensities and we have to say that also and maybe they even if we were using a focus beam that it was one micron our yak and our optics were not allowing us to have that resolution so that is something to try to improve no and then we can also see what happened in the laue that is the same but the interesting thing is the same no we really have this kind of symmetric effect of the laue that is quite a gorgeous from my point of view so the excitation of the two surfaces of diffraction to remember that is everything all the study is done in forward diffraction as we were interested in the implementation for the self seedings and in the case of the energy comparison we have that both the simulated and the experimental data they match to each other so now we have the position of the simulation and the experiments that they also match is more or less one so we are happy and we publish it but we saw this other information that is actually interesting what happened when we have a strain crystal and what would happen when we have a perfect crystal so if you see in the strain crystal as I saw we have a perfect fit into the simulations and so the theory but actually in a strain crystal what we can see is that the position of these echoes are much more extended and even they are part to be like extremely far away and we can see that for example the seventh maxima is 34 microns away while we were expecting to be 20 microns away and with that what we can improve to try to get into a collecting the data correctly or to get better resolution so problems that we have it was the big size of the bin five microns in the in our samples the high divergence of the bin that is somehow overlapping in the case that we are not on the focus so that means that we can never bring the sample to the focus because the detector should be on the focus and we have also a broadening due to the yak and the pixel size is the effective pixel size is around 650 nanometers so what is the solution by the lethichography use as a virtual detector so the idea is that we are using the p-hole as the real a detection plane and we are actually doing the collection of the data and a detector that is located farther a four meters away from our sample with this we are actually like cutting the intensity of the echoes and scattering or diffracting from the pinhole to the position as mariana presented the idea that we want is that after when we reconstruct the dichography in our pinhole we can propagate it back to the focus and with it we win in resolution because our transmitted bin that is not been diffracted that has a different case set that is not in the conditional diffraction that will be as small as 100 nanometers that is the pin size of a nanomax and we also increase our pixel size to 30 nanometers and we have a resolution of around 50 nanometers with that this is more or less more highest looking the geometry that we were doing we were using the sample that it was located in the center of rotation of our nanometer and then we were having the pinhole located as close as possible between 3.5 and 4.5 millimeters and we were having a detector to set the diffraction condition and maximize the location of our crystal with respect to the diffraction and then we were having the merlin four meters apart going with a helium tube to try to minimize the air scattering what we have is that in this case what we are wanting to do is to actually scan the pinhole to have the overlapping as oversampling that it was what mariana presented just before we did the first study we go to a sample that we know that is a usual sample of a nanomax and we do just the normal tele dichography experiment using the focus beam in this case what we can see is that we have put the sample 200 microns in front of the focus this is how it's looking the focus beam we can see some effects of the 7 star and this is how it's looking the wavefront in the pinhole reconstructed and this is our pinhole reconstructed we have some kind of effects of phase that can be because the pinhole is not as perfectly cut but also maybe because of dust or things because the pinhole has been for many years around CISACs use so yeah if we just look to the sample and we look to the scanning x-ray transmission microscope we get this out and in a diffraction condition and as we see we already have a little bit better resolution that what we collected at a micro access it's important to say it's really complex experiment everything was too close the kb mirrors was here here we were having a microscope to locate the samples in the focus of the kb's and we were having the pinhole you see like three to four millimeters in the back that means that actually these models were fully saddling our diffractometer yeah our detector in the diffraction condition that was located in the robot if we do the typographic reconstructions we find this out and in the diffraction and actually we can see that if we propagate it into the focus plane we can see the nice shape of the focus of nanomax and also we can have a high resolution of our echoes that if we look to the simulations that are we able to do with the code I developed at Swissfeld we can see that actually in a projection of both signals they match really good one to the other so that is like the best comparison I have been able to see ever of this and the important is that also with this technique we are able to actually see how the echoes are propagated and we can see that they are parallel to each other and actually they are not getting all the divergence of the transmitted beam or the incoming beam but just of the photons that are actually in diffraction condition and here is a presentation of how it's looking the phase together with the amplitude as you see it's quite complex probably also because we have some kind of convolution between the echoes phase together with the kb phase so the changes in the wavefront of the phase due to the kb so this makes quite complex to subtract information in the temporal domain from the phase of the echoes if we look now to a sample that it was invented by the group of Magnus Koriander at Chalmers University of Technology what we can see is actually when we are far away from the strain area that it was yeah we have actually some diffraction or forward diffraction that is looking similar to what we were having in the strain free if we look to the two indents 25 millinewtons and 75 millinewtons what we can see is actually that the echoes they look to generate a different positions and they look like to divide and in the case of the 75 we see that there are clearly like many small peaks that are actually generating or related to the strain if we look to this in the propagation we can actually see that all these look to be also like parallel to each other as we were expecting by the echoes no or by the theory if we try to simulate this using the takabi and takabi coupon equations and we introduce some kind of exponential, inverse exponential strain on to them we can actually get into some results that they match quite good but not perfect our experience no but actually this can even be translated to the world of time because our simulations take into account the time domain and see how it will be the delay of the echoes due to the strain and then we can see that actually in the strain sample that is this one so this guy is this guy this guy is this guy that guy that guy and here we have the strain flip we can see that even one millimeter away from our strain area we may have a little bit of changes in our lattices defined by these indentations and we also see how when we are in a high strain area what we are doing is like somehow compressing our echoes and we have much more echoes in the same eight microns that we are looking here and we can also see how in the case of the extremely sorry in the extremely strain sample we see much more beings that are appearing in these different positions so that is quite interesting because can tell us that we can also use this strain control to improve possible optics for our free electron lasers but you will say okay diamond and silicon who how that this help us apart from optics what about real thin crystals okay so then let's try to introduce the a little bit more of fiel science and we are going to do a micro focus or nano focus being as the one that we will cool have at mid so we have a hundred by hundred nanometers and we have such a pulse that is 10 femtoseconds long and what we can do is go to a gold a 111 reflection at nine kiloelectro volts and we can see what we observe no we can see that actually the bin is looking already split when we have 500 nanometers so that means that if we have a really thin crystal of gold the thinnest 500 nanometers we already may see effects of dynamical diffraction you will say may not because it's a strain and it's not perfect but it may and actually the indian timonade is showing that even the case of mariana shows that these things may be real and when we have one micron we can see clearly two spots in the projection also that we have here down and in the case of 500 microns we have like this kind nice butterfly butterfly shape where when we have 10 microns we already see here appearing a high intense peak and the absorption of the surface to be higher and here is where it's going to increase what i will call the borman center or borman peak in the case of the forward direction here is actually also what borman observed in the forward direction that it will be also in the center and we see that the absorption in the surface happened while here we don't have this high absorption so we have to go to a little bit thicker samples like maybe 20 microns to really see that this is completely absorbed and we still have the maxima in the center of the crystal to say and we can really see in the forward direction how we with 50 and 500 nanometers we can really see the effects of dynamical diffraction on them energy so these echoes they also have important dependence with energy and lower energies um higher energies we have a change of the black angle and as we have seen before uh the time delay and the uh transfer displays a position is related to the black angles and how so when we have a higher energies the echoes are much more spread around because they can travel in longer path in the crystal and when we are um at higher energies they are much more uh smaller so it's something to take into account when you also want to do experiments and if we look to the time signal actually we see that the gold so for the hundred nanometers we will have like here i have changed the simulation i have just simulated one femtosecond this is important to say it's not 10 femtoseconds but just one because this is it's easy to explain uh so here we have just one femtosecond and we see how we have in hundred microns just the one femtosecond that we introduce but actually if we go to the 500 nanometers we already see that we have win a few hundreds of other seconds that will be located in our uh signal uh if we go to one micron and we almost win a two uh to extend it to two femtoseconds and when we are talking about five microns of 10 microns we can see that is like four or even seven femtoseconds larger the signal uh but it's not just one femtosecond needed no uh we can see also the effect in the two femtoseconds five femtoseconds and 10 femtoseconds and we always see that we have this extension for the 500 uh so for the five microns we always see these three extra femtoseconds to the signal uh if we look to the BRAC reflection or the BRAC geometry we actually see that uh something interesting that is that always our intensity in the main diffraction is going to be of the same intensity but actually we are going to have also some kind of uh sorry uh or peaks uh or echoes that are going to appear and these are related to the uh back or rear surface in really thin crystal and the echoes are even appearing in the dynamically case of the BRAC geometry even up to 250 nanometers so extremely thin crystals much more easy to observe when uh we are looking to the forward direction where we clearly see the echoes appearing in the uh really thin crystal and uh also in the red snow and we see how when we are in the forward gold is absorbing a lot so of course our peak is going to decrease uh drastically uh as we say in Spanish it's not gold everything that signs and this also may happen a nickel galvanized night and indian demonite as was uh observed for a a mariana is also presenting so this is actually quite interesting no how we are living in this world where we are starting to uh to see the effects of dynamical diffraction in our sources and the high impact that this can have in our temporal resolution in the femtosecond world with that i would like to thank all my collaborators along this path and also like Anders Matzen that he has support me to continue doing these research while i am at the MIT and all the instruments i have worked at and especially i would like to thank Ken Finkelstein that was my uh the person that introduced me to dynamical diffraction who retired uh just a one year ago uh and yeah all the members of Max Hor, Sinek, Gerardina, Alexander and also the person that support me during all this year Maria Ljevi and remember and this guy is not a diffraction peak so with that i'm open to questions thank you so much Akal that was really nice enthusiastic presentation and quite a large overview i think that the most the most interesting uh uh message from this is that actually dynamical effects really appear in very very small samples and actually a few studies had already highlighted this actually i think that uh Alexander was in the audience was one uh part of one of the recent studies uh so the session is open for comments for uh criticisms for questions i see already Alexander says um Bushuev has shown that the pulse broadening effects due to dynamical effects 15 years ago this is why the reason why the beam splitter was not introduced in an expel so what's new in your study that's it i answered that we have been a splitter scene uh in uh expel or the european expel actually um we are using it in the splitting delay line of uh mid it's important uh to say that also this is also where we use Ben Christa for the spectrometers really thing once and yeah what is new in the study that is a really good question but maybe it's just the observation of the real effect and not just a theoretical workers who swept it so i think that is the experimental view that anyone was able to so until now yeah also with a with a very uh high resolution in this case Anna so i don't have to leave there very soon actually i'm really sorry about that and then i will try to follow you from the telephone while i am traveling but i just wanted to make the point that uh actually is actually using office knowledge on dynamic detection dynamical detection calculations in order to try to simulate uh exactly also the um how these waves propagating in the case of the pillars for example and taking into account all these uh echoes for these are just the effects that happen i'm putting them together he's starting to get yeah quite an insight into what may be happening in the experiment which could be something nice to follow up so just want to comment on that okay thank you Anna and thank you for for your presence today uh Vaslav uh this is just my ignorance and my curiosity how did you simulate it so did you use simple or standard takahito pen equations or in case of short pulses you have to add time as a as a variable how do you do it no actually we are not introducing the so so so what what this is part also of the collaboration that i have with the SRW people know so what you usually do is just you simulate the out of time and you just bring the different slice of the beam on time okay so it's uh that is how also it's done in SRW uh yeah other questions or comments i have one nice then i hi uh thanks for the nice presentation at the end it was a little bit too fast for me so you mentioned that you did some simulation with one femtosecond yeah then i saw some 10 femtoseconds uh so what will be the difference at the end no so so the difference is known as is presented here the the broadening of the signal is still the same no so so you can see that here we have one femtosecond and then we have all the broadening of this signal and here we have the 10 femtoseconds and the broadening of this signal yeah so okay so it's always the same broadening at the end and it's okay in another direction yeah okay now this size no always having this butterfly like in the uh transverse direction okay basalov raise your hand please yeah this is again my my my ignorance if you calculate the the the length of the wave packet for one that the length of the wave packet for one centosecond something like i don't 300 nanometers or something this is much smaller much shorter than the extinction length so how do you speak about dynamical theory in this case the length of the wave packet is some nanometers not 100 nanometer 300 nanometers this is much shorter than the extinction length and the extinction length is the critical value for distinguishing between dynamical kinematic theory so if if the sample is shorter than the extinction smaller than the extinction length which is in your case effectively then everything happens kinematically or i'm not right i'm not sure but this is not my this is my ignorance probably so actually this is just extremely a good and important question no because maybe the extinction length is not what defines fully the dynamical diffraction maybe we were thinking like that until now because we can look really in the time domain or not as you see if we have really big pulses like what we can have 150 microseconds we're not going to see this effect at all only when we have thick crystal thicker crystals that are of the rates of the extinction length may we see it even if it's smaller i think that this is possibly something a bit more intrinsic fast lag in the sense that at the end interferences with the the photons with themselves isn't it and and the beam still propagates through i'm not specialist in this field so i don't know so olexander olexander will know it well i'll just comment here vassilov i think it's actually for statical actually all this extinction lengths is what is commonly used and what is easy to use right but we still have effects even under those sickness as actually show here and we saw it before and here we are talking about really dynamical thing effects in sense of time constraints so here dina is also right completely so this is more complicated here and both viktor korn and bushu have calculated those effects for decades already so telly chukhovsky did some also yeah well you know a lot of people those people but actually i had another comment about the simulations so you have shown that you have some takagi to pen simulations for those kind of situations there was a paper by anatoly shabalian who made a program for simulating together with strain there is no complication to add strain if you have agreed in your takagi to pen questions so there are there was a paper published and i think cod is also available so if you are interested you can search for it it was like five years ago but there is also also all the work done with the phd from my instrument this called ilia petro no i don't know him oh yeah he is actually quite the boat and he's a master student from busue so i'm tough quite a big contact with this group with busue and liuba and all this and we are actually really interested on it no and he has presented really many works where he's actually also like integrating and using who was sorry that the person that you are anatoly shabalian he is now back he was in even vartenya's group and now he's back in daisy basically he was the one who actually did all the programming he's also talking about the relations yeah at the moment he's in with i just mean that there are there is a program and there was a paper about it and also strain is not a problem for those calculations so if you are interested yeah yeah yeah okay that's it yeah i have talked with him a couple of times i have a general comment um i mean the the main different one of the main differences between the previous the work presented by mariana and the walkie presented by angel is that in in the case of mariana is really imaging so that's the whole the setup is the same and the technique is the same but in one case there is a large beam that illuminates the whole sample so it's a real topography type of measurement so the final aim is really to get a high resolution image of the strain and understanding of the strain in this case we are actually measuring hangali is actually measuring a pure diffraction effect so we are not measuring the near field of of of an interaction between x-rays and matters we are we are effectively measuring a far field a full diffraction so there's no way to propagate back and get the information of the sample but it's just a pure diffraction so i have a question that comes back also to with virgini and ross and anna was talking before do we really need a coherent beam in entrance to apply the typography the principle coherent i mean diffraction is intrinsically a coherent effect we know that all this through highly ordered crystals only a full only a small part which is actually of the photons in in phase and they are propagated multiple times inside the inside the crystals so actually we're thinking with uncle do we really need a coherent beam to start with would it just not work even without human being open for i'd say you mix mix up transfers and temporal coherence so for any diffraction you need temporal coherence right mm-hmm but then if you want to reconstruct like shape or while size letter let lateral information about your object you still need transfer coherence but but but in this i mean we are coming in with a nano beam right so that's like a hundred by hundred nanometer beam and the diffracted field is actually produced by a temporal effect and it's extended over many microns 20 30 microns that that was my question if you actually refer to the diagram which was shown before like incident beam diffracted beam and the path difference that's not really applicable to dynamical effect because again extensional access was not mentioned it doesn't matter that you have 20 microns if your extension lengths is half a micron you don't care about everything else in depth right your beam is not going there you don't have this signal no i'm talking about the lateral extension of the of the signal in the plane of the fraction so in the perpendicular i perpendicular echoes is is like you know the actually the the plot that is here now shown yeah maybe may i may i have a comment i think in this case what matters is the size of the green function you have this green riemann function from the from the standard dynamical theory you know from from otter and this stuff and the size of the of the size the spatial size of this of this riemann function which determines in fact the size of the Bormann fan in in in diffracted beam in break reflect break geometry this is what can be should be compared to the to the coherent coherent bits so i would say if your coherent bits is larger than the size of the of the Bormann fan then you can treat the the incoming beam as locally coherent but everything but the multiple scattering evens occur in the Bormann fan so then if the Bormann fan is coherently irradiated then the multiple scattering can take place otherwise if if the coherent bits is shorter than the Bormann fan then then of course the the multiple scattering doesn't cannot appear fully in this in this volume so what what matters is the size of the Bormann which is which is dependent on the chi of the polarizability of the wavelength and on the geometry okay thank you and another question i had about the the dynamical effect showing up with a large beam they will superimpose each other you know we are capable of looking at these echoes and very well because we have a high resolution we have a very small beam so we claim that with a larger beam we would not be able to see it but in the case of the topography experiment that's a large beam and we still see the dynamical effect of course because in this in this case the beam is large so all these echo superimpose because in ideal case the the incoming beam is infinitely large in that in cross direction so all these echoes superimpose and what do you see is the interference of these plane waves they are plane waves so there is no lateral resolution there is no distribution of intensity in lateral direction the intensity in lateral direction is constant but you still do see the dynamic effect because of this interference of these echoes but you cannot separate the echoes in a real space yes yes i think so are there questions or comments this session has been a little bit long and heavy i suppose