 I think it's Uriel Morsan, Uriel, perfect. Okay, I think we see your talk. Okay, can you see it? So yeah, so next speaker is Uriel Morsan from ICTP. And Uriel will tell us about ultrafast electron ion dynamics around conical intersection. So please, Uriel. Just for a second, I want to move this. It's in the middle of the screen. Ah, yes. Okay. Whatever. Did you fix it? Is it okay? No, no, no. Okay, now it's perfect. Yeah. Okay. So thank you very much. I'm very happy to be here and thank you very much for inviting me to give this presentation. I will talk about today some very recent results that I'm very excited about those. So I will talk about the strong hydrogen bonds under connections with ultrafast dynamics around conical intersection. So, okay, I have to, okay. So my goal in this talk is to establish a connection, a general connection between the strong hydrogen bonds and the position and the shape of conical intersection. Again, I want to mention how this is related with some very interesting optical properties. But a disclaimer for this, I will talk about two very specific systems. So my claim or what I will try to convince you is that these are general properties. So these systems for me are more like a proof of concepts. All right. But yeah, I will talk about two very, very specific system. So this is more or less the overview of the talk. I will start introducing the strong hydrogen bonds. Then I will talk about non-aromatic fluorescence. After that I will speak about X-ray absorption spectroscopy and it's how it can capture the signature associated with strong hydrogen bonds. So here I will show a very, very, very simple and supervised machine learning method for studying the coordinates associated with conical intersection, okay. So we normally think about hydrogen bonds with the sketch that we have on the screen, okay. So this drawing represents a potential energy surface for a normal hydrogen bonds where we have two minima, one deeper than the other, and several quantum vibrational levels inside the deepest minima, okay. But it was this year it was shown in the group of Tokmakov that the strongest hydrogen bonds that we can find in nature, they behave quite differently. They really behave as more like a three-center covalent bonds, okay, because the zero-point energy is above this double-well structure, okay. So this really behaves as a broad and harmonic potential, okay. So recently in the group of Gabriel Kaminsky-Schreer at Cambridge, they found some protein aggregates that are associated with a neuro-relationative diseases. They found that these compounds can fluoresce without the presence of any aromatic residues, which is peculiar because we normally expect the fluorescence, at least in biological matter, to arise from aromatic stuff. And together with Ali Hassan Ali, here at ICTP, they observed that this compound contained very rich hydrogen bond, strong hydrogen bond networks. So this strong hydrogen bond that I just described, okay, this compound is full of them, okay. So in order to see if there is a link between this non-aromatic fluorescence and the presence of strong hydrogen bonds, this group at Cambridge, they synthesize a compound called L-pyroglutamin ammonium. Basically, it is a thermal decomposition of a glutamine. And the main difference between the two, the two are solid systems. The main difference between the two is that in the case of L-glutamine, the crystal contains a normal hydrogen bond network. In the case of L-pyroglutamin ammonium, we have a very strong hydrogen bond network. And as a consequence, while in the case of L-glutamine, the system is completely non-fluorescent, in the case of L-pyroglutamin ammonium, we have fluorescence in the visible range, which in principle suggests that there is a connection between the presence of strong hydrogen bonds and the non-aromatic fluorescence. So together with Gonzalo Diaz-Miron, Gonzalo is a PhD student in Argentina. He's a really, really bright student who was here until a couple of weeks ago. Together with Gonzalo and Ali, we did a set of calculations in order to try to understand the connection between the strong hydrogen bonds and the non-aromatic fluorescence. And here what I'm showing you in the screen is a set of non-adiabatic molecular dynamic simulations where we compute the non-radiative decay probability as a function of time evolution in the first excited state. But what we did was compute this non-aromatic decay probability for the L-pyro case, the system that contains a strong hydrogen bond, the system that fluoresces, tuning the hydrogen bond strength at different fixed distances. Basically what we see, for example, in the blue curve, we have a very weak hydrogen bond, and we see that the non-radiative decay probability increases really fast. But when we gradually increase the hydrogen bond strength up to the level of a strong hydrogen bond, we see that the non-radiative decay probability decreases quite a lot. So this shows that there is a direct connection between the strong hydrogen bond presence and the non-radiative decay probability. So, of course, if the non-radiative decay probability decreases, the radiative decay probability increases. We have a higher lifetime in the excited state, and therefore we have higher chances for the system to fluoresce, so higher fluorescence yield. So after this work, we continue studying the system in more detail. What I'm showing you here is, for the case of glutamine, the system that contains a normal hydrogen bond network, I'm showing you the potential energy surface. And this is basically the experiment. We excite the system from the ground to the first excited state, and then we let the system evolve in the first excited state. And what we find is that there is a conical intersection very close to the ground state geometry. So for those of you who are not familiar with conical intersection, a conical intersection is just a point or a seam in the potential energy surface in which two or more electronic states are degenerate. So, of course, these are critical points for non-radiative decay. These are the points where the non-radiative decay occur. So what happens here is that the system can reach a conical intersection, can access a conical intersection without a barrier. So really, really fast, once we excite the system to the excited state, it reaches a conical intersection, and it decays to the ground state without any radiation, without fluorescing. This is why this system is non-fluorescent. But what happened with the case of El Piro, the fluorescent system? So as I'll show you at the beginning, the potential energy surface for the ground state is quite broadened, and therefore the position of this conical intersection ends up higher in energy and further away from the ground state geometry. And therefore, once we excite the system to the first excited state, it takes much more time for the system to reach the conical intersection. So the life in the excited state is increased, and the probability of fluorescent is increased. So I don't want to deviate to another topic here, but I want to introduce this technique that I will use in order to study properties of the conical intersection. So this technique is a spectroscopic technique called PAMPRO UVX-ray Absorption Spectroscopy. Basically, it is a non-linear spectroscopy in which we shoot the system with two pulses, two lasers. The first laser excites the system from the ground to the first excited state. Then we let the system freely evolve for some time, and then we probe the system with a second laser, but this time with an X-ray laser. For those of you who are not familiar, the X-ray excite core electrons, OK? In our case, in particular, we use soft X-rays, which means that we are exciting core electrons to the valence orbital, OK? So again, without entering much into the details, I could just talk about this during this talk, but this is not my intention. I just wanted to mention that one of the most important advantages of this technique is that because of the high energy of the X-rays, it has a really, really high time resolution. So it allows us to study processes that are really, really fast, and they're impossible to study with other ultra-fast techniques. And therefore, with this time resolution, we can not only study the motion of the nuclei, we can also study the electron dynamics, which is really important if we want to study conical intersections. Why? Because the passage through a conical intersection is extremely, extremely fast. So it's so fast that right now, there are almost no experiment probing them directly. In general, there is some sort of conjecture, OK, in order to, I mean, you have some sort of assumption to think that they are in certain place in the potential energy serve. But just to motivate you about the study of conical intersections, conical intersections are essential to understand photochemistry, any photochemical process, OK? So processes so fundamental as the photosynthesis or the vision or the stability of DNA against radiation damage can be explained by the position and the shape of certain conical intersections. So these are extremely crucial. So for me, being able to study them and eventually to control them is a really, really exciting topic. So how do we study them with X-ray absorption spectroscopy? Basically, if we find an X-ray transition in which the final state is unperturbed by the passage through the conical intersection, then if we can probe the system fast enough to have some sampling before and some sampling after the conical intersection, we should see a branching in the X-ray spectrum, OK? We should see a branching in the X-ray spectrum because at that point, these two signals are degenerate, OK? So again, to do this, the high resolution of X-rays is extremely important, OK? So here, so one thing I didn't mention is that all of this is simulations. So the experiments, to do this experimentally, the technology is available, but it's still a really incredible challenge. So all of these are simulations of the spectrum. So what I'm showing you here is the X-ray, the pump probe UV-X-ray absorption spectra, OK? So this is the horizontal axis is time. The vertical axis is energy, OK? So at time 0, we excite the system to the first excited state. And so let's show it here. So at time 0, we excite the system to the first excited state, OK? And this band in the spectrum corresponds to the first excited state. And gradually in time, the system starts reaching the conical intersection and some population starts decaying to the ground state. So this second band is associated to the ground state. As you see, this band is decaying. This band is increasing. And here we have a branching in the spectrum where the two bands are degenerate, OK? So this allows us to probe in real time the passage of the system through the conical intersection, OK? And what happens in the case of L-pyro, the system that contains a strong hydrogen bond, the system that fluoresces, the system, yes, that fluoresces? Well, in this system, as I mentioned before, the conical intersection is much more difficult to reach. It's higher up in energy. So it takes much more time for the system to reach the conical intersection and to decay non-radiatively. And as a consequence, what we see in the spectrum is just one single band, OK? So this technique allows us to have a fingerprint associated to the presence of the strong hydrogen bonds, both statically and, more interestingly, dynamically. This technique would allow us to study the transient presence of conical intersections, OK? Of a strong hydrogen bond. So the last thing I wanted to say today is, OK, we would like to study, after this, we would like to understand what the system is doing when it passes through the conical intersection, OK? I mean, what are the nuclear motions or the electronic motions that justify the presence of this conical intersection, OK? So what do we have? What is the information that we have from our simulation? So here, this is a sketch, a very simple sketch of a conical intersection, OK? This is a coordinate that we don't know, OK? And this is the energy. This representation is called the adiabatic representation. It's just one possible representation in which everything that is on top is called the excited state and what is below is called the ground state, OK? But again, without entering into many detail into that, what we have is some points in the first excited state, some points where points are like, basically, some sampling in the MD simulation, OK? Some sampling in the first excited state. A few sampling around the conical intersection, just a few points because the passage is really, really fast, OK? And some points in the ground state, OK? So another possible representation in which we can write this plot is the adiabatic representation, which, again, without entering into the detail, basically is a representation in which the two states cross each other, OK? The two states cross each other. But if we can write a system in this representation, then the delta energy before the conical intersection would be higher than zero, and the delta energy after the conical intersection would be lower than zero, OK? So we can attempt, we can write a candidate for the coordinate of the system through the conical intersection in the following way. So the denominator in this expression is just a normalizing factor. These are just standard deviations. So let's look at the numerator here. This looks very much like a principal component analysis, OK? Because this is basically looks like a covariance matrix, OK? This is exactly the first term, is exactly the same as a covariance matrix. If I would write a covariance matrix here, instead of writing the energy in this side, I would just write another coordinate of the system. So the coordinate at time t minus its average, and this would give us a covariance matrix, OK? But this is more like a mixed covariance which yields a vector just because I'm using delta energy, which I have just one value per time, OK? So qualitatively, what this vector does is weight negatively all of the points that belong to the excited state and positively all of the point that belong to the ground state. And hence, this vector points from the excited to the ground state, OK? So this is our candidate for a coordinate associated to the conical intersection, OK? In this graph, what I'm showing you is the projection, the scalar product between the vector associated with our coordinate for conical intersection with each one of the vectors associated to instantaneous position during our molecular dynamic simulation. And what we see first in the graph of the left is that we have a bimodal histogram, OK? And one of the peaks corresponds to the first excited state and the other corresponds to the ground state. But more interestingly, if we compute this, if we evolve this histogram in time, we see a smooth transition from the excited state to the ground state, OK? So interestingly, if we take a look at this vector, this gives us information about what the system is doing around the conical intersection, what the system is doing when it passes through the conical intersection. And what we see is that, well, first of all, again, this is the glutamine. This has one hydrogen bond here, OK? And so what we see is that most of the motion associated to the passage through the conical intersection is located around this hydrogen bond, which is very important. And also, we can describe this motion like in three components. One of the components is an intermolecular separation between the two hydrogen bonded glutamine. Another one is a decrease in the strength of the hydrogen bond. When we go from the excited to the ground state, there is a decrease in the strength of the hydrogen bond. And the other one is a planarization of this carbon with respect to its neighbors, OK? So in the couple of last slides, I wanted to show you something that, for me, is extremely interesting, which is the following. Here, instead of the product of the scalar, sex scalar product of the coordinate of the conical intersection with the trajectory, we're writing just the hydrogen bond distance in each one of the frames of the multiple molecular dynamic simulations that we did, OK? And what we see is that the hydrogen bond distance in the ground state is higher. So the hydrogen bond is weaker in the ground state than in the excited state. But more interestingly, if we look at the points in the simulation that belong or are very close to a conical intersection, what we see is that when the system goes from the excited to the ground state, it first passes through the conical intersection. And hence, it has to compress the hydrogen bond, increasing its strength up to a level of a strong hydrogen bond. So basically, to go from the excited to the ground state, the system transiently becomes, the hydrogen bond transiently becomes strong, OK? And this we can see more clearly in this picture, where we have the hydrogen bond distance in the excited state. And then we see that it increases. At the beginning, it goes, it gets stronger before it gets weaker, finally, it goes to the ground state. And this has a very nice consequence. And this is the last slide, which is the following. This is the electronic structure of the system, OK? The balance, electronic structure of the system. So let's take a look at the first column, the left column here. So when we are far from the conical intersection, the HOMO is completely located in the donor molecule of the hydrogen bond, OK? There is almost no participation in the, there is almost no contribution in the other molecule, OK? And the LUMO is the opposite. It's completely located in the acceptor of the hydrogen bond, OK? But when we get close to the conical intersection, because we are forming a strong hydrogen bond, and because the strong hydrogen bond behaves like a tree-centered covalent bond, now we have a covalent bond between the two and the orbitals get mixed with each other. So here we see that the HOMO has some component of the LUMO and the LUMO has some component of the HOMO. And this is what justifies the generacy that is formed in the conical intersection, OK? So, OK, so to conclude, we show that hydrogen bonds, that strong hydrogen bonds can delay the passage through a conical intersection. And as a consequence, there is a retardation that increases the lifetime in the first excited state, incrementing the probability of fluorescent. We have shown that the PAMPRO UVX-ray spectroscopy can reveal both the electronic and nuclear dynamics around conical intersections, showing a unique signature associated to the presence of strong hydrogen bonds. And finally, the conical intersection coordinates in normal or standard hydrogen bond system implies a transient formation of a strong hydrogen bond, while in the case of strong hydrogen bond system, it implies a proton transfer, OK? So these are the people that participated in this work, mainly Gonzalo and Ali. So Gonzalo did most of these calculations, as I told you before. Gonzalo is a PhD student that is finishing his PhD. And yes, he did most of these calculations. And all of these calculations were done with the LIO project. LIO is an electronic structure code that we developed in the University of Buenos Aires. These are the things that the code can do. And if you want to, if you are interested in using it, the code is free and open source, you can download it in this website, or you can just send me an email, and I can help you with that. So thank you very much. Thank you very much, Uriana. We have time for a couple of questions. There are questions. We'll be also in the chat if you want. In any case, so I'll be in the afternoon in the, how is it called gathering? Gather.time, yes. Yes. So if you have questions later, I'll be there. Okay. If there are no questions, maybe we can wait a couple more minutes and move to the next and last speaker.