 Okay, so I think we can get started with today's colloquium. So welcome everyone to our colloquium today on this cold Bora evening in Trieste at least. So it's a pleasure to welcome you all to today. I'll start off with some words of introduction from our director Atish. Okay, welcome everybody. I think it's a pleasure to have Professor Julia Gully. As a speaker today I distinguished businesses that she's a new family professor of electronic structure and simulations. In the printer school of molecular engineering and professor of chemistry at the University of Chicago. She holds a senior scientist position at Argonne National Laboratory. She's the director of the Midwest Integrated Center for Computational Materials. And I'm told that she's not a stranger to Trieste. She got her PhD at CISA with Maria Tosi. And she has visited ICTP on a number of times as a lecturer at our workshop. Her visit was in 2013. So and she will be talking about materials and molecules through in silico lenses. Okay, thank you very much, Atish. Now hand over to Ralph Gibarro, my colleague. Yes, thank you very much. So welcome everyone. And I must also say so it is really an immense pleasure to welcome Julia Gully to our ICTP colloquium here. First and foremost obviously because Julia is a personal friend of many of us here in the Trieste community and we know her very well. And as Atish has just said, Julia is in a sense a daughter of Trieste and of the Trieste scientific system with her PhD which she got here. And also, I mean, in normal times I would say that Julia is perhaps here once a year or so for some workshop or some collaborations or some some conferences you name it whatever happens in Trieste, Julia Gully has probably already been taken part of it. So she is really a very closely and a dear person to many of us. And I would also like to say that as a person who work myself in the area of atomistic simulations and electronic structure calculations. So for me it is always a big pleasure and always very interesting to follow the line of work of Julia Gully. And this, well, because I think she has been working kind of in all aspects which go along with this science in atomistic simulations. And let me start first of all by saying the basic things is method development in our field and Julia has done a lot of method development so her name is associated with a lot of advances in GW and BSE, and she has created some Peter equation approaches. So these approaches which are based on many body perturbation theory. And there she has done a lot of really fundamental and method development work. But so, this is one aspect is whether and development which they are always interesting developments coming from Julia school, but then also Julia is involved in implementing her method. There are other kind of major codes, which come from Julia's group on or at least in which she has been also very heavily involved like for example the West code, which is a code which precisely implements all these many body perturbation theory approaches. Julia as Pfizer has also been very heavily involved in the Q box electronic structure code so these are method implementations which obviously require a lot of HPC knowledge and so on. And Julia is also kind of in this area and interesting contributions. And if you look at those codes you realize these are codes which scale very well to huge numbers of processors and huge computers. And this is because Julia is kind of a person who, who typically does her calculations on some of the fastest computers in the world. And she is also known for really very large scale electronic structure calculations and pushing the limits of what we can do with our methods and so on, large scales. And so, apart from methods and implementing the methods. Obviously, Julia is also very well known for her applications. So three life applications of matter, sometimes the matter under extreme conditions, and very often, as I've already said many body perturbation theory, it goes about interaction with light. So recently in Julia has been working a lot on photovoltaics and therefore kind of societally important application so I can, I would say from my perspective as someone working generally in the field of electronic structure. There is a lot of work which can be there, they are there in Julia's work so she's really a really very complete scientist, which is kind of also witnessed by all these prices and the vaults which Julia is I will not start and naming all of them. Artiches mentioned some of them. So, as an ICTP person, I would like to mention in the end one other aspect Julia has in the past also worked with one of our diploma students so one of our Vietnamese diploma students in fact when when Julia was still in California went to a group and did his PhD with her. As I know, both of them are very happy about that collaboration. So, hoping that today many of our current students are connected and are listening to us now so my message to all of you diploma students is aim very high, you can arrive you can achieve anything so aim very high work very hard. And if you do this, then you can perhaps even one day work with people like Julia Gali. Okay, so this is all I wanted to say and I give thank the floor to Ali and Julia. Thank you very much for that very thorough introduction. I guess before we get to the main finale. I have a personal anecdote I wanted to share. I'm not sure you remember this Julia, but we were at a Gordon conference once on water in New Hampshire. It is always stuck with me because I was a young student at the time. I remember you talking about your simulations of liquid water and getting the numbers right for the right reasons. And this always stuck with me. And until whenever I'm introducing my students to computer simulations of water, Julia's papers are really the milestone landmark papers that all students must read. So with that introduction, Julia, the floor is yours. All right. Good afternoon, everybody. And thank you so much for here to speak today and thanks for the very nice introduction. You know, usually I don't start a talk by the picture of myself, as you may imagine, but today I couldn't resist that this is the picture of our class of sister. We didn't have offices at that time, but we were occupying some rooms at ICTP and this is our class of 1983. And this is a picture that was taken on this pair of ICTP on that summer. And the one in the back with the red glasses. And, you know, since then, my activity, both scientifically but also personally has been very much connected to Trieste. And sometimes during the years, I would go as route to Trieste twice a year, certainly most often in summer and of course this year it couldn't come. Anyway, it's a real pleasure to be here even virtually today. So I'd like to tell you are some stories, atomic level stories of material that I think have changed society or about to change society. What do I mean by this? I will give you some examples of these materials and some very well known that we have not studied ourselves and some that we are studying ourselves that we think can be game changers. And then I will try not to give you any details, but an introduction on why we think that theory and computation are really indispensable parts of the discovery process of material and then I'll give you some examples of material for sustainable energy sources and I will define what I mean by that and also material evidence to quantum information. Certainly one material that everybody knows that has changed society is it's the material that it is that the transistors that power your iPhone, as well as for example the fastest high performance computer that you find in the world that this is summit I will mention this computer again at Oak Ridge National Lab. And I found this statistics and this is what Intel a little bit older. So, since 1960s, there have been a total of more than 10 to the 22 MOSFETs that have been manufactured worldwide as a number which starts being very close to the avocados numbers so, and of course, Brattain, Burdine and Schachtley won the Nobel Prize in physics for the discovery of the transistor so this is, you know, silicon interface with silicon dioxide is the material part of the system. The other that probably also is a material, you know, is an oxide is more complicated lithium oxides and which is in all batteries that are powering in the devices and this is much more recent year to good enough we think and you should know. So the examples that I will give you after a short introduction are what I consider, you know, emerging materials, some for sustainable energy sources and what I mean is a very general concept. Simply energy sources that do not destroy the ability of the human race to live in the planet. And materials maybe related to membrane that can make broadly a clean water. So these are two old picture and they still use them to make two points. This is an old paper that appeared in science in 2013 and the title was the closing door of climate targets and this is a very powerful graph in my opinion that tells you what is the concept was of inaction. So how much you would have to reduce your CO2 emission depending on when you start and you know we are already in 2020 and our price of inaction is, you know, very high. And the other is the map and again it's a little bit old but it hasn't changed much of water what's called water strength, meaning the really incredible inequality in the world of availability of water, you know, the blue regions are regions where water is available to at least to some extent to the average of population and the red. It is the opposite and also this tells you that talking about average number sometimes is not the right thing to do if you average the availability of water in the world you would find an average is not that bad. But what happens is how it is distributed. And as I mentioned, two minutes ago the other materials that we have been studying recently and that one day they will also be related for their use. So I'll get to that and I'll show you some examples of these very broad topics, but before I would like to tell you how we approach the problem from the standpoint of computation. And to main concept, there will be a problem and of course it's not by no means the only way but it's to look at materials at the atomistic level to look at how atoms and molecules interact and even, you know, trying to understand how the electrons in the nuclear that are in the material interact to make chemical bonds and using theory and computation based on the laws of quantum. And then we implement that just as Ralph told you into computer codes and we do computer simulation to as accurate as possible solution as accurate as possible to describe natural phenomenon. So this is the general idea and within this general idea of this very broad category, I will now tell you what, you know, I will zoom in a little bit and I will tell you what what we do specifically. In our community and also those in my group. So we are looking at systems that are not only in the order solids, but maybe they are complicated liquids or our disorder system and, as I said, we use the laws of quantum mechanics and we use the laws of quantum mechanics. There is nothing new about those and actually the equation that in certain forms we saw have been known for, you know, more more than 90 years and this is the Schrodinger equation and this is the Schrodinger equation. And most of the examples that I will show to you are examples where actually we saw the time independent Schrodinger equation. This is a beautiful equation that we don't know how to solve. Exactly, except for, you know, system like hydrogen atoms or some kind of helium systems. And so we do need approximations to be able to apply these equations. So this is a beautiful equation that we don't know how to solve exactly except for, you know, system like hydrogen atoms or some kind of helium systems. And so we do need approximations to be able to apply these equations. So this is a, I tried here to give you a historic perspective of the real use of quantum mechanics and the approximation that we're doing quantum mechanics came about and are used to study materials. So I divided the approximation into three main categories. One are so called mean field theory is this mean, it means that we look at electrons that are interacted with nuclear in a material in a mean field the one electron sees the mean field of others and you know, specific the mean field of other let me put it this way and this gave rise to the development of density functional theory that was, you know, the development started in the 60 and then specific approximation that now we use started being presented actually 20 years later in the big field is the field of quantum chemistry where instead of, you know, the basic variable density functional density of electron we look at complicated wave function, and also, you know, so called stochastic approaches but they will not go there quantum on the car also. So these are the approximation, but you know one thing is being able to make meaningful approximation the other thing is being able to translate them into practical solution that then apply to studying materials and really the ability to compute algorithmic and computational development. And this theory came in the 90 and you know, and of course, at the end of the 90s that it was recognized how is worth to study molecules and materials with a Nobel Prize to hold upon and jump over the chemistry in 98. And at a time that was also very important development circle important for my career, but also for, you know, the community in general, which is that the fact that. The ab initium molecular dynamics was invented or, you know, corporate method also very was invented in three aspect was invented system so this is a very much a three aspect that product of which we are all very proud. So what is this, this is the description of our density of electrons in the material looked at by dancing something with the Newton equation of motion so that we can study also materials at final temperature and I believe in this development contributed to really unifying not only, you know, density, you know, density functional theory and Newton equation but we find on this matter physics and physical chemistry in a way that in terms of the problems that people were were looking at. So what do we do. We have a material the material is a finite temperature the atoms move and in the first approximation we describe them by solving the Newton equation of motion at people and a and the forces that to be acting on these nuclear which comes from all the left turns around. We compute by doing calculation of answered functional theory in a smart way. Not of caution. The calculation is in this theory that in principle is exact that you choose an approximation of certain traction. So, for the expert in the audience when people tell you, I'm doing DFT asked them, which DFT, meaning which approximation of DFT. And I will not specify today which approximation I will be using in the example but that's a very important question to ask. This turned out to be a very general approach to study a number of system and some I will give you example from my own work, but many people in the community are tackling these problems like assembly of nanoparticles for sewer energy. For example, or very complicated system together to try and for example split water to produce hydrogen or to reduce your tool for clean fuel or defective solids and I mentioned effective solids in the context of information technology and even more complex system and this is an example that I took from the website of Ali Hassan Ali in Trieste, a system that are much more complicated like system of biological interest like amyloid proteins and you know these are, this gives you an idea of the breadth of problems, not all at the same level of accuracy but at the breadth of problems that can be studied with this approach. So, all of these equations. So, this is a very much like and of course you know there are people, people like John von Neumann who understands everything and in 1963, many other things that of course we need to use high speed computing devices that in the field of non linear partial differential equation will make a big difference. And this is exactly what we are sold. The equation of DFT or Konchan equation are non linear partial differential equation. The exact question is not that these are non linear. So, from 1963 to today, a few days was the announcement of a Golden Bell Prize, which was awarded by a team led by Roberto Carr in Princeton and Wayne and Hank in Princeton, with many colleagues around the world, the University of Berkeley, also in China, and were able to study initial accuracy at 100 million atoms with machine learning. And this was run on the largest supercomputer in the world that I just showed you in the first slide. So, what does this mean? So again, Roberto Carr is one of the inventors of the first response like the dynamics that we asked that. And in some now. So, what they did, they were able to generate the many configurations and by doing it by solving the DFT equation and then to use machine learning techniques actually quite, you know, both elegant and complex machine learning techniques that were deep neural network to solve the equation from a number of atoms that was totally unthinkable, even I would say five years ago. And it's not that all problems are solved, but I, and you know, there are problems of time scales, this cannot be done for any system right now, but I think that this is important development that I thought I would mention today. Okay, so what do we do in general? We use first principle molecular dynamics and density functional theory to study system. And what does it mean, all this machinery to get an idea of the structure of the system, how atoms are arranged. Then we want to characterize the system. And we want to understand how the system interacts with light. And so we do spectroscopy. And this is what actually Ralph mentioned at the beginning we, and many others of course starting in before us of course, develop method to understand how electrons react to light, because we want to probe a material with light at times because we want to set up devices. I'll show you devices for quantum information that are based on the reaction on the interaction materials with light. So, top of the developed method to look at integrated and also to probe spins, you know, spins, the probing of spins is very important again for quantum information. So, let me tell you about the development that I'm quite excited about that we have done in my group it's a little bit technical but I'll try to explain it as much as possible as a in general terms. So, when you look at the interaction of the electrons in your materials, the interaction is very well known. It's the cool of interaction, except that you know electrons are quantum mechanical power principle and all that mess but it's very well known. You can try to represent that interaction in a way which is convenient by the interaction for light. And this is the sort of screened cool of interaction so you sit on an electron, and you said, How does it interact with the other one, given that there is a sea of electrons around. So, you want to go from your what we call unscreen cool of interaction to a simple interaction. And there is a filter in between, which is called electric magic for the expert. And what we did recently is to actually use machine learning techniques, learn this filter for specific configuration of a material that then we use for many other configurations. So this is another example where machine learning can be used in very, very different ways that compared to what I just told you for just for for the first principle molecular dynamics and also be used to look at the interaction of electrons with light. Okay. So, before giving you examples of what we did the by studying. By studying, I would like to make a clear statement that, you know, there are approximation that you have to implement your approximation and then you have to have verify validated and optimized codes to really make some prediction and I do think that open codes are critical to innovation, there are many centers in the United States, including one at Oregon, where we are involved and of course, in Europe, dedicated to this and I think another, you know, famous open cold quantum espresso is very much related to Trieste, and he has so much contributed to critical innovation that again you have open code. Okay. So, with all this machinery density functional theory connected to the new Newton equation first like your dynamics. And then spectroscopy on top of this with many body perturbation theory method armed with all this we now study materials and I gave you two examples. So, what in my group, we have been focusing on in the 10 years or so our material to convert energy from the sun into electricity and material to produce clean fuel so to convert water into hydrogen and oxygen and much less we have worked on the second one but many people to convert CO2 to benign and useful. And we have also been recently in the last five years or so working on materials to harness the power of quantum information for quantum sensors. All right. So, why are we still studying solar cells and when, you know, silicon actually after so many years in refinement works pretty well. And it's a roof of many houses in the developed world. And the truth is there is no size that fits all, and it's not only a matter of efficiency. So this is a graph of cell efficiency as a function of years that is maintained by the United States and our year. And so why aren't we all using the most efficient solar cell because it depends on the application. It depends how it is, the cell, depending also on the country you are in. And so one of the materials that people are looking at is cells made of colloidal quantum dock. Why is that? Because these are very easy to make. And, you know, the hope or the dream is that they can be made everywhere very easily and they can be deposited wherever you want. Also in country that don't have all the technology to do, you know, to refine silicon into hand and the beautiful solar cells that we have on the roof of California, the houses or stuff like that. So what is a quantum dot solar cell? There are these big molecules that are made of a material that is good to write and to create electrons and create electricity. They can have organic ligands or inorganic ligands. So what is it that you have to do if you want to study this system? And I use this arrow to tell you that it's not just a matter of doing high throughput for specific materials that are the material that are inside this big molecule. You have to understand a lot of properties, the structure of interfaces, and of course the interaction with light, which is what I just told you about. And then you have to understand the transport. So when you have in mind of doing first principle simulation, you have to have in mind the fact that you have to do many different type of simulation and tackle different properties and when you put together the results then you really are able to solve the problem. So our contribution in the last few years has been to understand something that you know after the fact is kind of obvious, but we now have a strategy to understand this. And the thing is, you have this ensemble of nanoparticles like the one here on the right, they absorb light, but then you have a mass of ligands, right, other molecules around your nanoparticle. And these are very important, the use of these ligands around are the important thing to understand, it's not only how they absorb to understand experiment. The reason why I bring up this example is that at the beginning of many, in many simple models for this system, the ensemble of this, instead of this mass here of ligands, what is a very simple, you know, arrangement of these nanoparticles. And when I started talking many years ago with my experimental friend, they would always tell me, look, we just need a simple model, we don't need all your mass around the nanoparticle. Well, it turns out sometimes that simplified structural model are insufficient to explain even qualitatively the electronic property of this system. Sometimes, and I go back to what Ali said, and thanks Ali for reminding me of that, it's that sometimes simplified models cancer, but they give you an answer for the wrong reason and then you don't understand the physics. All right, actually the examples of water was a little bit different way. And that's, that's what it was. So this is a much simpler problem, absorbing light and the material and generating charges, then using light to store in chemical bond and triggering the chemical reaction. And this is what we have studied and what is of interest to so many people by looking at the water splitting in hydrogen production. So basically you have a material which actually is pretty messy. This is an example of one. It absorbs light, again, are just that are created in the material, but then you don't want to harvest those as electricity you want to use them and maybe make them go through a catalyst that to make to split water to do this chemical reaction. Very complicated problem, incredibly complicated problem. The materials are defective. There are issues all over the place. You have to understand a lot of properties and the properties actually are listed here. This is a complicated slide, but just concentrate on the left part in the right part. I think it gives some cartoons to explain the left part. So you have water in contact with the material, and you need to have a reasonable model of water actually salt water in many instances and the we and others and how your course was involved and many others have spent many years to try to have a reasonable model of water right to many very important for so many things and also for this energy application. Then you need to have an idea of the electronic structure of how the energy levels of the water and the outside. Behave because you wanted the charges to go from your material to water to make chemical reaction, and then you want to transfer the charges and so mobility design. Here I want to make a statement about the importance of connection with him. So these are incredibly difficult problem, or theory can indeed make a very big difference, but only if it's really integrated with experiment, and you cannot just have the, you know, a solution of a problem at least at this stage of our understanding without going back and forth with experiment many times. And the way I see the interaction with experiment is not like, you know, sometimes we go this arrow theory experiment experiment here and blah, blah, blah. It's very complicated. Yeah, there are many pieces of different properties and we go back and forth between theory and experiment. And it's only by setting up a really close connected strategy that you can make a difference. And here I'm just pointing out to you, one of our recent work in collaboration function choice with medicine in Wisconsin where we, I believe, accomplished at least in part this close connection between here and experiment to study this problem. And since Ralph at the beginning, one of the ICPP students, I would like to tell you that our very first paper on really the band offsets how the electronic system interface with the water came that was done in 2014, and on was a student at ICPP. He came to work with me when I was in California and we continued working together for many years. And he is a staff scientist at Lawrence Livermore lab right now and he really became an expert and has made many important contribution writing paper, acts of science and so on and so forth, studying problems similar to that type of just mentioned. Okay. So, are there open problems plenty. And I believe that one of the most difficult open problems is that a really predicting good catalyst, meaning you absorb light. There is a material with absorb light, but then you want to trigger a chemical reaction and you need a catalyst. There is no material that can absorb light that can really catalyze that reaction. There is an incredibly problem, in spite of many contribution by many people. So in case you have any that catalyst and understand catalyst is important. Let me remind you that a new catalyst can be a game changer in the world and this is the example that I think is probably well known to many. For Fritz Haber in 1908, the discovery catalyst would combine atmospheric nitrogen with hydrogen to form ammonia. Of course the catalyst is not exactly what we use today for the rain, not great, but yet what Haber did and then what Bosch did with mass production, based on a scientific discovery of the catalyst change the work, right, because the world is moving. So I don't know if we will ever get there that we will discover the water splitting and or to do CO2 reaction in that way, but certainly looking for catalyst is a difficult problem at that time, so completely experimentally, and it's a problem worth putting some effort on. Okay, so how are we solving complex problems so so far we are solving in from a theoretical perspective complex problem by trying to solve to solve this equation very high performance computers. And I showed you as an example 100 million atoms are done by Wayne and her about the car last last month. Showcase last month and then in the last five years or so. There are many complex problems that we are not yet able to solve not only material science but also, you know, in biology, this is another famous catalyst and, of course, for example, high to superconductors we understand something about those but it's not that we can simulate and predict high to superconductors really well. So there is a change of perspective right now in the world relative to which was started by Richard Feynman in 92 and also also by other. A beautiful lecture to find gave at Los Alamos in 93, like many beautiful lecture that he gave talking about tiny computers obey quantum mechanical laws. And he proposed the idea of creating machines based on the laws of quantum mechanics instead of the laws of physics right to have a qubit. And also to positional states instead of a bit like in the silicon that I showed in my very first slide, and also to exploit the fact that that can be entanglement of the mechanical object. And that you observe a system in a certain place, it is affected in another place and they continue to be entangled and then discoverable because of that entanglement. So I think that the story short, of course, I'm talking about the fact that people have today realize the quantum computers to some extent. There is no near term quantum computers, because we don't have error corrections yet. And so, as I will show you an application in in two minutes, quantum computers can can be used that they have the news for quantum chemistry but they still give very noisy results and certainly in the field of material science and quantum chemistry they cannot yet give you results that are better than classical computer and it's a problem worth tackling because of the potential that they will have going into the future. What are the major challenges one I told you a bit large scale for torrent quantum computers many many people and many, you know, industries IBM Google, you know, you name it to this problem. And then there is the problem of the coherence of encoded quantum information the fact that, yes, you can encode quantum information but depending on the system that you use the hardware that you use they can lose the coherence and then the information is lost. It's a hand waving explanation but let me put it that way. So what are we doing in this context as you know many other people in the world that we are looking at materials that can be good to realize new computing technologies but also nano scale sensors. Most of the quantum computers that you see today operating are built on superconducting qubits and but here I'm talking about having a defect in a solid very simple like a defect in diamond silicon carbide. This defect has certain spin states read on dressed with light. And again, we go back to the importance of understanding the interaction with life and that can form a QB. Where you encode the information. And this is an example of spin mass, not because spins wear masks, but because wearing masks. The reason for wearing masks is the same as studying spin is science. Okay, not probably maybe you don't need this reminder in Italy but we do in the United States. Okay, so we in again in collaboration with experimentalist we have studied these people. encode you know this spin defects in a material silicon carbide, and especially in silicon carbide that are realization in diamond a little bit expensive also you know diamond is the, you know, you can actually grow much smaller at the same cost, much smaller material than for example the silicon carbide. And we have addressed exactly the problem of spin coherence I will not explain this graph exactly but this is one of the major problem that are associated to all the quantum with the knowledge in decoherence and loss of information because of the decoherence between your spin. And we have looked at different type of materials and they're recently and this is in one of our also at nuclear memories and how the material and this plane in the material can be can affect the way spin. behave. I think that this is the ending failed the work, even with the approximate DFT calculation for material properties, coupled with spin Hamiltonian that we use to draw this graph on the left, you can understand, maybe only qualitatively but you can understand important properties in working with experimentalist that you can really contribute to to solve a problem. Okay. DFT can help you. However, there are states, especially for these spin state that cannot be described the DFT they are highly correlated states. And one thing that we have done recently and again this is a little bit technical now we have developed a quantum embedding theory. What is it. We have a defect, and this is complicated defect the red thing here, and we describe the states of that defect here with an active space electrons and then we look at the effect of the environment with DFT. So why did we do that. Well, we did that because, number one, we wanted to study the study of this complicated spin defect in a solid, and we wanted to do this for very large system because we want to have a realistic system with experimentally But we also did that for another reason, because if you simplify the problem to an active region of a solid, then you had a Hamiltonian that you can solve by complicated quantum chemistry methods like configuration interaction. And these are the methods that you can run much better on a quantum computer than on a classical computer. And so we have done this in this complicated slide. We have done this on an IBM quantum computer. So the part on the left is just to tell you that we check on a quantum simulation, you would have the same result as we have in a classical computer so this is just a verification slide. On the right hand side, you see the certain eigenvalues of our speed defect on a quantum computer. You notice two things. One, when you do a quantum simulator, you get the results that you want. We set it at zero, but the blue line gives you the results you want. When you have a quantum computer, you get the orange curve. It converges your algorithm, but it converges to a larger number because there is noise, and this noise is still there in the quantum devices that we are using today. And this is the big problem that we at the community need to address if you want to use quantum. Notice that the difference between zero and the orange curve is 0.2. This is a large number if you want to understand how electronic states behave. Okay. So, you know, the vision here, so this is a map of Middle Earth. If you are in a position of Lord of the Hobbit or something. So, we want to do materials prediction and design and for sustainability but also to do predict materials for normal computing and sensing devices on which we would actually then perform quantum simulations and you know this is a cycle if we do this then our prediction of course will improve and so on and support. But the caution in this field like in all new field to, you know, hide sometimes it's good to just, you know, bring, you know, the attention of people but as again, let me say five and again for a successful technology reality must take precedence over public relation for nature So, let's go easy on what we present and let's stick to science sometimes presentation of these, you know, potential also quantum simulation and a little bit out there. Okay. The last thing I want to say before closing is about data I told a lot of simulation all sorts worries, what are the data. One thing that we have been trying to push in the last few years is actually to make scientific papers kind of alive, instead of having only a paper where you present your data to make data on a paper by paper basis available to the public. You know, suppose that you instead of having a PDF when you click on the PDF in our PRB or whatever JCP, you also click on a link that gives you access to all the data that you have in the paper then you can mine it and then the ability of data also tells you that you can do machine learning much better or high throughput Anyway, we wrote a paper, you know, last year to to explain and we have a graphical user interface out there to help doing that if you want to have a look. Okay, so I would like to thank my my group this unfortunately is a picture of my group these days that we used to have a beautiful picture in the quad in Chicago but now we have pictures in zoom. And I mentioned my funding agencies and all of my collaborators, actually not all of my collaborators but I mentioned there and I will not read all the name, the collaborator that are part of papers that I cited today, including Anna. And I said, as I said, it's he's at Livermore today. So I would like to thank you for your attention and all that we have a number of position available both in Chicago and at Oregon, and happy holiday everybody and thanks again for inviting me here I'd be very happy to entertain any questions we have. Thank you very much Julia for this wonderful overview, the fantastic talk very clear. So, we will now move to the Q&A session. And so we have the way this will work is we can take some questions from the Q&A chat, and you also have the possibility of raising your virtual hand yourself. And then you'll be allowed to ask the question yourself. So I'll start Julia with a couple of questions from the chat that are coming in. So, the first one is, can we use electron excitation instead of photon excitation to probe defect states for quantum computation. So, we do. So, okay. This in a way is a complicated question. Shall I stop sharing? Okay, so, yeah. We wanted to realize the system which is a defect, which has this chronic states and the trend between these electronic states are what is important to probe the system and store information. So, the electronic excitation are related to these are off the patient and they are related to how we pulled the system with light or all lighted both frequency, I don't mean, you know, visible light to actually generate chronic excitation that we use to store information. So, in a way, the answer is yet because these are optically address systems. Okay. Thank you for the question. We have another question from Sara Fazio was asking, what is the experimental status of spin qubits so decoherence versus gate times. Yeah, yeah, yeah, that's. So, there have been really, you know, huge problems saying the last two or three years. There are people that have been published very recently in, you know, the journal that you may expect the science and nature both from Chicago, Harvard and, you know, many other places where they have done for spin qubits in semiconductors. You know, they have made so factor of 10 and now they have milliseconds for coherence time. Having said that, I think that my experimentalist friend and you know, they would certainly say that they don't think that the first big application spin qubit in semiconductor will be for quantum. But it will rather be for quantum sensing or communication. If you use, for example, spin in nuclei for as registers. So, people are still putting a lot of effort on super qubits for quantum computation, certainly Intel is working to do qubit based on silicon, of course, and also defect in silicon. So, a lot of the, in spite of the progress a lot of the qubits that have been developed experimental in diamonds to be conquered by people have at this point in time quantum communication quantum sensing application more than quantum. Okay, thank you, Julia. There's another question coming in the Q&A from Luca Grisanti. So Luca asks, why the noise affects quantum computing return a constant offset, respect to the corrected simulation. Yeah, that's an excellent question and of which I don't know the answer I wish I did. Okay, so the point that I showed you are, you know, I showed you an orange nice curve here, but, and I didn't put points but at each of these iteration points and these are averages over many, many measurements so called many measurements actually. A thousand of measurements. So, there are small error bars here that are not indicated within that small error bars, why the noise is constant I unfortunately do not know. By the way, these are results that are extrapolated using noise model that IBM gave to us based on the specific hardware. Yeah, I wish I knew the question, the answer and it would be an important answer to know but I don't know at that time. Okay. So we have another question in the chat from Christian, who asks, you know, can infrared the ramen spectroscopy be used to understand electronic properties or just structural properties and what is the connection between spectroscopy and structural properties. Yeah, yeah, that's another interesting question. Yes. So, it can, and it has been used by many people including our chairman here Ali Asan Ali, who has done work about spectroscopy of liquids and water and also ourselves in my group. And so, for example, let's take the example of water since I talked about water in contact with oxide, you can use first principle calculation to compute, you know, dipole moments of the molecules, as well as polarizabilities. Then you use those to do correlation function that gives for a spectra in the ramen spectrum and even you can do more complicated spectroscopy like some frequency generation. And this has been done by many people for, for example, hydrogen bonded fluids under many different conditions, but it can also be done in solids and you can compute polarization, you can compute from first many quantities that enter the definition of IR in the ramen spectrum. Yeah. Okay, we have another. Sorry. We have another question. Sorry, my connection was a bit unstable. We have another question coming in. So, what is the role of spin coherence in variational quantum icon solvers and how exactly it works on building a quantum computer? Yes, so, so I'm thinking here to give trying to think of an answer which is not too technical. Okay, so basically, you have a defect in this and the electronic states of this defect in the solids are localized in space and they are also localized in energy. You have a bunch of defects that's localized in energy for active space and then the environment. Okay, that has an influence on your active space. Or there are an occupation of electrons that can be with a certain spin, the spin can total spin can be zero, one, one half, whatever you want. And this here you have your spin. These are described by an effective Hamiltonian, which is diagonalized on a quantum computer, for example, with this algorithm. So, VQE, the variational icon solver that by the way, we haven't written ourselves and we have taken off the shelf from people who have written these algorithm to solve the second continuation. This is the algorithm, the connection with the Hamiltonian that was sold in one second quantization is the Hamiltonian that represent the electrons of the active space with a specific spin. And how VQE works, to be honest, it's a variational icon solver for and Hamiltonian, I'm sure I could tell you. And this is not something more than very vaguely, this is not something we have developed ourselves, but many people are working on and there are variational I can solver so that are made available to people by IBM and Google and others and we have used them. So, okay, let me let me say something, because, so this complicated connection of different codes, so you have, you know, your, all the DFT codes connected to GW because you need better approximation, then you have your Hamiltonian that you can do with pi fcf that you describe as pi fcf that is a code developed by Karnak channel and use variational icon solver using Qiskit, which is a solver provided by IBM. So there are several steps that I didn't explain in detail. Okay, thank you Julia for that. Since I'm chair, I will take this opportunity to ask you a question about, so you know you've been in this business of both ground and excited electronic states for a while for different types of materials. What's the status of non adiabatic molecular dynamics and it seems to me that there's not enough people. You know, maybe I'm wrong but it's not enough people thinking about how to accelerate those things and how to do those correctly and what types of materials, you know, and physical processes you need it for and you know, just like, you know, just like how you know, began, you know, many decades ago and being applied to many different systems. I don't see the same with non adiabatics and so I was wondering if you could comment a bit on this. Yeah, this this is an excellent question and I couldn't agree with you more than there isn't enough. There are a number of efforts in the community to look at this, but there are a number of systems and one is that really chemical reaction in water at the interface of an oxides that we and many others that happily neglect non adiabatic effects for, you know, what when we look at qualitatively at these reactions but we know that they are very important and we don't have a good way to address them. So, the development of this method before simple system has been going on in in some groups, for example, the group of Sharon Hammershipper now at Yale. She has studied proton coupled electron transfer phenomena for a long time where, you know, there is no way you can study those without looking at non adiabatic effects and that group has developed techniques to look at these effects. They are not going to where we can apply them to the kindness that I have talked to you about there are there are efforts, much more at the an approximate level in the group of work for us though at USC and I'm sure there are also other in Europe and in other countries. We are at the beginning and probably the community has not put enough effort as you said on these on these that are very very important and certainly when you have systems like chemical reaction in the interface that they will turn out to be even more important that we now we now imagine. So, I mean the short answer is that I agree with you about the importance of the need to put more effort and any of the of the effort that at least I know are giving their in say right now. Thank you. Any questions from the participants. You have the option of raising your hand as well your virtual hand and you'll be allowed to speak. So, or if you don't want you can still type them in the Q&A. I'll give you a couple of minutes. So just to remind you that at 530. We have a closed session with the diploma students with the Julia. If there are no more questions from the chat and no one wants to raise their virtual hand. Thank you everyone for joining. Thank you very much Julia for this outstanding talk. Wait, hold on. So, Julia, there's one one question that's just come in. And the question asks what your experience is with the pair of skites. Oh, yeah. Bit of a, I guess you could you could talk for days about that but anyway. So, I think that probably the question is about solar pair of sky. You know, related to, yeah, so we have done some work on it and we are, we are working a little bit on pair of sky. And the experience is that, first of all, for all the materials involved in pair of sky, you do need to do calculation taking into account to spin a little bit coupling and these are more complex than any of the things that I talked to you about. The excited state effects are much more complicated than just the in in nanoparticles. They are to the same old story. Interfaces and defects play such an important role that if you the interface and defects in your material really don't understand what's going on. And what we are doing now specifically is looking at the interface of this material with defects with protectively people are trying to put on these pair of skites to make them more stable and these are the materials that are not stable in there because they react with water molecules. So this is a very complex problem. It's also a complex problem because if you want to do first principle like your dynamics. For example, in organic perovskites, you have some molecules that are raffles inside of a very complicated and heavy, you know, gas materials and so doing first means more like a dynamic is not exactly a piece of cake. But there are many, many, you know, lots of progress. For example, in the Filippo de Andrages has done a beautiful work on so a pair of skites and there are some work that came out of Friesta. I don't know if he was interested in Friesta, but in Mari and other people who studied the pair of skites using GW and spin orbit coupling. So there is a problem. Okay, thank you. I think that concludes the questions. Thank you to our panelists for joining Attish and Ralph. And thank you, Julia. Thank you Julia also for my part and goodbye to everyone. Okay. Thank you. And so the session with the diploma students is in a separate Zoom room, Julia, I think you received the link for that. Okay, let me...