 Thank you very much for coming to this presidential lecture. Today we are happy to have Professor Beller from the ITMO University in Saint-Petersburg with us. I give you a brief background so that you know whom you are listening to. He received his degree actually from the ITMO University in computer science or IT in 2000. Then he did two PhD degrees, one in objects in Russia and another one in radio engineering which he did in Finland. In between he worked in industry with various companies, with Samsung in South Korea but also with Robert Walsh company from 2005 until 2012. He was working in the UK at Prince Mary University in London and currently he is the dean of the physics and engineering faculty and the head of the research center of nanophotonics and metamaterials at ITMO. His research interests include metamaterials, plasmonics, electromagnetic, nanophotonics and nanostructures. Today he is going to talk to us about the recent progress in nanophotonics for new optical technologies. Thank you very much for coming. Thank you very much for introduction. It's really a great pleasure for me to be here. It's my first time when I'm visiting Japan. It's my old dream. It's realized and I'm really happy to be here in OIST and Okinawa. I hope that you will get some interesting information today about nanophotonics. I will be talking about, but before I start about some scientific topics, I would like to show the place from which I just came. So I'm representing ITMO University from St. Petersburg, Russia. I would say it's right across from here at the other side of Russia. It's very, very close to the North Pole, St. Petersburg City. It's actually second city in Russia in terms of number of inhabitants and the people who live in there. And actually it's just six million of us. And it's used to be called cultural capital of Russia, Vindu to Europe and so on. It was built in particularly by Peter the Great in order to have connection between Russian country and Europe. And our university, ITMO University is actually sixth university in Russia in terms of optics and computer science in our ratings. So we have beautiful buildings in the city and we have also some nice science which I will be showing to you today. All there combined, actually our laboratory, our research center, which is called Center of Nanophotonics and Metamaterials, is located right in the heart of the city, right? This is a photo which is taken from when you go out from Hermitage and see the photos. If you cannot recognize it, I will help you a little bit, yes, so we have these beautiful bridges in front for those who have been to St. Petersburg, yeah? They are walking right at this beautiful environment. So the lab is located right there. But not only this lab, in particular, contributed to what I will be talking today about, yeah? We are working in very, very close collaboration with different groups around the world. I would like to acknowledge our very long collaboration with Australian National University, in particular with Yuri Kivchar, we are working with Boris Chichkov from Germany, Arseny Kuznetsov from Singapore and recently with Anvar Zafidov from University of Texas. So not everything that I will be showing today was done, particularly in St. Petersburg, but some of the things, some samples were taken from the colleagues, some of the things were measured in the different places, so we are working in this purely international environment right now. All right, in order to start, most of the things I will be talking today will be devoted to some nanoparticles, yeah? And actually, in the nanoparticles, what they can do, they usually scatter, yeah? And actually, scattering of nanoparticle historically is split into two different phenomena, first is relay scattering, then the nanoparticle is much, much smaller than the wavelength, and this is actually the reason of this phenomena, why the sky is blue, yeah? Because of the scattering and so on and so forth. But actually, most of the things today will be devoted to so-called mescattering. This happens then, the nanoparticles dimensions are comparable or a little bit larger than the wavelengths, and you have many, many different scattering peaks, yeah? And usually, even in the nature, you cannot distinguish them. For example, scattering from nanoparticles from the clouds, they create a white color of the clouds for us, yeah? But in reality, there are really, really nice resonance, depending on the dimensions of these nanoparticles, which we will be talking today about. Fundamentally, from mathematics, if you just solve mathematical problem of scattering of the sphere, like a plane wave is coming and scattering of the sphere, you can write some mathematical formula with some coefficients, but basically, you can identify different multiples and usually conventionally, the dominant scattering comes from electrical mode, then you have a certain electrical dipole and this electrical dipole dictates the scattering. Actually, the magnetic mode is usually avoided, because for small particle, for relay scattering, it doesn't matter at all. And even in optics, if I tell to anybody, okay, can we have nanoparticle with some magnetic properties? You usually will tell me, okay, no, it's impossible, because there are no magnetic properties for visible domain, for example, yeah? But actually, I will show you that it's quite possible, and it's possible to have magnetic mode, but it will be a little bit different. You will have displacement currents, which are flowing around this, and you will have magnetic moment of the whole nanoparticle itself. You don't have a magnetic response of the material, but you do have magnetic response of the nanoparticle itself. Let's further consider this, yeah? So usually, in nanophotonics, if we consider nanoparticles, people are using nanoparticle of noble metals, of gold or the silver. They're very simple. They behave as plasmonics, or plasmonic particles. And actually, there is a dipole or, again, electrical dipole distillation, as soon as you apply electric field in this direction, you will have electric dipole in the same direction, and it will scatter. So one electric field applied, one dipole is excited. But in case of dielectric resonators, which we'll be talking today about, yeah, you can have both electric dipole and magnetic dipole, and they're different. In the electric dipole, you have this kind of distillation of electric field, and in magnetic dipoles, the electric field will rotate around. Yeah, create magnetic dipole. Basically, in order to see all this nice phenomenon of nanoparticles, you need a material with very high contrast, with very high dielectric permittivity. Most of the results, which I will show, will be devoted to silicon, which has epsilon, which is very close to 13, 16 in the visible domain. Yeah, so index of refraction is around 3.5 or 4. But there are also other materials like germanium or aluminum, anti-monite, yeah? And they're being also used for this purpose. If you take a look at the scattering cross-section and the amount of power which is being scattered by nanoparticle as function of the wavelength, you will see that actually this silicon nanoparticle, in this particular case with radius of 17 nanometers, has two resonances. Some there at low frequencies at large wavelength, you have magnetic resonance, and after that, you have electrical resonance. And if you model this somewhere in, say, CST Microwave Studio or any software which allows you to see the fields inside, you will see that actually these two resonances, the fields inside of the nanoparticle, they're dramatically different. Again, it's either this electrical kind of the mode is excited when magnetic field is rotating around the electrical, always worse when you have electrical field which is rotating. You can clearly distinguish these two peaks. And now you have a completely different object. You have one nanoparticle with two responses. Usually, the people are considering nanoparticle with only one response, usually plasmonic and usually one electrical dipole is excited. Right now, we have competition of magnetic dipole and electric dipole from just a single nanoparticle. And you can play with this for different applications very, very, very seriously. Another important thing is that as soon as we have some metallic nanoparticle, you usually, if you create nanoparticle, say, from gold or from silver, you clearly know the resonant frequency of this nanoparticle and it's not really significantly depends on the dimension of this nanoparticle. The resonance is dictated by material properties in case of the plasmonics. But over here, if you consider silicon nanoparticle or the electric nanoparticles, the resonance is dictated by dimensions of your particle. Take a look. If you have chosen, say, 100 nanometers radius of nanoparticle, then electrical resonance will appear somewhere at 650 nanometers and magnetic resonance will appear at 900. If you would like to make a smaller nanoparticle, then resonance will shift to the low frequencies. So depending on application, for example, you would like to tune your nanoparticle to operate in telecommunication wavelengths. Then you have to find 1,500 and find the proper radius of nanoparticle to operate. If you need very low frequencies, say, small, short wavelengths, say, 500 nanometers, then you have to take small nanoparticle. So depending on application, it's very, very easy to find nanoparticle which operates. Basically, if you take a look at this cross section for small nanoparticle, usually you have magnetic resonance. If you enlarge your nanoparticle, then your magnetic resonance shifts with the frequency to low frequencies, to larger wavelengths. If you enlarge this radius even further, they again shift. So you can tune it. If you know how to manufacture this nanoparticle with proper dimensions, you can tune it for particular application without any problem. And this is dramatic difference. Previously, the people were playing with plasmonic nanoparticles, trying to modify the shape of nanoparticle in order to fulfill these requirements. Over here, just the size, just dimensions matter. All right. The simplest way how you can create a nanoparticle, it's like you are taking the femtosecond laser, launch the pulse onto silicon substrate. And if intensity is very large, you can make a large hole by this femtosecond laser inside of your semiconductor. And there will be an explosion of this silicon. It will be melting in this area. It's like eruption of volcanic, I would say. And some pieces of silicon will appear around. And these pieces of silicon will actually form nanoparticles of different sizes. And all these nanoparticles, because they have different sizes, they shine, they scatter the light at different frequencies. So this is the first experiment which was done. And actually, the people observed these different nanoparticles of different dimensions and observed these different scattering. And it was exactly the same like in the theory. So if you have a diameter of nanoparticle around 100 nanometers, they observed only magnetic resonance. And after that, if you have 140 or 150 nanometers diameter, they observed both electric and magnetic dipole. But such way of creation of nanoparticle, it's like a stone age, I would say, for development of the technology. Of course, nobody would like to create real nanoparticles for nanoparticles for nanophotonics like this. Other people are using different methods. But at the very, very beginning, the people were picking up these nanoparticles from this thing, by picking place technique, placing them at particular places. And studying, is it really so that this nanoparticle has magnetic response? We were quite happy to do it with near field scanning optical microscope for the first time in 2015. Then we measured the near field, scattered by this nanoparticle at the substrate, on top of the substrate, when it was located. And we were able to identify that really, that magnetic dipole, which completely corresponds to the theory, you have particular scattering geometry of the field, which is completely different as compared to electrical dipole. Because particles are very small. And theory, it's nice. Concept is nice. But it's always requires some verification. As soon as we have nanoparticles with magnetic response and you have strong magnetic fields there, you can manipulate the fields in a different manner. For example, you can have hotspots for magnetic field. Usually, the people we're doing in plasmonics with metals, the people that are doing hotspots for electric field. Because electrons concentrates on there, you have a small gap between of them. OK, and you have hotspot of electric field. Over here, because you have magnetic response of the electric nanoparticle, you can create magnetic hotspots. You can create, actually, both electric and magnetic hotspots. But in particular, magnetic hotspots are very interesting. For some quantum objects which require particularly magnetic field to be applied, this is the way when you can concentrate the field and create maximum of optical magnetic field without any problem. All right, if you have just one nanoparticle, you can collect it, do experiment. But usually, you need to create the real structures out of this nanoparticle. And actually, during the last five years, we have many, many different approaches how to fabricate them. I have to say that actually, there are three of them, which are mainly used right now. One is laser printing. Already not just burning silicon substrate, but you are taking a donor substrate, melt some piece of silicon. And after that, through the drop, you can create drops on the glass. You can use chemical synthesis. Then by chemical creation of the silicon nanoparticles, you can create even solutions of the silicon nanoparticles of necessary site. And of course, classical way of lithography, which is not really actively being used in particular for silicon, because we have to make sure that silicon is pure one without any doping. And with lithography, we are usually getting some not really nice, nice silicon in terms of the crystallinity. This is the way how we are using this. We are using this laser-induced transfer thing. At the beginning, you are creating donor substrate using some simple lithography techniques. Yeah, you can have fancy shapes of silicon, something. And after that, by laser pulse, they are melting these pieces of silicon. They create drops, and these drops fall down. And after that, in the receiver substrate, we have very nice silicon nanoparticle. Mother Nature helps us over here, because as soon as this melted liquid silicon is passing through the air, it forms ideal sphere. And already ideal crystalline sphere fell down to the substrate. And we have very nice ideal spheres without any changes. I will tell a little bit later about this fact. Yeah, that actually depending on the distance between donor substrate and receiver substrate, we can have a little bit different quality of the silicon nanoparticle, which we have started. But what you need to know, that actually you can create this without any problem in the lab and use it for different things. Okay, so this was a background story that actually just the silicon nanoparticles or all dielectric nanoparticles, they can be created, they create some enhancement, but still it was not clear how it can be used. And today we'll be talking about using of these silicon nanoparticles for different applications. Yeah, I will list now at the next four slides the most of applications which are known to me. And after that, I will go in details into some applications and some results which we have done. So the first one, the most amazing ones were of course done not in my lab, but I will mention them as well. One important thing that as soon as you have these electrical and magnetic dipoles, excited. Yeah, you can have a configuration then you have very strong back-scattering, very strong reflection. So basically if you make a ray of this silicon nanoparticle, you can create perfect mirror. Nothing is propagating, everything is reflected and the thickness of this mirror made of dielectric of silicon is very, very small. It's quite, quite nice for various applications. Because these nanoparticles, as we have seen, they can have colors at different frequencies. You can just use this technology for coloring. If in some areas you create nanoparticles, say 100 nanometer diameter, in another 90 nanometer diameter, you can literally make different colors. The people are just collaborating with people who are doing arts and making whole portraits and very nice arts with the help of this coloring right now at the different scales, by the way. Yeah? Another thing from the silicon nanoparticle is that actually you can not only control the frequency at which this particle resonating, but also you can control a phase with which this particle is scattering the light. And this phase is very important. For example, you can write different phases at the flat interface and you can create a holograms by placing different nanoparticles in different places and launching the light. You can create three-dimensional images, holograms without any problem. Another thing is that as soon as you modify in plane the distributions of this nanoparticle, you can create, for example, beam steering devices than they used to have propagation or refraction of the lights with a particular angle over here by modifying the surface made of this nanoparticle. You can tilt the beams on necessary angles, which are required for particular applications without any problems. Another thing which you can do, you can create, for example, flat lenses by manipulating the phase at the interface by these different dimensions of nanoparticles. You can create a flat lens which will focus the light with required numerical aperture, with required properties. Yeah? You just have to control this individual nanoparticles without any problems. And another thing you can make phase plates. Yeah? You can just put different pieces of the silicon nanoparticles with little bit different orientations with different phase. And at the output of this device when you're shining the light, you can have a vortices, you can have different orders and so on. There's a bunch of the people around the world doing this right now. Have to say that Federico Capas, Yuri Kivchar and other guys, family names are listed over here. They were doing all these things during the last five years quite actively and they continue to do it because this technology right now seems to be very, very convenient because it's silicon-based. Yeah? And so it's silicon-based. It means that we can use conventional technologies which we are using to process silicon in order to do this. Yeah? A little bit of melting, a little bit of processing and so on. This was the plane of the phase. Yeah? What you also can play, with losses and absorption in these nanoparticles. I will show you that actually you can have ultrafast optical modulations. It's like an optical diode made from just one single nanoparticles. And you can make a selective optical heating which is even more efficient than just in the case of gold or silver nanoparticles. And also you can even use it for data recording. You can not only save colors as I was showing in the previous slide, but you also can use it for recording the data, reading and saving. Interesting thing is that actually you can create non-antennas out of these silicon nanoparticles. I will show you again how we can do it. But what is important here is that you can have variety of these non-antennas done for different frequencies. And you actually can control interaction between quantum emitters through these non-antennas and far field. And vice versa, you can use effects of these quantum emitters inside of these non-antennas in order to enhance typically quantum effects. Like non-linearity, you can enhance second harmonic generation. You can increase Raman signal. You can increase photo luminescence with help of this nanoparticle. So the whole story of using of these large objects in terms of the quantum, like these silicon nanoparticles as a way of modifying the performance of quantum emitters, I would say. All right, now a few words about non-antennas which is pretty simple for me because I have radio engineering background, but during last 10, 15 years the people started to do optical antennas. Usually we were doing microwave antennas connecting it by RF cable and after that we have TV, we have Wi-Fi and other things. But in optics, we also do need it, yeah? Take a look. The conventional antenna at the right. This is actually matching device. You have some microwave or optical waveguide. You have antenna after that and it match mode of this waveguide with a free space. In this sense, you either convert mode of the waveguide to the free space or mode of the free space into the waveguide. So you either have transmitting antenna or you have receiving antenna, yeah? This is always used everywhere in all our mobile phones and everywhere. But in optics, there is a very small drawback. You don't have really tiny, sub-link waveguides. That's why in optics, the people do call non-antennas. Any devices, which either couple some near field of quantum emitters to the far field radiation or concentrate far field into the near field. It's a little bit different, but at the end of the day, all the antennas looks very, very similar to what the people have in microwaves, yeah? But applications are different, yeah? These non-antennas are used for quantum sources. They're used for being used for medicine for heating of the cancer in microscopy, in non-linear spectroscopy, in solar cells for better and efficient solar cells and for sensors. The people, as I told you, are doing this already for quite a while. And the simplest way which researchers decided to do to follow is just to mimic antennas which were available for microwaves. They took a dipolar antenna, monopolar antenna, bow-tie antenna and even rhombical antenna and Yaggy-Uda and just squeeze dimensions. Usually, they should have dimensions of, say, centimeters. Right now, they have dimensions of nanometers. And because usual conventional antennas are being made from metals, they decided to make it the same from the metals at the nanoscale. It's most quite successful that the problem is that metal at microwaves, it's nearly lossless. But metal in the visible domain, it's extremely lossy and a lot of fields are being dissipated into the heat in all these antennas. And they decided with silicon, with much smaller losses, that it will be much better. This is, again, this scattering cross section of single silicon nanoparticle. And right now, you have to take a look at these three points. Take a look. This is actually the points when you somehow have equal values of electrical and magnetic dipoles or electrical and magnetic dipoles with opposite side. This means that nanoparticle at these frequencies will either have a completely forward scattering in the forward direction or completely backward scattering. And by this forward and backward scattering, you really can create antennas easily. You put your quantum emitter next to this nanoparticle. And after that, you see that it starts to radiate directively, a single nanoparticle, either in the forward direction or in the backward direction. And by knowing this, how you have to operate this. So it's radiate either in forward or in backward direction. So it's either a reflector or it's either a director. And you can use it for creation of, say, Yagi Uda antennas. What is Yagi Uda antenna? You have some emitter. And after that, you would like to radiate power at the right direction. So you have to put directors, which will predominantly scatter in this direction. And one reflector, which will reflect the light. So the light will predominantly go to the right. And the direction of emission will be very, very directive like this over here. And all the power will go in this direction. It has been done, but for the first instance, numerically and experimentally, issues with substrate, issues with placing of the quantum dot over here, were quite dramatic. What we have done, we decided to make a scaling. We scaled this structure back to RF, to microwave frequencies. We found very nice ceramic material, which has exactly the same dielectric permittivity as a silicon, but not in the visible domain, but at 10 gigahertz domain. And we created this nan antenna with dimensions of centimeters. And in anechoic chamber, performed experiments and verified that everything is fine. After that, in one year, our colleagues around the world have done this experimentally in optics without any problem. We didn't stop at this point. And we decided to think, all right, so if we can now create antennas out of dielectrics, maybe we can create some different kind of antennas using this. And we started to think, OK, the main property of antenna is directivity. So it should emit light in one direction. How the people are doing this at microwaves? They're usually using either Udayagi antennas, or lenses, or reflectors. But all these devices are very large. And exactly the same idea is being now projected into the optics. Maybe in optics, we can do something else. Maybe we can have high directivity, not only due to large aperture of antenna, but due to something else. And if you dig something deep in the books, you will find that actually there is so-called effects called super directivity effect. Actually, directivity of antenna is dictated by area of electric field created by this antenna, but not the physical area of this antenna. And for some antennas, it happens that electrical area of antenna is much larger than physical area. And if it's so, then it's called super directive. And we observed this effect for single nanoparticle. If you have a whole of this nanoparticle, notched inside of this nanoparticle, and place some dipolar scatterer, you can see that directivity of this guy dramatically depends on the location of this emitter with respect to this nanoparticle. This is x-axis. 0 is the center of nanoparticle. This gray area, it's inside of nanoparticle. White area, outside of nanoparticle. If this dipole is located outside of nanoparticle, then we have very small directivity around 3, which I was showing you at the previous slides. But interestingly, as soon as this nanoparticle enters inside, you can have significant increase of directivity by three times. And you have already very, very directive emission but with single nanoparticle. So directivity is reaching level of 10, which is unpredictable. It's usually in order to get directivity around 10. Previously, take a look, you're required to have one, two, three, four, five elements over here. You have directivity equal to 10 with just one single element. How can it be? It happens because inside of this dielectric nanoparticle, you cannot do it in metal. You have to use dielectric. You can excite higher order multiples. You can have very crazy fields inside. And these crazy fields inside of this nanoparticle with broken symmetry create very large effective aperture which radiates. Basically, we're able to excite higher order multiples inside of this nanoparticle. So in addition to those electrical dipole and magnetical dipole, which we can excite from the far field by the plane wave, if we're placing the quantum emitter inside of our object, we can excite up to fifth order of multiples, which is something which is usually impossible. You need to use whispering gallery modes up to fifth order in order to do this in regular case. Over here, it's being done automatically. And as a result, all these multiples are forming very, very nice directive beams. These beams are so directive that actually you can use it immediately, say, for sub-rank resolution imaging. If you change the source location over here inside of this nanoparticle or nanolength, a little bit to the right, a little bit to the left, you have a rotation of the beam which is being emitted by this. Over here, this is the angle of rotation as function of shift in terms of the nanometers. So in the far field, you will be able to detect shifts related to 5 nanometers, 10 nanometers with respect to the center by just fact, but you will see that emission changed by 5 degrees. Detection in the far field, the 5 degree angle change, it's very, very simple. We verified this in microwaves without any problem, created this nice antenna and placed a dipole over there, moved it, and have seen an experiment that everything is working as I have shown to you numerically in the previous slide. But the real optical experiment, it's still a problem. You need the real quantum source we were thinking and continue to think about. And the center in the diamond for this purpose, you need a real silicon nanoparticle. You need to place it properly on top of this and the center. There are three groups currently working in the world on this. Yeah, all these are problems with quality of the and the center or the positioning. But at the end of the day, if you can do it, then you can extract by this nanoparticle quite efficiently the power which is being radiated by this and the centers independently on the substrate at which this and the center is located. I will probably skip these crazy figures, but I have to tell you that radiation efficiency of this antenna is about one. It means that most of the power which is generated by this and the center will go to the free space. So this antenna is working as a matching device between this quantum emitter and the free space. In regular case, if you have and the center on top of the substrate, it will radiate into the substrate. The power will be absorbed in the substrate. Over here, it extracted and passes to the free space. And actually, it's even more efficiently extract the power because it has very much personal factor. Personal factor reach level of the 10 or 15 over here. So it's double performance. One performance is directive emission of the energy. And another one is a personal factor generated by this antenna. OK. And also, you can use this antenna as soon as it's ready. And if you can put it on top of the manipulator for imaging purposes for different angles without any problem. Another interesting thing with these silicon nanoparticles is that, of course, we have a competitor like a silicon photonics. And you have a silicon waveguide and so on. But even for silicon photonics, we can have a very nice device if you make an array, a line, of the silicon nanoparticles. You can make a waveguide. OK. This waveguide is not better than just a silicon waveguide. But it has another nice property because there are so many bands supported by this waveguide. For one of the bands, which literally really using magnetic and electrical moments of these nanoparticles, you can create waveguide, which is not feeling any bending. So you can bend this waveguide by 90 degrees. And we will have total transmission without any scattering there. This happens because of this nice mismatching between electrical and magnetical dipole of these particles. And the light is just jumping by near field from one particle to another and being just passing through this H without any problem. OK. This, what I was showing you, it was linear, mainly linear phenomena. Yeah, now it's time to talk about some nonlinear phenomena. One interesting nonlinear phenomena, which we observed, was that, of course, if you shine light with very light intensity on the silicon nanoparticle, then you can change properties of the silicon itself. It's very, very well known. The dielectric permittivity of silicon depends on concentration of free carriers. And concentration of free carriers depends on the intensity of light which is shining on nanoparticle. If epsilon of this dielectric sphere is changing, the scattering will change. And what we have done, we have done so that, for example, if you're shining low intensity light on this, in the nanoparticle, you excite only magnetic dipole. So it will radiate both in the forward and the backward direction, like this gray curve over here. But with high intensity, epsilon is changing. So the property of nanoparticle is changing like we effectively change the dimensions of this. But dimensions remains the same, but dielectric permittivity of the particle is changing. But the scattering is being changed, and you have this blue curve for scattering. So the particle starts to scatter more in the forward direction if intensity is larger. If intensity is small, then not. And they performed this experiment. What we have seen, we have seen an experiment that actually ratio of nonlinear reflection with large intensity and small intensity of this nanoparticle was about 15% from one single nanoparticle. So in the sense, you already can start doing some logic over here. So it's really nonlinear logical device because it's like a diode. If some intensity, you can have a scattering in one direction. If another intensity, you will have a scattering in the opposite direction. In terms of publicity, this result was called like a smallest optical dipole. But right now, the problem is integration of this dipole into real devices. This is still a problem. Because as soon as you have silicon, which is changing dielectric permittivity in one nanoparticle, it's a very bad idea to use silicon waveguides to put the power towards this nanoparticle. And this problem is not solved yet. Probably, I mean, you can put this into some circuitry with required intensities. And I'll really use it like optical device. I was mentioning something about recording of data and changing the colors. We have done this thing as well. We decided that, OK, silicon, melting of silicon, it's too high power. Yes, so in order to melt silicon, you need about 1,680 kelvins. We decided to make hybrid systems. We decided to hybridize our silicon nanoparticles with plasmonic nanoparticles. They placed like a mushroom over here. The cap of this mushroom is made of gold. And this part is made of silicon. The good thing is that these two guys are melting at different temperatures. So what we can do is we control the intensity of light which we are shining. We can melt the head of this mushroom. And the remaining part will stay there. So with small intensities, it can be like this. With very large intensity, you can make a sphere on top of it. And we have done this quite effectively. This is our experiment, depending on amount of power which we applied. We can either make this sphere over here. And after that, even we can burn this sphere. It will go out. And there is a certain theory over here. Good thing that as soon as they mentioned, as soon as geometry is changing, the optical properties of these nanoparticle hybrid nanoparticles is also changing. And we have seen it. So basically, you can record some information there. And from this state to that state, sorry, from this state to that state, the difference in scattering is more than 100 nanometers. It's more than enough for any devices. So this is the way how you can write data or change colors with the help of these hybrid nanoparticles without any problem. We started to think a little bit further. And the third, OK. So we should have some nonlinear effects in these nanoparticles. What about second harmonic generation? And second harmonic generation is very useful because we know that in the living tissue, it passes very nice infrared radiation. But after that, we would like to have some kind of colors, light coming in the visible domain for detection of different things. And for this purpose, you need to convert infrared frequency to the visible domain frequency. For example, by second harmonic generation, by increasing frequency two times. It's quite well known for many nanoparticles made from different materials. And it's observable everywhere. But all these materials require not centrosymmetic high index dialectics. So they require high two. The problem is with silicon. It's completely symmetric crystalline lattice. So we shouldn't expect any second harmonic generation out of silicon. But in experiments, we do. And this is very simple, actually. It's also known from the literature that this happens because of the interfaces. So if we could create a silicon nanoparticle, which is idealism spherical, which is very complicated, and completely crystalline, that we shouldn't have any chi2 and nonlinear effects. But in reality, some interfaces will be a little bit rough, so on, so forth. And we will have chi2. And in real nanoparticles, which we are using, chi2 is very dramatic. Now I'm returning to this nice picture. Then I was showing to you that we are melting the silicon, cool it, crystallize, and after that, depose. We analyzed what are the nanoparticles, which are being obtained over here, depending on the phase, and depending on the distance between donor substrate and silicon substrate. Take a look. Over here, you have distances 220 nanometers, 250 nanometers, 300 nanometers, and 350 nanometers. You see that actually optical properties are very, very different over here. It means that nanoparticles are always very different. Why it is so? Take a look at second harmonic generation. We have seen that second harmonic generation is very different here, and it's much larger, at least 20 times larger, even more than just a silicon amorphous silicon film. And this, because we have different state over here, depending on the distance at which all these guys are flying, we either have amorphous silicon, or nanocrystalline silicon, or polycrystalline silicon. We do see it from the Raman signal from the silicon. So we either have a maximum over here, which is corresponding to amorphous silicon, or sharp resonance to crystalline silicon. So just these nanoparticles, they are in the different state. And for nanocrystalline silicon, we have many, many, many islands of the silicon. And from these many islands of the silicon, we have very strong second harmonic generation, because these islands do have a lot of interfaces. And all these interfaces generate second harmonic. So we've got very nice efficiency over here. And of course, because these are nanoparticles, this efficiency is increased at the resonances, electric and magnetic resonances of these nanoparticles. And this is quite nice, I think. After that, we decided, all right, if for such simple effect like a second harmonic generation, we have improvement by our resonators. What about Raman signal? And we decided to see what's going on with Raman. This is usual way how we measure Raman scattering with objective and Raman spectroscopy. And what we have seen, we actually have seen that the fields inside of nanoparticle have been enhanced at our resonances. And there was a very strong expectation that Raman intensity will be also increased. And it's really happened. We observed that largest, so largest intensity was at magnetic resonance in this case. At the same magnetic resonance, we've got enhancement of Raman signal 140 times. And interestingly, it's not just enhancement of Raman signal. It's Raman signal coming from silicon. And Raman spectrum of silicon depends on temperature. It's very well known that the Raman shift as a function of temperature is dictated by some formula. Literally, you can see that shift is just linear with respect to the temperature. So by the shifting of the resonant frequency of Raman, I can tell you what is the temperature at the nanoscale at some parts of this nanoparticle without any problem. I can make a picture. Right now, everywhere, it will be already temperature in this scale in Kelvin's for different nanoparticles with different sizes. We compare this with conventional gold nanoparticle. And we see that actually, depending on the frequency and dimensions, you have usually even higher temperatures which you can obtain there. And you can really map the temperature distributions over here. So this is the temperature as function of the location for this particular nanoparticle, taken from Raman signal. Yeah, we compare it with gold nanoparticle. Gold nanoparticle has very flat response. There are maximas for the silicon nanoparticle. And these maximas are quite nice. It means that you can have really better termometry as compared to the same thing with gold, for example. And this is example then with some oligomers of these nanoparticles. We measure it temperature distributions at different intensities. So I would say it's one of the new ways of termometry at the nanoscale. It's a little bit tricky because it involves silicon nanoparticles. But at the end of the day, it gives you information which is usually not really accessible. Because with some resolution about nanometers, you can get information about temperature. OK, last among all these nonlinear effects, which you can see again related to Raman scattering, we obtained the spectra of some native protein molecules in the system when we again combine our silicon nanoparticles, which is very good resonator on top of the gold mirror. How we did this, we of course measured this and found the areas where the field is very strongly localized. And got improvement of Raman intensity. And in addition to thermosensitive line of crystalline silicon, we are obtaining a spectrum of the protein on top of that. So we can both control temperature of the area from which we've got a signal and the spectrum of what is in this area. So we basically can trace molecular events. So we can know the temperature of our silicon nanoparticle. And we can see what's going on with these organic molecules there. So what we have seen here, there was some chemical reaction there. And depending on the temperature we have seen, it's a different stage. So this is new area for me and for my group. But the people who are doing this, they're saying that this is really kind of not really breakthrough, but step forward for understanding of this. Another method of understanding what's going on at the nanoscale with these chemical reactions and really molecular events. Because in this gap between nanoparticle and the mirror, there are only a few molecules. And chemical reactions there are quite of good, quite of large interest for biology. Maybe the people who are doing biological studies can help me with this later in the discussions. OK, last thing which we have done. We call it nanolamp. It's ultra-broadband nanoscale light source. And we have created again it from silicon nanoparticle. At a certain moment we decided that it's not enough to make mushrooms from silicon and gold. Let's mix silicon and gold completely in the nanoparticle. And then melt it, silicon and gold under certain concentration. And after that, at the output, we've got nanoparticle. There you have partially gold and partially silicon in different areas with certain concentration. We have done it just for fun at the very beginning. But as a result, we observed extremely huge photo-luminescence over here. It's very wideband. If you have silicon nanoparticle, then you have this kind of photo-luminescence. If you have gold nanoparticle like that. If it's silicon gold, you see it's very large and very wideband. So it's really nanoscale, very small light source, like a white lamp, but nanoscale size. Photo-luminescence happens because of free photon photo-luminescence and second harmonic generation over here. How it looks like. Actually, we can take this nanoparticle and put it onto the metallic tip. How we are doing this? Oops, it should be video. Video is not really working. Very, very pity. Sorry for that. It's here, but for some reason you cannot see it. Okay, so actually by this probe mechanically, we are coming over here, picking up this nanoparticle and placing it to another place. This is pick and place technique. And actually as soon as this nanoparticle is already being placed on top of some tip of this, this lamp can be used as a source for near-field scanning optical microscopy. So you can put it on top and after that, scan it on top of some structure. As you can see here, it's also supposed to be video, which is not working, sorry for that. All right, as a result, you can get a pictures of near-field at very wide frequency range without using different frequencies of the laser. In the lab, we have tunable laser. We scan the near-field with the tunable laser. And after that, we scan the same near-field pieces for our dimer antenna, which we are very familiar with. With the help of this tiny white light source nanoparticle, we've got nearly the same results. I would say this is a new way of doing the nanoscale sources, which was quite successful and we are continuing to do this. There are some limitations, mixing of silver and gold, sorry, gold and silicon for fun. It's one story, but we started right now to think about this more systematically. And most systematically, we really can create highly efficient sources with tunable emission, with some broad range in the visible. Should be very good in terms of quantum yields, tunable. We can modify it by mere resonances, so we should have high refractive index and low cost of fabrication. Gold is not appropriate for this purpose then. So we decided to switch materials and what we are doing right now, we are using pair of skites. Pair of skites, they're actually new kind of material with strange geometry like this. So it should have two cations, A and B, and after that one onion forms this kind of geometry. The name pair of skites is coming for Lev Perovsky, Russian minister who has nothing to do with research at all. He just gave a lot of money to researchers to go to rural mountains in order to dig. And they found this in the nature with pair of skites. And now these minerals are being called pair of skites because of him. And actually right now we can easily synthesize this pair of skites quite easily. We can choose this A, B, and X without any problem with different cations and onions. And you can create this pair of skite by just that chemistry processing. You mix these things, you rotate this, hit it up a little bit at the end of the day, you have this pair of skites. What is good with pair of skites is good that you can really control the spectra of this pair of skites depending on the materials and concentration of materials you put there. And it's extremely cheap. It's very low cost. It's that chemistry and very simple synthesis. And we were lucky that actually it's very low, low-high index dielectric, yeah? These are three kind of pair of skites, this formula, this formula, and this formula, even I don't know how to tell this. But good thing is we have different colors. We have different excitation peaks, yeah? And in experiment you really can see the different free colors of this pair of skite, non-antenous. So we have made this single nanoparticles out of this pair of skites. And they're very happy. They're also very good light sources, okay? So it came to the conclusion. What I wanted to tell you that actually in the area of nanopartonics we have just a single nanoparticle. Still there is a room to investigate very many different phenomena for different applications, yeah? With simple silicon nanoparticles we've managed to create many different antennas and find many different phenomenons. For example, nano-heating, second-harmonic generation, nanothermometry, and even create white light snoms. And it's still not the limit. We still have to search for different materials around us and in particular we, we are trying to study pair of scarlet perovskite now because we believe that it's a very good, nice semiconductors with a lot of features. And they are much cheaper than, say, silicon which has to be processed in a different manner. Yeah, and all the electric non-antenous, they really beat plasmonic ones. It may take a while for use of them in the real practice. But right now it's already clear that this is next step towards creation of nanantenous. A lot of groups around the world are doing this, silicon nanantenous right now. Okay, thank you very much for your attention.