HORIBA, Ltd. (6856) Earnings Call Transcript & Summary

Hi, everyone. Welcome to today's live broadcast on Raman and PL at the nanoscale, why it really matters for 2D materials. I'm Joanne Lowy, Marketing Manager at HORIBA Scientific and moderator for today's event. While we're waiting for the rest of the people to sign in, I'm going to do the introductions. So we're delighted to be bringing you these webinars under our virtual information program, or VIP, and hope you find them informative. HORIBA Scientific, global leader in microscopy and spectroscopy systems, provides an extensive array of instruments and solutions for applications across a broad range of scientific, R&D and QC measurements. These instruments include a host of spectroscopy solutions, including Raman, fluorescence, custom spectroscopy and particle characterization. In today's webinar, we will discuss Raman spectroscopy as an important technique for characterization of 2D materials, such as graphene, and how TERS and TEPL imaging can probe a number of nanoscale heterogeneities in 2D crystals. We'll also discuss ultra-low frequency TERS, which opens the gates to a wonderful area of nanophononics. Today's speaker is Andrey Krayev, U.S. AFM Raman Product Manager at HORIBA Scientific. Andrey received his Master's degree from the Moscow Institute of Physics and Technology in 1991. In 2001, he started to use SPM while working as an application scientist for QPT. Since 2008, he was the CTO at AIST-NT Inc and was actively involved in the development of the TERS technique and its implementation for real-world applications. Since the acquisition of AIST by HORIBA in July of 2017, Andrey has held the position of U.S. AFM Raman Manager and continues to be active in the development of TERS-related applications for advanced characterization of 2D materials and beyond. Once the webinar begins, you'll all be muted to optimize the speaker's audio, and you'll be able to post questions or feedback by using the question box on the right-hand side of your screen. If you don't see it, go ahead and click on the orange arrow on the top to expand the view. We'll have a Q&A session at the end of the webinar, where you can ask questions anonymously. And right now, I'm going to turn the presentation over to Andrey.

Thank you, Joanne. Okay. Let's show my screen and I'll start the -- okay. Can you see my presentation screen or the presenter screen?

I see 3 things, the presenter screen. I don't see the presentation only.

Okay. Just -- let's see the settings. Now should be correct, right?

Right. And then just use the orange arrow to get rid of that screen of people attending. Perfect.

Okay. So let me turn the laser pointer on. Okay. So thank you very much for attending our event today. And today, we'll learn about how Raman and photoluminescence can be helpful for nanoscale characterization of two-dimensional materials. In other words, we'll learn how TERS and tip enhanced PL can shed the light on the exciting flat world of 2D materials. Right after the discovery of graphene, scientific community quickly realized that there are many more other materials that can be isolated in the form of single atomic or single molecular layer. And according to the study conducted by a group in Stanford, there are over 1,000 of potential candidates for exfoliation in the form of single layer. Out of those thousands, only about 70 have been actually realized experimentally. But even within that family, the diversity of the properties is really amazing. We have the materials from -- all the way from the metallic or even superconductors to the materials with different band gap semiconductors and all the way to the electrics, like well-known boron nitride, for example. And of course, when we have this amazing diversity of the properties, we can use those materials together, stacking them or staging them either in lateral configuration or stacking them vertically and, thus, acquiring really amazing new properties, which are not available in any other material known to the human kind. Very early on, it was realized that Raman spectroscopy can be extremely useful for characterization of two-dimensional materials. Having the luxury of performing Raman measurements with different excitation lasers, we can probe the resonance versus nonresonant response, which greatly expands our knowledge and understanding about the properties of the photons in these materials. Another important aspect of Raman is that using Raman imaging rather than just single point measurement, we can probe the distribution of properties across the two-dimensional crystals, therefore, probing the heterogeneities in these materials, which are almost inevitably present in every -- 2D crystal that we create. The important thing to understand is that quite often, the degree of or the scale of the heterogeneity in 2D materials is actually at the level of few tens of nanometers, which is beyond the spatial resolution of conventional optical microscopy. And in order to address that level of heterogeneity, we need to switch from conventional Raman to tip enhanced Raman scattering, which addresses the heterogeneity at proper scale. So what is tip enhanced Raman, or TERS? It's a hyphenated technique that combines the advantages of surface enhanced Raman scattering and scanning probe microscopy. So when we position a plasmonic structure at the apex of the scanning of SPMT and we illuminate this tip with external Raman laser, the launched plasmons that got confined at the apex of the tip, here. And due to this confinement, the illuminated volume actually becomes only a few nanometers across. As a result, when we scan the sample under illuminated tip, we can obtain the Raman image with spatial resolution, which is very similar to the spatial resolution of the conventional AFM imaging. And due to the dramatic enhancement of the incoming and scattered field, we have strongly improved quality of the Raman spectrum in addition to improved spatial resolution. Therefore, we are creating the Raman images of our materials at the nanoscale. HORIBA has a broad line of products that enable tip enhanced Raman scattering or tip enhanced photoluminescence imaging: our flagship OmegaScope, which is a very high-quality AFM that allows tip enhanced Raman scattering or tip enhanced PL imaging on nontransparent samples with either top or side illumination, showed here; our TRIAX platform, extremely flexible, that allows top, bottom or side illumination. And with this platform, you can run all sorts of crazy experiments with split excitation collection, adding your own optical elements and so forth. And both of these platforms can be coupled to any of the Raman spectrometers that HORIBA manufactures, either compact and portable XploRA, midrange LabRAM Soleil, very recent addition to the family of Raman spectrometers from HORIBA or our flagship in Raman LabRAM evolution. So any of the AFM side instruments can be coupled to any of the Raman spectrometers on the right, and by this combination, we can enable the tip enhanced Raman scattering imaging or tip enhanced photoluminescence imaging. We are very proud that the very first TERS imaging of transition metal dichalcogenides, molybdenum disulfide, actually, it was done on our equipment by the group led by Marlan Scully, and this particular research was led by Dmitri Voronine. Now he is a young professor in University of South Florida in Tampa. And since that time, every year, number of high-quality papers appeared in very well-respected journals. Almost every year, we have at least one Nature paper. And I hope that we'll continue this suite further on 2021. And with that being said, let's go over the outlook of the further presentation. So we'll first discuss the -- how TERS enables Raman and scale Raman imaging of MXenes. Then we'll talk about TERS and tip enhanced for characterization of growth-related inhomogeneities in CVD-grown transition metal dichalcogenides. Then we'll discuss characterization of strain-related states in nanobubbles in tungsten diselenide, characterization of nanoscale variations in TMDC substrate interactions, characterization of lateral and vertical heterostructures, then we'll briefly discuss a very recent discovery, ultra-low frequency TERS, and why it's so important, why it really opens the gate to the nanophononics, not just nanophotonics. And as the cherry of the pie, I will briefly discuss the TERS characterization of reconstructed Moire patterns in twisted bilayers in -- of graphene. Okay. TERS of -- TERS imaging of MXenes. So what are MXenes? MXenes are a fairly recent addition to the family of 2D materials. It's a layer material, which consists of alternating layers of transition metals, marked here both -- in red, and carbon or nitrogen. And usually, these MXenes are obtained by etching so-called MAX phases, where these layers of transition metal and carbon or nitrogen are connected with a layer of any of these elements marked in blue. Most frequently, it's aluminum. And these materials were discovered 10 years ago by these 3 gentlemen, Yury Gogotsi, Michael Naguib and Michel Barsoum. Due to the extreme diversity of the composition of these materials and the possibility of alloying using several different metals in MXenes, the range of properties and possible application is really amazing. And almost every area of advanced research or advanced applications in the industry, whether it's sensing, optoelectronics, catalysis, biomedical sensing, MXenes can be used or can be explored for potential use in very diverse areas. Of course, optical characterization of MXenes is very important for understanding of their properties. And Raman and this UV–vis absorption spectroscopy is one of the most important characterization tools for these materials. The bad thing about MXenes is that their Raman scattering efficiency is fairly low. And usually, it takes tens of seconds in a single point in order to acquire Raman spectrum with reasonable signal-to-noise ratio. Just increasing the laser power on the sample does not really work because with excessive power, we can alter or completely destroy the sample. So, so far, before we tried TERS imaging of MXenes, there were no reports on imaging of these materials using Raman. But with tip enhanced Raman scattering, we were able to obtain very decent quality of Raman -- TERS spectrum of these materials with the power and the sample, which is about an order of magnitude lower compared to the usual power used in the confocal experiments. And here, you can see the AFM image of the several crystals through bilayer, monolayer and an overlapping area of kind of [ stat ] tri-layer of MXenes. And we can see that we can obtain a nice TERS signal from all 3 parts of this construct. And another interesting thing is that the 202 wavenumber peak or 203 wavenumber peak is the most prominent Raman band in this spectrum. But as the number of layers increases, the relative intensity of the side bands at around 730 and 120 wavenumbers, they start to go up. It's really well seen in this pattern where we normalize the spectra from mono-, bi- and tri-layer by the intensity of the main 202 wavenumber peak. And you can see that these resonant peaks of 120 and 720 wavenumbers, they keep increasing as the number of layers increases. So due to the dramatic enhancement of Raman signal, we can, therefore, using TERS imaging of these materials, clearly differentiate the number of layers in MXenes. Another interesting thing that we observed in the course of the investigation of these materials is that it's almost inevitable to get wrinkles or nanowrinkles when we deposit MXene flakes on the surface from -- the reason of that is because the deposition occurs from the solution, and it's always inevitable that there will be some folding of the single -- especially in single-layer crystals, and folds and wrinkles are practically inevitable. What was interesting, when we imaged the area around the wrinkle, we saw that the spectral response of the wrinkle was very clearly different from the spectral response of the adjacent flat area. And we can see here that the wrinkle is characterized by strongly increased intensity of the side 720 and 120 bands, while the intensity of the main band at 202 goes significantly down. Here, I would like to stress that it was not just the relative intensity of these side bands that went up, actually, the absolute intensity was strongly increased over the wrinkle. Therefore, we can clearly see that wrinkles, they do have special properties compared to the flat material. And why is it important? In the relatively recent study presented here, the researchers were investigating some crystals also with 785 layers, so the same layers that we used in our TERS studies. And the size of the laser spot of the crystals was approximately the size of these red spots here. And you can clearly see that within this laser spot, they got both the wrinkled and flat portions of the crystals. Therefore, in these spectra presented here, there was inevitable averaging of the flat areas and the wrinkled areas. And we can -- as we could see from prior TERS data, this sort of averaging is not actually very correct because we clearly saw that the flat areas and wrinkled areas, they do have different spectral response. Another interesting construct that we imaged with TERS using -- they were the vertical heterostructures that included as one of the component titanium carbide, the MXenes that we've worked with. So in this picture, we see a TERS AFM image and TERS image overlaid on top of the topography of the vertical heterostructure of molybdenum disulfide on gold. And we see clearly the crystal of titanium carbide, and we collected TERS imaging in this area where we have monolayer -- stand-alone monolayer of molybdenum disulfide. We have monolayer of tungsten carbide on top of molybdenum disulfide, and we have the bilayer of -- most probably folded bilayer of tungsten carbide on top of molybdenum disulfide. And we clearly saw the signatures of each component, including the signatures of the -- both titanium carbide and molybdenum disulfide in vertical heterostructures, whether it's bilayer or heterobilayer or heterotrilayer. And this actually gives hope that TERS can be applied not only to the characterization of stand-alone MXenes, but also to the heterostructures based on MXenes. And this dramatically expands the usefulness of this technique for this class of material. Another type of the vertical heterostructure that we managed to create and successfully characterize using TERS was the graphene oxide titanium carbide vertical heterostructure. Again, similar to what we observed with molybdenum disulfide within the TERS image, we could clearly identify the stand-alone monolayers of graphene oxide, the green spectrum here, the vertical heterostructure of titanium carbide on top of the graphene oxide and stand-alone graphene -- stand-alone titanium carbide as well. Okay. So I hope that I convinced you that TERS can be very useful for enabling Raman imaging and, even better, nanoscale Raman imaging of this important class of new two-dimensional materials. Now let's discuss how TERS and tip enhanced PL can be applied to characterization of growth-related inhomogeneities in transition metal dichalcogenides crystals. This is a relatively simple example, but still, it's quite important. So in this panel, we see a topography image of a predominantly monolayer crystal, but you can clearly see that on top of the monolayer, there are tiny islands of bilayers and, in the center, even thicker number of layer crystals. And when we collected tip enhanced Raman and tip enhanced PL image of this crystal, we clearly saw that in the areas where the bilayers were observed, the signal went significantly down. And the reason for that is the following: with the red excitation that we used in that experiment, it was 633, we have resonant response from the monolayers, but not from the bilayers. And due to the dramatic enhancement with resonant Raman, we observed nice tip enhanced PL signal and nice TERS signal, but not as much of that response over the bilayers or even thicker number of layers. And this is a good example how TERS and tip enhanced PL can differentiate number of layers or even tiny defects with increased number of layers in predominantly monolayer crystals of CVD-grown TMDCs. Another example -- and these results were obtained, gosh, already, like 5 years ago, was the characterization of CVD-grown tungsten disulfide crystals. And what we've been able to observe was the difference in the PL intensity and the PL peak shape or PL peak position across the crystal. As you can see here, along the edges of the crystal, the PL intensity was actually different compared to the intensity in the central part of the crystal. And also the PL peak was shifted to higher energy. And the reason for that is actually different stoichiometry of the crystal during the synthesis. Once the central part is synthesized, the stoichiometry is pretty much the same. But when the flow of the components for the synthesis of this crystal is being stopped or the temperature is being dropped, actually, the chemical composition of the crystal starts to change. And as a result, about 100 nanometers away from the edges of the crystal, the chemical composition is slightly different compared to what it is at the core of the crystal. And we could see that both in tip enhanced Raman imaging and tip enhanced PL. Okay. So now let's switch to a fairly exciting set of experiments, the tip enhanced Raman and tip enhanced PL characterization of strain-related accidents in nanobubbles of transition metal dichalcogenides. One interesting property of two-dimensional materials or at least some of two-dimensional material with specific band arrangement is that when we mechanically strain the material, that can lead to the accident funneling to the area which has been strained. Thus, the properties of this strained area may have -- may change quite dramatically compared to nondisturbed material. And it was found quite some time ago already that actually these strained areas, they are even more interesting. And at cryogenic temperatures, they can serve as single photon emitters. And as probably you know, single photon emitters are being considered as one of the possible candidates for qubits for quantum computers. So there was quite significant interest in investigation, the properties of these localized strain in the crystals of transition metal dichalcogenide. And I was lucky to participate in this long project, led by James Schuck and his group in Columbia University, and this work was published last year in Nature Nanotechnology. And in this work, we worked -- we characterized the optoelectronic properties of the bubbles in crystals of tungsten diselenide -- monolayers of tungsten diselenide. What was observed quite early in the course of that research was that in addition to normal peak, the nanobubbles, they also showed red-shifted peak that was absent in the adjacent flat area. And the degree of the diversity of the Raman spectra across the bubble, it was really amazing. Literally, each new pixel produced a different spectrum. So it was very obvious that the strain in these nanobubbles or the distribution of the strain strongly affects the Raman or tip enhanced or photoluminescence response. And here is another illustration of this extreme inhomogeneity of the properties, even in small nanobubbles. As you can see here, the [indiscernible] is only 15 nanometers. And this large bubble, we can see that there is not that much response here in the center, but there is nice, circular, bright area in this nanobubble. The smaller bubbles, they seem very bright and quite localized, but we should, again, remember that the size of these bubbles is only about 20 nanometers. And we can see that the response from the small bubbles is very strong, while the response from the flat area around the bigger bubble is not as strong. But the response of TERS signal from the inevitable organic contamination are always present on the samples, it's actually the same for flat area of the bubbles. And what it tells us is that the difference in the response over the nanobubbles and the flat area is not due to the variation of the enhancement of the TERS state. And the reason for this significantly different response is definitely related to the mechanical strain in this strain formation in these nanobubbles. We were lucky that in the course of the work, a very nice theoretical paper appeared in Nano Letters that actually describes the behavior and the distribution of the mechanical strain and how it affects the spatial localization and the optoelectronic properties of these structures. And apparently, a collaboration was established with that group. And it was amazing that their calculations, based on the actual AFM measurements of some nanobubbles, they correlated reasonably well with experimentally observed results. So we could claim that we established reasonably good understanding of how the mechanical strain in these nanobubbles is being distributed across the bubble and how it affects the optoelectronic properties in these strained nanobubbles. Okay. So that chapter is over. Let's move to the next one, the TERS characterization of nanoscale variation in TMDC substrate interaction. Back in 2018, a very nice paper appeared in ACS Nano, journal describing the use of freshly evaporated gold for exfoliation of very large areas -- monolayer areas of transition metal dichalcogenides. And it was demonstrated that using gold-assisted exfoliation, one can obtain almost square centimeters of predominantly monolayer material. What was also observed, that the freshness of gold was extremely important. Literally, 15 minutes after the gold substrate was taken out of the vacuum chamber, the exfoliation efficiency went down the drain. It was, therefore, apparent that the gold-TMDC interaction is actually responsible for this efficient exfoliation of the monolayers. And it was showed pretty much at the same time that the Raman spectra observed in these gold-assisted exfoliated monolayers, they were not exactly normal Raman spectra usually observed in materials exfoliated to, let's say, silicon, silicon dioxide. And there was -- in some areas, researchers observed the presence of this red-shifted peak next to the normal A prime peak to the right. And there were some suspicions that actually, the presence of this peak is an indication of strong interaction of gold with the transition metal dichalcogenide. And in the follow-up research, actually, this was confirmed. So in the materials exfoliated using freshly evaporated gold, quite often, we observed the presence of this red-shifted peak next to the normal A prime or A1g peak, in this case, in molybdenum disulfide. And we would like to understand why we observed that strongly red-shifted peak or why, in general, we could observe the variations of the Raman response in the monolayers of this gold-assisted exfoliated molybdenum disulfide. Naturally, we applied TERS imaging of these materials, and we observed several interesting features in these crystals that were exfoliated using freshly evaporated gold. First of all, it was very clear that sometimes in the predominantly monolayer areas, we may have tiny islands of the bilayers. They are really tiny. So you can see here, this whole image is about only 800 nanometers. So the size of these bilayer islands in the predominantly monolayer crystal is only a few tens of nanometers, 20, 30, not more than that. But of course, they can contribute to the Raman spectra -- Raman spectrum, especially when you do it with conventional confocal Raman when the laser spot is approximately 0.5 micron across or even more than that. Another observation that was made is that within the monolayer, we may observe the areas, which demonstrate strong interaction and therefore, strong shift of the A prime peak. And literally, few tens or few hundreds of nanometers away from such areas, there may be areas with less interaction and more normal position of the A prime peak. So we came to the conclusion that in the process of the exfoliation, the very tight contact is established not across the whole surface, but it's kind of -- it's in a patchy manner, and there are areas with really strong interactions between gold and transition metal dichalcogenide crystal, and these areas actually are responsible for the efficiency of the exfoliation. But also next to those areas, there may be areas where the contact between the TMD crystal and the gold substrate was not as good. And the Raman response from these areas was more or less regular Raman response that people usually observe in exfoliated or CVD-grown crystals. Okay. So next -- and to some extent, this chapter actually continues the story of the previous one. To some extent, we could say that in gold-assisted exfoliated TMD crystals, we had the lateral heterostructures with higher and lower doping level or high and lower level of interaction between the gold substrate and the TMD crystal. But in this chapter we'll discuss the characterization of the vertical or lateral heterostructures or vertical/lateral heterostructures that were created on TERS. Before we continue, I would like to spend some time talking about the importance of the proper sample preparation. So unfortunately, TERS is not as universal as conventional Raman is, not yet at least. Proper sample preparation is the key for successful TERS or tip enhanced PL imaging. Since very often, the TERS and TPL -- tip enhanced PL measurements are performed when the tip is in direct contact with the sample, clean sample surface, free of organic contaminants is very, very important. For tip enhanced PL measurements, silicon, silicon dioxide substrates work fundamentally better compared to [indiscernible] I do not have really good scientific explanation of that fact, but it's kind of experimental result that we've observed multiple times. And wherever possible, it is advised to use the advantage of the gap-mode TERS, preparing the sample on metallic substrates like gold, silver, chromium, gallium and so forth. And for this purpose, actually, we developed a fairly straightforward and fairly simple technique, gold-assisted transfer from -- of transition metal dichalcogenide crystals that were either CVD grown on the carrier wafer or exfoliated to the carrier wafer using a layer of metal. Usually, it's gold, but you can use silver or other shiny metals. So we cover the crystal with gold layer, then we glue -- transfer wafer with thin layer epoxy. And after epoxy pures, we can take the gold together with the TMD crystals embedded in the gold layer away from the carrier wafer. Therefore, we flip the crystals pretty much [ belly up ], and this is the ideal configuration for tip enhanced Raman scattering measurements, at least on thin layers. And as an example, I can show you the results obtained on the samples synthesized in the group of Professor Xiangfeng Duan in UCLA. These are -- so here, we see contact potential difference or the surface potential image of the bilayer. So this is a complicated structure. So we have bilayer and monolayer of tungsten disulfide. And along the edges of the bilayer of tungsten disulfide, a layer of tungsten diselenide was grown. So we have both vertical and lateral heterostructure at the same time in this material. And here, we see TERS image of this area that clearly shows the characteristic Raman signature, sort of monolayer tungsten disulfide, bilayer tungsten disulfide and the heterobilayer, where we have tungsten diselenide monolayer on top of the monolayer of tungsten disulfide. What was very interesting, that along the edges of this heterobilayer, the intensity of the signal of tungsten disulfide was significantly increased, and it was pretty much pixel limited in terms of the spatial resolution. So apparently, something very interesting is happening at the very edge of the vertical heterostructure that affects the response from the monolayer. And to better illustrate this, we can once again revisit the spectral response or TERS response from corresponding parts of this vertical and lateral heterostructure. Okay. So another illustration of the usefulness of TERS for characterization of synthesized heterostructures is the work that we conducted together with the group of Patrick Kung in the University of Alabama, Nick Borys in Montana State University and the group of [indiscernible] So in these images, we can see topography contact potential difference and the capacitance image of the molybdenum disulfide, tungsten disulfide hetero monolayer. So we have molybdenum disulfide core and tungsten disulfide shell around. And these crystals, they were characterized first with conventional confocal PL imaging, and we could clearly see that in the shell, we see nice response, which is typical to tungsten disulfide. In the core of the crystal, we see corresponding PL peaks that resembles closely the molybdenum disulfide peak. But in between, the position and the shape of the peak was different compared to either tungsten disulfide or molybdenum disulfide core. So it was very interesting to understand the structure of this transition area. And for that purpose, we used the gold-assisted transfer and we characterized, using TERS imaging, exactly the same crystal that we characterized using PL imaging of the original crystal. So in these TERS maps, we can clearly see 3 components of this structure. So we see molybdenum disulfide core. We see tungsten disulfide shell. And in between, there is this alloyed area with Raman response clearly different from the Raman response of molybdenum disulfide or tungsten disulfide. And it's characterized by strongly increased intensity of these 2 peaks, and also the position of this peak changes compared to the position of this peak in tungsten disulfide or molybdenum disulfide. So we have 2 characteristic peaks that actually identify the transition area. And this TERS map was obtained using 638-nanometer excitation, which is a resonant for both the tungsten disulfide and molybdenum disulfide. And another important information that we obtained using TERS imaging is that the lateral extension of this transition area is very different along the borderline between the molybdenum disulfide core and tungsten disulfide shell. In some areas, this transition area extended over almost 300 nanometers. But in some areas, it was pixel limited, so less than 25 nanometers. Then we repeated the same measurement using a different excitation. This time, it was 785, which is below the band gap for both the molybdenum disulfide and tungsten disulfide. And as expected, the TERS spectra looked very different. So for example, we don't see -- pretty much don't see at all the presence of the E2g peak, which is in-plane vibration in tungsten disulfide or molybdenum disulfide. And the reason for that is because in gap mode, with nonresonant excitation, it's the out-of-plane modes that get preferentially enhanced. And therefore, using -- or comparing the results of TERS measurements obtained with resonant and nonresonant excitation, we can make some judgment on the nature of the peaks that we observe, whether they are in plane or out of plane. And again, as we observed it with red excitation, we saw that the position of this not-really-well-understood peak in the heterostructure is actually somewhere in between of the position of this peak in molybdenum disulfide and tungsten disulfide. And again, we repeated the observation that the extension of this alloyed transition area varies greatly as we move along the borderline between the core and the shell of this heterostructure. And to illustrate that, we can take a look at the digital zoom in on the TERS image here. And with green and red spectra, we have the averaged Raman -- average TERS spectra that were averaged over 1 pixel, white line along the borderline between the molybdenum disulfide and tungsten disulfide. So correspondingly, red corresponds to molybdenum disulfide, green corresponds to tungsten disulfide, and the single pixel white line in between clearly shows the position of the peak is typical for the transition alloyed area rather than to molybdenum disulfide or tungsten disulfide. So the transition between these 2 materials can be very sharp. But literally, 200, 300 nanometers away from this sharp transition area, we observed very broad transition area that extends over several hundreds of nanometers. So it's kind of interesting observation, and I would like to stress that this sort of measurement is pretty much impossible to -- this sort of information is impossible to get with any other technique except for the tip enhanced Raman scattering. Okay. So hot, fresh recent results on ultra-low frequency TERS. First, I would like to start with the demonstration that actually ultra-low frequency TERS is possible. And here, we have 2 spectra obtained from exactly the same location: one, the blue spectrum with conventional confocal Raman; and red spectrum is TERS spectrum, obtained under the same illumination condition with the same integration time. So you can clearly see that we observed low frequency bands and very strong increase of this broad background. The nature of this background, again, still needs to be understood really well because it may be quite important for ultra-low frequency measurements. But the good news is that we can observe with TERS even the ultra-low frequency Raman bands, which is very important for characterization of two-dimensional materials and their heterostructures. And to illustrate that, let's take a look at these spectra. So we have injunction of the monolayer, bilayer and tri-layer of molybdenum disulfide that was exfoliated using the gold-assisted transfer with freshly evaporated gold. And in the tri-layer, we observed the wrinkle so we collected TERS map over the wrinkle and the flat area. We collected TERS map over the flat area in the bilayer and over the flat area in the monolayer. And the spectra here -- so we see clearly a difference -- clear difference between the monolayer signal and the bi- -- tri-layer signal, but there is not that much difference between the signals of the bilayer and tri-layer in this so-called fingerprint area. But if we look at the area of the ultra-low frequency bands, we clearly see the difference between the tri-layer and the bilayer. And also in the wrinkle, we clearly see this band at 10.4 wavenumbers, which is quite impressive in terms of the capabilities of Raman imaging at all and TERS imaging in particular. We repeated the same measurements using higher-density grating just to make sure that there were no artifacts and the results are repeatable and we got very similar spectral results. So again, we observed this 10.4 wavenumber peak in the wrinkle and 42.9, 53.7 peaks in tri-layer and bilayer correspondingly. The interesting thing was that when we imaged exactly the same area using helium-neon laser, which is resonant, as I mentioned before, with molybdenum disulfide, the results were quite different. So some peak state, like, for example, the peak in the wrinkle, it was pretty much at the same position as the peak obtained with 785 excitation. But the 53 and 42 wavenumber peaks were mostly gone. And we observed different peaks located somewhere around probably 40 wavenumbers. But clearly, the spectral response with red and infrared excitation was quite different. And here is the summary of the response -- TERS response or ultra-low frequency TERS response of the same areas with 2 different excitations. So we can see that for the wrinkle, the position of this ultra-low frequency band at approximately 10 wavenumbers pretty much stays the same. While the behavior of the tri-layer is quite different or the Raman signature of the tri-layer is quite different for 785 excitation from the Raman signature at 632.8 excitation. So there are many things here that still need to be understood, but the good news is that TERS is capable of ultra-low frequency measurements. And this really opens the door to the nanophononics. I really like this term that was proposed by our good customer in PNNL, Dr. Patrick El-Khoury. And we are working on putting together a manuscript, summarizing our recent results on ultra-low frequency TERS characterization of different transition metal dichalcogenides. Okay. And as we are coming close of the promised cherry of the pie, the TERS characterization of Moire patterns in twisted bilayer graphene. An amazing work done by the group of [indiscernible] and it definitely deserved the front-page Nature place. So this was covered, big Nature article. And here, what this group -- what they did, they produced the twisted bilayer graphene sample. And as a result of this twist, a special Moire pattern or reconstructed Moire pattern appeared in this twisted bilayer. And they were able to image the optical properties of this reconstructed Moire pattern based on the TERS signature. They did it with their homebuilt TERS instrument, using very special, highly efficient TERS probes. But the good news is that they clearly demonstrated for the first time that TERS is actually applicable for characterization of twisted bilayers. And probably many of you know that this twistronics is a very hot topic in 2D material community because these twisted bilayers or twisted multilayers, they possess really unique and fascinating properties that can be very useful for advanced applications even in the nearest future. Okay. So to summarize, TERS and tip enhanced PL provide unique possibilities for this heterostructures characterization of nanoscale features in 2D materials, often unachievable otherwise. In addition, these 2 techniques allow straightforward course correlation of observed spectroscopic features with other channels provided by scanning probe microscopy topography, surface potential, conductivity, friction, photocurrent, et cetera. Next step, time-resolved and nonlinear tip enhanced PL and TERS, and this is already happening. And I hope that with our recent advances with ultra-low frequency TERS, we can start thinking about tip enhanced brilliant spectroscopy, which will be really a fascinating further step in the nanoscale spectroscopy -- vibrational spectroscopy. Another thing that I would like to say, that TERS and tip enhanced PL research, it does require dedication and commitment. But the reward would be there. Thank you very much.

Thank you, Andrey. [Operator Instructions] We are running out of time, but we do have a few questions.

So our first one is, is titanium carbide the only MXene on which TERS imaging was obtained?