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

Hello. My name is Margaret Harris, and I'm an online editor of PhysicsWorld. I'd like to welcome you to today's webinar on using Raman and photoluminescence microscopies to address challenges in 2D materials research. Joining us as our speakers today are Agnès Tempez, an application scientist working in the nanoscopy group at HORIBA France; and also Thibault Brulé, a Raman application scientist also at HORIBA France. Agnès has a background in time-of-flight mass spectrometry for materials characterization. Thibault did research on proteins and blood characterization based on dynamic surface enhanced Raman spectroscopy before he joined HORIBA. Agnès and Thibault welcome your questions, and you can submit those questions at any time during the webinar using the questions facility you'll see on your screen. We'll have some time for a question-and-answer session at the end. And I believe you should also be able to see a link to some handouts providing more information. Before we start today's presentation, we've got a poll for you to fill out, which should come up on your screens very shortly. And we'll give you a few seconds to read the poll and to make your answers. [Voting]

Just a few more seconds to give you time to vote. I think you can vote for multiple options if you like. All right then. I think we'll close the poll and carry on to the first slide of the presentation. Now I'll hand over to Agnès and the presentation.

Today, to give you a presentation about optical microscopies for the study of 2D materials. Our goal is to show you a number of results opting on the large number of materials. These results provide chemical and physical information and the microscale as well at the nanoscale. Let me first give you a short introduction. What is a 2D material? A 2D material is a crystalline material with a thickness of a single crystalline unit. The first 2D material discovered is graphene, which consists in a single carbon layer of the hexagonal graphite network. Why there is so much interest in 2D materials? Because of the exceptional properties: electrical, they have very high electron mobility; optical, high transparency for graphene; thermal, very high thermal conductivity; and they are also mechanically very strong. Graphene is already used for industrial applications, present as composites with enhanced mechanical and thermal properties in photoconductors, chemical and biological sensor and battery. But graphene is under a lot of research to be integrated as an active layer in a lot of optoelectronic devices, sensors, super capacitors, solar cell and neural interfaces. Nevertheless, there is an issue about graphene. It is a zero band gap material. So a family of 2D materials with semiconducting properties meaning tunable band gap has emerged. They are the transition metal dichalcogenides, consisting of a transition metal highlighted in blue in this periodic table and a chalcogen element: sulfur, selenium, or tellurium. Other 2D materials are made of 1 element of the fourth column, like graphene with carbon. There is also black phosphorus and a dielectric material, hexagonal boron nitride. I'm pretty sure you're aware of the huge funding delivered to research in 2D materials right now and for the last few years. Things are moving ahead, as you can see with the front cover of this issue of Nature Materials entitled 2D materials towards industrialization. This issue addresses challenges because challenges to integrate this 2D material into final nano devices are big. First, fabric heating film in a reproducible way and another large scale; second, developing processes of patterning capping etching that will maintain the integrity of the 2D material. As it is clearly said in the sentence, this is the biggest challenge to preserve the intrinsic properties and device performance of 2D materials that are extremely sensitive to structural disorder that may be generated during synthesis or processing. Reproducibility on a large scale can thus hardly be achieved without structural control at the nano scale. And of course, we've recognized the 2 keywords for us in this sentence, disorder and nanoscale. Because this is what we do and we do very well, characterizing disorder at both the macro scale and the nanoscale. To you, Thibault, now to tell us how we do that.

Thank you so much, Agnès, for the brief introduction. Let me now introduce you about the different phenomena that we are looking with our systems like the Raman, photoluminescence but also TERS and TEPL phenomena. The Raman effect is an inelastic light scattering. When you shine a molecule of laser, most of the light will be scattered by the molecule. In that case, what can happen is that the full energy of the laser is scattered as it was at the excitation. So this is the release scattering. It's inelastic light scattering. But you can have also a part of the energy of the laser that can be taken in order to do some vibration in the molecule. In that case, this is what we have in terms -- this is what we call the Raman effect. This effect has been discovered by Professor Raman in 1928, in India, and can characterize fully the different molecules that are under the laser shining. Raman spectroscopy can talks about molecules and can give a lot of different information. First, looking at the band position, you will have information about the chemical structure of the molecule. If you look also at the band density, you will have information about the polar visibility of the chemical bond. That can be useful because this possibility will depend of the environment of the molecule. You can have also information about the crystallinity of the molecule. Indeed, the bandwidth will give you some information about it. And if you have a shift of the band position, you can also have information about any deformation in the crystal structure or any stress that has been applied on the molecule. The polarization will also induce some effect on the band densities and the band ratios. In this, the orientation of the molecule will be sensitive to the laser polarization. And so you will have different band intensities depending on the orientation of the molecule. Finally, the band ratio can give information about relative concentration for different kind of molecule in the same laser spots. For 2D materials, Raman can give a lot of information. Indeed, it's a nonbasic technique and an nondestructive technique. It has a high throughput, and it's substrate independent. It can provide you a clear identification of the different 2D materials that you have on your sample. For example, for the graphene, you can easily distinguish between the different levels of -- and the number of layers between, for example, single layer graphene or graphite. You will not have the same Raman spectrum. It can go -- so to a different point of information like the disorder in your structure, but also the number of layers, the presence of defects and can give you also information about the presence of any doping. The photoluminescence is a physical effect that provide information about electronic properties of semiconductor materials. In this, when you shine your semiconductor with a laser, the electrons that are present in the balance point will move up to the conduction band. In that case, when you will create an electron all pair and when the electron is present in the conduction band will combine with all in the balance band, you will create what we call an accident. This accident is -- will have an energy characteristic to the band gap between the balance and condition band of your semiconductor. With that, you will have also looking at the spectrum of emission of this -- of this accident. You will have information about the disorder of your structure, but also the number of years and the presence of the effect in your structure. And so you will have the electronic information about the electronic properties of your material, where, in Raman, you have the information about the chemical structure of your material. So photoluminescence is a complementary technique to Raman spectroscopy.

Let's move on -- or I should say, let's move down to the nanoscale with Tip-Enhanced Raman Spectroscopy. How do we break the Rayleigh diffraction criterion, limiting optical spectroscopy to submicron special resolution? How Tip-Enhanced Raman Spectroscopy, or TERS, works? In TERS, the Raman excitation laser, label as 1 on the left side of the diagram, is focused at the end of the AFM probe coated with gold or silver indicated as 2. Matching the wavelength of the Raman laser to the natural surface plasmonic frequency of the novel metal generates an intense localized evanescence electromagnetic field or hotspot, and the prop tip drawn as #3 on the diagram. Since the Raman intensity from the sample is proportional to the local electrical field, bringing the hotspot close to the sample significantly enhance the Raman signal, often by a power of 10^5 or 10^6. You can think of this plasmonic effect as the lightning rod effect. The antenna is the coated sharp tip. This instrumentation is defined according to 3 factors imposed by this plasmonic resonance. First, the polarization dependence, the angular dependence and the spectral dependence. And of course, we have taken these factors into account in our instrumentation that you will see in a few slides. Let me show you right now how this huge game environment signal localized under the nano region defined by the tip allows us to access the nanoscale information. I am going to illustrate this from graphene and graphene oxide flakes. On the left, you have a combination of Raman map giving you a number of layer of the graphene and underneath, an AFM topographic map. If we focus on these small 3 by 2-micron area, here, it's the microgram and map, how it would look like. And if we perform tours on the same area, this is what we get. You can see clearly now the graphene ages tends to this great improvement of the optical resolution. You can even see some carbon nanogen surrounding the graphene oxide flake. Quickly, another example to compare in this case, photoluminescence, macro photoluminescence or far-field PL versus Tip-Enhanced Photoluminescence or NanoPL. On this nice, polygonal flake of tungsten diselenide, you can see how many more details are invisible on the right map. These details are in homogeneity such as point [ defect ], grain boundaries or wrinkles that you couldn't see with conventional optical techniques. This data I acquired from a instrument coupling an atomic force microscope and the Raman microscope. And now I'd like to point out something really important. We find the benefit of achieving nano resolution for photoluminescence and Raman spectroscopy. This instrument allow to correlate this chemical information with the physical properties given by the atomic force microscope at a nanoscale. So we get both type of information, chemical and physical at the same time, hubs and spokes at the nanoscale. And we will integrate this for a number of examples in the next slides. Before Thibault gives you an overview of our Raman microscopes, I am showing you one of our NanoRaman platform. In this case, a version with a compact explorer Raman microscope on the right side of the picture. On the left side, it's the SmartSPM, the AFM microscope. In between, you have the optical mechanical company called OmegaScope, which is bringing the laser from the top for colocalized AFM-Raman measurement, all from the sides for optimized tip-enhanced measurement.

In the HORIBA portfolio, we have different micro Raman platforms down to a compact and automated [ system properties ] up to the HR Evolution, which allows you to go from deep to a highest resolution. But our main platform is the LabRAM Soleil, which is our brand-new micro Raman platform that allows you to do all kind of micro spectroscopies that you are looking for and which is a very fast and robust platform with everything integrated in less than 1 meter cube system. All these Raman -- micro Raman platforms, allows you to characterize your materials at the microscale. However, you can also combine all of them with the NanoRaman capabilities that we have, coupling this micro Raman platform with the AFM. In that case, you obtain a NanoRaman system, which is able to do nanoscale characterization, encouraging microscopy with high quality and high reproducibility. Let's move now to some examples about application of these different techniques on 2D materials and start with carbon-based materials investigations. Carbon-based materials are materials that are based on carbons. So these materials can be from 0 dimension like [ fullerenes ] up to 3D materials like diamond. They are structured materials with specific organization of the carbon atoms in order to have crystalline structures. In that case, we will more focus on the 1D and 2D materials like the nanotubes that can be of interest in semiconductor industry. The graphene, that is the typical 2D material for carbon-based. And the graphite also because it's a derivate of the graphene which has just more layers than 1 as you can have in graphene. The different applications that you have for these carbon-made materials are in automotive, are in the different structures of the different part of the automotive. So carbon-made materials can be involved. Also in nanoelectronics where you can find some kind of carbon-based materials because of the semiconductor capabilities of the graphene of the nanotubes. In the display also because these materials have had the capability to be very robust and to be flexible. In biomedical do some drug delivery, but also in the battery to be used for the anode of the battery because of the high reproducibility and the high stability of these different structures and in more other applications. With our system, we have the capability first to characterize carbon nanotubes. At the microscale with the micro Raman, we have the capability to characterize these carbon nanotubes. Looking at the few characteristic bands, but as the D, G and 2D bands are in the Raman spectrum for this material. We will also have information about the quality of these nanotubes and a number of walls that we can have in the nanotube. Moreover, looking at low frequency. We are able to access to the briefing mode that will directly inform about the width of these nanotubes. Moreover, you have an example here where we demonstrate that with our micro Raman, the LabRAM Soleil system, we have been able to reach the limit of resolution that can be achievable in a micro Raman, here around 315 nanometers. If you want to go below, you will have to use the NanoRaman capabilities.

Here are 2 maps of our single carbon nanotube, showing the nano resolution capability of our NanoRaman system. In this TERS map, the signal from the D, G, 2D bands are super imposed. On the left side, The map is 400 by 400 nanometer with a pixel size of 4 nanometer. The map on the right is 100 by 100 nanometer with a pixel size of 1.3 nanometer. This is allowing us to measure the width of 8 nanometer of this single-wall nanotube. I would like to point out that this data were obtained on a conference booth, so not in the lab, not in the ideal condition, and this really proves the stability of the system. Now if we look at 2 Raman spectra from 2 adjacent points along the tube, shown in this graph in red and blue, you can see that the band signal drops dramatically from one to the next. Along the tube and not across indicated nearly absence of defect in the second point. I don't want to get into the debate of the definition of optical resolution achievable with TERS, but I think it's reasonable to say that we can chemically distinguish 2 points separated by only 1 nanometer. I'm now going to show you some data on graphene oxide flakes deposited on gold. And so this work, we've collaborated with the group of Naresh Kumar of the NPL. They actually work on the data treatment. On the left, you have a topography map with a height profile on the right corresponding to the dotted line. This profile reveals thick layer and pure area of graphene oxides. Now I'm showing the TERS map corresponding to the D bands and the chip intensity of this graphene oxide on the same region. The resolution can be estimated by doing a TERS intensity profile. As you can see, we can reach an optical resolution of 10 nanometers. Let's take now a deeper look at the Raman data to extract meaningful chemical information. We can see very nice contrast in the D band and G band TERS map with some enhanced Raman intensity at the edge or at the wrinkles. Actually, we can distinguish 4 types of Raman spectra. And the 1 to 4 points marked in the B label map correspond to example of these 4 types of spectrum. First, you have the low contrast area with a D to G band ratio of 1.2 that can be attributed to the graphene oxide letters. Then some high TERS intensity area with a D to G band intensity ratio of 1.5 corresponding to area with a large number of defects. Basically, let it at the wrinkles, edges or steps. Finally, you have some new and very sharp additional Raman bands that can be found quite different to the typical D to G bands, and these bands correspond to carboxyl or metal groups bonded to the graphene oxide letters. This distribution suggests that the chemical groups are preferentially localized at step and edges. To conclude on this data, we can say that the D to G ratio can be used to locate chemical groups on the carboxyl graphene oxide surface. I have just shown you how the Raman signal of graphene oxide can be related to topographic features such as edges, steps or wrinkles. Let's now see how we could link the chemical information with electrical properties. Here, you have a contact potential difference map, or CPD map, obtained from Kelvin probe measurement. From this, we can measure the Fermi level. And here, it varies from 5 electron volt to [ 5.14 ] electron volt. Below, you have a TERS map corresponding to the D band of the same graphene oxide flakes. And here, actually, there is no clear correspondence between the D band intensity and the CPD response. But if now we plot the D to G band ratio instead of the D band intensity, we can correlate it to the surface potential. As you can see, the high CPD area, which is the high D to G ratio area, like, for example, the area noted B, and the low CPD area always have low D to G ratio like the A area. So what we can say is that the Fermi level in an area of graphene oxide increases with the D to G ratio, which is also an indication of the density of defects.

Graphene flakes can also be characterized with micro Raman. Here, you have an example of the characterization of the graphene flake during the Raman mapping at the micron scale with the LabRAM Soleil. With this system, we have the possibility to use the SmartSampling algorithm that allows to go very fast in order to do the mapping. Here, you can see on the left, a mapping that has been done in around 3 minutes and 30 seconds comparing to one doing in 41 minutes with the same level of details. And that has been acquired using the classical fast mapping mode that you have on all our micro Raman systems. In that case, doing this kind of characterization will give you a lot of information about the number of layer in this flake and so characterize the quality of geographic structure, but also we'll give you some information about the presence of defects and the electronic properties of it, because we will be able to do some photoluminescence measurement on the semiconductor structure. Let's move now to TMDCs investigations. TMDC materials are semiconductor materials based on transition metal, coupled with 2 chalcogenides elements. The physical properties of these materials depends on the disorder, but also the number of failures, the presence of defect and the heterojunctions. The heterojunctions are the association of different TMDC materials in lateral or also in vertical crystal association. In order to characterize such material, Raman and photoluminescence are very powerful because they will give information about these different physical properties of this material at the micron or the nanoscale, depending on how you will characterize them. The application of these materials are in photovoltaic because they are highly used for photovoltaic structure building because of their high capability to combine between the light and the electrons. Also for nano electronics because of the high capability as semiconductor materials. They can also be used for biosensing. Looking at the electrical properties of this material couples with some biosensors for drug delivery and also for catalysis of some elements because they can be used as catalyzers and they can also be applied in more other applications. The first example about the characterization of such material is here a colocalized measurement of photoluminescence and ultra low frequency Raman measurement. That allows us to access to the number of layers of this difference of this structure. Here you have a flake on the top, you can see that we have -- we did the photoluminescence characterization. So in purple, you see the photoluminescence profile when we have a monolayer area and in yellow, when we have a multilayer area. Based on that, doing the Raman mapping, we have been able to see the distribution of these 2 different things between monolayer and multilayer. However, we don't have any information about the exact number of layers in that case. But when we look in ultra low frequency in Raman, we are able to distinguish between the different modes that appears due to the vertical junction between the 2 or more layers that we can have when we have more layers at the micron scale. In that case, we have been able to distinguish between 1, 2 and 3 layers based on the switching mode of the vertical junction between the different layers of WS2 that we have here.

Let's now look at correlating nanoscopy from tungsten diselenide that has been transferred onto a gold surface. The video shows TERS data overlay with topography and surface potential maps. CPD is measured and mapped with the AFM TERS tip. It shows interesting variations in electronic properties through here, the formulary position. One can see the nice correlation between Raman signal of A1g plus D to G bands with the contact potential difference. Zones with low CPD coincide with on low Raman intensity, and high CPD zone exhibit higher and broader Raman signal. A way to confirm the correlation between Raman and KPFM signals is to take profiles across the same line in the TERS map of the signal center at 240 wave number from the A1g plus E2g bands and in the CPD map. Again higher CPD regions exhibit stronger TERS response and new Raman peaks. Correlation between surface potential and resonant tip-enhanced Raman signatures indicates the presence of inhomogeneities within interfacial electronic properties, which are attributed to variation in the local doping of tungsten diselenide. Indeed, in the area with low CPD, some metallic phases form and suppressed the TERS response.

Association of 2 different chalcogenide atoms can also happen as it is the case with this tungsten surface selenide structure. With photoluminescence, we are able to characterize the electronic properties of this material at the micron scale very easily. But we won't have any information exactly about the ratio between the sulfur and diselenide elements. In that case, the Raman can be a very good tool in order to control this ratio in the flake and to control the evolution of this ratio in order to see the evolution of the electronic properties associated to these different ratios, and also to characterize the quality of the synthetized material. At every bilayers of 2 different TMC materials can be also of interest. In that case, we have a vertical junction between SnS2 and WS2. The photoluminescence microscopy becomes a perfect tool for the characterization of the electronic properties of this material, looking at the distribution of the photoluminescence intensity and the photoluminescence band position of this single flake. However, As you can see here, when we reach the vertical heterojunction, we lose the information about the electrical properties because there is no band gap anymore due to the vertical heterojunction. In that case, the Raman will give the possibility to confirm the presence of this structure. Looking at the different spectra at different applications in the surrounding of this vertical integration. Looking at the position of the band and the presence of new specific bands, we can still have information about the kind of heterojunction, which appears and what is present in that case, under WS2 flake and confirm that the SnS2 is well present in the bottom corner of this flake.

With [ Nat Raman ], my colleague in the United States, in collaboration with researchers in UCLA and [ UNAN University ] has measured a sample of tungsten disulfide, tungsten diselenide vertical heterostructure. On the left, you have the surface potential of the end of a flake. On this 12 by 8 micron, it is easy to distinct the monolayer, the bilayer, tungsten disulfide and the hetero layer with a stack of tungsten disulfide and selenide. Within this area, a TERS map was collected in a smaller area of 2 by 1 micron, including monolayer, bilayer and the hetero bilayer. On the 3 colored map shown on the top right, the middle hetero layer appears in green from the E2g bands signature at 250 wave number as you can see in the corresponding green spectrum. In blue, the intensity of the sulfide A1g band is more intense in the bilayer than in the monolayer. In here, you have a typical spectrum from the bilayer; and in red, a typical spectrum from the monolayer. Below the 3 colored map is a map of the sulfide A1g band intensity only. And interestingly, at the border between the monolayer and the heterolayer, there is an enhancement of the intensity.

Lateral integration can also be characterized. This structure will have specific electronic properties and need to be highly characterized in order to see the change between the different crystal phases. Micro photoluminescence is also a perfect tool for that. Here, you have an example of a map that has been done in micro photoluminescence and that shows the distribution of the WS2 in red, WSSe in blue and WSe2 structures in green. So we can see the different crystal phases that are ordered and we can have so information about the junction between these different elements. However, due to the spatial resolution limitation that we have in micro photoluminescence, we don't have a full understanding of this junction. That's why it has been good to move to nano photoluminescence characterization for this material.

Before I show you the NanoPL map that Thibault really wants to see, I am showing you the topography and phase images of 1 end of the same 60-micron beam flake. As you can see on both images, we can see a distinct material between the inner sulfide and the outer diselenide. It is the ternary alloys, the tungsten sulfide selenide. The idea of having an alloy is band gap engineering and having sharp and well-controlled interfaces. Here we go with the true NanoPL map. I've taken the same color as Thibault. So in red, you see the PL from the inner disulfide around 630 nanometer; in green the diselenide with a PL center at 760 nanometer; and in between, a PL at the center at 660 nanometer for the alloy. As you can see, the 2 alloy bands on each side of the end of the flake exhibit different intensity PL. On the left, it's quite dark. And on the right, it's a bright blue. They also have, if you remember, a different wave on the topography image. This type of information, some micron PL response variation would not be visible without the tip enhancement. Well, I'll encourage Thibault and you all to download our application note showing more data on this particular sample, but also on other 2D materials.

To conclude, we have seen that optical micro and nano spectroscopies are very nice tools in order to characterize 2D materials. We are also able to have information about the electronical properties looking at the photoluminescence signal that they can give, but also about the chemical properties and the number of failures or the presence of the defect, looking at the Raman features. Moreover, coupling the micro Raman systems with the AFM. We have even more information about the topography of the sample, also the capacity and the distribution of the sample, the phase, et cetera, all the kind of information that you will be able to obtain with an AFM. So in that case, using our combination and unique combination between an AFM and a Raman system, you have the full package in order to fully characterize new 2D materials. We would like to thank you for your attention, and we will be very happy to answer to all of your questions. Thank you.

Well, thank you all for listening to the webinar, and thank you both to Agnès and Thibault for a great presentation. We will have a chance to move on to questions now. So please keep sending your questions in. The first question, I think, have we got Agnès and Thibault there? Just check to make sure that they're coming in.

Yes, we are here.

Excellent. Wonderful. Okay. So the first question that's come in is, sort of why is it important to probe 2D materials on different substrates?

I'll take that one. Well, these 2D materials will be integrated in electronic devices, such as transistor. So the semiconducting material like the transition metal dichalcogenides and the layer based on those materials would be interfacing with the electric, but also with contact with metals. So it is important to study both types of interfaces with silicon oxide and gold, for example, that we've shown in this presentation.

And Thibault, do you want to add anything to that?