 I have experience in doing the cleanup talk, and so I give what I plan to promise to talk about, which probably won't be all that interesting because it got a little bit specialized. But then I thought I would say something about what I got out of this conference to help stimulate the discussion. So a little bit of an overview, this is, I think you already know, the history of the subject goes back a long ways, so about 50 years is for me anyway, but the subject is about 10 years more. And at the time, early time when the subject started, when graphene was first discovered, there was nobody working in the field. So the work went almost unnoticed, and even the author mostly didn't notice it. I don't make a comment about that. And then we went through all these other topics, intercalation compounds that was discussed a number of times, fullerines, nanotubes. During this period, we had a lot of people that joined the field. It got very, very exciting, and lots of people came in on the nanotube period. About 1995, and a big community was built up that understood something about Linear E versus K. They, of course, didn't have graphene, but they, in their minds, they had graphene. So that was going on in this period, and then the big paper came in 2004, and this field just exploded. And now we have approximately, what is it, 1,000 papers per year? More? 4,000 papers a year? It's a huge number. Nobody can read so many, or maybe Andre can, but the rest of us probably can't keep up with it. So it's really a challenge, but it's nice to be, to find, for a person working in a field when you started, there was nothing there, and then we have so much interest. It's quite amazing. So I wanted to focus on the graphene, but I'll refer to a little bit about Raman scattering, because that wasn't really discussed, so that was my topic, because I'm the only one really talking about that in a more detailed way, so that's what I'll do. So we go back to the Wallace paper, and that was 1947, so that was roughly 10 years before I got into the business, and so when I was a very young researcher, my first independent career thing, I was told I couldn't work in what I had been working before, which is super connectivity, so I had to switch fields, and people were doing all kind of magneto optics at that time, and I had access to high magnetic fields right where I was working, so that seemed like a good thing to do. And that is the direction I took. But everybody was working on semiconductors, and one semiconductor seemed rather similar to another semiconductor, especially though there were three, five compounds, and people were making very good two, six compounds yet, so the choice was not so great, so I decided that I like to do something really different, and I thought I would do carbon. And I didn't get a lot of support or interest in this, and that was quite all right because I thought it was interesting, and all the people that are here now, you think it's interesting too, so you agreed with me. So the field was set for me at that time to enter. We had a theory for my classmate, Joel McClure, we went to graduate school together, and he worked out the E versus K for graphite, and we had it only in 2D for two dimensions by Wallace, and he made it in three dimensions, and I needed it in three dimensions. That was the paper in 1957, and in 1962 when I sort of entered the fray, he finished his paper on the magnetic field dependence. So he was just ready for me, and 1960 was also a celebrated year because that was the year that HOPG appeared here in London in Imperial College, and so I went to visit Imperial College. I gave a lecture there and told them that they started graphene in a way, but that is in fact the case. So I was able to get a sample that I could have on orbit, a cyclotron orbit large enough in the sample, so I wasn't getting scattered by the boundaries of the sample. So I had a theory and I had a sample, so it was time to get to work. So that's a history, I'm not sure that you all know about that, but that's how I got started in the business. What I'm going to talk about today, because that's what I promised to talk about, was how we use Raman spectroscopy. And as you can see here, there are all kinds of peaks in the spectrum. I threw in the spectrum for nanotubes, so you can see the radial breathing mode because there's been so much written about that in the last 10 years or so, 15 years. But the main thing for graphene is the G-band where the two atoms, carbon atoms, are vibrating against each other. And we have one layer and then a second layer, and that makes the unit cell, whether you have graphene or graphite, it's almost the same. So there's a great deal of similarity between graphite and graphene. So, and then there were a few other features. When you go to a conference in graphene, you hear about the G-band, and then you hear about this G-prime band, and I'll talk about it, sometimes called the 2D band, but 2D band is actually something a little different from the G-prime band. I'm not going to get into this because this gets into some difficulties, takes too much time, and I'm not going to deal with that. But I would like to speak about some other features between the G-band and this G-prime band. There's some scattering, and this M-band hasn't been discussed very much. One of my crummy students, Victor Brar, when he was an undergraduate in MIT, actually worked on this thing with me, but for this occasion, we did, we have some more work done with the people from Singapore. So that's developing, and Andrea Ferrari has done something recently on this very low-frequency regime at less than 50 wave numbers, and I don't have his view graphs, and I haven't seen his paper yet. I don't know if it's published, but that's a new result that we'll be looking for in the next few months when it gets published, because that's a new feature in this spectrum that we've been waiting for a long time. So this is, even though this is like many things we heard at this conference, part of this field is quite old, but with the reinvigoration that came with the discovery of single-layer graphene, the field was invigorated, and some of these old traditional areas that we've been working on for years are now a little different, because we started thinking about this much more seriously, and we have results that we didn't have before. And that will be the gist of my discussion for you today. So a little bit on graphene with a review. So this is the 1962 paper of Berm, and I went to a couple of conferences and people were questioning whether this was really a correct result. So I met Hans Peter Broom in 1977 when we had the first intercalation compound conference, and they had maybe two handfuls of people. It wasn't a very big conference, because conferences were very small at that time. There weren't that many people working in physics. But he was there, but he didn't talk about this. He was talking about very early work that he did in intercalation physics. He was one of the pioneers in that area. And from the period that I knew him in the early 1970s when that field started until he retired, that he was doing, making new materials, and a lot of work in intercalation physics at never once during all that time that he talked about graphene. But I did talk to him very recently like the last half year, and I asked him what he thought about this paper. He's exactly the same age as I am. He retired maybe 10 or 15 years ago, and he doesn't travel, but he remembers this paper and he says he sticks by it. So there you are. So it has about the synthesis of a single-layer graphene, and he has a table that shows something about it, and he believes that he made graphene from graphene oxide, and some other people have been reproducing that work. I wanted to pass this on because there were some questions about this work, whether he believed in it, what happened to him, and so forth. Some people don't know this history. Okay, so here's graphene, and we've seen this picture many times. There's not much more to say to atoms per unit cell, and the B atom can be in this position or it can be in this one or in this one, so we have some choices. And we could make ABC stacks with different configurations, but the AB, AB, AB is the ground state, and that's what we normally have. This is ordinary, Bernal graphite, and that's what we have a lot of data on. There's also the ABC version that has been known for many years, but there's almost no data in the literature on that. That's a very uncommon form, and there's very little work that has been done on it. And it's gotten interesting now that we have graphene because we can make these different combinations and maybe many more from the discussion that we had here. So that opens up another research field that we might be thinking about. Okay, so this you've heard, and so we have a single layer, and we have linear E versus K, and we have bi-layer very close to the direct point. We have a quadratic form, but when you look away from it, it doesn't look terribly different from this, except that you have two layers rather than one. And I have a little more to say about that. So this is a picture of monolayer and graphene, and two layers, and it was discussed numerous times. This is the famous paper that I think broke the field open, and after this came, it started the thousand paper, and then we were up to, maybe we'll be up to 10,000 by the end of the decade. So anyway, the field took off after this point, the quantum hall effect, integer one in graphene. It was different than Gallium arsenide, so people got excited. Soon thereafter, there was Raman spectra, and this was also taken by the Manchester group in conjunction with Cambridge. So this is Andre Ferrari's work. And we have known that for a long time that the second order effect, this G-prime band, is very large compared to the first order. Usually in physics, the second order effects a second order, and they're small. That's why we call them second order. But this material, because of the linear E versus K, you have all K vectors in the Bruand zone, resonant in a sense, so you're always doing triple resonance. And in general, for graphite, we have a double resonance effect that's responsible for this G-prime process. I'll tell you that a little bit more about that later. But every possibility that you can have for the monolayer is resonant. So it gives you this huge signal at the second order effect. And the bilayer is also quite large, but it's much less than the monolayer. I have been waiting, I'll throw this out to people in the audience. I've been waiting for people to consider this system as a metrology system, because it is kind of simple, and in a way everybody can reproduce it. And we could use, as Philip Kim and his talk was talking about doing the resistivity as a function of temperature and seeing what the power of law and checking the old formulas that we have going back to ancient times. I believe that it is correct to say that until you did it for graphite, nobody checked the T to the fifth power that occurs in the resistivity. I don't know if you know that, but if you look in Quintels where he has all these curves superimposed, yes, they're all similar, but they're not T to the fifth. So it's really great that you could show that reason. You did T to the fifth? Yeah. Oh, okay, but Quintels' picture is not T to the fifth. Yeah. Okay, all right. So you did that, and I didn't know that. Okay, I apologize to you. No. Well, you know, we have to advance the concepts of physics to even if they're boring. So this is the process that that's... He said his thesis was boring. Not that my top is boring. His thesis is boring. I'm happy. I didn't follow that because I kind of stopped teaching at one point. So anyway, I'm going to explain what happens with the G prime band because we'll be using that in the talk. So this is the second order process. Electron comes in, it excites an electron from the conduction band, from the valence band to the conduction band, leaving a hole behind, so we have an electron hole pair. So that makes a neutral system. One of the new things that I learned that's just coming up now, but I didn't hear it at this conference, but we're supposed to talk about what's happening in graphene, is now people are not only exciting electron and hole, but they make an ion charge. So you have two holes and one electron. And they call that a trion, and that has a different spectrum. But I didn't hear anything about that here. I think that's an interesting new development in this field. So anyway, let's go back to this story. So we have the electron excited. Hole is left behind. The excited electron gets scattered, and it scatters from the K to the K prime, which is essentially an equivalent point, except that there's time reversal symmetry between the two. But when it gets to here with the scattering process, it emits a phonon in the stokes process, so it loses a little bit of energy. And then it comes back here, loses a little bit more energy, and then it recombines. So this is a two phonon process. There are no elastic events as stated. If the elastic events happen as they do in the D band, then this would be 2D. But those elastic events, well, they could be there, but that's energetically not favorable. So that's the difference between 2D and G prime, if you really want to be fussy about that. So now we go with different photon energies, and so with different photon energies, we'll be having a smaller Q vector, and that means that the phonon will have less energy. So because of that, it will be shifted. So you have a dispersion relation. G prime is always a dispersive feature. And now if I come in with red light here, it's even a smaller wave vector. And so you could see that the G prime band has a very large shift, but I'm going to show you some others that have even more shifts later on, which is very unusual in second-order processes. So the shift here is 100 wave numbers per EV. That's a very large for Raman scattering. Because of Linear E versus K, every one of these processes is resonant because we're going from a real state to another real state, we scatter to a real state, we come back to a real state, and then we recombine to a real state. So everything is real states, and that's the reason for this huge intensity. In the Raman effect, you often make the transition to a virtual state, and therefore the probabilities are very much reduced. So that makes this line shape a very unusual in physics. We don't have, I don't know of any system that's like this. This is a unique system. So that was the first-order effect. So that's one-layer graphene. But now if you have bi-layer graphene, and we have a lot of papers and bi-layer graphene, so we should say something about that. So the photon can come in and make the excitation I've described before for the G-prime process from the lowest energy band, that's this one here. And so that we go here, back and forth, and I call that P11. And the total energy of the two photons, phonons that are created will be up here. So this is P11. So that's one point for some energy of excitation. Since everything is resonant, we have a choice of lots of points to do here so we can populate this whole line. But instead of doing the first level, I could do the second level. So here's it starting in the second level, going to the second level, scattering to the second level, everything back and forth second level. That's P22, and that's over here. But I don't have to do that. I can go to the... Here I can make a transition here to the first level, first level to the second level, and then come back and... This one is from the first level to the first level. Then I go to the second level here and come back on the second level and find myself a hole to go to. And so that makes the cross-term P12, etc. So that's where all this comes about. And since the phonon has a different wave vector, it'll have a different energy, and you could separate it by just very few small energies, just a couple of millivolts. But you can certainly resolve that. If you're looking at a whole bundle of... a whole bunch of bilayer, a spectrum that looks like this. So the individual peaks are not resolved. The line width is greater, so they overlap and you can overlap energy. But it has a certain line shape, and that has been one way that people have quickly identified this is bilayer and not monolayer. Monolayer has this huge single peak, and this one has a much broader peak. Now I'm going to show you something new, a new feature, so we have yet another thing that we could look at. And then I'll go on to talk about other subjects. So this is work done in collaboration with a group of young people in Singapore. And so they had a sample, and this was kind of nice. Maybe it has some deficiencies because it has a lot of different regimes here, but they have a region that's one layer and they have another region that's bilayer and three, four, et cetera. And it's a pretty big sample with some microns, so the optical beam can be housed in one of these regions. So they did a scan to find out what's in the sample, and then they came back. Doing the kind of thing that you've heard here several times of seeing what the absorption is at different spots. You can sort of monitor where you are. And what we do in these kinds of samples is we put a fiducial mark on the substrate so we can always come back to exactly the same place on the sample and repeat, and do the spectrum repeatedly on the same point. So that's the way the experiment is done. And so this is usual. You know this. So this is the G-band. It's discussed many times, and that's G-prime. I discussed that as well. In single layer, it's very, very big. And then we have another region in here that you don't see a whole lot. But if you are very persistent and you look in here, you can identify Victor Brar's M-bands. I think Michael Crummy will be happy to see his students' M-bands. And then I'll show you all the stuff we missed because we didn't maybe spend enough time looking for all these other little features. So this is the place that we're going to look now. And so this is the region here between roughly 1750 wave numbers and 2100, 2200, where there are other features. If you look carefully and spend a lot of time looking, you will see more things. So the first part I'm going to show you is what happens between 1700 and 1800. And this is the region of the spectrum that we had studied before with Victor Brar. And at that time we identified two regions, M minus and M plus. And when you look at this, we saw this looking like a letter M. Can you see something that looks a little bit like... We thought it looked like letter M, so we called that the M band. We didn't have M band and graphene before, so that's what we called it. But then when you look at it a little bit more carefully, you notice that this has a non-Lorensian line shape. And this one is mostly Lorensian, but this one has a weird line shape. And if you look at all the places in the sample from the thickness measurements, et cetera, there is no signal whatsoever for the monolayer. It's absent. And you begin to get something in the bilayer, and then you get some more. And if you sort of look at the features, this is maybe the most interesting, so you have maybe three features. And if you think about this as three Lorensians and you apply that to the whole series, then you get something that I'm going to show you in the next view graph. So I'm going to show you a little bit of results for what happens in this region where the M band occurs, and that one we knew something about. But then I'm going to show you something about this region as far as I know nobody has studied before. And maybe we should go back and look at all different kinds of regions because maybe there are more things like this that we've missed because we haven't looked carefully enough. So I guess last night we had a philosophical discussion about science. Some of you were present. And somebody was criticizing the science community that we don't do careful work. And I think that in defense of the science community, I think that the first thing that you do when you do science is you do what's apparent and what's easy and accessible and you have a chance of understanding it. And then maybe something gets more interesting and you come back and see if there's maybe more to it. And I think that the art of the fifth dependence of the low-temperature reasons to have any is maybe something in that department that we wanted to get it right and it was wrong for many years. And I think there are a number of other things that came up at this conference that are also in this category that people call boring, but they've been wrong and not quite right in the literature for a long time. And now with graphene, we can straighten them out. So I think that we should say that that's a positive thing that we can do. We'd like to have everything right. And sometimes getting everything right is difficult, but if now we have material that is enabled, maybe we should go ahead and do it. Okay, end of philosophical comments from last night. Well, I think it's the last speaker. And I was told I was supposed to summarize what we learned at this conference. And that was one thing I learned from the conference is that the people in the UK like to have philosophical conversations. No. So now I am here talking about the M-band and I'm going to show you in a few view graphs how we did a little more careful study of the M-bands. So one layered has nothing, so it doesn't appear here. It's absent. I'm sorry to explain that mine is no M-band in monolayer graphene. But in bilayer graphene, because looking at all the different number of layers, you have the idea that the lowest term here, this one here, should be a doublet. So you identify it as a doublet that works rather well and then you could follow all the components. And so then we have the second part of the doublet and then we have this piece here. And usually when we do these kinds of things, we like to identify which phonons are involved. And the frequency for two out of these three work rather well with the M-band, which is identified with this phonon. So you see that there are three acoustic phonons and three optical phonons and there are two atoms per unit cell, so that all makes sense, right? That's elementary, boring physics, but we have to keep track of those things because we might get right. And the out of plane very soft mode that comes in the C direction we know almost nothing about. So even though it's boring, it's kind of interesting. And so we had some new insights about that feature from Andre Ferrari, but he's not here to give that talk and I don't have his data. So that's coming up and we'll read the physical review or nature or ACS nanoledders. I don't know where he's going to publish that. That is the analog of what happens with the GBAM. So anyway, the identification of this is twice this mode. So this mode comes roughly at 900 but a little bit less like 970, 870. That's a known mode. It's IR active, so we know that. We don't know from Raman spectroscopy because it's Raman forbidden, but it can show up if you have some kind of disorder or edge states or something. And it's a very weak feature. I was saying you have to really work at it to bring it out. So that's the MBAM and these two work with a combination of twice the MBAM. One for intra-valley scattering and the other one for inter-valley scattering. So intra-valley scattering, instead of going from K to K prime, you just stay on the same K or the same K prime but you don't go between the valleys and it takes less energy to do that. So they're a little bit shifted from each other and that all works out in the energetics of it. And then we have another mode here and so we made an identification of what that could be. So the two TO modes that we have. But then we have a combination mode that we're going to identify with this lower frequency mode. So what's clicking for you is the OTO and that's the infrared active mode that's made a rum on active by having the small feature and probably some defects. So we have a symmetric and an anti-symmetric combination of the two vibrations. The vibrations are the same with respect to each other within the plane but across the plane they have opposite phase and that's shown on the top here and that opposite phase introduces larger Q vectors and that's why we have small frequency shift. So that's the explanation of that piece. Let's see what we have. Okay, so the other, the three bits that we have here are the TO modes. The two of them and then we have another piece that I'm going to bring in a little while and that's this one here that's a lowest one, LO plus ZA and this is the frequency, the laser energy where all these pictures were taken so that will give you the identification of what it is, LO plus ZA. And now moving along here, let's see. Okay, now I'm going to go to the next little bit which goes from 1850 or so to 2100. And now we can have something happening for the monolayer. You see there's some peaks and so what are they all about? And you can follow those peaks. They go not only for HOPG you have them but you have them very nicely for monolayer. It's very much more helpful than HOPG and then you get what happens here as you go up the trail. And so now we have to do some identification of what these things might be. You see that they're all dispersive and then but there's some of them are a little bit asymmetric and you have to find some kind of identification what they are. Some of the possibilities that we came up with so we have three different possibilities here. And I hope we have some. Yes, we have some names for them. And so here's combination mode that works in frequency very nicely for this. So the ITA is this one here that's the ITA. This is out of plane mode and LA and TA and then we make combinations with these guys. LO and TO. ITO is the in-plane obstacle that's 1580. That's the one you know very well. G-band is what we call that. And so here are three and then we have three from the from 1700 to 1800 and I believe I have all the identifications. There are four identifications for the higher frequencies and three for the lower. So there are seven all together. And there's another group in Clemson University that's done something very much like that. And they also have the same identification independently and for the same group of features. So if you sum up the story, there are a bunch of wiggles here and they vary as the sample changes go from one layer to seven layers. There's some small difference that hasn't been worked out in detail yet but you see that what we're headed here. And there are some that show say this feature here is the band feature and there are others that require that we have to go to two layer, bi layer, single layer doesn't have it. And if the two layers are in commensurate, you don't have it either. So you have to have AB stacking and more than one layer. So that's what we've learned. Those are new results. And I don't know how interesting you are interested you are in this but maybe as a result of our discussion last night plus other things happening in the field that people will clean up all these little features. There are a whole bunch of features. There are combination modes of this sort. When we were studying fullerines we had 200 combination modes to put together. And now you have a C60 molecule you have a huge amount of symmetry some were infrared active, some were Raman active. That didn't have much broken symmetry because the fullerines are really very robust molecules. But here we can have vacancies as we found out and other kind of distractions. So this is to show you how the M band is a new thing and we have to do more theory with that. Okay, a little bit now going back to nanotubes. Nanotubes are different from graphene and we have been working on nanotubes maybe 10 more years than graphene and the field is still active. How come? Maybe that field should have been dead by now because usually 10 years is enough to exhaust any field. But the problem with nanotubes and why it is still active and maybe will stay active for I don't know of 10 years but we have so many experts now from graphene that when they get tired of working on graphene they may try to mop up the unfinished work and nanotubes is very easy for them because the physics is very similar. But there are many things in nanotubes that are still not understood and why is that and I'm going to try to say that because I think it might be interesting to this audience they do something and they want to see how it is in nanotubes. So a nanotube is a rolled up sheet of graphene and we have a single, we can have a single sheet and it was kind of funny that this whole thing was fantasized well before it was ever produced in the laboratory and maybe it would not have been produced in the laboratory anywhere as quickly if it hadn't been fantasized and the reason was that for various historical reasons I won't go into here we were into the idea of predicting what the nanotube would be like if we could ever make it as just the Gadokan experiment and the prediction was by three groups and they came out at the same time and the papers were published in different journals around the world different countries at the same time predicting that if we ever could make a single wall nanotube it would be either semi-conducting or metallic so that was 1992 and the first single wall nanotubes were made in 1993 about half a year, nine months after the first papers predicting them how interesting it would be to make them if we could so what's the reason for it what's the physics behind that so I'm going to go into that and try to explain that so if you take a graphene sheet and you roll it up then you have a very few atoms around the sheet maybe ten so that means you'll have ten cutting lines and so you have the graphene cones like this but since you don't have all the states available you only have ten states to manufacture from this and you have to space these cutting lines appropriately to satisfy quantum mechanics as you put them in and the geometry that goes into requiring how you put them in a cutting line might cut through this a direct point or it might not so it turns out that one third of the cases will go through the direct point and that will make a metallic nanotube and the other two third of the possible cases will not go through so we'll have a semiconductor nanotube and so that's just from the geometry of the triangular lattice so that was predicted and people didn't believe it I must say that and it really took until 1998 six years later after the first papers were written for people to say yeah we could have metallic and semiconductor nanotubes most of the people in this audience don't know that history because when you join this field probably it was all settled and everybody as soon as the experiment was done showing that actually in the laboratory we could make them and had spectra then it was all over and so that's what happened starting in 1993 okay anyway for the nanotubes we also have this radial breathing mode and so that's where all the atoms in the tube are vibrating in a radial direction and the frequency depends on the diameter so that becomes identification of the nanotube diameter and it's been used for that ever since and also we can a little bit more indirectly identify what the NM is by measuring that is the chirality both the chiral angle and the diameter of the nanotube just from this frequency and the theory behind it then we have the longitudinal vibration associated with the G band it has a transverse component we might mention that already and then we have the second order feature G prime that I already mentioned and then we have this M band that's new and we're just learning about it and then we have this disorder induced band here that shouldn't happen but if you break symmetry so we have a lot of symmetry breaking discussion at this conference and that's one of the features and one of the early features that we used in carbon science to study disorder, breakdown of symmetry we're getting more sophisticated and I think in the last year one of the things that I'm seeing is that we're not just satisfying that saying that we have a defect but we want to know if it's a vacancy if it's a die vacancy how the die vacancy is ordered if we have an edge and edge is a discontinuity in the infinite lattice so that could be treated as a defect and it produces a discontinuity and we can pick that up so the radial breathing mode is here and you see it's at different frequencies and that indicates that we have a different nanotube that's in resonance with a different diameter so therefore a different frequency and then we can get the NM values so if there are many ways now there's Rayleigh scattering and other ways to determine NM and I think the interesting thing that I've been picking up right this year and it's pretty new is many labs are now doing multiple experiments to correlate information either on structure on electronic properties optical properties, magnetic properties, whatever and it used to be that when you publish the paper you would have one technique but now some of the really interesting work that I heard in the Bordeaux conference before coming here was of this dual different kind of techniques so I throw this out to you because not many people from this conference were also there so I summarize a few things at this conference that I learned from them there so the earliest work like here was done on bundles of tubes and so with bundles you just get an ensemble information that's the earliest work that we did this very first paper the first one on first Raman in nanotubes was 1997 and that was done on bundles but it became obvious right after this experiment we did another experiment to show that the spectrum looked very different from metallic than semiconductor tubes and that meant that we really had to do this at the single nanotube level to be able to in detail understand what went with what so this picture on the right is spectroscopy taken only on one tube and so we have one semiconductor tube and that's this one and one metallic tube that's over here so this is the first paper on single wall spectroscopy in the nanotube department but now people doing it with many other kind of techniques it's not only Raman it's many other things spectroscopy good for so this picture has a lot of people's work and it shows it allows you to distinguish between different kinds of carbon so this is just maybe a little not everybody realizes that diamond and graphite have very different spectra so there they are frequencies are totally different but they're all carbon so SP square bonds and SPQ bonds they have different separations and they have different coupling one is a planar bonds and the other one is tetrahedral bonds so it's obvious that they'll be very different and so you see different spectra but not only that when you have different kinds of SP squared so this is graphite and this is single wall nanotube amorphous carbon and everything looks very different and that has to do with defects and many other issues so we can use Raman spectroscopy to take other allotropic forms C12 and C13 that came up at this conference you can tell them very very well by Raman spectroscopy because vibrational frequency of almost everything will depend on the mass of the atom and C12 has a very different mass and it's almost 10% different from C13 so it'll be way separated and you can distinguish so some of the experiments people were talking about here would be relatively easy to do and there's one paper that we recently published but I don't have the U.S. for it where one layer is C12 that's on the substrate and then we have another one that's C13 that's actually the other way around C13 is on the substrate and the C12 is the positive on top and we can study substrate effects directly because one of them has a different they show up differently in the spectrum but the interaction is almost the same those kinds of experiments I throw out to you as being very relatively easy to do nowadays it's easy to get isotopes and they give you a lot of information so that's an addenda to this and some of the modes are dispersive and that's what I have here to show you that you can do studies here as a function of issues like size of a domain and that changes the amount of D band that you have or you could just do experiments to see what the dispersion is to study what the wave vectors are because with Rotman effect you could measure phonons and the phonons will be different in the second order process depending on what the photon energy is so that's one thing that you can do easily and this one I threw in here to show you that you could study issues like doping so if you have an undoped sample you get one kind of spectrum a P-dope sample and so the top one here is undoped and then you have endoped and P-doped so the endoped pushes the frequencies down to lower frequencies and the end a P-dope pushes the frequencies up to higher frequency same thing happens in intercalation compounds so this is sort of general throughout the carbon literature so if you know that you can do a lot of things you could study pressure dependence and you could see here for different amounts of applied pressure you get shifts and so this is part of so this is uniform squeezing we had stretching mostly in the talks you could do stretching here too and that gives you something similar effects that's another thing these frequencies you change the separation by a little bit of the carbon-carbon distance not very much as necessary and you get big changes that you could see in the spectrum so temperature dependent phenomenon could be measured you could measure equilibrium or lack of equilibrium between electrons and holes because you could probe the phonons and then probe the electrons at the same time just that are chosen and the phonons come in by the phonon frequency so on the very same shot you can study electron-phonon interaction in this kind of way and then we have edges edges represent a breakdown of symmetry so group theory symmetry tells you that if you're an internal point here there should be no reason to have sort of infinite lattice so you have no D-band but if you come over here close to an edge this is an armchair edge which we found out by doing Raman then you get a big D-band signal but symmetry tells you that the zigzag edge has a vanishing matrix element so if the sample had been a really good sample with a perfect edge with no defects then this signal here should have vanished it's pretty small so that's the best at that time this is pretty old work it's seven years old and we make better edges now but that's just to show you an idea about edges how you could study them so this is doping and defect so if you add more a quantity of a dopant you broaden the lines so that's what's shown here and then this is well, donor and acceptors this is mostly what I put together here to show you that doping and defect can be studied and this is now size of samples domain size grain boundaries we saw a wonderful picture from Cornell I think that was showing an individual grain boundary we saw five, seven defects along the grain boundary the boundary between two grains I think was it you Philip that showed that we saw that today somebody own up yes, yes, yes Dr. Hong from Korea showed that but I believe that picture was taken at Cornell was that right? yeah that's the one I know beautiful picture so that shows us that we could understand what happens at the grain boundaries and now we know exactly just a whole series of stone whale defects along the grain boundary it's really a very beautiful maybe we'll bring that again if anybody's interested in seeing that I think we can learn a lot from that and this is something new it's submitted but not published electronic spin resonance not spin resonance electronic Raman scattering so what is this effect? this effect has been seen in other systems but this is the first time it's been seen in the carbon system so a photon comes in and makes an electron whole pair electron whole pair can be generated for any different kind of wave vector so that's the reason that you have this very broad band it can be seen if you have a rather pure sample so that the Fermi level goes here because you have to go from an occupied state to an unoccupied state to excite the system so if you dope the system you'll kill this transition because this is not that many MEVs so you can quench it with moving the Fermi level this is observed from metallic tubes not for semiconductor tubes so it's unique to that that's the prediction and that's the observation that's a new topic double wall tubes I'm going to say something about that this is sort of to me a little bit analogous to bilayer graphene so I show a little bit about that this will be very brief whereas single wall nanotubes there's a uniqueness if you have the 6.5 tube and you show the 6.5 tube the Raman spectra and you do it in one lab and you do it in another lab you get pretty much the same spectrum but if you do a double wall tube that's not the case and the reason for that is that if you take the two tubes you can move them with respect to each other so the hexagons, relative positions of the hexagons change and then you can change the chirality this way so you have two ways to alter this spectrum so you get a range of values for any one tube so be prepared for that so at the conference in Bordeaux they had Fen Wang and he and his group were working hard to get kind of a standard to tower up plot for single wall nanotubes and once that's established so if you have good data to send to him he's putting everybody's data together and taking into account environmental effects and everything else so that we get some medriology going and he's the one that's really championed that program so many of you know Fen Wang, he's a very able young professor at UC Berkeley you might want to cooperate with him because that would be very interesting. Now with double wall nanotubes we have four different flavors that you could have and the Kataura plot was going to be different from all of them so the situation gets a little bit more complicated but Raman spectroscopy could be very useful for characterizing all of these different tubes the bottom line is that the double wall tubes do not quite behave if you take a 6-5 tube and it's a single wall tube and you put it inside a double wall tube its spectra will change but not that much it doesn't change so much that we lose identification you can mostly identify it but be aware that it may change even its identity in that part of the space where you have many tubes a high density of tubes that's large diameter tubes so here's the double wall tubes here so this is a bundle and you see with the bundle you have a mass you have tubes that you can identify exactly so you never know which outside tube went with which inside tube so really spectroscopy is not possible but if you do it at the single double wall tube level you can identify the various features so I'm just showing you this to encourage you that if you want to do spectroscopy you have to work at the single nanotube level so that's enough said on that the spectrum for graphene is similar but not identical so this is a topic that has to be researched and it's I would say this a little bit of work but not a lot of work so this is still something for a young person to take on with the double wall tubes you can be in a situation that the inner and outer tubes are resonant at different laser energies like here and working at the individual nanotube level is essential to separate all of that that stuff so a few labs do that now and sometimes you're lucky and with the same laser energy you can see both the inner and outer tube that's unusual but I show you sometimes you can be lucky now people are doing triple wall tubes then this has gotten interesting because people are interested in trilayer graphene so the correspondence and you can see I'll show you some spectrum where we were lucky and we managed well this is the triple wall tube you can see here this is where the tube grows and that's the two and three so here's it sort of looks a little bit like a mess but there's a radial breathing mode on the right and this is the G band a lot of activity but I think triple wall tubes we'll be able to separate out and see what the shifts are but I'm not interested in doing more than three I think I'm going to stop at three I think three is about my limit just a couple of comments about ribbons ribbons ribbons are something and edges are something that are unique to well I don't know if they're unique to graphene but they're part of graphene and so there are the edges and whereas nanotubes they can have any chirality and any angle works with almost equal probability for the I'm going to show you that the edges in graphene are very special they're not like nanotubes at all they have very great preferences I was surprised by this myself but I'll show you so this one here the edge is zigzag and the edge here is armchair and we label the ribbons in that way by the long edge you could say in a sense it would have been a little bit more symmetric if we had labeled them by the short edge but that is the way we did it historically I won't take the time to explain why that is but that gives you more symmetry between nanotubes and ribbons but historically we didn't do it that way and I think we better not change and I'll create too much confusion we have enough confusion with g prime I'm not looking forward to any more confusion so here's armchair you see this is now no spin orbit interaction this is our 1996 paper very ancient but you have a ribbon with four rows of carbon atoms so we call that n equals four everybody uses that notation you can see that's a semiconductor and n equals five is linear e versus k ribbon and here now we have a semiconductor again the zigzag is different n four five and six and anything has a high density of states at the Fermi level and that makes that one quite reactive so the chemistry is really quite different and when we do experiments on these edges and folded edges and so forth they really look different in the microscope so we've gotten used to that and the reason for it is there's difference in density of states which really pops out at you when you do the experiments so Inoki a long time ago started making ribbons and even before we knew how interesting they were and so here's an ancient ribbon 2004 it's off the page but it's 2004 and so you could see this is 80 nanometers and you could see that's very small compared to 80 nanometers so you were in making quantum ribbons and this is the heights you could see that this piece here is one layer tall so this is a mono layer ribbon and this is just to show you that Raman spectroscopy satisfies the selection rules as it might and so we have a ribbon sitting on graphite so the upper frequency is graphite and that's the substrate the ribbon is this one, the lower frequency why is it lower? you put a laser on it and it's a very tiny ribbon and even if you put a small amount of energy it gets hot because the C-axis thermal conductivity is very poor so it has very poor contact so laser energy heats it and so that's what produces the downshift so we can tell them apart which is nice and then we do polarization experiment that was predicted in 2003 I have a cosine square dependence and experiment gives cosine square dependence so we were happy with that and so this tells you we can understand a little bit about ribbons, identify them with Raman and get some information so other people have shown this U-graph also this is a 2004 paper that distinguishes the D-band itself the D-band has an elastic process in addition to the inelastic process it makes a phonon but it also has elastic process so 2D would have two elastic processes but people call the 2D usually has two elastic processes it gets a little bit confusing it makes a small difference in the frequency if you want to be technical about it ok so this is now how to clean up the edges and then I'll end so we have a TEM microscope and we put our little piece of graphene ribbon here and we put electrodes across so we can make IV characteristics and we can also do jewel heating I just want to show you so the ribbon that this is a ribbon that was made in Mexico and Mauricio Turronis' group and it's over here is the ribbon they brought it to one of the meetings and showed me a picture of it and I said come to my queue let's look at it I think it would be interesting so we did and so we did some jewel heating this is the sample in the beginning there's no there's no clear demarcation of where the edges are here at all and then a little bit of jewel heating starts out linear I versus V so that's no temperature no heating and then it gets non-linear and then at this point the whole thing takes off and gets extremely non-linear and that's when you start having motion of ions particles platinum to measure what the temperature is and I'll show you this in more amplification so you can see what was happening and so you see that this is the ribbon that we start with and after some jewel heating this is this is SEM picture so you can see low magnification and here are a bunch of edges with about 20 layers of arrays that interests me and the more you heat them the more these arrays line up with a constant distance between them and interestingly the edges are almost entirely either zigzag or armchair edges and we don't have anything in between so that is very different from the situation of nanotubes where all the NMs occur so it would be nice to have more data on that because it is a black and you see very few chiral ones and the probability of having the armchair and the zigzag is small because it's a very special orientation but almost everything is armchair and zigzag and that sort of thing doesn't happen with nanotubes we have 6, 5 and 9, 3 and everything else but with graphene graphene likes to have asymmetry directions and theory says that in equilibrium state the armchairs should be more frequent but when you do make them with dual heating the zigzag seem to be equally or more and you can actually watch the motion of these junctions between the different kinds of edges and we've been moving them and measuring the velocity motion and so on this is going back to yesterday when we ended our discussions what's next we've had graphite we have intercalation compounds intercalation compounds have infiltrated graphene people put stuff between the layers and is very much related to the work early work of intercalation compounds and it's very interesting what happens with graphene because you study the effect for bilayer even single layer that's sort of putting surface atoms on and that field I think can become interesting after not being too active for many years and well we've had fullerines now fullerines are important in the energy business and we have nanotubes still going and graphene is taken off and what are we having next linear chains have been looked at and that hasn't turned out to be a whole lot but maybe somebody has ideas what's coming next maybe we'll go with moulins little new molecules that might be the next direction I don't know so let me leave it at that and discussion how would you distinguish it for example from fluorescence from hot luminescence or jaminite recombination how do you know that it is Raman or why do it's maybe terminological question but why is it Raman not fluorescence if it all real states that is a good question and it bothered us too time resolution is one thing how you excite it that's we have all these different processes with it initial final state and lifetimes when we first saw this for the monolayer and bilayer and it's always resonant that's troublesome so you think that probably is fluorescence but time resolved and intensity measurements are the way we usually do these kinds of things and that's what we did you had a picture of these various interfaces zigzag and so on yeah I mean do you have them that's right I was trying to light them up yeah that was nice because you seem to have quite a high density of zig tag but that's because we have 20 layers and what happens is that they break up a one layer in the next in this ordered way we didn't expect that but it always does that maybe you have a theory for it no not at all but I mean there is this theory which has been mentioned that the zigzag edge is ferromagnetic yeah but that wouldn't show up here you have to do magnetic measurements I don't know but I'm suggesting that if you could get a lot of edges not just on the edge of ribbon who's going to be in some more more bulk ferromagnetic carbon the idea is very good the next step of your suggestion I'm just taking your suggestion one step further as you can make those edges like I showed here as part of a narrow ribbon and the number of rows determines whether the relation of ferromagnetic and anti ferromagnetic on the relative edges and you might have very interesting magnetic array that's the next step to your idea but you have to kind of arrange that this sample isn't exactly right you have to cut the sample to make the right shape so that we get exactly what you're talking about but that's a great idea I mean there have been mentions in obscure places in the literature of carbon samples of various types of carbon where they claim they see ferromagnetism with very high curie temperatures now I mean I don't know how they go up to 100 degrees well I've seen there are a few are odd they may not be right but there are a few in the literature where they've got much higher ones not in graphene in various weird samples but this would be very exciting if you could have something TC well above room temperature and something like I think carbon this is magnetic this is a TC for superconductivity high TC it is it's a good idea nobody had but people have studied the magnetic features of the edges the person that's done the most work on that is Inoki in Japan you want to follow that in the literature what determines the Raman shift of the peak in the electronic Raman scattering of your metallic nanotubes oh well you could see that we have different Q-vectors so that's different energy required to go from the initial to the final state so that's different wave vector and different energies so that gives a very very broad line because that contribution is seen over many millivolts it's not a very strong feature we really have to work for a long time to get that spectrum and that's why it's been missed all these years because it's been seen I'm sure by many people they weren't looking for it the C2 nitrile and there was another molecule to dope the nanotubes and look you showed the Raman spectra for that I wondered whether there was any signature of the dopant molecules themselves which come out or enhanced by the nanotube very good question you have to work at that when we were doing intercalation compounds the one compound that we used to see very good spectra for the intercalin that we put it on was bromine I don't know why that one worked so much better but there were other molecules like ferrocluoride that we had a huge ton of data and we never got the vibrations of that molecule and I don't have a theory theoretical explanation why bromine is so wonderful but bromine doesn't do stage one and I don't really understand that either it only starts working on stage two and it's the only one only intercalin that I know that has that characteristic so I just wondered whether it might be the shift is amazing huge the vibrational frequency goes from something like 300 to 200 wave numbers that's almost unheard of shift in Raman spectroscopy there are some molecules where the absorbed molecule Raman signature is very strongly enhanced by graphene oh yeah that's another thing so is there any connection between the nanotube work and the graphene work yes, yes graphene is it's called GERS that's what you're referring to there's a phenomenon called surface-enhanced Raman scattering which will enhance the signal of a molecule by many orders of magnitude ten orders of magnitude is a huge effect with GERS this is a graphene substrate the enhancement is less than one order of magnitude but it's very large so that is to say let me repeat because maybe not everybody knows about this phenomenon this is something pretty new if you take a graphene surface and you put say a molecule sitting on the graphene surface the spectrum Raman spectrum of the molecule will go way up in intensity but it doesn't do it by many orders of magnitude it does it by maybe a factor of a few but that's a lot so I mean it's very, very noticeable and the work on that the best most complete work is done in Beijing but anybody can do it it's a very direct experiment