 I don't know if you can hear me, though. Same song, but I don't know if you can hear me. OK, it's perfect, so it's good. You should be muted for a while. I should be muted, yeah, I am, so it's fine. OK, so OK, it's fine. So OK, it's nice to see all these friends again after these two years, so it's been a long time. We tried, so that we think we can thank Erie, Andrea, and Erie for the endurance to endure and keep these two years and trying and not bail out in and then stop the conference. So it's nice to see all of you together and be here together, so. We're going to try to present what we are currently doing on this field of fluids and solid interaction and friction. And most particularly what we are really doing lately on interaction between fluid and electrons, fluid molecules in the fluids and electrons in the solid, in particular in carbon and the material that are easy to understand and probably also easier to measure. And we're going to try to describe what this interaction can imply in terms of transfer of fluids in this material. So basically on the effect and the impact of this fluid-electron interaction on the fluid-solid friction. And on the other way, it's also how we can in a certain way use fluid-solid interaction to create electronic current in solid materials. We're going to be basically dividing in two parts. We're going to follow this idea of seeing how this fluid-electron interaction can play this dual role. And in the first part, it's going to be very fast. Just basically recall what we know about fluid transfer in carbon nanomaterials, a particular fluid-solid interaction in carbon nanomaterials, and how we can in a certain way revisit these results in view of this fluid-electron interaction. So basically everything started in 2009, when measurement on membranes made of billions of carbon nanotubes show a very large water permeability. So with water flowing at a speed that was order of mind to larger than what expected or what predicted by standard of dynamics. Since then, it resulted in being performed on any kind of carbon materials, being graphene, graphite even, or more recently, graphene oxide. So different material, all based on carbon, and all of them have this common behavior that presents a very, very large permeability. So basically a very large flow and very large speed of water that is in a certain way can look like a consequence of a very low friction for water and for liquids at the interface with the solid. So if we keep in mind this standard way of understanding permeability enhancement or permeability in nanofluid devices, the best way to try to rationalize these results is basically introduce no-slip boundary condition for flow. And this is mathematically just the idea of introducing this B parameter in the Navier-Stokes equation and in the boundary condition for the Navier-Stokes equation where we know that B is a slip blend that is basically the inverse of the friction coefficient for a depth fluid solid interface. So if you introduce this parameter, you immediately have an enhancement of permeability that is depending on this B parameter divided by the radius or the confiner. So it is clearly here, short or ready, that going small in size is helping because it's going to be more and more for the same slip blends or for the same kind of friction. You will have a more measurable effect if you go to the slow dimensions. So if we keep now use this, and we try to be a bit more quantitative, for example, for the case of 2. Then we try to be a bit more quantitative, for example, for the case of carbon nanogyms. So we can basically extract the permeability, compare this to the permeability or the flow for the water compared to the standard aerodynamic. And then we get from this enhancement factor using the format that we showed before, we can extract a sort of a slip blend. And in the case of nanotube, the first result were basically all in the order of hundreds of nanometer up to microns. So very large value compared to, for example, the radius. And even more important, this value were always one up to order of magnitude larger than the best case that we could have served experimentally on other standard material, like, for example, code like hydrophilic or hydrophobic material. So it looks like for real carbon is something special. And that was the starting point. So the first part that we did, and it was very fast, was to try to produce this result in more control and configuration in order to be able to compare to a theoretical modeling that was, in a certain way, more easy to apply. So what we did was basically try to do the experiment on individual nanotubes, so this case carbon nanotube or boronite, for example, for the moment on the carbon. So we introduced a carbon nanotube at the end of a glass capillary and we used this technique with this device to measure the flow induced by a pressure applied to one extremity of the nanotube. And basically we were able to measure the velocity of a box of colloid that were at the other extremity of the nanotubes. And from this, with this technique, we showed that we had enough sensitivity to detect the flow coming from an individual nanotubes. So experiment had been performed to be quantitative by changing the parameter, for example, the pressure. So we were able to basically perform this experiment for pressure increasing from 0.5 bar up to 1.7 bar. And again, we collected the velocity of the colloid. Basically we analyzed the colloid trajectories and from that we were able to extract the velocity of these colloid colloids and finally the flow streamlines. And from this streamlines we were finally able to obtain directly the nanotube permeability. So from this we were able then to compare this, we performed this experiment and compared to the reference, it was a standard thermodynamic. We performed this for different nanotubes, this year in green carbon nanotube and blue the boron nitrate. And we did this for increasing, for changing of various ranges. And we observed all the time this behavior with an increasing enhancement for carbon nanotube while you go down to small ranges. And from this we were able then to extract the slip length and we observed that in the case of boron nitrate basically there was no slip. And we're clear in the case of carbon nanotube, the slip length was very important depending on the, of course on the ranges and going up to several hundreds of nanometers for hours smaller nanotubes, it was 50 nanometers. And then we tried to compare this with the standard theory basically based on the molecular dynamic simulation based on the standard, I mean commensurability effect and how water can interact. And this was basically the theory based on the result of simulation that the first style looks quite in agreement at least in the terms and values. We obtained for small nanotubes the large slip length in order of few hundreds of nanometers and that look at the first time quite good. But actually if you look more into details and a bit more carefully, you realize that the theory is completely wrong with the experiment and especially in the X axis where you look at the radius. So basically for the theory if you just consider commensurability effect you suspect to have this enhancement of the slip length just for radius that are in the order of few nanometers. While we observed this increase in this announcement already at 20, 30, even 40 nanometers. So this was clearly a gap between what we observed and what the theory seems to suggest. And then of course what was also important on the standard that first thing is that maybe our experiment are wrong in a certain way, so it's not extreme but then we compare to what was observed also in other group and we realize that basically everything is more or less the same experiment result that you have in literature is always showing very small slip length for graphene or very large nanotubes but you'll start to observe immediately an increasing of this permeate and increasing of this slip length for relatively large nanometers within the order of 30 nanometer radius and that is goes very high even when you're going. So we had to go a bit farther with the theory and that was what had been performed by Nikita Kavakin last in the next year in our group and the idea was to look to other channel of the dissipation and basically start to look with a sense of non-classical fluid solid friction and in particular start to understand if there would be possible to have an interplay between the fluid transfer and the electronic property of the materials. We observed here for example before there was a very big difference also between carbon nanotubes and boron nitrate that are basically it's the same apart from the fact that they are one insulated and another metallic or semi-metallic. So the idea was is it possible to introduce or to take into account explicitly the role of the electronic property of the materials and then that was Nikita developed and I suggest to look at this paper for the web more details on the very nice and the big effort he has done. So basically what to introduce this taking the specific account is electronic effect in the fluid solid interaction that is basically this friction force that are so here but what is important that if you look here basically just in the we can look very easily what this form says is that this electron fluid electron interaction so basically interaction between the fluid and the electron as opposed to be important only when the collective excitation that are basically given by this imaginative part of response function for the electrons and for those of the fluid meet or are in the same region. So if you are in a configuration where these two collective excitation can meet you expect to have a very strong fluid solid interaction and finally a large fluid solid friction but if you are not you expect to have basically low friction. So if you look first is that in general we don't expect to meet because as always the same electrons are high energy and molecules and fluids are low energy. This is a standard. So basically it's comparing KVT that is few tens of milli electron volt with the electronic excitation plus basically plasmons and so on that are more in the electron volt but there are actually some material this is not like this there are some cases where you have the diswoct excitation can meet and this is particularly the case of carbon materials. So if you look for example on this on this center graph you look at the here we compare basically the the collectivity for electrons or basically the electric response for graphene and for graphite. So graphene is basically characterized by a response with only a few modes that are the right energy to meet and to interact with the fluid molecules in the water. But on the other hand graphite is more characterized that is the dot in the center graph is more characterized by sort of dispersionless response where there are a lot of modes exactly at the right energy. So they can meet interact quite efficiently and effectively with the waters. So then if you put this just for this point of view you can expect that in the case of graphene you have a larger as a slower lower friction than in the case of graphite. So then you can then let's say generalize this result and goes to the case of more like for example carbon nanotubes that was interesting then consider what is the response of carbon nanotubes from an electronic point of view and also from let's say optical point of view. And then you observe that when you go down at the small as when you decrease the radius you change the properties of the nanotubes. And then if you put this in this formula you extract this nice curve that you can obtain where you observe that this slip length or the friction dramatically change already a large radius. So this is the three curves are a theoretical model that compares very well with our result. And basically says that if it's more radius commensurability plays a major effect so the standard friction is what is playing a major role. When you go into intermediate you are entering in a region where non-classical friction is the key ingredient. And a part of this what is interesting is that also this offer another new way to tune a bit the friction between a fluid and the side but also open new way or can give us the hobos to do the opposite way. So if we see how electrons can modify the fluid flow because basically this is what happening we can also expect or hope that we can in a certain way use fluid to modify the electronic response. And this is basically the second part is how we can measure or show the phonon drug or basically the movement of the generation of current because used by a phonon drug at the fluids are in the interface. And this experimental evidence presented in a few, I mean already some years ago so basically what was shown that the flow of fluid along carbon nanomaterial were able to generate current or electron potential. So the first result was in 2003 for EGOSH. They observed that the flow of fluid at the exterior of a bundle of nanotubes was observed to generate a voltage drop between the two sides. The origin of this has been for example, I know that you have done a lot of thought about thinking about what was the origin and this logarithmic dependency show, I mean suggest quite immediately sort of a stick slip phenomenon. Later even in 2014 another experiment was been proposed where there was a movement of a drop of salty water on the top of, microscopic salt water on top of graphene and they observe again a generation of a current in phase with the movement. And here there were mechanisms was completely different it was a sort of a change of capacitance because of a change of the shape. And finally two or three years ago it was observed another very nice experiment where a flow of ions induced by a voltage drop inside the nanotube were observed to generate a voltage drop at the exterior of this nanotube. All of this seems to suggest the same thing that we are seeing but so basically how this fluid friction can generate electrons but the mechanism exactly plays still a bit confusing. So what we decided to try to was to do again another sort of control experiment. In our case what we used was again our tuning for base atomic force migas but in this case we didn't use an AFM at all. We just used it as a sort of object to control for a control deposition of a small amount of liquid. In this case we have touched a pipe at the extremity we filled with the liquids, ionic liquids to charge liquids and neutral upholer liquids. And we put this, we put this on top of a graphene sample connected with two electrons. The electrons are connected on top of the substrates of sort of the graphene, sorry, so right on the interface that touches the liquid. So the liquid can be moved and displaced by the tip by applying a movement on the tuning fork and the contact and the interaction is controlled by basically measuring the frequency shift. So as I said, the flow, the drop is put in displacement by the electrons and in the same time we measure the current that is generated or recorded between the two electrons and then this generated by this movement. We performed experiment as I said on two different liquids, ionic charge and a polar neutral silicone oil and on the very large different kind of samples changing of the thickness from one nanometer up to 17 nanometer, but also what's important with two completely different surface states. One very rough with a lot of wrinkles and folds and one basic atomic reflux. And what we show here is that all the time we observe here, we compare it for example in black, the wrinkle and in red, the flat. That if you compare samples with exactly the same, everything is similar, so electronic properties are the same and thickness the same, size the same, everything the same, we just one very wrinkle in other words, the very flat. The difference that the wrinkle you observe current and in the flat you observe always nothing. We go a bit farther and we start to perform experiment in this time as a function of the velocity. So in this case we perform this experiment on four different sample with the high density of wrinkle and all the time we observe that this velocity is linear so that the current is linear with the velocity. So there's a just that the six may not apply in this sense and also what we observed here is that we were able to show that this conductivity was directly related to the corrugation, so the number of wrinkles and the height of wrinkles. So and we performed this experiment again for liquid, ionic and non-neutral results were always the same. So there were basically two options for us to look was one was the Coulomb drug. So the direct generation of electrons movement because of Coulomb economic interaction or the phonon drugs. So basically the idea that one you have a direct Coulomb interaction that generates in the second is more like collision of water molecules and then finally friction of fluid molecules that with the solid. So this the fluid is basically transferring momentum to the solid and then this momentum transfer generates phonons and then because of phonon and phonon interaction you have a second of cascade effect on the generation of electrons. So basically the idea is more precise is that this can be generalized for every kind of liquid but you can apply for example in the case of ionic or neutral and put the numbers there and you immediately rule out the Coulomb interaction. Neutral and non-neutral effect to the neutral liquid and charged liquid would have a spec to have very different behavior and we observe exactly the same. Also we do not expect to have many difference with the presence of wrinkles. So that was helping to rule out the Coulomb drug. So what stays there and looks to investigate more was basically the phonon drug. So the idea here is quite exactly the same. So just to have a sort of cartoon picture is that, okay we have a liquid, this liquid is put in motion and this motion basically the liquid interacted with the solid because just basically of viscous force and because of this friction in the world. So this friction basically is the friction force between the fluid and the solid basically is translating generation of phonons and this phonons has also an energy shift that is given by the fact that, so the population is shifted by the fact that you apply this constant flux of momentum because of the motion of the fluid and then at the end you have a shift of this population that is if we look for example in the case of, if I assume that the movement of the liquid is on the right basically is changing and there is a fringe shift on the right band of the energy band. And then this is basically because of scattering of electron with the crystals, this shift in the phonons and this generation of phonon wind is translated in a generation of electronic movement that is of course that takes the same velocity in a certain way of defective velocity of the electrons. And so if you go with more details and it's been developed by Baptiste Coquino and again Nikita in full details, you can get a rough estimation of the current you'd expect. So here is the current density but it's more or less the same and what you offer to serve is that okay the actual current is basically the number of charge basically is in the charge carrier velocity times the number of charge carriers. So the charge carrier velocity in this case I assume that to be the firm velocity and the number of charge carrier that what matters is basically given by the density of states and the induced frequency shift that what we'll serve is given by this fluid solid interaction generation of phonons and relaxation that goes to the difference shift in this and then change in the population. You can go very going to the details and then try to estimate properly this effective velocity and then finally the induced frequency shift and then you observe that it's basically the effective velocity of the phonon wind is only determined by balancing the momentum fluxes in and out the phonon system. And if you look at this you observe immunity what is interesting here is that this effective velocity is first proportion to the velocity of the drop. So basically the fluid and then since this momentum flux as all this transfer of momentum is mediated by stokes forces and friction of fluid at the solid you observe that rather is the surface rough and higher is the friction force. So and this is this parameter W that we introduced here. So basically assume that more a subset is rough more the friction is important to some highly and easily you can transfer momentum with between the liquid and the solid. So seems to be all in quite a good I mean at least a qualitative agreement between the experiment and the result but the question is would it be possible to test it more in the two details? And if you look that is that basically there was still another parameter that we can place that here we didn't speak right now so far about this but it's basically on this density of states. So the question is is there a way to tune a density of states or a number of these things and the best way to do it is basically introducing an external parameter is basically gating. So add a voltage gate and if you then you have to know exactly what is the bond structure of your surface or your system but basically we know that okay we are using multi-layer graphene so multi-layer graphene can be approximated quite fine with a graphite with two double parabola that are touching because they are not exactly uncoupled in case of graphene and in the sense you observe that you expect to have a direct linear transfer so basically what we are here with the experiment we perform this measurement again with as a function of the gate. So we perform between okay we change from 40 millivolt down to minus 40 millivolt we change and what we observe is that we see that this gate has a direct impact on this electronic current so you decrease the gate, you decrease the current that you measure and when you inverse the sign of this of the gate you switch, you change the sign of the current so basically you're going in phase opposition so this basically means that you are moving from a region of reconfiguration where the electrons are the main carriers to a configuration where that holds the main carriers and here we observe the conductivity as a function of the bias we see there is no discontinuity but basically we go continuously from positive to negative current that this basically means that there is no band gap opening in our system that's quite understandable. So at the end all together this results convinced us that we have been able to observe this full on drug and this electron generation and just to conclude so we are now quite convinced that fluid electron interaction impact the transport in carbon material when you look at the fluid transport but also the V-spacer you have the opposite effect so where you have this interaction can also be translated in generation of electronic current induced by a flow of a liquid and all comes from this strong interaction that is basically the strong fluid side friction at the interface and then now we can also start to look or hope to control the fluid side friction by tuning the matching between the collective response of molecules in the liquid and those of the electrons in the soil. So this is more or less what we would like to do right now but it's gonna be taking us a bit of time and to do so I can finish here. I thank you all for this attention. I thank you for this opportunity and the pleasure to be here all together again and I'm ready for any questions. Thank you. Thank you very much for the questions. Ariel. Thank you very much. So what about the famous logarithmic? You have linear response, right? In our case we have linear response. We expect that in our case was we were basically dominant we were a very high viscous liquid and with quite a rough material. So in our case we were completely dominated by sort of mesoscale friction. So we were not able to enter in what you observed in logarithmic it cannot be anything else than basically stick slip in a certain way or it's very hard to understand that the same thing. We do not because we choose to be in a sort of intermediate regime where viscous flow is the dominant fact and basically you just smoothen everything and you do not see stick slip. Yeah, that's also true. Can I ask about the flow into the carbon nanotubes? What type of tubes were they? You just said nanotubes. We were using carbon nanotubes and boron nitrates. So carbon nanotubes were very slippery and boron nitrate. Yeah, but the nanocarbon nanotubes were multi-wall? Multi-wall, sorry. Yeah, absolutely. Okay, because if you look at the electronic structure if you have single-wall tubes you could actually realize semiconducting and... Yes, and that is exactly what is, for example, and we are currently discussing with this and we have some results with the Ming mind exactly about what happened when you go and you go with the double-walled nanotubes in particular and exactly what it seems that more you go with the normal number of wall and more this effect is even more important because then you can also observe this, the coupling between the different walls and basically you end up with a stronger dependency on the electronic with the radius of the properties at the end. We would expect that if you were able to measure single-wall or double-wall, you'd even be able to observe larger or larger slip and lower friction. Okay, one final question. Thanks for the nice presentation. You had a slide where you compared in Coulomb drag versus phonon dispersion, I think. Could you give some idea how these two are compared together? I mean, in which system, which one is dominant? This is, okay. They play it all together, the two are together in a certain way. It depends in, for example, if you consider purely graphene, very free graphene where the, I mean, if you look here, for example, in the phonon drag, you have different, if it goes, there are different parameters that are in a game. So if you have completely flat and the friction force can be low, you end up being less sensitive to phonon drag and more sensitive to Coulomb drag. Also, if you are in a perfect graphene where the effective mass of the charge carrier is lower, then you're going to be more sensitive to the Coulomb drag. The two are playing together, but let's say it's more efficient Coulomb drag, phonon drag can be way more efficient than Coulomb drag, but you need to be in a configuration where the friction force is very high. If you are perfectly slippery, for example, I would expect that you would have, in other hand, a very large Coulomb drag. Okay, thank you. So we now come, thanks again. We now come to the final talk.