 what's the nature of dynamic friction in a humid hydrophobic nanocontact. So, the floor is yours and I... I would first like to thank all organizers for inviting me to this conference. It's a really pleasure after corona time and this work was done together with colleagues from France, Olivier Noël and Pierre-Manuel Mazran and they performed experiments and I tried to reproduce and to explain these results with my simulations. Okay, so a short overview of this subject. Actually, problem with this subject is that it is very easy to generalize everything but it is a very, very, very various topic with many details and basically, when I started doing this with Olivier, I realized we are filling a puzzle. So, some pieces of the puzzle we know and some pieces of the puzzle we should discover. So, basically, it is really nice to start some early works on sliding and contact in two phases and that's capillary friction. You have air phase, gas phase and liquid phase and that was one of the early simulation done by Mark Robbins. And then I should point experiments of Elisa Riedo with capillary condensation where it was first time proposed the mechanism. Then, recently, relatively recently, there were experiments with tuning fork AFM where they measured mechanical properties of the capillary and then there is a whole, I would say, cluster of research done where people took AFM. They went through or over a layer of water and measured the frictional characteristics of the systems and most recently, a work where humidity was changed and friction was studied in simulation and in experiment. So, what Olivier is doing and in order that you understand how we came to ideas in this talk is circular mode AFM. So, circular mode AFM means that his probe is making a circle and he can measure both lateral and normal force simultaneously and that means that he never interrupts movement by going back and forth with like in classical AFM. And just briefly that you know what is the plan of the talk, I'm going just to tell you something about results by AFM, then talk about results in AFM where we try to see what is happening in these experiments and give just short conclusion and outlook. So, basically what Olivier is doing, he is turning his probe and while he is turning it, he is approaching it to the substrate. In one moment probe touches the substrate or a capillary bridge is formed and then he continues to press increasing the normal force and in opposite direction, he starts to... He reverses direction, the probe goes slowly out of the contact, then he can measure adhesion and then in one moment capillary bridge will snap. So, we can introduce all kinds of different properties like adhesion force, it is different in normal force between going into the contact and out of the contact, you can introduce coefficient of friction, you can also measure jump in friction, so the friction when contact is created. So, first thing which you can already see if you look his experiments on different velocities is that for... A H or O P G that adhesion force doesn't depend on velocity because this distance between approach and retraction curves stays constant when you change the velocity and now I would like really to enter into details of this talk, so it is a model of friction which was proposed by Andres Vernes in the Diagon form, so we have this jump in friction and friction coefficient then connected with the normal force and we could extract all this data and we saw that with velocity this jump in friction so that this left figure depends... It has a power law with velocity very close to linear and coefficient of friction depends linearly on velocity. Of course, adhesion force, the blue points on the red diagram, it doesn't depend on velocity. Okay, so it was a little bit contraintuitive, if you think on condensation, if you move your probe faster, condensation should have less time to act so basically you would expect that friction should fall with velocity and that's something what Eliza Riedl in her work has obtained. So we created these simulations and basically I had a big substrate and I have put water on this substrate and the water has created droplets so we started to move the substrate and then these droplets got attached to the probe and then you could see with the displacement or that these droplets, they accumulate under the probe, it grows and at certain point because the surface tension of the droplet is going to pull it out of this point of the closest approach between probe and the substrate. So now I'm going to introduce you into my simulation. So we had a substrate which was fixed, we had a probe which was attached to the points so that we could measure different forces. We had a relatively large amount of water, at least the most of what I could simulate in some final time and I tried to make a probe to correspond to AFM tip radius which Olivier was using. So here basically I have fixed the view on the probe so substrate looks like as it is moving but actually probe was moving. It rises as it rises there is this snap and the probe is detached from the substrate and before they were connected by capillary bridge. So just once more and then I'm going to go to data. So of course you can evaluate normal force, dependence on contact angle and if you have a hydrophobic system you are going to have a first normal force which is positive and then when you stretch the droplet enough the droplet is going to start to resist to the deformation and you will have adhesion also in hydrophobic systems and in the hydrophilic systems you're going always to have adhesive regime and I made a very simple model for this. I treated capillary bridge as a cylinder and these lines which go through are results of this model so it is also one of the results of Mark Robbins who showed that if your droplet is large enough you can basically describe it very well with continuum results. So basically you see two lines and one line was for moving probe and the other line was for fixed probe where the probe was not moving just going up and down. So basically this is the snapshot so you can see the compressed droplet in hydrophobic case. You can see the stretch droplet in hydrophobic case where it produces adhesion and then the droplet detaches. Okay, so what further I could do is I could calculate dependence of lateral force on the surface and I could see that lateral force depends linearly on the surface. That's an interesting conclusion because some other proposed that it depends on the contact perimeter so just a contact line between substrate and the water liquid. And what we also obtain is that lateral force depends with a power law on velocity. It increases as it experiments, but the power is smaller than what would be expected than what was obtained in experiment. So basically at this point I would like to just get a little bit philosophical. So basically if you have a nice simulation I did these simulations with the constant volume. So basically that could be a lower limit of this dependence of lateral force on the velocity but of course in experiment as in the first slides I showed there could be accumulation so with time the droplet could increase the volume. So basically if you have a higher velocity it would maybe increase larger and then it would be limited, this increase would be limited by other process maybe evaporation but it remains to be investigated. And of course what is interesting is that in experiment we see also that coefficient of friction is dependent linearly on velocity so there is a question how this jump in or initial lateral force is connected to friction coefficient which are processes which are connecting and linking two of them. So just in conclusion so in this talk we introduce another mechanism for creation of calipelar bridge and that's by accumulation of water in a moving contact. So if you have asperity it could come along to a small droplet which is already there at your surface and it just collects her and then around this droplet the capillary bridge could be formed. We show that lateral force in simulations depends on contact surface and increases with velocity and we show also that adhesion force in systems has a complex behavior which can be captured by a model and also we could show that it doesn't depend on velocity. Thank you very much. But let's start with Bo. Why do you have adhesion when you have hydrophobic interface? Well, even in hydrophobic systems you have van der Waals interactions between water and the substrate. It is not completely repelling and that if you stretch the droplet enough it will create adhesion. So it's just van der Waals forces which are playing the role. So if it is hydrophobic that doesn't mean it wants to throw away water from itself and the fact that there is van der Waals force. But you will have a positive plus pressure in the droplet if you have hydrophobic surfaces. You should get repulsion. I understand if you have van der Waals interaction between the two solid walls then that could help you to get attraction. No, but you also can stretch the droplet. So basically deform a droplet and then it turns in. Then it's some dynamic effect due to viscosity and things like that. Yes. Thank you. I had exactly the same question. I'm wondering to follow up. I mean, can you get from your simulation? You have the interactions in your simulation so you can see what is the van der Waals interaction that the market of dynamics is producing, correct? Yes, I can do it. Yeah, so I guess my follow up question is does your system reproduce the fact that there is a Laplace pressure there? And how does the force from that repulsive Laplace pressure compare to the attractive force from the van der Waals? And also I guess you would get some attraction from the surface tension as well. So I would assume you could separate all those contributions because you have the simulation, you have the atomistic detail. So could you break down those different contributions? Actually in this model I did exactly that. So I had a contribution of surface energy of this droplet compared to air. I had a contribution of surface energy of these interactions with the substrate. And then if I calculate energy and calculate the force, I get in the model negative forces and adhesion in the hydrophobic system. Okay, I follow up questions. You said it was a constant volume assumption. The AFM experiments, although it's fast AFM, it's still much slower than the thermodynamic timescales of the system. So a constant chemical potential assumption might be better. So can you do that? Oh, well, I could also turn it on in lamps. I mean, it would be possible. It is just that me personally, when I was designing this simulation, I was aware of your publication with Ashley Martini. Was not quite sure because I see just a few molecules flying around in the simulation. And there should be some equilibrium around in this volume. So I'm not quite sure if adding molecules would just inundate the whole surface with the water. So I decided to design simulations like this. But yes, it is one thing we should consider. So I was more looking into partial pressure inside and partial pressure was okay for this humidity. I have a question. Water at surfaces often ordered over this length scale of a few nanometers, something like one or two nanometers. That's about the size of your droplet. Is the water inside of the droplet ordered? Or can you determine an order parameter or heat capacity of the water? Would it change when you pull at the droplet? I must say, I didn't look at it, but indeed it is a very interesting question. So I should look at the structure of the water at the surface, but I didn't look at it at this point. But yes, it's legitimate and very interesting. Okay, thank you. Thanks a lot. And now we switch to Guillain-Baville-Heine.