 call the afternoon session to order. And our first speaker will be an invited speaker. So this is a 30 minute talk with hopefully about five minutes, including five minutes for questions. And so Martin Muser joins us from, I believe you're, I assume you're in Zabruken. I'm in Zabruken, yes, yes I am. All right, well welcome and the floor is yours. Okay, wait a second. I need to get my PowerPoint working. Okay, so can you see my screen? Yes, we can. Yes, you can. Okay, I can get my laser pointer now. Oh wait a second, my laser pointer I can get here. All right, so thanks for all the invitation. Sorry that I couldn't make my way to Trieste. It's always an exciting place to be. And actually the last time I went, I had a very pleasant encounter with somebody I was asked to share a few words about. So Erio and Andrea asked me to make a few comments on Mark Robbins and his contributions to the field. Now I find it very hard to do that in a few minutes because Mark had so many, many outstanding contribution. Yet it may be worth emphasizing a few. I think the first really big milestone he left was on the melting or sliding induced melting of a boundary lubricant, which would basically pull out energy from motion, but the recrystallization would not propel the system. And so he suggested a very important mechanism, which is responsible for stick slip. In later work, I also on boundary lubricants where I had the privilege to be involved. He emphasized once more the importance of boundary lubricants and explained how the presence of a boundary lubricant would make two solids pin, which may otherwise be superlubic. And moreover he showed that in a case where you have two stiff solids in contact, where you get large, normal stresses, where at the microscopic scale, you would see friction loss that very much would look like those that you have on the micro scale. And I think these topics pertain very much to the theme of this conference and have left a deep impression in the field. He also opened up a new field, the simulation of rough contacts where we got very important information on what are the stresses, what actually happens when we have nominally flat surfaces, how does it affect the t-tion? He did that together with Lars Pastefka, who is in the audience. But he also made contributions outside of tribology where he always kept asking seemingly simple questions with a very, very big impact. For example, why do colloidal systems so frequently condense in BCC, body-centered cubic systems, and not in dense faces as we would suggest? So this was also one of his milestone contributions. In my opinion, he was the most prolific modeler in the tribological community and certainly a front runner in the bigger field of condensed metaphysics, stat-mack and materials physics. And actually when I first got to know him, he gave a talk at Rockefeller University in New York, and he was introduced as a computational physicist who does not produce numbers but insight. And I think that really describes Marx's scientific contributions very, very much. He also had a deep personal impact on people's lives because of his generosity. And I just mentioned one that affected me personally. Mark and I, we hadn't talked I think in two years, not because we were hostile, there was no opportunity. We didn't exchange emails. And out of the blue, I get an email where he says, look, in your hometown that is a job opening, today is the deadline, don't you want to apply? And so he thought of me and I don't even know why he remembers my small little pathetic hometown, but so he did and I moved. And that really changed my life because I got back from Canada where I was personally not very happy. I moved to Germany. I met my second wife, and so now we have two children. And the reason why I mentioned this is that I show one picture that is a nice transition to the thing I would like to talk about, which is tribal electricity. So my son has fewer hair, so I couldn't show him and he looks like me and I wanted to spare that for you. But here you see something that is very well known actually since the time of Thales, where the rubbing action between solids induces electricity, which then makes our hair stand up if we jump on a trampoline in a dry summer day. And the classical experiments involved an amber stick and biological matter. So you have an amber stick and a cat and you rub the amber against the cat and the amber becomes negative and the cat plus iron gives a cat iron so the cat iron sits on the cat. And it's a bit controversially discussed what is the origin of that electrostatic charging. In principle, you could think it's simple because there can only be two mechanisms. There can be electron transfer or there can be iron transfer. I mean, these are the two ways how you can basically charge a surface. And you would think figuring out which one it is it's a lot simpler than figuring out what friction mechanism dominates. I would argue there are roughly let's say 10 to 12 or 15 different friction mechanisms but for charge transfer, it's simple you have electrons or ions. And as long as you get to metals in contact it seems an easy story. The electrons go to the solids with a greater work function and this is also experimentally observed but it becomes a bit more tricky as it comes to dielectric. And the big trouble here I'm quoting George White's sites out of an article that was the first review I read on tribal electricity. Daniel Lex also has very nice articles in the recent past but this was the first one that I read. And he basically says, look in a lot of the systems that charge that get tribal charged you have a big band gap so you wouldn't really expect electronic flow but if you look at the ions they're also immobile. And so not really clear what is the mechanism if we think of Taley's cat. Though apparently Taley's didn't do the experiments himself he only knew about them. Even more so there is a big riddle where often you get what is called a cyclic triboelectric series where basically zinc picks up a negative charge from glass picks up a negative charge. So silver picks up a negative charge from zinc and so on and so forth. And it's a bit like in the painting by Escher where you keep going downhill, downhill, downhill but that doesn't really make sense. We know we cannot really go in circles. And so I got curious in the question of modeling this to find an answer to these type of questions. Yeah, so the mechanisms are known as in friction. It's more the question of which one is more important under what circumstances and what parameter favors one over the other. And that calls for modeling. And so I'm by nature a modeler who likes to use potentials because then we can use a few thousand or a few hundred thousand atoms. And so when we do want to model triboelectricity we must be in a position to assign charges on the fly. We cannot say you are a lithium atom or ion you have to charge zero or plus one. The atoms have to figure out themselves what charge they want to be at. And the next few minutes I spent on stuff that I did eight years ago but I mean the process of getting back to it. So I thought it might be nice to talk about it again. So what we need to do from a modeling point of views is we have to chop up our nature into finite elements or into atoms. And then what we need is we need a description that says what is the potential energy as a function of the atomic charges that we assign and usually what you try to do or what you assume when you do an electronic structure calculation is you assume that the system tends to find the minimum of this function here. So what has been traditionally done is that you write down your assign charges to atoms, QN is the charge on atom N and you write down something like an energy expression where you say an atom with a lot of electronegativity wants to attract negative charge. So this is what drives the charge on an atom. Too much charge on an atom starts to repel itself, right? So we cannot put a charge of minus three on an oxygen atom. It doesn't like that. So this term would be called the chemical hardness. And then we have the Coulomb interaction which may be screened. And then we minimize this thing with respect to the charges and assume we can be happy. Now, DFT does something similar except that DFT doesn't discretize space towards atom but it's in a charge acceleration model that I'm talking about now. The energy is a function of discrete charges and then DFT, it's a functional of a continuous electron density but the philosophy is very similar. Now, if we do that, we run into a big trouble which is that if we separate atoms from each other, let's say we separate a methane molecule from a water molecule, the molecules end up with non-integer charges. We have a super linear polarizability of polymers with a chain length. Our dielectric constants are always infinitely large so we are formally corresponds to a solid and the dipoles of an alcohol chain are linear in the length of the chain which is not really meaningful either. So this problem we managed to fix by saying that if we assign charges to a molecule, here's a hydrogen terminated silica molecule, we can donate a partial charge only to a neighboring atom and what we have to do is paying basically pushing partial charge across a bond needs to be penalized with a bond stiffness term. And so this was a combination of potentials that existed before that were atom-based or bond-based and this model was a split between the two and now on the fly, we determined the charges using this method and it works great. It produces meaningful dissociation limits. It leads to more transferable charge assignment steams. It has to correct size-dependent polarizability and this parameter of the bond stiffness can be easily related to a dielectric constant but the correct association limit is not only a blessing but it's also a curse because whenever we separate two charges, eventually when we separated say a chlorine more atom from a sodium, eventually the sodium says give me my charge back because my ionization energy is greater than your electron affinity and so at very large separations, we always go back to a uniquely charged state. So here you see in the movie, what happens is any physicist will recognize that this is an amber rod and this is a cat, right? So it cannot be mistaken. So we approach with a rod to the cat, it picks up negative charge, the cat picks a positive charge but as we go back, the charge goes back and DFT actually suffers a bit from the same problem because like any other methods that assigns charges or charge density on a unique minimization principle, your charge state depends exclusively on the atomic positions but it has no history dependence added to it but we do have a history dependence. For example, if you separate an NACL molecule and you do it very slowly in a non-polar solvent, you will end up with two neutral atoms but if you separate them very quickly or in a polar solvent, which you later evaporate, you are going to end up with a positive charged cation and a negatively charged chlorine anion. And simulations need to reflect that so that we can model tribal electricity. So what we did is we introduced a new variable to the classical potentials where we said we have to develop potentials relative to the oxidation state. So a Cl minus atom is a completely different species than a chlorine atom and the force field has to develop for each of those fields. And basically this oxidation number is not subjected to a bond hardness. And otherwise we do everything as before and now when the ember rod approaches the cat, it can pick up charge. And so here we basically model things with an electron charge transfer where there is no ion that crosses the interface but there is charge because the charge wants to swap. When the ember rod and the cat are close to each other, the charge is happy when it swaps and when it's pulled too far away, it's kinetically hindered to do the transition. And in a way we could argue we have a dissipation mechanism because it costs more energy to pull the tip away than to approach it. So whenever you have hysteresis that leads to energy dissipation and tribo electricity certainly is one of the mechanisms that we haven't really studied so much and it might be worth to do so in the future. Now we can do the same thing for metals. When we switch on metals you see the charge of the metal wants to sit on the surface you get much more charge flow. And once we did this, we actually noticed, now we can model actually the discharge of the battery which we did. So we have an anode and electrolyte and a cathode and whenever you're close to switch here you basically get charge transfer and that is irreversible. So in a way when you close the switch you get tribo charging and if you open it again the charge doesn't go back. And so once we had the model down to simulate tribo charging the first thing we actually did was to look a set up a simulation of a small scale battery where the anode could basically pass electrons through an external resistor to the other side. The switch would be closed. The charge would flow proportional to the voltage between the anode and the cathode which DFT would always level to zero. And then we could measure the flow and we could basically see the discharge curves for different resistances. And that actually matched experiments fairly well which we only see one year later. So this voltage time was really a prediction and not a post-diction. So we didn't have this Duracell battery in front of us when we did the calculations. Now the trouble of all these calculations that are presented is they're all done with Poir model. So we take Leonard Jones and we add a complete redox scheme on it. And we want to move to what's more realistic chemistry. And one of the steps to do that actually also took a long time ago. So let's go back to tribe electricity and ask the question, what are the ways to actually produce ions? And one of the things we noticed when we worked on decomposition products or what we expected to be decomposition products of underwear additives was that stress could produce ions. And I'll show you the slide in a minute and I will elaborate on that also a bit more. So imagine you have neutral molecules to begin with. You add stress and then what can happen is when you remove the stress you get a modification where one molecule is negatively charged and the other one is positively charged. And we found such a reaction and triphosphoric acid which you see here where at a pressure of 3.5 gigapascal, a hydrogen would move from one triphosphoric acid molecule to the other, leading behind a negatively charged group here a positively charged group there. The reaction was endothermic but still it was very persistent up to I don't know how high we heated up 2000 kelvins. And so this very stable compounds even though it's a bit strange to have a protonated acid molecule. And in recent work, when I went back to look in one of my favorite topics which is the functionality of underwear additives we basically asked the question what is the role of stress in chemistry? And we discovered very much to our surprise that it really affects the way how ions are generated. One step back actually there's a great paper by announced in 1896 where he basically argued if you have any gradient at an interface be it stress be it temperature of course concentration and you get a gradient across an interface you get what he called touch electricity and the arguments he put forth are exactly the arguments that we should discuss again perhaps more often what Nance didn't know at the time he didn't know about electrons yet and he didn't know about the symmetry of the on saga coefficients but otherwise my impression is that Nance had it down and there was another fellow that I can talk about Knubloch is his name a German Jewish scientist who disappeared and in the 30s he has great contributions to the field which are absolutely worth reading and astounding given that they didn't even know about the existence of electrons. Anyway, coming back to real life systems we have here two molecules of what collaborators of mine and I expected to be decomposition products of zinc phosphates that are used as lubricant additives the functionality I don't want to discuss here I just want to focus on the ions and we add to that two triphosphoric acid molecules and once we densify this moderately we actually get a long molecule with the zinc phosphate backbone and kind of funky side groups which however pack very nicely into a cubic simulation cell. And now we ask the question what happens to the system depending on a deformation mode? So we can compress it isopropically as it would happen in a diamond anvil cell but if you think of a tip if you think just of a mechanical contact where you touch down with a tip the material underneath the tip wants to basically push to the side but it can't because there's other matter already sitting there. So we took two extreme point of views where we said let's make a uniaxial compression where the material is not allowed to breathe and then let's make a compression where the system is allowed to breathe to the extent that the system is volume conserving. So I would argue that real life is somewhere between this area conserving and volume conserving condition. And what you get is you get implicit shear. So you don't necessarily apply a shear force on the surface but because the stress tense is not diagonal anymore you have a deviatoric stress or for mesa stress or shear stress whatever you want to call it. And then we made the interesting observation that depending on the type of compression mode the system transformed in a very different way. So as we added isotropic compression we had a truly exothermic reaction if you wish where the energy gain per zinc atom was 0.7 EV or I forgot the unit shear so 0.7 EV and the system was very, very stiff. When we did the same with uniaxial area conserving we got a similar stiffening but we gained much less energy during the compression and when we did this in a volume conserving fashion we barely gained any energy and the bulk modulus the stiffening was not as big. And here not surprising actually what the systems are trying to do when they're under non isotropic stress they're trying to become isotropic and then once you release the stress again you'd be soft along the direction parallel to which you have compressed. And so I think I'm convinced that happens in the tribological contacts and if the chairman of the session ever manages to cut out a zinc phosphate tribal layer and rub if you ever managed to measure the stiffness normal and perpendicular to the old compression direction I would cook you a great dinner. So what do we see on the charges? We see something very, very interesting I find. So here I just show you one case for the area conserving compression which is the Femisa stress didn't really go up so much but at a strain of 0.1 or 0.15 we found 10 mobile hydrogen atoms or protons and what do I mean with mobile hydrogen atoms at one value of stress we run a simulation of roughly 20 picoseconds and we counted the number of protons that had changed the oxygen that they had been bonded to. And what we see is actually the stresses were not very big so the hydrostatic stresses were one giga Pascal or two giga Pascal something that you can easily achieve in a tribological contact between two stiff materials there was this whole large bunch of mobile protons in the interface and then again Nant's argument you have a gradient across the interface be it chemical, be it stress and boom you get a charge transfer. As we kept compressing further again you would see that the number of charges go up and even in a small cell like this on average 10.5 ions or protons change the oxygen number, oxygen atoms that were bonded to then a coordination change of zinc occurred and the Lewis structures, yeah and after that the number of mobile ions decreased. So stress and actually not only hydrostatic stress really the exact shape of the stress tensor appears to have a tremendous effect on how the atoms charge there. And here you see the structures after decompression. So after we decompress the samples we see a lot of Twitter ions where in the hydrostatic compression we get a negatively charged oxygen here a positively charged there. So this group is only bonded via coulomb interaction to the zinc. So you could argue we have produced a cation which is a protonated phosphoric acid which then in an interface might decide to go to the other side of the interface. And under shear we get actually here the shear is the lattice we get a very large number of ions we get two cations produced nearby and then a doubly charged counter charge here. And again, the type of chemistry we do under shear really depends very much on the shear stress. And I think this has important ramifications or implications for the prior electric series which is a, we observe that Twitter ions are produced at very low contact pressures. The nature and the number of the stress induced ions depends sensitively on the shape of the stress tensor not only on the compressive stress or the hydrostatic stress it depends on shear stress and the hydrostatic stress. The larger the deviatoric stresses the easier it seems to be to promote the ion formation. And any system in which charge transfer is triggered by ion transfer. I'm not saying that all of them are I'm not saying I explain everything but those that do tribal charging in a system that has a very stress dependent number of ions simply cannot be assigned a unique positioning in the tribal electric series. I would say that for the class of tribal charging that I was looking at what you should generally have in a tribal electric series is at least two stiff materials where you get very high contact pressures and one where you are soft and then it's almost unavoidable to have this type of series. So actually I was faster than I was supposed to be I was supposed to hurry. I would like to thank my collaborators Saggy Suhomlinov he ran the simulations on the zinc phosphate Wolfdup Joint supported me in the journey doing the beginning, the battery simulations and the tribal charging of this cat. If anyone is interested continuing this type of work I'm looking for postdocs who like this type of research. Also my big thank you goes to Mark Robbins for introducing me to the field and I would like to thank you for your attention. Thank you very much. Thank you, Martin. So we have time for a few questions. Thanks for that inspiring talk. I wonder for my whole life why Amber gets charged when you rob a cat but I still don't understand it. Why does it happen? We don't we can say it either yet. So our plan is to look at the response of keratin to stress if it pops off positive charges. So what I said today was again a little bit of a proof of principle study and I think one has to be very, very careful in the selection of the materials where we try to understand how does stress build up. But actually our skin appears to be soft but the top layer is actually fairly rigid. So the epidermis has a Young's modulus I believe of close to four gigapascal. So you can get very high pressures in the very top layer at very small scales and Amber is also not exactly soft. It's also is about the four gigapascal. So I wouldn't be surprised if when you rob Amber against an organic substance that is keratin based that you get stresses locally in the order of gigapascal where you could create mobile ions and once you have mobile ions I suppose you agree with me. It's pretty obvious that you are going to have high end transfer most likely protons there around because they're the most mobile but making a really quantitative prediction that is the challenging part. Yeah, thank you. Hi, Martin. As far as I understand that the contact electrification is what triggers from the beginning the all the charge separation and all the charge related ionization. And so, but all the dynamical effects so you are taking into account is what makes this thing complicated. In particular, you have to go beyond the Born-Oppenheimer approximation and beyond all the static as you show. Well, I think if you have ion transfer you get away with Born-Oppenheimer. So the systems that I looked at, the phosphor systems the band structure or the band gap barely depends on barely depends on the stress. So we always get a band gap of about 4.5 to 5 EV in the realm of DFT. So which is always to be taken with a grain of salt but we really don't see that. So I think we can live for a long time with the Born-Oppenheimer approximation. And I think at the point where the ions are separated over a distance that is large enough for no tunneling to occur. That's the point where we basically do the conical intersection if we wish or where we change the Landocena level if you're a physicist. But then it's too late for the electrons to cross back. So I think if it's ion transfer you can get away with Born-Oppenheimer but you have to be careful that when the surfaces are really separated that you don't get an unphysical that we call it unphysical electron tunneling current that DFT can find because DFT can do a global energy minimization but nature has to find a way through the vacuum where the electrons don't want to be, right? So this is why this is such a hard problem for DFT and perhaps with classical simulations we can get a bit further where we put in a bit of the phenomenology by hand and where we don't get a tunneling current. Okay, and you know what? There is one question in the chat. So Martin if you're able to take a look and perhaps if you can try to do a quick answer to that one. Oh wait a second. But then I have to, can you read it to me because I see only my... Is there anybody on Zoom who can read the chat? I don't have access to the computer myself but hopefully someone in the room does. Is there anybody else? Or someone, some of the other remote guys can you please read it to me. The question is what about the temperature dependence of tribocharging? And this is from Anders Vurnisch. Does tribocharging have a temperature dependence? Thank you. It certainly will but I don't want to make any predictions at the point. So what I really want to do is I want to parametrize force fields and then I can start to answer these questions. But currently I have the tools but I don't have the potentials yet to really look into that. Perhaps a little bit... No, actually I don't know. I think if temperature is higher perhaps I can give an answer after all. Once you pull the surfaces out of contact there is an energy barrier for the ions to go back where they want to be. And if temperature is high that is going to happen more quickly. So my belly feeling says if temperature goes up you reduce the tribocharging because ions have an easier time to go through a barrier to the place where they actually want to be. But that answer may depend on the specific system. You referred to some work we and others are doing and that is, and I wanted to share, yes actually we are starting to look at sulfur free and even some zinc free phosphate esters and others as tribofilm additives. And the second point I wanted to share Laurie Marks is going to be giving a talk in this symposium which I believe is focused on work he has been doing both experimentally and in the calculations of how flexor electricity stress induced charging from bending occurs. We have to move on to the next speaker to stay on time. I see there was a question from Laurie but maybe we'll have him address it when he gives his talk. And let's thank... I hope I can make it to Laurie's talk because I have to teach, I have a busy week but I'm trying to make it to Laurie. I'll read his comment. It is for some cases, e.g. sliding shot key contacts, the temperature dependence follows thermionic emission. So comment about the temperature dependence. Okay. Very good. Great discussion. Let's thank Martin for a very stimulating talk. Thank you.