 I know too, so. So, good morning, everybody, or good afternoon. I'm Nicola Seriani from the CTP. I'm chairing today's session. So it's my pleasure to welcome Professor Robert Schlegel from the Fritz Haber Institute of the Max Planck Society. So he's an expert in catalysis. He studied at the Ludwig Maximilians University in Munich and spent some years between Cambridge and Switzerland. And now he is director at the Fritz Haber and also founding director of the Max Planck Institute for Chemical Energy Conversion. So, and he's so well known that he has a number of honorary professorships and awards, including the Anya Award for Energy Transition and the Arvind Mittas Prize of the German Catalysis Society. So, Professor Schlegel, the screen is yours. Thank you. All right, good morning, everyone here and outside. Thanks for the kind introduction and also thanks for the opportunity to contribute to this very interesting series of lectures. I have skipped the introduction largely to what we are going to do because there are so many excellent lectures on this that I can go straight into the details of what I want to tell you. So it's on operando for the mission from electrified interfaces and there is only one motivation slide that I have after my acknowledgements. Of course, the acknowledgement is to the group because I'm doing nothing rather than entertaining you and the group is doing the work. There are a few postdocs who I'm mentioning here who have essentially contributed the work that I'm going to report to you. These are the names of my group leaders who operate the whole thing here that you see with these many people. And I would like to in particular mention also Peter Strasser and his team but he will be also on this conference later as I have seen in the program. So that is not necessary to go any more deeper in here. So what are we going to discuss about? We talk about electrified interfaces and it is enough to say these are the critical science elements in our energy transition because however you deal it whatever you wanna do with renewable energy you will not be able to do anything if you don't understand what electrified interfaces are. The traditional concept that we have is very similar to heterogeneous catalysis you have a solid liquid or usually a solid liquid interface and the solid is considered as being static and there's a liquid outlayer that is of course dynamical. There's an electrical potential drop across this interface and that reaches from the interface into the liquid and there's the double layer concept that is also something that is heavily discussed on this meeting. So I do not need to go any deeper here. And what I mentioned only here is that is identical in fuel cells, electrolyzers and accumulators. So you can also study the same thing everywhere and ideally of course you would like to know also the operando experiments from all these three units but what I'm going to concentrate on is just on the electrolyser part. The capacity that you get when you have such an interface because there's two half reactions is of course well detectable. We don't have to discuss this but the chemical nature of the interface to what this capacity relies to that is a matter of debate and this is one of the motivations why it is good to do operando experiments because we are interested indeed in the interface and we are interested in the top atomic layer where this charge that you put on the system is actually seeing the other phase. And one hypothesis that I want to bring forward to you is could it be that the interface is not rigid and is gradually changing its chemistry in order to accommodate the potential drop rather than there is a potential drop only into the liquid? There could be of course also a potential drop in the solid and the solid could change under reaction conditions. Now in particular from theory side you would say this is possibly not the case because for example you have an oxide or a metal and then the limit that there is no chemical conversion you would not see why that should change its chemistry. But I would like to draw your attention to the point that you have seen and you will see also in other presentations from this conference that we want to do of course production there and if we do production then we are far away from equilibrium. So we look into non equilibrium states of the system and then it is not so clear what the chemistry is. Only to illustrate the first sentence that I have on this slide is when you look at this diagram then you'll see whatever you won't use as a potential fuel you will always have this interface between green electrons and green molecules. And if you want to switch between green electrons and green molecules then these electrified interfaces are the critical part of it. And that is the reason why it is very worthwhile that we really have a deep understanding of how that works. So now we look into polymer electrolytes, electrolyzers. This is my example here. We could also use another example. We also working on other electrolyzers. In fact, we do our larger experiments not with spam electrolyzers but that is something where we can really dig a little bit deeper and create some understanding. And a good catalyst is of course iridium oxide that is no longer to be debated here. It is also clear that this is nanostructured. It is not so clear that this is iridium oxide. This is why I put the question mark there. The catalyst gives a stable contribution to the total over potential. Also that is something that is not usual in the case when you look in the operation regime of such an electrolyzer then you'll see that the catalyst has a significant contribution if you increase the current density then you also see that the catalyst is not perfect. The degradation of the system is, you would say for academic reasons, it's perfect 30 millivolts per month in dynamical operation. But what is interesting is that the degradation is much faster if you do static operation of such an electrolyzer. We use electrolyzers of a couple of kilowatts of power and we started them for a couple of thousand hours. So this is not something for 24 hours on a rotating this electrode experiment where you get this data from. And this indicates to us that there's materials in the dynamics in the play and this calls for in-depth analysis and that was also the motivation why we started about seven years ago with looking much deeper into what is this iridium oxide. This is just an experiment where one can see this. What you see on the X axis is the current density on the Y axis the potential and one sees the different contribution. So the light blue one is the catalyst itself then the catalyst with its catalyst layer so with its contact is the dark blue part and you see the membrane. The membrane is one of the weakest things by the way. Anyway, when you look into degradation studies because these membranes are not so stable as one would think and even the current collector hardware also contributes to the potential drop and one sees there's a significant contribution of omics things and there's the catalyst itself and this is much less than what you would expect from just an omic contribution of a non-conducting system. That is also understandable because we know that iridium oxide or iridium metal bolts are metals. So chemical spectroscopy, what we do now first we would like to know what's the chemical constitution of the material that you actually use in a catalyst. So iridium oxide in inverted commas is after synthesis and mixture of the so-called black material and iridium metal. It's not iridium dioxide. And the reason for that is deep into the chemistry of iridium because iridium does not precipitate from an aqueous solution but it only is being forms a solid by disproportionation. Almost all other metals in the periodic table they really form a compound with OH for example or with water and then do a polymerization reaction under the liberation of water but iridium does not do this. This only forms by disproportionation. That means you get something that is lower in oxidation state and something that is higher in oxidation state that is the reason why you get this binary system and that is also the reason why all iridium materials that you make are core shell structures where you have a metal in the core and some undefined material in the shell. This already by itself is a strong hint to the suitability of a water splitter because if it doesn't polymerize and form a polymer material as an oxide then you can also learn that the ligand field of water and of OH-minus is too weak with respect to more complexing species for example nitrite, chloride or perchlorate or organic N ions. And that is a very good sign because the interaction between iridium and oxygen is very weak. Now some experiments that you can really see this if you precipitate this iridium black that is also commercially used to make the electrons then one finds if one reduces that with hydrogen to iridium metal that this iridium oxide is reduced at very low temperatures between 50 and 100 degrees Celsius. And what you see in this red box is that would be iridium oxide, iridium oxide by itself. If you make it, you can make it by calcination then you can also see its reducibility and the material that you use in the electrolyzer is very easily reduced in iridium oxide. That's the first indication that the material is somehow different. Now when you look at the chemical constitution and for that you need to do a series of analytical experiments then you'll find that the material of choice is a mixed material with a strange average oxidation state of 3.5. And then you see if you go to other conditions of precipitation then you'll get something where you approach iridium dioxide so to higher oxidation states. But this is interesting because if you go to formally higher oxidation states then you'll find from analysis that the contribution of metal is increasing. This is because of this disproportionation reaction. But this is not always considered by people who actually do the synthesis of this material that if you try to attempt the composition iridium dioxide then you get a maximum of iridium metal and a minimum of iridium dioxide and you have to do high temperature calcination in order to get rid of the iridium metal. So what one can also see if one looks into the operation of such a system with electron microscopy then one finds that the successful catalyst indeed consists of nanoparticles. You see it is about three, four nanometer particles and one sees the lattice image of the iridium oxide so that is perfectly fine. But after operation you'll find this cloudy stuff that in this image you see on top of it and you see also that there's the impression there's some holes in this cloudy stuff because there's dark spots in it. If you change now the imaging conditions in your microscope and you have to be extremely careful to be able to do that because this cloudy stuff is extremely beam sensitive. And if you just focus the beam on it then you see another iridium metal. And this is also sometimes you find in the literature the result which is perfectly okay after some electron irradiation then you find iridium metal on top of iridium oxide and this is a beam effect that is not the true constitution. The true constitution is this quasi-molecular system and one sees in this seven angstrom little circle there one sees the constitution of the iridium atoms that is something that the crystallographer knows immediately that is the typical motif of a hollandite structure but that is not hollandite because hollandite is a beautiful crystal. This is just a molecular analog of hollandite. You can do this also by diffraction and when you do this by electron diffraction then you find something which is also interesting iridium oxide in its root tile form should of course only have corner sharing iridium oxygen polyhedra but in the successful catalytic material you find quite a bit of edge-sharing polyhedra and that is an indication that this is certainly not only root tile and of course we can do the structural analysis and then one finds indeed there is a mixture of this root tile and of this hollandite structure but again don't be misled this does not mean that there's a crystal of hollandite in this material. You can of course synthesize hollandite and test it and it's a lousy electro catalyst in this form but why is this interesting to detect this as a molecular species because this makes it very easy that you change the oxidation state of iridium between three plus and four plus as this is not a metal you of course if you change the oxidation state then you have also to change the coordination of the system and that is done most conveniently if you change the mixture between edge and corner sharing octahedra because that's an easy and mobile interaction between several polyhedral if this does not form a crystal but if it's this kind of oligomeric cluster or molecular species, however you want to call this and we think this is a very important property of the working surface. So we could say that the well-working system of iridium oxide has something that we mean in accumulated science you will call a solid electrolyte interface and the solid electrolyte interface is also consists of this octahedra like you have them in a rutile but they have a much more loose connectivity and the looseness of the connectivity is given by OH and water that is coordinated to the system that is probably the SEI interface of the iridium oxide active electro-catalyst. Now what's the oxidation state of iridium that is also an interesting question in particular if you look at this black material because if it's a mixture then of course the oxidation state does not exist this is always a mixture and that one finds also what I show you here is the performance in blue to the bottom and one sees depending on how you have prepared your iridium oxide you get quite different performances. You can also see that performance and lifetime are well connected to each other that is also normal in electro-catalysts that is not something special but if you do the standard for the emission analysis in order to determine what is the oxidation state and then you would expect that there is a difference the good and the bad catalyst should be electronically different in their structure but the difference is extremely small between iridium 3.75 and iridium 3.82 this is a very small difference. Now, when you look at the analysis of this iridium oxide this iridium for the emission spectrum then you'll find there's a lot of contributions I have no time to go into that this is not simple and this is absolutely not straightforward because iridium as a metal has a different shape of core level lines than iridium oxide and this iridium oxide induces a satellite structure that has not been considered by quite a few early analysis of the for the emission spectrum and unfortunately it is the case that the shift the chemical shift of iridium 3 plus is not at the expected position but overlaps very strongly with the satellites from the iridium 4 plus that is the reason why it is not easy to analyze this asymmetric line study you get from working catalysts I have no time to go into that if you want to discuss this any further then we can do this maybe in the question and answer section so now a short deviation on how do we do actually this experiment so as I said it is very useful to do this in Operando because we learned that this SEI is quite fragile and it contains a lot of species and if you don't do this in Operando experiments you might be misled what is the real composition of the SEI so synchrotransfer the emission of course is a useful tool that is now very well understood and is being practiced also by other people who occur on this conference here but there are some challenges that one has to overcome and this was the reason why it took us about seven years until we could come to successful measurements there is something like a pressure gap also in electrochemistry and if there is not sufficient electrolyte there then you cannot pass enough current through the system and if you can't pass enough current through the system then you cannot see the full evolution of this non-equilibrium state that I mentioned in my introduction so if you do this with two small currents in the limit of very little conversion then you might be misled what is the true chemical nature of the SEI and then of course like in the electron microscope also in the photoemission experiments there is a massive problem of beam damage and the radiolysis that is also something that I would like to point out here to the synchrotron users the idea that for Operando experiments you need super brilliant beam lines this is not always useful because sometimes you create your system by the brilliance of your synchrotron beam rather than you study the system that you're really interested in and that's the problem of information depth of course it is much easier to do synchrotron experiments with high energy x-rays because then also the radiolysis problem is not so large but the disadvantage that you have is that you lose the surface sensitivity and as we have seen in the chemical analysis that all this at least this iridium electro cut list there are very delicate core shell structures or has to be very careful do you now analyze the average of the whole thing or do you analyze more the bulk of the material or do you really analyze the interfacial system that you are actually interested in? That means the reaction environment in which you do your Operando experiment is critical and one Operando experiment and another Operando experiment in different reaction environments even if it's the same system might give you different answers so when you see these distinctions in the literature that people disagree possibly what is the nature of this SEI this can well be simply a consequence of the use different reaction environments. Now how do we do our experiments? We use of course our high pressure analyzer this is this golden part by now everybody knows how that works and I do not need to go into that then of course you need a light source this is this green part and then you need your electrochemical cell and everything has to come together pretty close in order to be able to do works that means this is a matter of a millimeter or less distance and that gives you also a limitation to what kind of Operando experiments you can do. There are many suggestions in the literature that you should do multiple Operando experiments and you should look with several spectroscopy simultaneously to that but when you then really think about on what is necessary in order to perform the electro catalysis experiment then you'll find that your special accessibility of the area that you study actually is very limited and I have to be very careful with the suggestions that you should do multiple Operando experiments simultaneously because then you're very easily compromised the reaction environment. This is a quick look how that looks in reality at our synchrotron all of you who do such experiments have seen this the interesting thing is that they have a synchrotron experiment and there's a cylinder of liquid water there that you're pumping that that is something that of course a standard UHV scientist would never do. Now another representation of that we understand a little bit better what it means for the pressure gap is this is a very nice representation that I stole from Juan Velasco who is in our team and what you see is the planet Earth and this very tiny overlay which is the atmosphere and then when we look into the height the pressure dependence as a function of height over our planet then we can get a quite nice feeling of what do we do actually when we do an Operando experiment. So under vacuum conditions where one normally does for your mission then you would find yourself a couple of hundred kilometers above our planet and then of course you can see what do you learn on our planet if you are say at the ISS then you look outside then you know this is not exactly very characteristic of what happens at the ground of our planet. Now what we do is with the instrument that I showed you before we come then into say the stratosphere so about one millibar pressure but then you're still in the area where high flying airplanes are flying and you know as well you cannot survive even in that area but that is the experiment that we normally would do and then if you are very careful then we can go to higher energies and you will get talks also later in the conference on this kind of technique and then you come a little bit closer to reality but you're still far away from the ground of the earth. So one has to be extremely careful when one does such Operando experiments as I say that you do a reaction environment in which you at least under kinetic conditions you can say you have liquid water in the system and you pass a current through the electrolyte and they have all ions of the electrolyte present and only then you can expect that you actually really see what you want to see and the techniques which we probe our systems they are extremely invasive with respect to the reality that we are interested in. I have here a little list of what are the different techniques that are at the moment being used. You know them very well you have them also again here in the conference there's the very famous dip and pull technique which is simple to do but which has this disadvantage that your liquid is open to vacuum you do not know whether you have really a steady state under these conditions and of course you have the salt formation and all kinds of other things that are an overlay over the reaction you are interested in. Then of course there is the graphene-covered whole silicon nitrate membrane so with the holes where you look with your system through holes of the electrode that has the advantage, correct electrolyte and it is perfectly surface sensitive but this window technique is very critical and you have lots of beam damage areas and then locally you have an evaporation of the electrolytes through a little hole and then of course what you see in your spectrometer is not the one that is in your sample but there we are there as the hole and this is of course something different. Then what we also brought into operation is the PEM as a support where you put the catalyst on top of the PEM as in a real electrolyser and then you put this into vacuum and I show a little bit more in detail how that works that is easy to use but it has also the disadvantage that in reality there is no liquid water presence, only water vapor and water vapor is not exactly an electrolyte and then it was only necessary to put a lid on top of the whole system and when you do this then maybe you can approach the real system but this is a very delicate experiment and it is not easy to do. So this is the standard as I say early pioneering work from many people where you see the working electrode for example of iridium and you see an electron micrograph where you see the nanoparticles as before and you see the gaps in between this is what people call mud crack type of electrode this is perfectly fine but unfortunately we have to discover that there where we do actually our experiment at the interface there is no liquid water there is only water vapor and this is a serious disadvantage. The consequence also of that is that we find some transformation of material here for example also platinum and we found that the platinum was slightly oxidized it formed a surface oxide and at the time we thought surface oxide is then the true nature of our electrode and this was nicely reversible with potential but it turns out this is only the beginning of the transformation this is not the whole story. So just a short excursion to that as we started this extensively with platinum and platinum with of course reference material in electro catalysis I show you only one slide on that as a function of potential when you now apply a real liquid water layer on this the retox state of the electrode changes instantaneously and drastically and of course any post-mortem finds only platinum metal and confirms the noble character of the metal because unfortunately the thing is fully reversible only seen in under operation conditions. Now I summarize this here you see in this box B you see a potential change in the operando experiment in the box number A you see the response of the total oxygen that we measure and once these as the potential increases you come to OER conditions the amount of oxygen identifies enormously at the interface. So that means there's a lot of oxygen there and there's also water you have the impressions really compressed at the interface. Now when you see is this only does that something change in the electrolyte or is there also a change in the platinum metal? Then one sees there's a drastic change because at open circuit potential you see in the platinum for F it's platinum metal as you would expect everything fine then when you put it a couple of minutes under OER conditions then you get this green spectrum and this is a very indication from an hydrated platinum oxide that is not PTO2 this is PTOOH2 that is something different and then when you switch off the potential and go to open circuit potential as fast as you can measure a spectrum which is a couple of seconds this thing goes back to platinum metal and you see the blue spectrum is indistinguishable from the red spectrum so if you would take it out you would never find that there's platinum oxide you would say as you expect as a chemist platinum is noble so it doesn't change under reaction conditions but unfortunately it does. Now when we return now to our eridium then of course we know already there's a coexistence of metallic and oxidic eridium and then we found at the beginning okay we find in our operando experiments the same so we would be happy and say yes this Cauchel particle survives under reaction conditions but we ask ourselves of course after some time how correct was the assignment and where are the protons in the picture and then we identify an interesting species this is this oxygen minus one by now I do not need to introduce this anymore I think you have this all now on the screen but the interesting question that we asked ourselves is this really part of the reaction or is this a side shrug or is this maybe a salt formation so is this something that does happen but it's not related to the kinetics I only give you a short indication what you see if you take this this reaction environment in which you have only water vapor as an electrolyte then one sees indeed yeah there is an oxidation and one sees this coexistence of eridium 3 plus 4 plus and eridium metal that is fine you see also the massive change in the oxygen K H spectrum it gets much larger and it gets also a quite different pre-H feature that is because you see the oxygen minus one occurring and whenever this oxygen minus one occurs one sees also in the mass spectrometer that's oxygen formation so that is somehow related this we can easily verify you can also verify that that is reversible this is now a three electrode cell so you have now a real a proper potential scale not as we had only a two electrode cell at the beginning and then we find indeed yes this oxygen minus one this gives this broad small feature that you see in the data is dependent on the potential that we apply it's reversible if a switch of the potentially goes away immediately so we would say okay it is related to it fine but what we were not convinced about is the width of this line because this is a very broad and diffuse line and we were not sure whether this broadness and diffuseness is correct with what we would have actually expected because there are many oxygen species in the system so and this is makes it a bit awkward to actually analyze this so that's of course largest oxygen yes there's water as a reagent there is water as a solvation species which is not necessarily the same as liquid water there's of course OH as a ligand and there is OH as a reactant again these are two different species and there is also oxygen in this terminal form in various oxidation states so this oxyl species and that for example is of course as we all know not a constituent element of the iridium oxide structure so that requires the structural transformation that you get this actually in and then you have this pressure gap and then you ask yourself are these species dependent on the pressure gap or not always the contribution of the species dependent on the pressure gap that you have I illustrate this here with iridium oxide the middle plan panel is actually the oxygen kh spectrum simply of liquid water this is good to know this is our reference now if you take now this iridium oxide and measure it then you see is quite different from water you are okay that you would expect this is fine and it has a distinct very nice pre-aged feature that also has been discussed in the literature by now extensively but if you take this iridium oxide and put it now in contact with the electrolyte and measure it then this is the top spectrum then you see oh there is a significant change so it is not sufficient to say this is iridium oxide or iridium oxide rock side and you measure it in vacuum and then say you understand what are the oxides and species that are present now they do change as soon as they see liquid water and now here is just a re-analysis of what I have shown in the experimental flights and this colorful fit that you see they come from theory now we have now the chance to do theoretical calculations where we would expect the different species and then line up very nicely so we would say there's a very good agreement between say computational spectroscopy on one side and the experiments on the other side and that is of course absolutely necessary when you see this whole two of species that are present and unfortunately it is as complicated as it looks here the question that we have is is it now really necessary to put liquid water into the system and the answer is yes it is absolutely necessary and an approximation to liquid water as the electrolyte is giving you wrong answers or incomplete answers so what is the real active form of the iridium oxide the active form is of course not really iridium oxide but this iridium oxy hydroxide and in collaboration with the Max Planck Institute in Stuttgart and Bettina Lodz team we were able to make such material to measure its properties this is a very nice layered material and you can change the redox state by the same mechanism that I introduced to you with this hollandite by changing the interlayer hydrogen that you have there then you break up the interlayer order and you have essentially isolated layers of this material and they can then have any kind of oxidation state between 3 plus and 4 plus depending on the protonation or deprotonation reaction the spectroscopic response of that is very little you have to be careful in order to see this protonation effect in photo emission this is not easily detected when you look at the oxidation state change this is given essentially by this complicated issue with the satellite structure that I introduced but the electronic structure that comes from theory again is showing you big changes when you go from a package of layers to a single layer and when you go from a protonated to a deprotonated form protonated and deprotonated stands here for oxidized or reduced because when you take this is always meant by proton plus electron and proton plus electron in or out and one sees this changes from essentially a metal to a semiconductor backward and forward and the only way to discriminate this is the degree of protonation of this terminal oxygen and that fortunately is very well accessible by the oxygen K H spectrum that you see in this panel B and one sees just by watching it the spectrum you see already whether this is protonated or deprotonated now because we had all the computational spectroscopy support if we would have not had that it would be very difficult to analyze this data so now when we put now liquid water into the system then we can see what happened this is now a cartoon that shows in proximity what we do so we have a real electrolyte we have this iridium on the polymer then we put a graphene lead on top of it and this polymer electrolyte allows now to have confined layer of liquid water and this confined layer of liquid water is not only water but it contains all the ions of the electrolyte and we have looked and verified that all the ions that are in the electrolyte are permeable through the PM and really accumulate into this confined layer of water I have almost no time to go through that when you look into panel B of this slide this is the photo emission spectrum that you get from the system and what I only want to show you is this very tiny peak at around 535 electron volts this is the contribution of the confined liquid layer of water so this is a very small contribution to all the many oxygen species and unfortunately you would have said okay and more pressure helps more this is what you see in panel C that is not true because as I said you have to be very close in your experimental arrangement between the cell and the detector because you have the attenuation of the electron the photo electrons from the gas phase and if you increase the pressure in the gas phase that you get more liquid water then the attenuation of the electrons is very surpassing the effect that you have more water in the system and get a better electrochemical response so this is a highly to be optimized system and we found this back pressure of one to two millibars is the optimum where we have high sensitivity we have still liquid water and we can still detect proper spectra so does it really work? Yes it does work I have to hurry up a little bit it's interesting to see that such an electrode system really produces very small amounts of oxygen but we can very well detect it you see the potential in the middle you see the current response in green and the mass spectrometer response in red and you would say yes this really works this is now a working electrochemical system and we can assure that the system is actually doing what we want so we can study it carefully so this is what we do now very quickly is we see what is going to happen when we go from open circuit potential to OER and then you all know that as a redox wave before we come to this point this is at about 1.25 volts this is this yellow pattern that you find in our spectra and at the beginning everything is fine you have a transition from eridium metal to eridium oxide that happens around 0.8 or 0.9 volts and there is no special oxygen species everything is as you would expect it but when you come to this 1.25 volts then you see immediately that there is the occurrence of these two extra features that come from the deprotonated terminal oxygen species and this is the first indication that the active centers for and again without having the computational spectroscopy at hand we would not be able to assign this we would not know that the higher energy species is from new tools or from bridging oxygen that is protonated or deprotonated and that the 528 is the terminal oxyl species and what is also without computational spectroscopy not obvious to one that there is a huge change in the position of the lines if you protonate or deprotonate the hydrogen again this is a very large expected normally the spectroscopist would have said that doesn't make a big difference it does make a very big difference and again I praised this here without theory no way that we would have anticipated that so what happens during the reaction we do this now in O'Parando and as you have now seen the OK edge is the real thing to look at and what we need to study here is the question what happens with the OK edge when we really actually do this do the reaction so in the top trace you see the current that is flowing through the system as a function of potential on the x axis and then we do not typically spectroscopy but we just stand at one of the binding energies where we have either the protonated or the deprotonated forms in B and Z and then we follow what happens to the intensity of these lines when the current increases and you see very clearly that there's an agreement between these things so we have now we have a very good O'Parando experimental evidence that if the current is flowing then we see the deprotonation of this Mu1 and this Mu2 species and we see what we would have expected that the Mu2 species is first deprotonated but this is not very efficient and the Mu1 species is the one that really does it so that's indicated here now comes a long debate where which I have no time to go into about what is the oxidation state of this this is a complicated debate because there is a strong hybridization between iridium and oxygen that you know very well by now and if this would be a normal metal and not so heavy in the periodic table then of course this iridium-oxygen single bond would not exist but this would be an iridium-oxygen double bond which is iridule and there's titanule for example chromium, ferrules, the 3D elements all make these double bonds but iridium doesn't do this iridium actually just shares then the so-called whole state or the oxidation of the oxygen by forming these radical species the chemist has no problems with assigning this he calls it tautomary and so tautomatic forms are very well known in many parts of chemistry but of course this is something which is a little bit unusual because normally that doesn't exist in a material that you can fill into bottles this is only something that is happening during reactions but as I say chemists have no problems with this tautomerism but of course physicists might be a little bit surprised what that is so whole states and the distribution has been debated a lot and is possibly still debated here I think this is a this is a difficult thing if you really have to debate this very extensively because there's a strong hybridization and this is one of the key reasons why iridium is so good because of this hybridization and the hybridization of course changes when you do the oxidation reaction but it does not fall apart so now comes this word is this now electro catalysis or is this not and this is the question what comes first comes first the iridium oxygen species that is nucleophilic that is oxidizing or is it really necessary to apply the potential that the reaction at the water occurs and the iridium oxygen species is simply as I said is a spectator species and we think this is not the case what happens if you apply a potential to the system then of course you begin to oxidize this iridium oxy hydroxide or hollandite phase however you want to call it the sei is being oxidized and as you oxidize the sei you create active sites and these active sites become more dense in number as soon as you reach the potential that deprotonates the whole system because then you go to the higher oxidation state and as soon as you go to the higher oxidation state whatever that is iridium 4 plus delta then you reach the nucleophilicity that attacks actually water molecules so the potential induced oxygen evolution reaction occurs through a surface chemical process with a rate proportional to the number density of active sites created through the activation potential and that is exactly the same what happens if you take a chemo catalyst a normal heterogeneous catalyst and heat it this is at room temperature doesn't do anything because you know active sites and first you have to heat the whole system that the active sites are being formed and then you get a reaction so this is of course our standard theory how the reaction works and you have two to the chemist unknown things so what is the star actually and we have now a good indication that this is this deprotonated iridium OOH which has this this possibility to change the coordination between edge and corner sharing octahedral and what is O-Arts O-Arts is probably this also species minus one deprotonated you can do lots of experiments to verify again that this is directly related between the occurrence of this oxidized or deprotonated form is correlated to the OAR activity and this is not a spectator and this is not a poison it's really the system that actually does the reaction and here I just want to point out that this is an interesting methodology how one can directly verify this by doing this potential dynamic experiments where do these pulses and you can see how the spectroscopic response really mimics exactly the potential change that you actually apply to it so one sees there is a causality between these things and of course you can again apply theory to that and many thanks to this theory people who were able to do all this and one sees there's a very nice qualitative agreement one sees it of course first this mu two species so the the bridging oxygen is deprotonated that we would expect because it has a higher stability and then comes this tautomeric form that people call mu one oxygen and the mu one oxygen is the one that really does the breaking of the of the water molecules because that is the nucleophilic form now we bring this to a final picture and this is now it's very primitive but I think it shows exactly what is important in this reaction so we have learned we have an iridium oxospecies and this the oxidation state is somewhere between three and four and there are many reasons to to be careful to say what is the exact number because this is a core shell material there's always iridium metal present that makes an error in this and of course there is this question how what is the protonation or deprotonation form of this iridium oxospecies and that is so to speak our active sites then you do oxidation so you apply a potential and then you have a debate that is either iridium five plus oxodouble bond which doesn't exist with this heavy metals or it is this oxyl species and what one can of course say it is a mixture of the two so that means the oxidation the say the corrosive potential that you apply this 1.5 volts is actually shared by the two atoms so by the central atom and by the ligand and that is something that is untypical because normally if you put such an oxidizing potential to the system then of course the transition metal that takes the burden so that is being oxidized and the ligand stays always the same and here we have the situation that we have a so-called non-innocent ligand that means the ligand contributes to the redox property and not only the transition metal atom and that is something which is very special in this iridium oxosystem so this probably is our active species that we have from theory and then we do normal chemistry this is extremely corrosive the system it forms this peroxide and this hydroperoxo species is still quite corrosive that is also the reason why I have for example the deactivation of burning holes into organic membranes that you have in the system because there's always a little bit escape of this peroxide species but you want to get of course if possible as quickly as possible get rid of the whole system that happens by the last part of it but the critical step is indeed this red one is also oxyl species and that gives you an explanation why iridium is so good because it has this possibility to share the burden of this highly corrosive potential between the transition metal and the ligand itself so this is something that is not occurring very often in chemistry final slide what do we learn from all of this the chemical constitution of iridium dioxide is way more complex than a rutile or a hollandite so this approximation there's a crystal structure of iridium oxide and that is actually our catalyst this is an approximation that is good and necessary for modeling but the real system that is working is a little bit more complex because it has all these additional oxygen species and it needs dynamic structural flexibility so we recaptured the sequence of reactions of course very well without however showing the chemical realization of the catalyst what operant really contributes is now what is the atomistic realization of this active science the chemical dynamics of the iridium is induced by sharing of the corrosive power of the oxidation potential in the terminal iridium oxygen bonds that is the secret of the whole thing as good as a metal can do this innocent non-innocent ligand play as good actually is this as an electrocatalyst this is no structural element so the sharing of the of the corrosion potential is no structural element of iridium oxide as rutile nor of the iridium OOH or but only of the hollandite motive because you need this polyhedral exchange so the active state forms during the reaction and protonation regulates the current flow also that is an interesting feature you can make now a nice analogy to cobalt species because in cobalt species you have the similar thing or in the manganese cluster that does it in the green leaf that also has exactly the same properties it has also this possibility to do this polyhedral sharing and that is a critical issue here so the easy structure switch between layers of octahedral and molecular species of the iridium or to acid is the secret of the stable high performance operation and that is this is so easy for iridium because iridium in the language of chemistry is a very soft high and that means it's very large it has a diffuse electron gel and by changing the local coordination the energy of the system doesn't change very much if you go now to harder metals like ruthenium or to cobalt then you cannot switch so easily the polyhedral and then the disproportionation reaction disappears and then you have polymerization or depolymerization reactions and this is actually not what you want because the depolymerization of the electrode leads to dissolution or you get recrystallization into an insulating oxide with a wide band gap semiconductor that is also not what you want and iridium because of the softness doesn't do this because it does this deproportionation reaction or in other terms as we have tried to to illustrate this the unique property of the good catalyst is actually that you can really interpenetrate the states between the ligand and the metal when you do the potential drop and you get this corrosive potential then the system does not respond by changing its constitution but it really does change its hybridization and this hybridization just shares the burden so this yellow part that you see in the OER part of this this is the secret of it however you can make this possible then you get a good electro catalyst but unfortunately nature limits this quite a bit because as I said this 1.5 volt is extremely corrosive and there are not many chemical bonds that withstand 1.5 volts my final slide is I would like to thank you for your attention Albert Einstein said there is no fundamental law requiring simplicity in natural processes that means you were such little this was very complicated why do we make it can't it be simpler no it can't be simpler it is as complicated as it is and I found it appropriate to say with Leonardo Da Vinci knowledge is the daughter of experience so it is a good idea to do a real operand or experiment and then draw the conclusions what is the nature of the catalyst rather than to make a picture first and try then to verify that your picture is correct you better first look how complicated that is in reality thanks a lot for your kind attention and I stop sharing the screen thank you very much I clap for everybody I see Simone and Maria Lorenz clapping as well so now the the stage is open for questions so is anyone please type in the Q&A or raise your hand if you want to ask a question so I will start myself so so do we have to understand this is a truly dynamic state so these these octahedral change their their coordination during even at constant potential yes I think this is the case because this this possibility as I say allows to form this terminal oxal species and this terminal oxal species they can we can take this role as deprotonated in non-innocent ligands so they take part of the oxidation power is all minus one so is this in a sense liquid this this structure oh that is a that's a very good question I don't know I would say it is definitely not crystalline it is not in the translational it has no translational symmetry of course when you think liquid is something that is in time dynamical so it changes with time then one has to think about what is dynamics of course if there's no potential there then it probably recrystallizes and that is only reason why it deactivates but if you have a potential gradient then it it behaves like a liquid but this is not dynamical because dynamics is fluctuation without an external potential and I think whether this fluctuates without an external potential I'm not sure okay and just to be understood to be sure I understood correctly so even the crystalline form of the oxyhydroxide is not active or not as active as the as the it's only the defects you get always a minimum activity because there's nothing that is free of defects okay but in the ideal case also for example we we have done this years ago with Peter Straszler when you take iridium oxide then go through a calcination series and make it more and more crystalline then you find that the catalyst gets worse and worse so even your approximate is to a perfect crystalline activity or activity of crystalline iridium oxide is zero okay thank you so I see in the meantime people have started raising their hands so Axel Grosz would like to ask a question please yeah thank you very much Robert for this very nice talk I have a very general question with respect to the oxidation numbers because we have also dealt with oxidation numbers in electrode materials the very concept of oxidation numbers of course is a very well-funded concept in chemistry but as a quantum chemist I have problems with integer charge numbers so if you talk of non-integer oxidation states do you mean then a mixture of different parts with integer oxidation states or is the whole concept of non-integer or integer oxidation states should be revised that if I would this would be now a chemistry conference probably I would be shot down for what I'm saying now because this is a constant debate in between chemistry ever since chemistry has introduced this concept I would say I would follow you and say oxidation states are just an order in principle of chemistry and the physical meaning of oxidation state is rather limited so when you're as a spectroscopist when I try to measure this I have no oxidation state spectrometer that doesn't exist so you assign certain spectroscopic properties to formal oxidation states and if you are correct then you would also say formal oxidation states and formal oxidation states are integers of course because you saw not half electrons there are electrons but of course electrons as I said are shared between bonds what we call non-innocent liquid and at that moment the concept loses its meaning completely because if you're shared between two atoms then you have to find out how you divide it and then this is a problem of crystallography and whenever you do whatever you try to measure the electron density between atoms you always find finite electron density there this idea of an integer oxidation state is an idealization of chemistry it's a beautiful order in concept but it's not worth the debate yeah I mean when we want to calculate them of course we have to put a sphere somewhere and then there's where does the n-in or cation starts and the other one ends and there's no fixed rule for that so that's why we also can't give integer numbers at all yeah okay like crystallographers also do they do it the same way and of course every crystallographer has its preference is this the minimum or is it the half of the bond or whatever and then they get different oxidation states this is an ordering concept and not a physical reality okay thank you very much thank you there is now a question by Karina Faber so please you can unmute yourself yes hello so first thank you very much for this really nice presentation it was it's just a short question to if you could give the reference of the most relevant publications on that topic from your group so that we can read in detail if you go the last year's nature paper has all the references that we have ever published okay so from there you find them all okay thanks okay thank you now there is a question please also say your affiliation there is a question by Naveed Hagmoradi so please unmute yourself and say your affiliation thank you so much for the question can you hear me yes very well yeah thank you so much you used the operanto analysis to kind of analyze the performance of your electric catalyst you know your conventional electric catalyst and says that different oxidation states and composition maybe crystalline structures can give different performance can we use this operanto analysis in this a kind of special synthesis procedures to produce a highly active radium oxide instead of synthesis synthesizing it then after that analyzing which one is better yeah you raise a very good question this is what is the practical use of all of this and of course the practical use is you can now make synthetic approaches for example use alloys or put the iridium into some other oxidic matrix and use spectroscopy to look for the signature of this oxo species or oxyl species the ability of a system to get deprotonated this is the one that you need and we use this also as an indicator for successful synthesis strategies and what you also need to learn is from the operanto experiment the single atom iridium catalysts people sometimes try to advocate is most likely not going to work because you need this polyhedra change thing and if you have this polyhedra exchange experiment between different elements then one will find immediately as I said that the propensity to form a poly condensation or depolymer depolymerization reaction is higher than this redox change this proportionation reaction both can be measured and you can make one synthetic experiment then you can find does it have this property as I said you look at the oxygen k-H and then you can see whether you find the right features of that and if it doesn't have this feature then you don't need to continue that thank you so much I'm sorry I forgot to tell my application I am from Sabanc University from Turkey okay thank you so much thank you very much so now Simone Piccinini Simone Piccinini from Trieste the question I have is regarding this layer of the hollandite molecular units that we have at the solid liquid interface so do we know whether this is metallic or not and if it is not metallic does it contribute significantly to the upper potential that you measure of course this is now an interpretation that is very difficult to answer because the metallicity of the system depends on the degree of deprotonation from the calculations you have seen that the gap opens and closes whether you are protonated or deprotonated and of course now it is an interesting question is this always deprotonated or is this dynamically and this is only a feeling I would say it is it tries not to be deprotonated that means it is mostly metallic but it becomes it becomes a semiconductor at the moment when it performs the reaction and then it falls back into a metallic state this is my interpretation of it but this is Schleuchel's meaning there's no evidence for that the point is this question whether this is metallic or non-metallic really depends on with what time resolution do you look at it it's given by the deprotonation okay and I would guess in average it is metallic because this is much lower in energy if it is then it is deprotonated but as it is fluctuating sometimes it's terminal and then it deprotonates because there's the potential then it just the reaction falls back into the metallic state and this is the resting state of the system that is would be my answer and can we assign a thickness to this let's call it amorphous layer yeah about two unicell thicknesses half a nanometer or a nanometer don't make it too accurate but this is it is significantly smaller than the particle size of the iridium system that you have it's really this is a core shell system thanks thank you so yeah yes please yes this is really a challenge for calculation I can see I have a question is any of this visible in photometry or in impedance spectroscopy yeah one can do that so in impedance spectroscopy I think one can see this but of course one has to be aware that there are so many parts also contributing to that and in time resolved experiments at least we could not see this because the time resolution is given essentially by the exchange of the electrolyte and the oxygen bubbles and all these things and if you run this in the limit of very low conversion then you would not have all these transport issues then you do not see this very well you really need high conversion that means I do not give you the chance that this is time this time resolution is probably not a good way of looking at it but impedance spectroscopy does show such phenomena and should show this but I'm not an expert good enough to say what do we really need to do to really decipher this because it vaguely remember reminded me of what Michael Lyons always has been saying in Ireland right and that was all based on impedance spectroscopy yeah the problem impedance spectroscopy is of course the model so how do you get all the components out of it and certainly impossible for a theorist also for an experimentalist this is really difficult thank you thank you so last chance if you have a question or a comment so I see no further questions so let's thank Robert Schlögel again for the wonderful talk and thank you very much and we reconvene at 2 p.m. three at the time so thank you bye bye and thanks so much bye bye thank you thank you