 Good morning. Good afternoon and good evening, depending on where you are in the world. Welcome to the Storage Act International Symposium. I'm Yi Chui, a professor in material science engineering at Stanford University. I'd probably like to welcome you to the second event of Storage Act Symposium last launch. Our first event successfully with two distinguished speakers, Professor Stan Witingham and Professor Jun Liu. This week we will continue this exciting chat with two leading scientists in the world on energy storage to join us to give the two seminars. Professor Dr. Kler Amin and Professor Peter Blues. Kler Amin is a distinguished scientist in Agar National Lab. He just won so many awards I don't really need to repeat, but I would like to mention two of them. He's a fellow of Electrochemical Society. Recently, he won the very prestigious global energy prize for his contribution and lithium ion batteries. And Peter Blues is a professor and University of Oxford. He also does not need more introduction. Just mentioned a few things, a few honors about him. He is a fellow or lawyer society as well as a physical secretary. He is also the founder and director of Faraday Institute in UK. As you know, Faraday Institute is perhaps the most important energy storage program in UK. With that, I would like to welcome if Peter you can turn on your camera and let's switch to you and please share your screen. So thanks very much. It's good to thank you for the introduction. It's great to take part in this process. I think it's a super initiative that you've taken here in these challenging times for the world when we are a lot of us are locked down because of the COVID-19 experience. It's great to be able to connect with people across the globe and still talk about science that excites us all. So grateful for you doing this. So I want to talk a little bit about lithium cathodes, picking up on one of the themes that Carl mentioned. I should say the picture you see in front of you is an image of the skyline of Oxford. It doesn't have my labs in any of those nice historic buildings. My labs are in rather newer buildings but maybe a little bit more functional but not as attractive. So the title of the talk is Walking Back to Happiness, Oxygen Redox Lithium Battery Cathodes. And what's behind the talk is that the title is that I want to look at these lithium rich cathodes which exploit redox reactions on the oxygen as well as the transition metal iron. So in a conventional cathode and we'll touch on this in the next slide because I'm conscious we have a wide spectrum of expertise. I'm sure in the audience. In a conventional cathode and lithium ion battery, one exploits redox reactions on the transition metal to store charge. But to boost that charge storage and one of the biggest challenges in lithium batteries, again picking up on a nice introduction Carl gave is if to increase energy density. And one of the biggest challenges in doing that actually relates to the cathode. So boosting the charge storage on the cathode is one of our big problems and going beyond transition metal redox is one of the potential solutions to do so. The Walking Back to Happiness really refers to the fact that right back about 20, 25 years ago when people like Mike Thackeray and Jeff Dan first reported these lithium rich materials. They showed that there was oxygen loss and concluded that that was the main mechanism for the extra capacity on charge. It turns out that they were perhaps more right than we thought for quite a number of years and and molecular oxygen seems to be at the heart of most of what's happening in these in these materials both reversible and irreversible. So just so everyone's on the same page and I'll go through this very quickly because after Carl's talk it's maybe a little bit unnecessary but just to remind you of how a lithium ion battery functions but particularly to focus on this issue of the charge storage mechanism and the cathode. What you see in front of you is schematic of, I guess a first generation lithium ion cell with lithium cobalt oxide and of course on charging lithium ions removed from the lithium cobalt oxide particles, and this involves oxidation of cobalt three plus to cobalt four plus If you are more expert, you'll know that that's a bit of a simplification, even the case of lithium cobalt oxide but if I just take this as a simple example to get the concept over that conventional lithium ion cathodes the type we still use today, essentially involve charge compensation for loss of lithium on charge by oxidation of the transition metal ions. And then on discharge lithium ions return from the from the graphite back into the lithium cobalt oxide particles and one reduces cobalt four plus back to three plus. So the point is we're storing the charge on the transition metal ions. If we want to boost the energy density the energy storage and lithium ion batteries we need to attack this problem on the cathode and figure out how we can store more charge per unit mass and volume in that cathode. So what you see now is a plot of the potential versus the specific capacity in milliamp hours per gram for lithium cobalt oxide and we're going to look at this was the evolution of intercalation cathodes for lithium ion batteries over the last 20 to 30 years so Li coo two is the sort of first generation cathode material and calls talk focused a lot on the sort of nmc type materials. So mixed metal nickel cobalt and manganese on the transition metal sites of these layered compounds, which boosts capacity at around the same potential. And then to increase energy storage in another way you can go to high voltages like the nickel manganese spinel or the lithium cobalt phosphate. That's very challenging from the electrolyte stability point of view. So you can consider whether you can invoke two electrons per transition metal such as in these lithium vanadal phosphates, although they do go over a very wide voltage range. And one of the other options and light blue there are these lithium rich cathode materials which are of course are based on the lithium cobalt oxide the nmc type materials, but are increasing the capacity by, as you see on the right hand side redox reactions on the oxygen to store electrons as well as on the transition metals and the benefit here is not only that you increase capacity but you do so at a relatively high potential of around four and a half volts. So that means you increase energy as well as not just by increasing the number of electrons stored but also the energy at which they are stored. So these materials have a lot of challenges and I think some of those are captured on the, on the next slide here. So on the left hand side what you see is the representation of the crystal structure of one of these classic lithium rich materials this one with the formula that you see here the so called lithium rich nmc and this is the one I'm going to focus most of the talk on. There's lithium ions as one does in lithium cobalt oxide, and then we have transition metal layers with the nickel manganese cobalt and lithium present in those layers. And then on the right you see here the first cycle the voltage versus composition or capacity on the first cycle. So charging here and then discharging along here and immediately you see one of the problems of this material is this voltage hysteresis you first charge this material you oxidize the transition metal ions the nickel and the cobalt because the manganese is already fully oxidized and this is just conventional intercalation chemistry nothing unusual here in this green region. But then when you go beyond that into the plateau, you're now oxidizing the oxide ions. The decision is happening at four and a half volts but unfortunately the reduction process is occurring one volt less 3.5 volts. And so immediately you're losing energy storage energy density if you have to operate along this line. You'd really like to go back and forth along this one here. So the first cycle voltage hysteresis is one of the problems. You can see that there's a lot of voltage on cycling. It's represented by these series of black curbs you see here. You can see that the hysteresis is much less but there is still some hysteresis. And then there's this voltage fade problem you see the shift in the positions of these lines here which represent the problem of a creep in the voltage a bit of voltage fade on cycling. One of the problems that one has to face and overcome with these lithium rich materials. The other one which we'll see in the next slide is of course that as you oxidize the oxide ions there is indeed oxygen loss from the materials at the surface. You'll find out that of course this significant change in voltage on first cycle really signals the fact that there is a substantial and important structural change taking place in the material and understanding this hysteresis is in part about understanding that structural change. And of course the challenges can we preserve this high voltage and high capacity and retain high energy. But there is oxygen loss. It's not the dominant process in these materials as was originally thought, but it is still present and we can see that in these in this mass spec experiment here so we take a cell we charge it and we pass an argon and carrier gas through it into a mass spectrometer. And then we look at the gases evolved. And this is the load curve charge and discharge unfolded this time for NMC for lithium rich NMC so we're charging here, just charging here, and we're using 18 labeled materials so that any species containing oxygen that we identify, we know that these oxygen species derive from the cathode and not from just decomposition of the electrolyte. So if you look at the oxygen gas release first, you see oh to gas is released towards the end of the charging process, and it is substantially in oxygen with 18. So it is coming from the cathode material. And then if we look at the CO2 we see that there's carbon dioxide released across the plateau, as well as just beyond the plateau. Now I think some recent work. Plus, but by others such as uber gas tiger has shown very elegantly that in fact the CO2 is also derived from dioxygen released at the surface. But it's singlet oxygen so dioxygen can come in comes in two forms the triplet form which is the stable one we're used to, but also the singlet excited state form, and it's particularly reactive you get a small amount of that in the oxygen. It can attack the electrolyte decompose the organic carbonate electrolyte and give rise to CO2 evolution. So basically all of this is about largely about oh oh to loss from the surface. But the key take home message from these studies is that the loss of dioxygen far from being the dominant mechanism by which these materials can charge is actually the sideshow it's the it's the minor problem. It's still an important problem and you have to mitigate it and you have to stop oxygen loss for all the practical reasons again some of them that call mentioned, but in terms of understanding where the real fundamental origin of where this extra capacity comes from when you start oxidizing oxide ions. It's not the main event. It cannot explain the charging plateau in the subsequent discharge etc. And that significant effort by by some great efforts by a number of groups around the world and it still continues today I'm going to show you some of our recent results. One of the techniques that has been particularly valuable at understanding the nature of oxidized oxide ions when you charge these materials is ricks resin resin and elastic x-ray spectroscopy. So ricks basic ricks experiment shown here the incident radiation that you impinge on the material causes excitation in this case from the oxygen one s states. So it's K edge ricks to high energy states here empty states here, and then of course the electrons fall back from these oxygen to p states into the option one s. This is a different technique and you can look at the emission at various excitation energies which is what one does ricks. So to show you the results so here's our load curve again here's our first cycle charge and discharge and what to understand the processes that are occurring in here. So the first thing we do is take some Rick a Rick spectrum at this point so we've charged through the normal region. This is the transition metal oxidation region. We want to understand which is this oxygen or oxidation of the oxide region, and this is the Rick spectrum before we've embarked on our transition across the plateau where the oxide ions are being oxidized. We then collect a spectrum in the at the end of the charge it looks like this so there are changes here. And in particular in this region, if we then look at the discharge again so back at the end of the cycle, we see a Rick spectrum ricks emission spectrum which is very similar to the to the pristine material. And so in this region here and here lies a clue to what is happening in the bulk of the material when you oxidize the oxide. So this is not oxygen loss from the surface electrochemically because ricks is a technique carried out in high vacuum and can't detect the oxygen that we were commonly used to thinking about the di oxygen which is the di oxygen lost from these materials at the surface. So this is probing the nature of the oxidized oxide ions in the bulk. And it turns out if you look at a spectrum of molecular or to a Rick spectrum molecular or to it essentially has the same sort of features as the spectrum that you see here at the top of charge. So the long story short is that this is indicating that we seem to have molecular oxygen actually trapped in the material at the end of charge, and that that molecular oxygen is reduced again back to oxide ions at the end of discharge. So just to summarize, oxide ions are oxidized to all two minus in charge the oxide, the molecular oxygen is trapped in the lattice and then reduced back to two minus on discharge. Now of course if you're going to generate molecular oxygen, and it's going to be trapped in the solid. So that's not going to be possible without some structural reorganization. And so that's very much in concert with the, the change in the load curve that we do, we do see so here's the show you some structural data now to indicate the structural changes that we observe in this material. So here's our our crystal structure again, looking along the layers of this nmc lithium rich nmc. And then this is the plan view we're looking down on to our transition metal there, where you see the cobalt in blue the manganese and pink, the nickel and blue spheres here. What you see on the right is the powder diffraction pattern with the highlighted region corresponding to the honeycomb superstructure. So this is the nice honeycomb arrangement that you see of metal ions in the transition metal layers. This is the pristine material before we we charge. So if we now look at the data from scanning transmission electron microscopy, the, the image you see on the left is again looking along the layers so looking along in this direction, 010 direction, and these sort of dumbbell features are classic image that you expect from the honeycomb arrangement this is the what the honeycomb arrangement looks like in stem looking along the layers. What you see at the end of charge, if you look along these layers is that you've lost that variation and contrast that you see here. And that's very much indicative of the fact that the transition metals are no longer in a honeycomb arrangement they've disordered, but largely within the layers. Okay, so the white represents the heavy scatterers the transition metals very little between the layers, they're staying in the layers primarily, but they're disordering intra layer disorder. And you can see if you look at the powdered fraction pattern that these peaks associated with the superstructure. As you go from the beginning of the plateau to the fully charged region you lose those superstructure peaks that's in keeping with losing the honeycomb, and they don't return on discharge. So there's more evidence for the change in the structure on charge and discharge you can see here. So this is the lithium six NMR. And this represents the city center of gravity of the lithium peaks in the spectrum of the pristine material. And when you've completed your cycle back to the discharge state, you see that the lithium do not return to the same environment. That's actually on average, less than interacting with the manganese than before so they're more they're more diamagnetic environments than they are here. And this again is basically a signature for the fact that they're moving back into positions that are different from the pristine material in this disordered structure. So all of this is basically telling us we lose the honeycomb structure largely through in plain disorder there is some out of plain disorder but it's only a few percent. And the lithium ions are returning on discharge to different sites we're not we're not reverting to this ordered structure that you see here on discharge. What is happening throughout this cycle if we put the evidence for dioxygen formation together with the structural arrangements and indeed a whole lot of DFT that we've also carried out that I don't have time to go into in detail. So here's the transition metal layer in our pristine material again with the nice honeycomb arrangement you see here of the transition metal ions I've just used one color for all the transition metal ions to make it simpler here. The lithium's in in light blue. So when you charge up the material you take lithium out of the material you remove some of the lithium. As you reach the beginning of the plateau, but not all, and this is charged compensated of course by oxidation of the transition metals. We then charge across the plateau, we remove the rest of the lithium or most of it. And now we've oxidized the oxide ions. You see those the red spheres of the oxide ions so you see them flashing here representing the oxidation. But when we do that of course we we can we auction manganese bond as we oxidize oxide ions. And that promotes this disorder of the transition metal so these two black arrows just represent all that has to happen is one manganese here undergoes one hop to a vacant lithium site. Another one here to the next vacant lithium site and you've created this small vacancy cluster. You now have these two sort of orphaned oxygens here, which can form molecular oxygen oxygen bonds with these oxygens here to form dioxygen trapped in these vacancy clusters and of course strong driving force for this structural reorganization is that by forming these strong covalent oxygen oxygen bonds that helps to drive stabilization. You don't really want to have oxidized oxide ions hanging around. There are two common species of oxygen in the universe molecular oxygen and then in solid state oxide ions and so it's not unreasonable that you move from a stable oxide ion when you try to oxidize it to something that is molecular oxygen if it can do so. And then on discharge you reduce these dioxgens back to oxide ions. The lithium ions come back into the structure, but they don't go into the same position. We now go into these filling up the sites and these cationic vacancy clusters that you see here so you can contrast this the pristine material with the discharge material, a different arrangement of lithium and crucially as we'll see in the next slide, a different arrangement of lithium around the oxide ions that have reformed here. So lithium returns to different sites, the structure on discharge reductions different from the structural change on charge and that's in keeping with the voltage hysteresis that you see. So to get a sense of the main energetic driving force for these changes. We can look at this diagram of the of the oxygen to P states and how they, they, they, they change. So here's the an oxide ion and our pristine material with transition metal there above it with two manganese I've just used manganese for the transition metal ions to keep it simple and lithium up here and then free and this would be a sort of valence band of oxygen here representing the oxygen valence states which are formed by interaction with the surrounding metal ions. Now if we pick out these up this oxygen to P orbital in particular, it's interacting with lithium ions, this is an ionic interaction, it's a very weak interaction and so these oxygen to P states will tend to be relatively high in energy because covalency drags them down there's very little covalency here so they will tend to sit at the top of the, the top of the oxygen valence band. And these are the ones of course that are preferentially oxidized when you charge. If you contrast this with where we end up at the end of discharge which are oxygens surrounded by lithium entirely these are the oxygens that reform in that in that vacancy cluster now they're all surrounded by lithium's when the lithium's go back in. So these two P states are now surrounded by lithium, all of them are very weak interactions you have this sort of orphaned oxide ion if you like. And since all of them are in this weak interacting state, these oxygen to P states will tend to be higher in energy. Higher energy means more voltage. And that's part of what gives rise to this lower voltage at the on discharge. I mean, you must remember that these very isolated molecular orbital pictures are not really a true representation of a solid since you never get isolated orbitals like this. There will always be a degree of interaction hybridization so it's the net effect of the interaction of these option to P states with their environment that is important and shifts this this up. So what happens on the second cycle, where we have smaller hysteresis, as you see here. So on the second cycle you oxidize, you go back to molecular oxygen and we can see this in the ricks reforming molecular oxygen. But of course you don't have to reorganize the structure to do so because these vacancies are here. When you drop these lithium ions out, we form the structure and when you discharge again, you will come back to the same basic structure. So this is in keeping with the fact that you come back to the same energetic point here that you started from on the second cycle which is different from the first cycle. There's no difference in the endpoint from the beginning here and the second cycle there is in the first cycle. There will still be some structural changes because going from this to accommodate molecular oxygen will be some structural strain involved. So you would expect some small hysteresis but it's in keeping with this being less than the large one volt change here. So more reversible on the second cycle but it's still you see molecular oxygen formation and then it's reduction back to oxide ions. So this is the mechanism by which oxygen redox happens, the reversible part of the oxygen redox not the oxygen loss. How can we think about maybe mitigating this because we would like to stay on that voltage plateau? Well you can get a bit of a clue by looking at these materials. Now these are sodium intercalation compounds, they're layered compounds, a different structure. Let me just show you the structure here. This time because sodium is in the alkymetal layers, sodium prefers trigonal prismatic coordination and that changes the stacking sequence here but otherwise it's very similar to the materials we've just been looking at. But the interesting thing here is these two materials, very similar composition, same crystal structure, very different electrochemistry. So once as voltage hysteresis, the 0.75 material, this one exhibits what we really like to have would be going back and forth pretty much along this plateau. This is the 0.6 composition. So similar crystal structures, same crystal structure, very similar compositions, very different electrochemistry. So what is different about these two materials? The difference actually turns out to be the superstructure. So here's the superstructure. This is a powder diffraction pattern for the 0.75 material. And this is the transition metal layer again. And you see again this honeycomb superstructure. When we look at the 0.6 material, the one that shows the reversibility that we like in voltage, here are the superstructure peaks. It's a different superstructure. We've called it a ribbon superstructure. We like the ribbons here. So not the honeycomb structure that we're used to. The honeycomb structure is almost ubiquitous and lithium rich compounds that of course again all show this first cycle of voltage hysteresis. This is a much less common superstructure. But this seems to be a major reason for the ability to retain the crystal structure to not undergo those changes that allow molecular oxygen to form. And you can see that here. Here's the stem images again. This is the pristine material. It's a different arrangement here, not the dumbbell because it's not honeycomb. On charge, you see the retention of the structure and it's the same on discharge. You can see in the PXRD, the superstructure peaks here. You see these in these grade regions reappear on discharge. There's the pristine material. They're still there on discharge. And here's the lithium NMR, pristine and at the end of discharge, very much the same position for the lithium resonances. Everything is structurally stable as one would expect with the voltage retention. And it's the ability to retain the ribbon structure that is responsible for maintaining the high voltage cycling. And so if you look at the another type of spectroscopy, actually absorption spectroscopy that you see here. It's another way of probing the oxygen states, the OK edge spectroscopy. Here's the spectrum for the pristine and black, the charge material in red and the discharged in blue. And these states here that are represented by this change in the absorption spectrum at 531 EV, these are the states that we actually saw in the ricks. These are the molecular oxygen states. If you look at the ribbon structure, this is the honeycomb structure. If you look at the ribbon structure. There are still some of those molecular oxygen states present because as you will have seen, we don't get complete reversibility across the whole of the load curve towards the latter part of discharge, the voltage drops again. But the important point is you see this new signature in the XAS, which is not present here and is actually indicative of the creation of true whole states on oxide ions. So this is, this is the oxygen redox we want to see if we want to get good reversibility of the potential along that first pattern. So I think I should bring my talk to an end at this point here. As I say with DFT a lot of other data and I could talk more about the sodium compounds as well and show you more evidence for the honeycomb changing in the case of the 0.75. I think that allows me to give the main messages that it appears that the main mechanism that underpins the oxygen redox, at least the reversible oxygen redox chemistry, at least in these 3D compounds which we focused on is this formation of molecular oxygen on charge. And then it's reduction to oxide ions on discharge. And this, of course, signposts some possible directions that one might think about in terms of trying to mitigate that, one of which I've shown you in terms of the superstructure. Not to say by any means that that is solved. The problem of voltages to resources solve that's not the point. But but I think it does give us at least one way of thinking about the challenges that that presents. And with that I want to thank all the people who have done all the hard work. Thank you again and the team Stanford for the invitation to do this and of course all of you across the world for for listening. Thanks very much. Thank you very much Peter. Thank you for the deep dive into the oxygen redox and in the castle. A lot of good insight right there. And also many questions from the audience. Let me start from the first one. One person asking you to define the oxygen redox a little bit, you know, what exactly is the oxygen redox and this content right there. So, yeah. Okay, so if you take a conventional transition metal iron redox like nickel to plus, you move an electron from nickel to boss you form nickel three plus. It's of an oxide iron. It's very much the same. You've got an oxide iron with a completely filled out a shell. You take an electron from the oxide iron. If you took one electron per oxide it would be normally a minus species that you would form of course again in a solid you can't really speak of that level of localization but that's what we mean by oxygen oxidation. The reduction part is just putting those electrons back on when you discharge so just as we can think think of charging and discharging a conventional cathode as involving removing electrons from the transition metal lines and bringing them back on oxidation and reduction. It's the same process for the oxide. It's just that up until maybe a few years ago, it was widely believed that that couldn't be done with oxides with sulfides. It's quite common. You can go from s two minus with two minus charge in each sulfur to an s two. That's two sulfur atoms with an overall two minus. That's pretty common and well known. But the oxide is somewhat different. So I guess Peter maybe from this person. I guess the thinking is transition metal and oxygen, the P orbital always have certain degree of hybridization happened right there. I mean, it's probably the percentage issue when you do transition metal redox right there you are really affecting very little, you know you utilize very little contribution from oxygen P orbitals then once you go high and higher wattage, it will increase more and more of the it becomes become mainly oxygen less on transition So of course, there's no such thing as a pure D orbital or a pure oxygen to P orbital in these systems. Absolutely right. I mean, there is always a mixing and some kind of degree of covalency but it's true in the 3D systems in particular, which are largely because the D orbitals are concentrated and you have a high so called onsite Hubbard u parameter you have a big energy difference in the different oxidation states. And so, in a situation like that you have much more localization you do in the heavy transition metals. And so it's, it's, you know, it's to a first approximation reasonable to talk about, at least in some of the transition metal lines, you know, redox processes happening mainly on the transition metal line. And then when you get down into the deeper oxygen states processes that are mainly on the oxygen. So, yeah, there'll still be a certain degree of transition metal character in these oxygen redox process as I mentioned some cases I would say if I remember from the DFT it's maybe about 20%. But you can still think of this as extensively an oxygen process. And I moved from that from the load curve as I showed, you know, you, the behavior changes when you've you've exhausted all the transition metal oxidation. And you move into the other one so there is a difference it's not a continuum process. So I think you can think about transition metal and oxygen redox. Yeah, okay, very good. Peter, our next question. When you're describing this from the audience, I just replace it. You start to charge. Once you charge these manganese move to the neighboring site. And the question is, and on the transition metal layer, you know, these manganese can move in plane inside the layer can also move out of the plane going down to the lithium layer. So why not lithium layer. I mean, they, there is some, and we've measured it, and it's, as I said, about 6% of the transition metal ions do drop into the to the layer below. I guess there are a number of potential reasons why it appears to be dominantly occurring in the in the transition metal layer. One of them is of course that that they this if you like the size of the sites in the transition metal layer are already in optimal for the manganese to migrate into, you know, because if you look at them that the lithium sites in in the transition metal layers, although we describe them all as octahedral the dimensions are quite are somewhat different. Right. And so, and so it's, it's, it's energetically more favorable I think for the for the manganese to move into neighboring site in the transition metal layer that it is to move into the into the to the layers below. I mean we have explored both of these things. And of course experimentally as you saw from the stem images. You can see that there's very little metal scattering from the alcohol metal layers. So if you potentially it appears that's what's happening in terms of interpreting why it happens. I think that would be one of the reasons it's one, one hop, if you like, it's not very different to go into the out the metal layers to be to be fair the distances are not greater. But I suspect it's the matching of the sizes that maybe makes it easier for that to happen. Yeah, good. Next one. What's the difference in your rigs the resonant and elastic x-ray technique right what do you expect what's the difference. If the oxygen O2 species, they could be charged per oxo or silver oxo like rather than neutral oxygen so what will be the difference in the spectrum. That's a very good point and I probably should have put it into the Rick spectrum so apologies for that but it the peroxo species produce a Rick signature that's that's very clearly different from the from the molecular oxygen spectrum that you show. So that and I don't know if we can go back to it but if we can't don't worry about it but but I'll actually let me maybe if I've still got control of the screen I can quickly. So that's great. So let me just quickly go back to the relevant spectrum. These peaks you see here are actually vibrational states for molecular oxygen. So this is the elastic peak of ricks and this is basically the set of vibrational states for oxygen and they match very, very closely the spacings that for molecular oxygen. Now if you have a peroxo species, it will appear in a different place in the in the in the Rick spectrum and really the reason we've identified this is that this is high resolution ricks. Most ricks is is doesn't offer you the ability to resolve what would just look like an envelope in here but it's the ability to resolve these and identify these spacings as fitting the vibrational spectrum of molecular oxygen, which helps to tie this down. But the real short answer to your question is the peroxo species appear quite differently in the Rick spectrum. So there's no, it's not like you're just looking at a, you know, a small and somewhat ambiguous difference. Okay. So, so Peter next one from the audience. The oxygen loss of lithium rich cathode occurs at a wattage or 4.8 watt. But the amount of oxygen loss is also very limited from the dams result DMS. His or her question is how much oxygen loss in total. That caused the wattage fade during cycling. So this is oxygen gas loss these person indicate. Yeah. Yeah. This depends on for different materials depends actually on the the coatings of the material. So one of the efforts to make these materials useful is that a number of groups done some very nice work applying coatings and surface modifications to reduce the oxygen loss. And you can get up to, you know, 90% capacity retention on the first cycle. But in the case of the data you saw there. I think it's about 30 milliamp hours per gram or something even less than that I think it's a it's about 10% of the oxygen redox capacity is associated with the actual oxygen loss at the surface the irreversible oxygen loss. And the rest of it is all associated with the ground called the reversible oxygen redox that's that's related to the oxygen trapped in the solid. Yeah, you know, Peter very early on you started doing lithium oxygen, you know, and really motivate many people to go in and Cal you also have very exciting work doing oxygen chemistry, your nature paper I think. So what's your, you guys comment on the future of lithium air and in general the metal air batteries. And you can open up as far as you want and make comments. Yeah. Well, it's a big, it's a that's a big question. We could, I'm sure Kelly and I could give you an hour of all this stuff. Two minutes each and most okay. I wasn't I wasn't warming up to take the rest of the 10 minutes. I think the, I suppose the first thing to say is that some, you know, despite everything that call and I have talked about this afternoon in terms of the, what I would call the conventional type of with your mind battery and even go into lithium metal anodes and things. I think that's what's going to need energy storage that goes beyond what we're ever going to get from this technology. And if we've learned anything with it with maybe with the COVID pandemic in most countries is hey wait a minute guys if you just ignore the possible dangers in the future. They'll come at you much more quickly than you expect and you're not ready to deal with them. So, I think it's wise to plan for a little bit longer term than next year. And so I think the, what lithium air lithium air battery can do a recent principle is give you a specific energy in particular at a whole systems level that will be very, very hard and probably impossible to realize with the lithium ion. I mean, if we ran some numbers recently actually using the modifications of the argon software. And, you know, you can show that with the knowledge that's been developed by people like Carl and others including some of the work we've done over the last few years, you can predict that you'll get about 600 watt hours per kilogram from a lithium air battery. That's at the full system level taking account of the balance of plan all they're handling etc. And one of the reasons for that is we now know that you can work with concentrations of one molar water in the electrolyte. You know, you don't have to get it dry to one ppm. So that just illustrates two things and illustrates that there's been advances in understanding the lithium oxygen chemistry, which has actually removed or reduced some of the challenges that people thought would be really very hard to overcome. It's been shown you can cycle these things at about one or two milliamps per square centimeter, that's rate at several milliamp hours per square centimeter capacity. Whereas, you know, 10 years ago we thought it would be microamps and nothing like that. The, the, the use of redox mediators has been able to bring the voltage gap between charge and discharge down. So there's a lot of advances that have come through from the basic science. And there's still a lot of problems, right? I mean, one of them is the stability issue of electrolytes, probably the biggest remaining problem. But again, through the recognition of single oxygen formation on charging that I touched on in fact in terms of the lithium rich cathodes. So that's now our major factor in the decomposition of the electrolytes. So if one could remove that problem, and it's largely removed by reducing the charging voltage, you could, you could perhaps reduce the single oxygen generation and much improve the degradation. So we just run through a few things there just to give flavor of some of the things that have happened that if addressing some of the challenges that people recognized, but at the same time, recognizing that there's still a long way to go. There are still important problems to be solved. But I don't think we should. I think there is still potential in this technology, difficult and challenging though it is. The goal of being able to achieve something around 600 watt hours per kilogram at the whole systems level. That's to be compared with a full battery pack of a lithium ion could be so important in some areas, maybe even an aviation that it is, it is worth maintaining an important effort in those kind of advanced challenging and far reaching potential technologies. Well, I think some rise pretty much everything's you know what one thing is really I agree with him when you talk about the energy density if you use it like oxygen tank whatever is going to be a significant drop obviously and that defeat the purpose. What we're doing at our gun is a little different now. It's not really a little mayor is trying to to stabilize the lithium superoxide, the crystal in phase, which is, you know, so to prevent this proportionation. And we put that concept working at very hard. If we can do that, then we have a closed system, basically like lithium ion. So you start with an electrode that cathode which is a lithium superoxide and a little metal and you shuttle it in between the two. Now this material is more conductive so the association of the lithium oxygen bond is much easier. And so if we can do that with the right electrolyte. I think there is an opportunity to have a closed system where you can use the same conventional process that the lithium ion is using, assuming that within the next few years, lithium metal stabilization could be achieved. Then maybe there is an opportunity, but otherwise, if you, if you talk about it, you may have like what Peter mentioned, there is a lot of challenges and the gaining energies will be also not very significant. And so the bottom line is that I think we should continue is a challenging, you know, for scientific curiosity is big, big challenge. So it's good to continue. And then if we can find a way to know if you look, there is several lithium, you know, you move from lithium superoxide to peroxide, reversibly, if you can do that, that's great. Then you can add one more lithium and increase further the energy, which is much more challenging. So that's the, at least that's what we're doing. We have a big effort in this area now. It's most of our focus in to enable the closed system, which very likely will be less challenging than lithium, because lithium as you have air or oxygen coming in, you have potential crossover that can affect your anode. You have many, many other, many other tanks that Peter mentioned that make it very, very challenging, but it's not going to be happening within the next 10 years. This may take longer. Thank you, Cal. So we have a few minutes left. Let me ask both of you one last question, because just many young faculties and young students, I think in the audience right there. Both of you are looking at your career. I may ask similar question and last time to Stan and Jin. Well, Cal, I look at you. You work on many topics and each topic of area, you always have come up with great ideas, you know, very impressed, very broad range of topic. Peter also, you as well, you know, a cathode, whether it's oxygen chemistry, right, this is a really big variety of topics. Anything you want to share with young students and young faculties, they're starting out their career, only going to a career for a short, short time, not very long yet. The lessons learned, you can share with them to build their career and energy storage space. Well, maybe Peter can go first. He's going to have more wisdom, I guess. I can second that. I don't think so. Go on. You go. I took the last one first. You can go for this one. All right. Well, you know, you know, one things that, at least in my case, first of all, the work we're doing at ARGA is not just myself. We have great team, my folks at work with me and my collaborators. So it's a group, an effort that is much more broader than just myself. So that's maybe one thing. Collaborating is a good thing. You know, initiating collaboration with the established scientists maybe can help drive the curiosity of young people. But one thing I would like to advise is, you know, relying on one topic usually you basically narrow your focus. I have to have a much broader area of your, you know, you know, people who work in their PhD on a topic, they cannot bring that topic to the energy. So just for example, in my case, I work on fluorine chemistry. So I was able to bring in some of the ideas from my past work to the energy storage, you know, activity. So the other things is that the young scientists should not just rely on established scientists work and try to just do an incremental change. I think you have to look beyond the box and try to be more creative. This is how you can make a name for yourself. Now I look for example, just give an example of sulfur. I looked at the paper in sulfur. I would say 70 to 80% of the work is just to incorporate lithium in carbon, sulfur in carbon, because some people have done that and show some good results. So big chunk of the effort has been on that instead of addressing all the key area that need to be addressed. So one has to not to just follow. You have to look beyond what people have done and try to come up with some of their own contributions. So those are my advice to young people. And Peter. Yeah, I mean, I think that's very, very, the last point that Cal made is absolutely right. I mean, the, the important thing to do is to look for, look for something new. I think that, you know, there's a danger and it's not a criticism of any, any one because it's to some extent the nature of the way things are, but there's a danger that we all sort of pile into the same problem, you know, and try to solve the same problem. And I think you need to have a level of pluralism in science, you know, you need to be looking at different things, trying imaginative and adventurous ideas, because, you know, it's not, it's not everyone should do that. There's a place for saying, you know, here's some very specific targets and here's a roadmap and you should follow that and some people should do that but not everyone should do it. Some people should really try to do some creative stuff, some off the wall things, some things that don't look promising that you can't predict, you know that if you were ahead of a, of an industrial R&D lab you would never let anyone do this because it would make sense it's not a sensible for a business. But that's what academia is there to do right to think, think the unthinkable to try new things. You need both these things and some people will be more suited to one than the other but both are important and both should be respectful and think of the other. So I would say if you're one of those people that likes that certainty in the roadmap then that's fine, but also it's good if you want to be creative and think out of the box and explore new things. I mean I started my career and sold electrolytes, I mean ceramics old electrolytes, and whoever thought they would become they would come back into fashion right. In a way I did a lot of work on polymer electrolytes and all these things and so, you know, you also see that what goes around comes around so don't assume that something has been done 20 years ago and therefore it's all done and it's all, you know, being shown to never work and will never happen again because these things do have a habit of coming around. So one thing though, a little bit to come back on the lithium, lithium air, lithium oxygen. Because I think Carl's approach is very interesting but I just wanted to reiterate that the recent modeling that we did based on what is known about lithium air now. And using air, not oxygen, not oxygen tanks predates a fully functional system with close to 600 watt hours per kilogram. Okay, now that is, that is, that is not working and I say not working on oxygen, not taking into account of tanks. What's the difference. The difference is when you use redox mediators, you form the lithium peroxide in pores in solution in the electrodes so you can have thick electrodes. Also, another change is that you're assuming you're using protected lithium metal, which we're going to need for all sorts of things, you know, anyway. And so, I just wanted to put that out there because I think there are some assumptions from the work of five or six years ago that these things are not possible. And that all the problems of lithium oxygen are solved. But I say again, you know, we've demonstrated one milliamp per square centimeter, which is a pretty reasonable rate. We've demonstrated several milliamp hours per square centimeter capacity. So for degradation, cyclability, plenty of things still to solve. But that model is using, you know, that modeling, that technical modeling is based on what we know now. Without any assumptions beyond the fact that we know, for example, that we can work with 13% relative humidity in the gas stream, and it doesn't kill the electrode, doesn't kill the cell. So we can have water there. We can have humidity in that inlet air. You don't have to have super dry air, for example. So these are some of the changes that have happened in our understanding that mean that you can predict those kind of potential performances. So I just say that because, again, it's exciting to work on these problems, right? And they may, this may never still come to anything, right? But I would want people to, especially new people entering the field that are not making a pitch, they should work on lithium oxygen, but they should think broadly, you know, think imaginatively, do something new. Yeah, very good. So what I get out from both of you, beneath a diverse background, expertise, diverse problems, be open, be persistent. That's what I get out from both of you. With that, I think I would like to conclude today's second event of Storage Act Symposium. I would like to thank you both of you for your, you know, great speech and also showing your perspective. Thank you very much. Now, next week's event, we have also two experts joining us next week, Yogan Janet and Linda Nassar next Friday. I look forward to seeing everybody next Friday. Thank you very much. Thank you. Have a good day or good evening. Bye. Bye now.