 Welcome to the bi-weekly StorageX symposium. My name is Will Chu. I am the co-director of StorageX Initiative here at Stanford University along with my colleague Professor Itui. I am pleased to welcome everyone back. Today we are very happy to be featuring two speakers to discuss advanced chemistry and architecture for battery technologies. Over the past couple of sessions we have explored alternative ways of energy storage. Two weeks ago we talked about heat and today we are returning to battery technology by focusing on the next generation chemistry. And for our first talk I'm very pleased to introduce Professor Eric Guaxman from the University of Maryland. And I seem to be saying this quite a lot lately. Eric is also a proud alum of material science engineering at Stanford. He received his PhD training here. And Eric plays several important leadership roles. He is the head of the University of Maryland Energy Institute. He's also on the board of the Electrochemical Society. Today Eric is going to talk about solid state batteries which is a very promising next generation technology that he's trying to make current generation as well both in the lab and at his startup. So Eric I remember vividly you presented on this work maybe six or six or so years ago at Stanford when you first came out with the solid state architecture with the cathode and the solid electrolytes. I remember that vividly and I have been watching the amazing progress over the past few years. And today we're really excited to hear from Eric on the latest progress for the solid state batteries in his lab and in his company. Eric please go ahead. So thank you Will and also you for inviting me. As Will mentioned I actually graduated twice from Stanford. I actually started in chemical engineering. Did my research initially with Kirk Frank and then transitioned to Dave Mason. Those of you who may not know Dave Mason he founded the chemical engineering department at Stanford and I was his last graduate student before he passed away. And then transitioned to material science where I worked with Dave Stevenson and Bob Huggins which I think most of you may be aware of. And so therefore my interest and my research has always been solid state ionics and that got it really got it start as grad student at Stanford. And so I'm going to talk about that. But before I go into the inside of the cell I thought it was important to talk about the outside of the cell. And in particular Tesla as an example. So I want to say this in all the best ways possible. That's a picture of my first Tesla in my garage. I was so happy when I got that thing. That really is a wonderful invention. I think Elon got it right in trying to go for a car that people wanted versus trying to make it just a cheaper car that everybody could afford. And you know it really was a technological marvel. And if you look at the batteries in that Tesla they would look like something like this. And each of those packs that you see there fit in the carriage under the car. And what he did is he really just took thousands of what are called 18650 batteries. And you can see the image of one there they're 18650s because they're 18 millimeters in diameter and 65 millimeters in length. And they're just put in series inside of this pack. One of the issues though by the way is that because of this there is a loss in energy density. So these are our NCR 18650 cells from Panasonic and you know they're 243 watt hours per kilogram for the cell. But the pack is about 212 watt hours per kilogram. And so there's what we call a pack overhead that's the loss or the additional mass due to all of the packaging to have all of these cells. It's about 15 percent. And that's on a grav metric basis. And why is that? Well in fact it's very important that you maintain the temperature of these cells. And so they have to have a cooling system in it to maintain the operating temperature. If it gets too low with the current liquid electrolytes you have a significant energy or power loss. You can see that in the plot there from the manufacturer's battery data. You can also see that there's a big ohmic loss at basically at zero capacity because liquid electrolytes will tend to freeze and reduce conductivity at lower temperature. And so you have much lower energy that you can get and much lower power you can get out of these cells. And so they have to maintain a high enough operating temperature so that doesn't happen. But the other aspect of it is the liquid electrolytes. And here you can see the typical liquid electrolytes. They have a flash point of 16 to 23 33 degrees C. So they can potentially catch fire. And you can see the vapor pressure and how that increases the boiling point of these these electrolytes. And that's why you've seen these types of things in the news. And so you can't also go at very high temperature. And you're really limited in the operating temperature of these batteries so that you don't go too low so that you get high enough power and you don't go too high so you don't have a safety issue. And to do that, not only do you have this mass, what they're doing is they fill it with the liquid which circulates through it, but then to flow that liquid there's a parasitic power drain. And so we talk about energies of batteries. We talk about columbic efficiency. And then that's true for the battery cycling itself. But the entire system efficiency has to include all of the temperature control. So if you put power into maintaining the temperature of the pack, that's an additional reduction in the efficiency of going to electric vehicles. So that's the mass basis. But let's look at volumetric. And so these are the dimensions. Again, I never took apart my Tesla battery. So I haven't measured this myself, but I went to a website and they gave the dimensions for that particular battery pack and I'm showing it here. And so now what I've done is I've looked at sort of the top end of those things. Each of those green circles corresponds to one of the 18650 batteries. That's the diameter of it. And you can see basically based on the way that they're packed in that pack that basically there's a 77% volume pack overhead. They're only getting a small fraction of well, they're taking more volume to enable the cooling than they would have if they were just packed spheres together, okay, or packed cylinders together. So what I'm also going to show you then is now what happens if rather than a cylindrical battery, we had either pouch or prismatic cells, which are the sort of planar configuration. And the area, the cross sectional area, see to the right of each of those little green rectangles is the exact same area as the area of the circles to the left. And so, you know, what I can do if they're flat like that, I can pack them and have the same spacing. And so this would be this same 5.3 kilowatt hour pack, but now configured with a planar either a pouch or prismatic cells. And that would allow the same area for cooling for that fluid to flow through between them. Now, if I didn't have to cool them, in fact, I could pack them much more tightly. And now you can see the dramatic reduction in volume of the battery pack relative to what it currently is just by removing that need for cooling to maintain temperature. And in fact, there are further reductions in volume and mass that can happen by going to what's called a bipolar stacking or put all the cells in series using the configuration ratio as shown to the right. So that's just to give you an idea about how changing the cell chemistry and therefore enabling different temperatures of operation can improve upon the overall system performance. If I go within the cell, there's been a major driving force for going to lithium metal anodes. And so when we talk about future generation batteries, this has been considered, you know, what needs to be done. And this is a paper by Paul Albertus and his colleagues when he was at RPE, which compares a typical cell on top to one that has lithium metal below. And the advantage of lithium metal are a couplefold. One is the carbon anode that's typically there is actually a lithium carbon six. And so if you take the mass of all of both the lithium and carbon, you get something on there about 339 amp hours per kilogram for the anode. But if you go to lithium metal, well, you get really carbon. And the capacity goes to 3061 amp hours per kilogram, basically a potential 10x reduction in mass. So I showed you the volumetric aspect. This is the mass aspect about why you would get better energy density by going to a lithium metal anode. And of course, you also then have this reduction in anode volume, which is shown there between the amount of lithium that would be sitting on the anode side of the separator, if it was just a pure lithium metal anode versus having a graphitic anode. But in addition to that, though, there is one issue. And that's this issue about infinite volume expansion. How do you address that? If you go to a flat lithium foil for your anode, and it cycles, and if you really do 100% depth of discharge on your lithium, you remove 100%. So the volume goes to zero of lithium, and then it goes to 100% on every cycle. That's an infinite volume change every time. How do you package that cell to enable that lithium to come in and out on a repeatable basis without re-depositing in places you don't want? And how do you seal that? So these are problems with just going to straight with the anode. Now, our approach to developing cell cell batteries is using a garnet electrolyte. And you can see the video in the middle. This is your conventional liquid electrolyte. This was done, I was interviewed by CBS News after some certain battery fires. And you can see how the liquid electrolyte just burns up. It's a flamble liquid. But the garnet doesn't do anything. We center this in a furnace at over 1000 degrees centigrade. You see some carbon deposits, but it's a completely non-flammable electrolyte. So that takes away that safety concern. It has quite reasonable conductivity. And also has a voltage stability all the way from lithium metal all the way to over six volts. It's made from inexpensive elements. Latham and Zirconium are pretty cheap. The most expensive material in the garnet is the lithium itself, which of course you're going to have in any lithium battery. Plus, of course, you can have some dopants in there, but they're there in a smaller concentration. But the other aspect of it, again, is that enables that lithium metal anode. It's stable, again, with the lithium metal. But also, if it's processed in the right structure, it can block lithium dendrites. And so the figure to the left with the red, you know, through it is your typical polymer or organic electrolyte. You can see lithium dendrites propagating through, causing a short. In the case of the garnet, which is seeing right now, is a pore in the garnet filled with lithium metal. And I'll come back to that in a minute. So what's been the living, the successful development of solid state garnet batteries, since they have all these wonderful advantages? One is a high specific solid solid interfacial impedance. And you can think about it by stacking stones on top of each other, right? You have a gap between them. How do you get a uniform on an atomic level interface between two dissimilar materials? The other is typically these things are pressed and centered discs, which are then polished. And you can see a photograph of one to the right. And you have a planar interface. And because it's thick, the the omic is very large. And because it's planar, the interfacial area with the electrodes is very, very small. Both of those result in a very high impedance with typical garnet electrolytes. And as a result of those two things, there's been limited demonstration of high energy density cells. So we developed a unique structure that we believe solves all of these issues. And there's an SEM photograph here. And I'm going to go through some of the advantages of why it does what it does to address these particular issues. So the first thing is getting lithium metal to wet the interface. And so if you see the photograph in the middle, what we did is we took lithium foil, we put it on a garnet disc, we heated it up above 180 degrees C where lithium melts. And rather than being a flat foil, it balls up and forms a ball of lithium metal, which will just roll off. That's classical non-wetting. And you can see the SEMs of the interface where, in fact, you can see the gaps that occur. So if I'm doing a macroscopic current density, really the microscopic current density at the points of contact is much higher, dependent upon the actual interfacial contact that we have. We developed a simple atomic layer deposition technique. This case, it was just alumina. And by putting about five nanometer thick alumina across it, you can see on the figure on the right how, in fact, lithium metal now wets. And we get a conformable coating of lithium across all the surface non-uniformities. The data to the left is the impedance spectroscopy. The data in red is the one without the LD treatment. You can see it's well over 3,000 ohm centimeters squared, very, very high impedance interface. And by doing the LD coating, it drops down to that little teeny black semi-circle in the lower left corner, which we blow up to see what the impedance would be with the LD. The reason that we determined that's the case is, in fact, that all of these garnet materials that are lithium-conducting oxides, if exposed to even the smallest level of CO2, even that in the glove box, it likes to form surface carbonates. And lithium metal just does not wet the carbonate. So we need to find ways to avoid forming a surface carbonate. And the LD does that. Okay, so here's cycling data. The data to the left is now a constant current galvanostatic cycling. The data in red without the LD treatment, you can see where we're getting over 6 volts in different cycles. Again, towards the upper limit of the stability of the garnet. Whereas after the LD, it basically looks like a flat line. So we've blown that up so you can see the symmetric square wave cycling we can obtain with that LD process. The data to the right is just showing you, again, at a higher current density, how uniform that is and how it can go on for hours and hours without any problem, a very, very stable cycling. And if you subtract the electrolyte impedance from the total cell from the symmetric cell, you get an interfacial impedance, which is on the order of about one ohm centimeter squared. We've done this by a number of different techniques or a number of different materials, I should say, all with the same effect. Basically, this one here, you're seeing the effect on using an additional layer here. And this is just a coated one of silicon. This is another one. This is aluminum. Again, the same thing, it wets, reduces interfacial impedance and allows the thing to cycle high current densities without any degradation. This is zinc oxide. Now we're starting to see a structured interface, and I'll come back to this again. So overall, this is a critical thing. Your two solids need to wet each other, and this is the case of lithium metal. And we've done this with oxides, we've done this with alloys, we've done this with liquids and gels, all of these overcome that intrinsic interfacial impedance by allowing a conformal coating of your electrode material on top of your electrolyte. But the other thing you need to do or we've done is develop a structure that allows the dense layer in the center to be thin and provides an extended three-dimensional network that you can then have higher contact area for your electrodes. And this is done by very conventional surround processing called tape casting. We just mix the powder and appropriate solvents and binders. We cast them on a sheet. We laminate three layers in this case. Dense layer in the center. On the outer two layers, we put a pore form, which is the polymer bead, which burns off in the furnace. We put them into a furnace environment one step and we get the structure lower right. And this is just showing you a magnification of it. Again, you can see how the dense layer is extremely dense. You'll see no grain batteries on it, no pores. It's about 15 microns thick, a very low, small thickness that is supported by the pore support on either side. And so because it's so thin, the ohmicasr is quite low and because it provides now a continuous, extended three-dimensional structure, it increases the surface area for electrolyte interface contact by about a factor of 50x on both sides, therefore reducing the interfacial areas. And these can then be interfiltrated with whatever electrode material you want. And so the one to the right is a cartoon where we're putting lithium metal on the bottom and our cathode material on the right. Coming on the top. And this is the case, now we're filling with lithium metal on both sides. And so you can see the dark gray is lithium metal and the white is the garnet. You can see how well it fills that pore structure. So it's easily filled. And this is now showing you that single pore that I showed you before. Because we wet the surface, you can see how the lithium actually prefers to touch the garnet surface versus itself. And so this enables us to cycle this lithium metal back and forth by a pore filling emptying mechanism, removing that issue about infinite volume expansion that you would have if you had a lithium foil for your garnet. So now combining these two things of that three-dimensional structure with the surface coating, we're able to go to quite high current densities. And so the data on the left is the center of that is the current density at 1, 2, and 3 milliamps per square centimeter. And the data on top is the corresponding voltage and at the bottom is the ASR of that cell. And what we did is we then at 3 milliamps per square centimeter went to more and more time on each cycle corresponding to more and more capacity of lithium cycling. We're going up to three and a quarter milliamp hours per square centimeter, which is 10 to 15 microns of lithium that's cycled with no degradation or performance decay. And the figure to the right now is a plot of the area-specific resistance for these symmetric cells versus the thickness of the dense layer. The dashed red line is the conductivity of the garnet and then the blue data is various different thickness samples that we made. And you can see how well they fit that line. The basically what you're seeing is just the OMIC ASR of these things. With the tape casting it's quite uniform thickness so you see the error bars are quite small whereas with the other ones are those polished discs and they tend to have greater thickness non-uniformity and that's why the error bars in thickness is greater. And the other aspect of it with this type of structure you'll notice the dashed line in gray without your typical ASR for an 18650 battery. So we now are achieving OMIC ASRs for both the electrolyte and anode interface that are less than that of your typical 18650 battery enabling the potential for high power density with the solid state cells. So we've taken it beyond that. This is now data at 10 milliamps per square centimeter. Again very stable cycling. And what we did with this one is then we ran this for hundreds of hours. We switched over and did that same thing where we increased the amount of lithium cycle on each pass but in fact took it all the way to 7.5 milliamp hours per square centimeter. And this is what's called an exhaustion experiment. The pore volume is 6 milliamps per square centimeter. So you'll notice very stable cycling until we get to the point where we are literally stripping 100% of the lithium out of the pores. When all the lithium comes out of the pores it no longer has that conductive path and now the resistance goes up. And these are one of the experiments we ran for RP to confirm that there were no dendrites with the lithium cycling cells. I'm going back to this paper by Paul Albedras and he plotted here the plating current density that had been achieved in the literature versus the cumulative capacity how long this has been cycled. There were a number of other circles I've grayed out because they weren't solid state. They were liquid ones. And it shows the targets that they have. The green one at 3 milliamps per square centimeter was the RP-E ionics goal for his program. And then the one at 10 milliamps per square centimeter was the part of energy's vehicle technology office fast charge goal. And you can see there's just a few orange or yellow red data points in the literature as far as lithium cycling. And the one that's at the highest current density in his paper is reference 11. And that in fact was our work because at the time he was my program director. But since then we've exceeded that. And as I showed you at 10 milliamps per square centimeter. So in fact we're as far as I know we're the only group that's achieved both the ionics and vehicle technology fast charge goals for current density for lithium cycling at room temperature. I do want to point out again the room temperature aspect of this. We can cycle even faster higher temperature but this is at room temperature. So this now is a platform for all kinds of cell chemistries. And so we have lithium metal in the anode. We have a garnet structure. And now we're trying different cathode chemistries to show the performance. And so here you can see some of our earlier cells using LCO as a cathode material. You can see the cobalt inside of the garnet pores. Here's the corresponding cycling data. The data above is the current density. The data below is the corresponding voltage. And this went on for well over 450 hours of cycling without any problem. I'd also want to point out that then what we did is we had a couple of different holds for 24 hours to demonstrate that it held 100% state of charge, no voltage decay, again confirming no dendrite sharding both at the the initial cycling and at the end of the 400-something hours of cycling. The corresponding coulombic efficiency and capacity is shown here. Again no capacity fade. As one of our earlier cells was a little bit more noisy but again demonstrating the ability for this type of technology to cycle with a lithium cobalt oxide cathode and lithium metal anode. We later on tried some high voltage spinels. These are four and a half volt spinels. And you can see now much more stable data. This is 480 cycles. Again great coulombic efficiency and no capacity fade for 480 cycles. We've even demonstrated this with the oxygen. An advantage of our electrolyte is it does not get oxidized which is one of the issues with the liquid electrolyte. So this really does open up the potential for solid state lithium air batteries and something that I'd like to pursue further. Most of our work has been done with with lithium sulfur. We had a project funded by NASA where we did this. And so now we would just fill sulfur carbon for electronic conductivity and a little bit of liquid electrolyte to help with the interface on the cathode side and lithium metal on the anode side. You can see how we can get about 1200 milliamp hours per grams because of the sulfur utilization we're able to get out of this. Again great coulombic efficiency and less than 20 percent capacity fade for over 600 cycles. A lot of people worked on lithium sulfur batteries. They have a problem it's called the polysulfide shuttle where the sulfides form on the cathode side and diffuse through the liquid to the anode side. Again because of that dense ceramic electrolyte we blocked that polysulfide shuttle. And so theoretically we should have very very high stability with these types of cells and this is some initial data to show that. The other aspect of that lithium sulfur data is we also then measure over wide operating temperature range. And you hear now you can see cell data from minus 10 to 90 degrees centigrade. It just falls a linear trend with no drop off on either end. A significant increase in energy density as we increase temperature. We achieved 280 watt hours per kilogram at room temperature and that's a total cell mass both the both electrodes and the electrolyte at 90 degrees C. We're up to 350 watt hours per kilogram. And in fact we could have gone higher. The issue we had is the packaging can't take a higher temperature. So we're looking for right now for developing packaging that'll enable a wider operating temperature range. The battery itself is perfectly fine with that. It's just the packaging. And going back to my initial part of the presentation this dramatically reduces the cost complexity mass and volume requirements of current battery technology. If you don't have to cool the batteries or you don't have to heat the batteries you don't have to put in all the infrastructure to take up the mass and volume of doing that and you remove all the parasitic power requirements of maintaining the temperature you just let the battery equilibrate at whatever its operating temperature would be. The other aspect of that is the safety. Again these are non flamble but now you're seeing the ability to just cut these things open and there's no degradation to continue to perform without any problem. We've also done this with titanium sulfide which allows even higher temperature and now you're seeing 1C cycling at 90 degrees C without any problem and in fact talk about safety. This is that titanium sulfide cell under an open flame. Absolutely safe, no flame and in fact it just burns. It just shows a higher energy density because of the LED lights up farther. So we're also now making bilayer structures which are a dense porous layer and the advantage of this versus the trial layers this allows us to use commercial cathode materials like NMC. Here's some initial results showing that and in fact we're able to achieve on the order of well over 300 watt hours per kilogram on a gravmetric basis and about 1000 watt hours per liter using a conventional NMC cathode on one side with the lithium metal and the dense supported garnet layer on the other side. So these are a transformer platform for a wide range of cathode chemistries and we're scaling the fabrication and trying to commercialize it through a company I founded iron storage systems and I can see the larger area of ceramics that we're doing this with and these are some of the specifications of the cells that the company is developing using this platform. The one on top is this bilayer where we're using an NMC cathode we're achieving 265 watt hours per kilogram but now completely safe because there's no non flamble. It's a ceramic electrolyte with all non flamble components to it so it removes that safety concern. I'll show you some of the data for lithium sulfur which again we've achieved sort of 200 watt hours per kilogram but we are on track to get towards the 500 watt hours per kilogram and now these are our no nickel no cobalt very very cheap cathode materials and then again the possibility of making really high temperature batteries these can go 350 degrees C or so so there are a number of unique applications that this would be again uniquely capable of addressing and also the fact that we get so much higher current density available at the higher temperatures. This is the one slide of some of the data for the NMC cells that are being made by the company we're now currently scaling these things to 10 megawatt hour per year production levels. You can see the data for the pouch cells the number of cycles and repeatability between cells from this this more you know of a manufacturing process and we do plan on having prototypes available sometime Q1 of 2021 for a variety of interested partners. So with that I want to thank my my colleagues of Bing Hu at the University of Maryland who I've done a lot of work with Venkat Thangarai University Calgary he's actually one of the inventors of the the Garnet material Yifei Moe and Chen Chengwai and also at the University of Maryland a whole list of students and postdocs more than a listed here but there are ones that contributed these slides. ARPA-E for for the funding as well as DOEERE multiple awards I want to thank Chen Dong for that the battery 500 which again that that's one of the things that that Stanford is taking a lead in our ceiling contract NASA Lockheed Armour Research Lab and I need to do my point and point out that I'm a founder and equity holder and iron storage system so just identifying my conflict of interest and I want to thank you for your attention. Eric thank you very much for that insightful talk we have received more questions than we can cover I thought I'd just start with a very high level question one of the images you show the cross-sectional SEM looks exactly like a solid oxide fuel cell which you're also very well known for maybe can you just briefly talk about the learnings that you have transferred between the two fields. So actually you're right I mean it looks a lot like a solid oxide fuel cell when I did my graduate work at Stanford I was working on solid oxide fuel cells and I went through this same learning curve where you saw the one image which was that polished discs that's where I started with right I was making those polished ceramic discs myself and everybody else working the solid oxide fuel cell field we're trying to figure out ways that we can make that electrolyte thinner and so by depositing a thin layer on a on a porous support that was enabled and that's why the majority of solid oxide fuel cell companies make a anode supported cell. So as it turned out Venkat Thangarai who I mentioned from Calgary did a sabbatical with me at the University of Maryland and and as one of the inventors of the the Garnet material we started talking about well what else can we do and so we put our thoughts together and decided you know what that same structure for Garnet would work perfectly well we submitted our proposal to RPE which was a little seedling and you know you can see what it can't result from that so it really is a direct evolution of the work that I did with solid oxide fuel cells. What a great story really really great um so let's dive into the technical questions Eric. So the first question is on lithium metal penetration so there's been many many reports talking about lithium fracturing the solid electrolyte LOZO there's many discussions of the critical current density so the question is what is the current density in your composite electrode after you normalize for the surface area is it I presume far lower than that of a planar electrode and is that the reason why you're not seeing lithium penetration? So so there's a whole variety of things that we've done um and and that's one of them right the first thing was was getting lithium metal to wet so I showed you the SCM if I didn't wet the surface then what is your actual surface area it's only a fraction of the actual surface area and the vast majority of literature out there with the dendrites did not get lithium to wet so you know I I question the value that they have for current density if they weren't actually achieving that way okay now more recently people all have been doing that they've been trying a number of different techniques to get lithium metal to wet and they still are getting you know current densities you'll see in the literature it's been going up and there's some recent ones besides ourself that are saying that they've got to 10 milliamps per square centimeter on a planar surface by simply uh removing the the interfacial impedance by polishing and other techniques although I you can't polish the inside of the pores right um the other is as which you pointed out was a surface area so that's a factor of 50 increase in surface area so if I'm you know having 10 milliamps per square centimeter going through the cell I've only got 150th of that going through the interfacial area and the dendrites form at the surface so it's not it's not a matter what the current density is through the dense layer in the center it's what it is in all the surface points and so you just take that 10 and you divide it by 50 and you're like at 0.2 milliamps per square centimeter instead of 10 okay um there are other aspects of it also um you know a lot of the the dendrites observe because they're surface flaws and and there's papers out that you can see where basically they polished it with a 400 grit versus a 1000 grit the sanding paper that the grit of it would change the propensity for forming dendrites so it's a surface flaws in our process with with the tape casting and there is no we don't touch the surface it's buried in in that pore structure right so we don't have any any uh microscopic surface flaws um and the other is that that we you'll notice from the SEM there's no no grain boundaries observable so another uh a propagation mechanism is through the grain boundaries if you can proper if you can process your ceramic and avoid grain boundaries you you've taken away one of the the paths where dendrites can prop up great Eric so maybe a follow-up on that question so you use the composite structure to lower the current density and better wetting but you do pay a price um in the volumetric and the graphymetric energy density can you comment a bit on how big of a hit that is when you introduce the composite structure to the anode right so um on both sides whether it's on on just the anode for for the bilayer or on both sides if it's a trial layer the uh volume fraction of the pores is between 50 and 70 percent it's only a 30 volume percent or so of the the garnet material um and so the energy densities I I reported to you were the actual then measured energy densities that includes the mass of the garnet um yeah I mean if if I was to go with just a lithium foil on the anode versus having that coarse structure um I would have uh uh you know um a greater volume metric energy density but it's only a factor you know 50 percent hit and it's a 10x improvement over the the the carbon right so it's it's it's significantly better than than than a graphitic anode it's not as great let's say as it would have been with a planar lithium but then how would you accommodate the lithium cycling every time and going from zero volume to 100 volume on every cycle you have to build some structuring so a lot of salt say bad is what they do is they have a a pack pressure right they they put their cells under pressure to maintain contact but that's now pack overhead because it's an additional expense it's an additional structure it's additional mass additional cost we don't pressurize our cell everything there is just sitting um in its pouch without any applied pressure to it and so that's a significant savings at the system level by by doing it that way so Erica my understanding correctly that at the systems level the porous structure is highly competitive um with other approaches for solid state batteries as well yeah okay again if you can get rid of having to apply a pressure against your cell that's a significant savings um in again massive complexity of the system yeah no i i i really think this is the right way to look at it you have to compare at the systems level rather than the cell level so um continuing on on the lithium metal question so um you mentioned the importance of the wedding angle and you showed a picture of um melted lithium wedding or non-wedding a solid electrolyte how does one go between the wedding angle and the surface energy between solid lithium how do you translate between the liquid and the solid lithium in that case in terms of the surface energy right actually i think that there was a paper by you where he looked at a variety of different coatings and then looked at the wedding angle we not actually tried to quantify wedding angle we just were looking at what's the the different coating materials what is the impedance we get and does it wet or not we've not yet gone through and analyzed the wedding angle so i i can't answer your question because we you know all things we've done that's just not something we've gone and analyzed for each material what the wedding angle is right eric i think the question was actually a slightly different one i think the question concerns you if you observe the wedding angle for liquid lithium oh is that how is that related yeah i don't know the the solid interface is obviously going to be something different than the the the liquid interface but but it it does show you and again if you look at that that pour that i showed you with the lithium inside of the pour it's a matter does the lithium want to bond to itself or doesn't want to bond to the garnet and and that shows you in fact that the lithium actually the surface energy is such that prefers bonding to the the the garnet than it does to itself terrific maybe now moving to the cathodes um how big is the chemistry place space for the coatings you showed a few and you also showed a few for the anode as well is there a massive amount of coatings available how much are you trying to identify sort of the unusual chemistry that can be the coating um we're pretty much open to any cathode chemistry you can think of we have the advantage that that now we're stable to much higher voltage and so i'm really hoping that somebody out there working on cathode chemistry is going to come with a greater higher voltage than anything else out there right so five six volts or potentially higher we don't have an upper voltage limit that i know of yet um but we've been trying all of them uh from a company perspective we would rather use you know the nmc which is sort of the state of the art because they're commercially available we don't want to recreate the supply chain for for for providing the cathodes but you know they all work again lithium sulfur lithium air you really can't get any at least theoretical energy density higher than lithium air um and that's quite potential um you know i've started now looking into some of the conversion cathodes right so that also is a possibility that that those can work right um so along another related question about the cathodes eric so how you didn't talk too much about the volume expansion and shrinkage of the cathode um and i imagine that's quite important in the context of a solid composite as well um can you speak to a little bit about the mitigation strategies for maintaining contact so that that is the the biggest issue um is the cathode luck light interface right um and as you pointed out that they're going to expand or contract within the pores so one in terms of just making sure you don't crack your your your cell is you don't put more cathode material in the pores than it would be once it was fully lithiated right so with sulfur for example you put the sulfur in um it's not lithiated and then it expands as it gets lithiated so you got to make sure you don't put more volume of sulfur in then the pore itself can maintain once it's fully lithiated in the case of some of the other cathode materials which are already lithiated it's not going to expand beyond that size because it's already lithiated okay but it's also more difficult to get some of those lithiated compounds into the pores and that's another reason why we're going towards the bilayer structure for nmc and other other cathode materials like that but the other issue then is is how do you maintain good um interfacial contact between the two phases right and and and how do you accommodate these uh the um expansion that occurs uh by the chemo mechanics of cycling back and forth um i'm happy to say that that the one of the most recent uh e area awards i've got is is focused on addressing that and we're looking at different interfacial layers that will overcome the the impedance between that but for right now we've just been using a very very little uh catholite to bridge the gap between the cathode particles and and the garnet so that if the cathode expands and contracts in the pores then in fact um that liquid kind of accommodate um that that difference in in expansion but ultimately you know people would like to get to fully all solid state um and there it's not just that it's like can you co-center these things with having out having a reaction at the interface between your cathode material and garnet and so we're making some progress i think on how are we able to address that also eric one last fundamental question i'd like to ask you a few more questions on manufacturability um so the final fundamental question is uh there's been many discussions on partial electronic conductivity in l zero very small but there's some um as being the culprits for um lithium penetration and deposition uh beneath the surface um is this something that the coating can help address yeah um the the coating can help address it and again there are a variety of different garnet compositions right so you know again this is all psionics can you can you adjust the composition the defect to glibria to minimize the type of electronic conductivity that you observe um and i've seen you know some of the results where they show some small level of electronic conductivity and the way when i presented when they presented asm is that n type or p type right because it's a different way of addressing electronic conductivity depending upon which side and and they don't have the answer to that now and i i've not yet looked into it more but i mean we've done this consistently through again my back and sole oxide fuel cells how do you address n type conductivity in syria as an example right so is this n type or p type that's going to give you a different method of adjusting electronic conductivity to minimize it with the correct open that work has not yet been done but but again it's it's very doable you have to understand the defect to glibria of the garnet materials and then you can develop the right doping to address it great thank you eric so on to manufacturer ability maybe just one or two questions on that can you talk about your learning so far if you're able to uh what are some of the biggest manufacturing issues that has to be addressed to take the scale up one level so um making sole oxide fuel cells is is you know i wouldn't i'm not going to say it's easy uh but you know we've been able to scale those things up to 100 square centimeter cells in my lab um uh we made lots of these things and tested them we're not yet at that level with the garnet um because the the centering process is more complicated okay um when you center the ceramic for sole oxide fuel cell oxygen you know comes out when you raise the temperature up that's just your your entropy and then as it cools back down the oxygen comes back in when you heat up garnet lithium oxide comes off but it doesn't go back in right so there's an issue about how much excess lithium you have to have during the centering process um the other aspect is is how do you manage the the the warpage right so again we've overcome that with the the sole oxide fuel cells as we scaled it this is something we're trying to do right now where again that's one of the advantages of the trial layer where you got a porous dense porous structure it's symmetric so when you when you heat it up and cool it back down it will maintain flatness but if you go to the bilayer structure which is the one we want to get to now you've got a dense layer on a porous layer and so we have an asymmetry in the in the shrinkage during centering so maintaining flatness um and maintaining the exact microstructure and doing it over a large area are the primary issues we have for fabrication but but being able to to make large quantities of the material the tape casting the laminating those are all very scalable processes now i think one maybe crucial difference between solid oxide fuel cells and solid state batteries is the environmental sensitivity during manufacturing do you think this will be something that would limit the cost of the processing at some point um actually one of the advantages of this over conventional batteries is we don't use a dryer right so all of the processing of the garnered material is just done in air um even through the the centering process and the centering process we have a controlled atmosphere um at that step um but then we take it out of the furnace you know uh we we we have the the techniques to be able to to have it sitting out in in atmosphere if we'd like although we probably wouldn't want to because just we want to protect it um but but all the processing of the ceramic it can be done in ambient atmospheres only when you start putting in the electrodes that you actually have to transition to a dryer well eric there's many more questions that we don't have time to answer today i hope we can talk about some higher level points uh in the panel discussion so eric thank you so much again for sharing this great work and i'm going to hand it back to you thank you great well thank you eric for the great talk uh now let me uh bring zhennan to the stage let me do a short introduction uh it's a great honor to introduce professor zhennan bow at chemical engineering at stanford university right here um of course i have a long history of uh collaboration with zhennan um zhennan is the the word expert in polymer and inorganic materials her claim of the flame all has always been uh skiing electronics soft electronics about a decade ago zhennan started to move into the battery field and clearly after she joining the field has she has uh she has been making a very innovative contribution coming up a lot of new ideas to help the whole field and really move a lot of progress going you know she started the self-healing type of idea i was fortunate to collaborate with her on that and uh so many of you have seen great work coming out of zhennan's lab so today is absolutely a great honor to to have her to to speak um also due to her outstanding work over the years on many things she just won so many many awards so let me just mention you know she's a member of national academy of engineering so uh without further redo i will let zhennan take you from here okay thanks yi so much of course yi has been my collaborator for so many years on the battery side and has been really great working in this new area and as yi mentioned that i come from the kind of polymer science polymer chemistry side so our interest is in the battery area is really to to see what are the unique contributions we might be able to make by using our tools on molecular design so in this talk i'll show two examples of rational design of electrolyte solvent molecules and also artificial sei for stable lithium metal anodes you have already heard from the previous talk that the future direction for high energy density battery is going to be moving towards lithium metal as the anode but of course compared to the conventional graphite based anode there's no longer a support that lithium can interpolate into and this becomes a layer of lithium that has to grow from zero to a huge change in volume and and also in addition to that there are potentially during this growth all kinds of side reactions might happen and lithium might grow into the stand right and then cause shorting and a lot of issues with the stability and the safety issue so that has been the main challenge in kind of bring lithium metal anode into practical application and then to reiterate the two modes of instabilities that are really important to address to ensure stable lithium anode is when the mechanical instability the huge volume expansion for the growth during the growth of lithium metal and the other is the chemical instability that is the side reactions that are taking place at interface so these are two key to modes of instabilities that we need to address and indeed there have been a number of approaches that have been investigated in literature and some have shown great promise but still no approach is able to meet all the requirements yet for practical application and it's likely maybe all the approaches have to be combined eventually to solve this big challenge. There are activities on modified electrodes and coming up with protecting layer artificial SEI to coat on the lithium anode electrode. Host material is a very active area of development and also in the solid-state electrolyte area here I'm only talking about the polymer-based and in this case people try to design tough polymer gels as the electrolyte to hope to suppress the dendrite growth. So in our case we want to choose some specific areas where we see a lot of opportunities for molecular design and try to understand what are the rational molecular design rules so that then we can with the gaining of the understanding then we can over time come up with better and better improved systems. So two specific areas I'm going to be talking about are the electrolyte solvent design and the polymer coatings as the artificial SEI. So traditionally the electrolyte solvents that have been used for lithium-ion batteries are these two classes one is carbonate based and this type of electrolyte can tolerate higher voltage for example NMC type of cathode but the problem is it has significant reactivity with lithium metal and then the other class is ether based they can give reasonable stability for cycling with lithium metal but the oxidative stability for this class of electrolyte solvent is poor it tend to have oxidative reaction decomposition at the cathode side. So there have been a number of approaches in literature to try to solve these issues so for example the very promising approach one of them is high concentration electrolyte so basically using very high concentration of lithium salt and basically all the solvent molecules are participating in solvating the salt and then therefore reducing their reactivity at the at the electrode side and in this case one of the drawbacks is because of the high viscosity of this kind of high concentration electrolyte the ion conductivity can suffer in this case. So the improved method reported by the PNNL is localized high concentration electrolyte so basically there is a solvent that doesn't solvate the lithium salt is added and the lithium salt is being hypothesized to be still highly concentrated in the region and the solvated by the carbonate or ether but then these solvated regions high concentration solvated regions are basically dispersed in the non-solvating solvent and slowing down the reaction and the increasing the stability towards lithium metal. So these approaches have shown great promise we decided to approach this from a systematic way so here we started by first investigating the how does how do we first tune the salvation of lithium and how does that impact the deposition. So here in this case we take the DOL ether-based solvent and then we added different non-solvating solvent such as hexing, cyclohexane and toluene and then we use lithium NMR to study how these solvent can solvate the lithium so towards the right side is the upfield shift of the lithium peak in NMR and that's a more solvated case. So we see that the salvation of lithium is very sensitive to the composition even the ratio between the DOL and the hexane can significantly change the salvation and in general we found that when we have a more solvated case then we will have the higher over potential but the less solvated case we have lower over potential and we have longer cycling time and the more smooth deposition in that case. So that provides one information that is less solvated lithium might be desirable for improving the stability and also another aspect is the oxidative stability the first one is more lithium metal reactivity the second one is oxidative stability and here we made molecules that has fluorinated Cf2 groups incorporated into the ether this provides electron withdrawing capability and then the acetylene oxide units substituted are used for solvating the lithium ion and allow transporting of the lithium ion and here we found that as we systematically change the number of Cf2 groups incorporated then we will change the oxidative potential of this molecule and the electron withdrawing effect can indeed help to make the acetylene oxide to be more stable towards high voltage cathode such as Nmc compared to the one without any fluorosubstitution. So with this information we start thinking about how can we design electrolyte molecules that can potentially provide both high stability at the lithium metal side as well as oxidative stability at the cathode side. So the traditional ether solvent has this DME structure and then we ask the question of if instead of two CH2 groups we extend that to slightly longer what will happen and we found that this actually improved the oxidative stability and then further we added Cf2 groups which provides electron withdrawing groups which supposed to further pull away the electron from the oxygen and how that how is that going to be impacting the lithium metal based electrode. So when we prepare these solutions using these three different solvent immediately we can see there's a big difference when we use the FDMB with electron withdrawing fluorogrups substituted. You see a dark color here and we were able to grow single crystals by using the lithium Tf as the salt and using the single crystal we were able to see that in the new molecule interestingly the salvation of the lithium is actually through the chelation of oxygen and the floral atoms instead of the typical two oxygen chelating to the lithium together and this unique structure and also the incorporation of the floral atom is responsible for the color change and the structure for the lithium FSI is shown here this is based on modeling basically again the oxygen and the floral are participating in the salvation and then in each solvated lithium ion we see that there are two around the two FSI molecules involved in the in the salvation and the floral groups participation in the salvation can also be seen from the shift in the floral NMR to confirm that it's indeed participating and this is also can be explained by just looking at the electron density this is modeling done by Jianqing's group in Stanford chemical engineering department and you can see in the FDNB the negative electron cloud are localized on both oxygen and the floral while the other ether case are entirely localized at oxygen so therefore in other case only oxygen two oxygens are used to chelate and in terms of the voltage stability we observed the FDNB can achieve significantly higher stability compared to the other previously reported solvent as we expected and then in terms of the lithium metal growth you can also see that the growth morphology is significantly changed in the case of FDNB we see more plate like growth while in the other two solvents which tend to have instability in chemical reaction with the lithium metal you would see those dendrite like growth of lithium metal and here is the cryo EM to look at the SEI interface with the new solvent we see a very smooth homogeneous and amorphous SEI that's quite thin it's only six nanometers and the left side is comparing to the traditional the carbonate based electrolyte and also I didn't show the XPS here but the SEI we found is more anhydride the SEI and contains quite a lot of the floral groups and lithium fluoride and this is likely due to the fact that the FSI is participating in the solvation in the lithium there are two FSI incorporated in the solvation structure that might be incorporated into the SEI and this kind of smooth SEI is and also the growth of the plate like lithium is consistent with what's previously reported by Shelley Mone's group where they found that when the lithium growth in this more plate like structure it tend to have higher columbic efficiency in the in the cycling and indeed this is what we observed so here we build full cells with lithium metal on one side and this is a thing lithium metal we tested 25 micron lithium metal and NMC as the cathode so this is the high voltage cathode and the full cell columbic efficiency we can achieve is one of the best reported in literature 99.6 percent and with very good capacity retention here even after 400 cycles also we were able to to use this solvent to build the anode free pouch cells so in this case the anode free cell basically doesn't need to start out with a foil of lithium and that's that can potentially reduce the the volume and also the weight of the overall cell and it's a very challenging type of battery to build because if the electrolyte is not stable then the lithium will be quickly consumed the lithium supply from the cathode will be quickly consumed and will not be able to maintain stability but this electrolyte can reach high columbic efficiency after one or two cycles and therefore is able to to be one of the very few solvent that actually enable anode free battery with high cycling stability and here we have built such cells with several different kind of commercially available high energy density NMC cathode in the pouch cell geometry so with just the electrolyte kind of approaching the electrolyte side is one approach and just with improving electrolyte might not still might not be sufficient to really achieve the long-term stability that's needed so at the same time we're also investigating the rational design of artificial SEIs for achieving more stable lithium metal anode and this SEI is very important because it needs to accommodate the volume expansion of lithium metal and the normal SEI formed by the reaction between the electrolyte and the lithium metal tends to be more rigid and potentially during the volume expansion may form cracks that leads to the dendrite growth and also the SEI if designed properly we hope it can prevent the reaction from the solvent molecule decomposition and also be able to help the chemical instability problem so almost a decade ago when we started to work in the battery field the very early work that we did in collaboration with EASE group was to use a self-healing polymer basically this polymer that's formed a network through hydrogen bonding and since hydrogen bonding has a weak bonding so it can break very readily but it can also reform very readily at room temperature and this provides a flowable and dynamic network and at the time we thought that this binder for silicon anode could potentially accommodate the large volume expansion of the silicon particles and result in more stable cycling and indeed that give us quite promising results for silicon anode so that motivated us to continue work in this field and explore the potential of polymer chemistry so in the lithium metal side we thought that the self-healing polymer could also be interesting to be applied as the protecting coating for lithium metal because it's flowable property which we characterize using rheology and this polymer the microscope image you show you see on the lower right side basically we poke a hole in the polymer and over 60 seconds you can see the polymer already kind of move around and flow over the hole and the sealed hole so we thought that this unique property of this kind of dynamic polymer could potentially seal any pinholes on the surface of lithium metal and prevent the cracks in the layer and prevent and the result in the dendrite growth and the other question we asked is maybe if we have this kind of polymer that may provide more uniform ion transport and there's no hot spot of ion transport it might change the or the lithium metal deposition morphology altogether so to our delight that we found that our first try actually got very exciting results that without polymer we saw the typical dendrite growth and with polymer we saw a very dense and a uniform layer of lithium metal that's grown on the surface and indeed it seems it's changing the morphology of the lithium deposition and we were also happy to find to see that another student a postdoc of Ease took another flowable polymer dynamic polymer that's a silly putty and this one has the boron oxygen bond that's dynamic and again this polymer shows that with the coating it grows the nice plate-like structure and to give much improved cycling for the lithium metal so with those we started to study what is really the role of the chemistry and also the mechanical property of these polymer coatings and we found that the one very important role of the polymer design having the dynamic cross-linking is indeed macroscopically we can achieve much more uniform lithium deposition without having the non-uniformity that may arise from the if we have a rigid polymer even though the polymer chemistry is almost the same but the non-uniformity of the polymer and the subsequent SEI formation could lead to in macroscopic regions non-uniformity of the lithium growth even though in the local region we see same chemistry give similar nano structures and the other thing at the same time that our colleague Jianqing has been using modeling to investigate the role of mechanical property of the polymer coating and his modeling also suggests that with the viscoelastic polymer coating on the surface of the lithium metal tends to give growth of more densely and also uniform layer of lithium metal to give the desired morphology so this information also coupled with our systematic study of the surface energy and the polarity and salvation of various different polymer chemistry to investigate their impact on the columbic efficiency that led us to conclude that for the desirable coating we should keep the self-healing and flowable property that comes from the dynamic nature of the polymer we design this will be desirable for addressing the mechanical instability in the in the coating and also the mechanical instability in the lithium metal anode and then we need to have good lithium ion conductivity and even better if single ion conductive but at the same time in order to prevent the reaction from the reactive electrolyte solvent then we will want this layer of polymer to be non-swelling by the electrolyte solvent but yet still able to transport ions through these single ion conductive channels so that sounds like conflicting kind of requirements but indeed we were able to come up with the appropriate design of the system so here we have the aluminum center coordinated with four oxygen so that to create a negative charge so then that would become the site that can solve the lithium ion and allow the transport of lithium ion in the single ion conductive mode because these sites are being localized in the polymer network and then we choose the polymer backbone to have these floral ethers and if it's all floral ether based then it doesn't solve a lithium ion in that case so then and also it doesn't swell by the solvent so this provides a network that doesn't swell but has the single ion conductivity from the aluminum side and then the aluminum site is also a dynamic site to compare to this we also synthesize a polymer that has the boron oxygen side that's a single ion conductive but it's not dynamic it's a strong bonding and also silicon oxygen side this one doesn't carry any negative charge so it's a non-single ion conductive and also non-dynamic and we are able to determine these aluminum oxygen side as dynamic by using both NMR to characterize the association constant and also DFT calculation to calculate the bond energy so compared to the boron oxygen and silicon oxygen you can see much lower bond energy and carbon carbon bond typically is in the order of 300 so this is a half of that and therefore create a not not very stable bond so that's why it has the dynamic nature so here I think my time is running out just go to the final stability of the system so you can see that with cycling with lithium metal if we keep the bare lithium metal without any coating in the electrolyte then we will see that the impedance at the surface will keep increasing this is because the formation of the SEI and it's keep growing because of the reaction and the impedance keep increasing but with our single ion conductive dynamic network this impedance change is very small over time and indeed it gives stable cycling by using this coating and this coating compared to those reported there are a number of coatings reported in literature and this is one of the best currently without if we don't use any other tricks such as a 3d host or other kind of special electrolyte and here we're just simply using the conventional carbonate electrolyte and we are achieving one of the best performance for full cell cycling okay and again growth is this plate like structure so to summarize here I've shown two examples of rational molecular design of electrolyte molecules and also artificial SEI our finding is that it's very important to tune the salvation of lithium in all of these systems and the weakened salvation in the solvent we design and also the incorporation of the anion into the salvation structure lead to anion enriched SEI this is the reason we think is enabling stable cycling and then in terms of the artificial SEI the approach of using dynamic polymers and then combined with chemistry to design a stable polymer network is the key we think to achieve stable lithium metal artificial SEI and ultimately these approaches probably we're investigating one at a time one parameter at a time and ultimately I see that potentially they might need to be combined together to realize the stable lithium metal end node so finally like to thank the support from DOE, ERE, the BMR program and also the battery 500 program my collaborators are mainly e-scroup, e-trace group on the battery side and also Jen Ching's group on theory and these are the students and postdocs who are the key players in our battery work and thank you very much. Well thank you very much to Nan for a very nice talk on using organic materials to improve the batteries you know we have a number of participants in Zoom right here but we have a lot more they are watching and through another channel I think the questions started to flow in let me start by asking the first question first set of questions related to the organic electrolyte there's a question related to FDMB the audience asks FDMB and also a few other research groups recently demonstrated up to about 99.5 kW efficiency so this of course is excellent and then we still need about roughly about 0.5 percent better kW efficiency to go to go get to 99.9 percent I will arrange so what's your thought about electrolyte engineering how do we get there how do we get there this is an excellent question is it the only approach or we need to combine this other approach to get there. Yeah I think the getting to a stable electrolyte liquid electrolyte that's the more near-term goal to enable the lithium matter based battery and many groups have been working on different approaches and also our understanding has been grown over time and I think the approach of for example high concentration electrolyte and the localized high concentration electrolyte then those case the solvent molecules are based on the commercial molecules or non-molecules and in our case we are focusing on the rational design of these molecules to make them stable to begin with so potentially one can imagine that these approaches they don't really conflict with each other and they can definitely potentially be combined to design more stable electrolyte and I think the challenge is that the lithium salvation is very complicated and it's extremely sensitive now as we start working with these different solvent molecules a simple change one atom change in the molecule can change the lithium salvation structure completely and that can have a huge impact on the stability and what's the nature of the SEI so I think to make the long answer short I think a combination of different approaches would definitely be a promising way to go and further understanding of the salvation structure and how we tune the salvation structure I think it would be very important to other direction as well yeah so in the still in the organic electrolyte domain of this question fluorine is you know all this organic once you add in fluorine something oftentimes something good happened in the history of the lithium ion battery the electrolyte FEC for example fluorinate ethylene carbonate is often used you also shown FDMB fluorinated this either it has been very very important any thought about how to design fluorine based molecules so what will be the guiding guiding principle right there to design fluorine based molecules maybe doesn't need to be fluorine would chlorine make sense or other other atoms you know and the electrolyte space there's many trying arrows right there there's a little bit of intuition I think guiding things moving so what's the thinking right now how can we do better design based on what we know yeah I think yeah that's very interesting the fluorine atoms seems to have some special ability to result in promising electrolyte solvents and the fluorine atom indeed is quite unique because it's strong electron withdrawing groups so because I come from the organic electronics field we use floral atoms and floral substitution very frequently to tune energy levels of organic electronic molecules and the floral atoms are typically entirely electron withdrawing and very strong even through multiple bonds it can still result in the electron withdrawing impact but in the case of chlorine based and also fluorine based molecules tend to have special solubility behavior many of them only dissolve in polar solvent or other fluorinated solvent because of the very electron activity but with chlorine atom they can have both electron withdrawing impact through bonds but can also have electron donating effect through resonance effect so then their behavior would be very different from fluorine because the lone pair on the chlorine can actually donate to potentially to lithium ion and also their solubility becomes much better in many different solutions yeah so I'm not sure whether chlorinated solvents have been investigated much but I expect it will completely change the solvation with lithium ion and and the other thing is I guess the fluorine is special is lithium fluoride has been found to be very beneficial to form stable SCI if incorporated into the SCI and that might also be the other reason that the floral solvent tend to give such good results so far yeah that's great since we started the storage axon posion our first speaker Professor Stan Weitingham a Stan alum of the Nobel Prize winner has been a you know first speaker in Norway speaker and also a strong supporter of symposium he's here every time or listening to the symposium including today Stan asked a question about fluorine again the fluorine you know seems to be good you know to to the bad space but bad for the recycling and environment what will be the solution if if we consider you know the fluorine issue down the road yeah can we solve this well the I think certainly making the molecule into into lower a higher boiling point would be helpful so that they don't just escape during handling into the air during handling and and also can be recycled much more easily and then down the road potentially looking at other alternatives now we understand that the electron withdrawing impact of floral atoms are very important there are other organic functional groups that also offer strong electron withdrawing effects but then the question is their stability chemical stability in the battery that's still unknown but they haven't been explored much so I think those that could be the other directions to go yeah so now let me move on to the polymer side of the questions about a decade ago you showed these assault dynamic bonding you know it's good self healing we all know in the in the polymer side of of the battery field a long time ago John Nealman and Anitas Bracero in Berkeley show you know using strong polymer mechanical force you know module needs to be higher than you know certain gigapascal in order to suppress dandruff so the about a decade ago this demonstration of self healing you show you know it's an the opposite philosophy to solve with your metal problem it would be great for audience to learn about your thought or the opposite way to to solve the problem that polymer needs to be solving dynamic yeah just give you a chance to maybe explain compare these two two approaches yeah yeah well the prediction by Nealman on the high modulus coding that can provide prevent dandruff formation has that model takes into account of the kinetics of the reaction but then there are also other aspect the diffusion of the of the lithium for the reaction that's also another important consideration and then more recently linden arches group has been developing model that looks mostly at the transport issue of lithium ion and then in our case and in reality the transport as well as the kinetics are all linked together and so our colleague jen chin is developing a model that takes into account of all these aspects and in his model it seems to suggest that the inconsistent with our observation that the viscoelastic polymer tends to be the one that gives the dense packing of lithium deposition and of course the lithium deposition process is very complicated when it's below the a polymer layer and this really depends on many factors as i mentioned the mechanical instability needs to be addressed and that is addressed by dynamic polymer that that that aspect is more obvious to see but then the chemical instability side it's basically chemical reaction at the lithium surface and it depends on so many factors the salvation of the lithium transport of the lithium depending on the ion conductivity and also that depends on the dynamics of the polymer which can also impact the ion conductivity and at the same time if the solvent molecule diffuses through this layer so not only the polymer could be reacting but then if the solvent molecules swell and go through this polymer then the solvent molecule can also react so in our coding design we kind of took into account the obvious problems we can prevent that is the mechanical instability by using dynamic polymer and then the prevent solvent from getting to the lithium metal using the polymer network that doesn't get swell I shouldn't say it doesn't swell at all still we observe less than five percent of solvent uptake if we leave the polymer in the solvent for long enough time so that's why the solvent molecule stability is another thing that we need to consider and then combine with this coding would be the kind of the multi-pronged approach to solve this big challenge that's great I think your explanation really helps also answering other questions the audience asking relate to mechanical properties, surface energy, conductivities, stability which one is important so I won't repeat that so with this thank you let me bring Eric and also Will coming back to stage let's have a short panel discussion Eric if you can turn on your camera so I will hand this to maybe to Will first to ask the first panel questions right E thank you and then I thank you for the wonderful talk as well so as I alluded to earlier in in Eric's Q&A both of you are coming from a field that is adjacent to energy storage when you started in the field I think this is a really great example of interdisciplinary research and maybe you can also talk about what got you started in the battery field coming from the polymer side well the I was working with skin inspired polymers so we have stretchable electronic polymers, self-healable electronic polymers and our approach is always when we design some new materials so we try to understand what's unique about them and then beyond the initial direction where we design them for then once we understand their unique properties we want to see where what are the other places where these polymers could be applied so between yeast group and my group we had some work prior to this looking at carbon materials and also conducting hydrogels for supercapacitors so then when we had the self-healing polymers then there are already there already had been some collaboration with yeast group so then it becomes something natural to think about oh we have this new type of polymer then can we do something in the battery field and then postdocs I remember I think our postdocs yeast postdoc and my postdoc they go to the dining hall together very frequently and the dining hall is an important part of the story so that led to the very initial collaboration of the first experiment and it worked so that we continued and becoming more and more interesting over time so now we are we really have a devoted effort to study this it has been really fun Eric I also noted in many of your work on solid state battery also involve a very collaborative team by Maryland in fact I think you are responsible for growing and nucleating that team as the head of the energy institute can you talk a little bit about your thought process and building that team up and and making strategic hiring for your institute well I mean it it was a natural thing that happened when I was brought to the Maryland actually Chenxing Wang was already here and he was really the first hire for the energy center and clearly you're familiar with his work in batteries not solid state but but the water and salt and the rest of those things then I was brought in and I was obviously most of my research was solid oxide field cells but just really solid ion conducting materials and then I had the opportunity to make a number of strategic hires both directly for the center which now is a state of Maryland institute but then also you know within multiple departments and so the first one I heard was Bing Hu who again another Stanford connection because he was working as a postdoc with you sway and Bob Huggins before that thank you for having him he's really taken off and and so that that was a great strategic hire and then we you know we we wanted to do more computational materials like like a lot of other universities we bought an e-fame oh who who had again had been working you know and computational on these things with dirt cedar and others and we just continued down that path most recently we hired Paul albertus who I mentioned his paper several times he was my program director at RPE but now he's the associate director of my institute so we just really you know it was a focus it seemed natural we had one of the DOE FRC's on energy storage before so it just seems like an opportunity to grow on a strength yeah I want the young students to hear about your comment that this is great and seeing how the young people you know go into the pipeline and really bubble up become super star in the field you know there's a lot of opportunity and now we know we're spinning off people that are going into other universities and and opportunities in the energy field so you know it's a great opportunity it's it's the place to be I think to me this is the most important work I could be doing and I'm excited every day to go into I would say go into the office but because of COVID it's my home office but but still you know it gets me up early in the morning I I literally get up at like 3 30 in the morning every day and look forward to work every day so I couldn't do anything else I'd rather be doing now I know you have one to find you this is a very convenient time to get to get to you so if I look at four of us well will you ask this great question you know turn on moving to the battery field you look at all four of us right here in the panel we all come from different field and start to work on battery so that's very interesting um now let me ask another question to both Janine and Eric um Janine you are doing organic you're doing organic electrolyte polymer Eric you are doing ceramics you know solid state I want to you know probe on both of you a little bit you know certainly I can see the solid state's value I can also see continuous organic electrolytes path if you compare these two organic electrolyte you know having the safety issue right there and as one of the main downside you know keep concerning the whole field the application solid state is safer so what if people can make organic electrolyte much safer would that be a you know strong contenders you know to say well solid state you know you will reduce the desire to do solid state well I want to give you guys the chance to express your thought this is not a conclusion this is just a I think discussion mutually how do not how do you think about solid state and Eric how do you think about liquid electrolyte like having a little bit of discussion I think it sounds like you're trying to inflame an argument it's to me that's exactly the part by the very friendly way um I about you know my batteries are a huge market and opportunity and and I'm not going to say that one battery is going to be the solution for all um you know there's start off with all kinds of batteries and lithium ion just took over everything and you know it's going to areas that we didn't expect it would it's doing grid storage that was not the expectation for lithium ion batteries it was because of its lower mass and into obviously initially consume electronics and then into the automotive but people thought that other things would be doing grid scale storage you know but because the volume has increased the cost has come down and it's competitive but there's so many other markets that we have not thought about getting into in the past and so for example with the solid state the high temperature capability is really unique now I personally want to believe that it's going to take over all markets but I'm not you know going to say that it is because there's a lot of competition polymers are great there's nothing against them and they have their their opportunities yes they're more flammable maybe you can make a non flammable polymer but you know the high temperature end of things I don't think there's anything you can compete I mean I showed you a battery operating under direct open flame in air and and so you know there's space there's military there's other types of applications downhole drilling things like that if you want to integrate a battery with a heat source let's say you want to actually put a battery you know talk about the 48 volts for for for automotive that you can put a battery under the hood and it's still internal combustion engine body provides a start-stop capability you can't put a lithium ion battery under the hood I don't care if it's polymer electrolyte or not that type of high temperature you can put ours there so there's markets for all of them and I think in all of them they're going to take the niche market where their performance makes them the winner and then we'll see how with time each of them expands in other markets depending upon cost and other factors yeah I completely agree with Eric that each approach has its pros and cons and but then the market is so huge there are so many areas needs a battery and to me I'm very interested in also batteries for wearables and implantables and where we want to have flexible batteries and stretchable batteries so those might be areas that organic can play a very important role and also the I think for the liquid electrolyte based on small molecules even though the current commercial ones are flammable but many of the newer versions of high performance electrolytes are more stable and non-flammable so it's very promising and that infrastructure of making batteries is already there so they if they reach the stability requirement then they can be incorporated into commercial battery very quickly then over longer term probably there will be all these new generations of solid solid-state electrolyte inorganic or could be polymer then they will each find their own niche to serve the space where there's a gap or there's a strong need that will pull them in these are great answers thank you back to Will well Zenon I was very intrigued by your your statement that many of the liquid electrolytes are dropping solutions for existing battery factories and you know maybe to you Eric I think the capital intensity of ramping up a different production system for solid-state battery must be daunting so since in our storage x initiative we also touch upon business and policy can you maybe discuss a little bit what might be needed in terms of support from government or regulations or policy that can help with the scale of challenges of something that's not dropping so um you know I'm going to give you the us centric answer to that one right we are competing on a global stage and we know where you know battery manufacturing is dominated it's in asia I mean that that's the case and so you make a drop in solution where's it going to be deployed first in the existing infrastructure so it's not really going to help us competitiveness but if you develop a new technology where now it's not currently being used in the bad industry like we're doing in fact it is an advantage for us competitiveness because it doesn't exist in the other place it's not a drop in solution it's going to be taken over by catl at least not immediately right i mean i'm not going to say long-term that how markets will pan out so that that is an advantage right that that it is a different infrastructure it does have different markets to address initially so again starting a new company going for the markets where our performance provides an advantage and we're not as cost sensitive then then it's perfectly fine how it goes long-term in the scale up that that's that's a volume thing um costs and all of these things always drops as as the volume of the market goes up as the percentage of the market you capture so it's not an inherently expensive process compared to conventional batteries again i mentioned before the ability to avoid the dry room for a large part of the processing is a significant cost advantage people make you know ceramics in large quantities technical ceramics the the industry the the the supply chain is there to do it it just hasn't yet been done for batteries so you know we'll find out how how how it all pans out in in the future but i still think that uh for the markets we're looking at as a company um the us manufacturing aspect of it that that we can make a profitable company and we will see how as we expand volume of market of manufacturing how that competes in the long term well the future is certainly very bright um well i think we have come to the end of our time here and uh Zenon and Eric i'd like to thank you both again for taking your uh morning to talk to our audience here and Justin if i can have the introduction to the slide please so i'd like to remind people that we do not have a seminar two weeks from now because of Thanksgiving early happy holidays uh to you that celebrates and um on December 4th we will return to our x equals question mark series this is going to be x equals longer duration storage and we are very happy to feature um two leaders uh working in the area of long duration storage uh Mike Aziz from Harvard University and George Crabtree from Oregon National Lab who also is a director um for the Joint Center for Energy Storage and i hope you will join us then and thank you again for tuning in today thanks for having us thank you very much Eric and