 I'm going to spend the next 25 minutes to share with you personal experience past 10 years with G-SAP. So I have a little story to tell later. But let's look at energy storage. This becomes very clear. But 10 years ago, it wasn't so. So we look at this, you say, well, every year we sell about a billion smartphones, each one of you carry one. And electric cars are coming in in a big wave. And California requires renewable storage. Just going from here to the car, roughly you need to double the annual production of lithium ion. And then going from the electric car to the grid, you need multiple times. So this is a huge demand right there. Then you say, I ask the question. Everybody asks, what kind of batteries do we exactly need to meet with those demands? Let's use lithium ion as the benchmark. For portable electronics, there's no question. It has to be lithium ion for transportation. Also, it looks like it needs to be lithium ion because of high energy density needed. So this is cell level and system level. Let's look at both. Cell level, we have roughly around 200 or so while per kilo. For the long term, we really like to get to 600, three times. But if we can double, that's already very, very good. But for academia research, we like to see whether three times is possible. System level, you cut by half. That's the number you are getting. So this is referred to the pouch cell. When you go to 1860, 50 cell, the cylinder cell what Tesla is using. So this number will be higher. This will be 250. Then you will say, I want to multiply three times. That's 750. This is kind of cell level cost. We are somewhere around 150 to $200 per kilowatt hour. System level somewhere between 300 to $500 per kilowatt hour, roughly this range. So we will like to get to system level 150. Cell level probably 70 or so. These two are really critical. These two for the transportation. For the Grease scale storage, of course, energy density will not matter that much because the footprint can be bigger. However, it will matter. It will affect the cost eventually. Cycle life, different people have different number right here. This calculation is based on you charge your car once per day. You want it to be 10 years. You need 3,000 cycles. For the Grease, you need a lot more. You need 30 years, perhaps. However, I have to say this number can be deceiving. If you have a car like Tesla, it runs 300 miles per charge. You only need 500 cycles. 500 times 300 gives you 150,000 miles. That's plenty. So depending on your range, how big your pack is, this number will change. Safety, this becomes more and more important. So sometimes no seven accidents, certainly really brought everything into consideration again. Can we really make our batteries very safe, fundamentally safe? That's a key question as well. So let's look at the grand challenge of the batteries right there. Giving those numbers. These are some of the questions we ask. Do we have the batteries chemistry to give us 3x of energy density compared to the current technology? How do we do that? Can we get the cost 3x lower? Can we make it safe? If this happens, these great things will also come. So past 10 years, I stand for right here after joining faculty. So we opened up a bunch of a number of research program trying to adjust those challenges. I will pick some examples based on the GSEPS research. If you look at these three grand challenges right there, one major thing is actually the energy density. Increase the range. If the manufacturing cost per battery is the same, you also reduce the cost of the batteries per kilowatt hour. That's really a key thing to do. So graphite is a material for 25 years now, used as an anode to integrate lithium to store lithium. We have new materials such as silicon and lithium metal can do 10 times higher capacity. So if you could use that, your energy density will go very high. Cathol side, this is the lithium metal oxide, no matter, it doesn't matter it's a lithium copper oxide or lithium manganese oxide or lithium myon phosphate. The family of these three different materials roughly get to somewhere close to 150 to 200 million per milligram. You also have a choice of sulfur become lithium sulfide. That gives you also 10 times. So if you can make this happen, this is a theoretical energy based on graphite, silicon and lithium metal. You have chemistry available to do three times. You even have chemistry available to do six times. So if you can make this practically working, then you are getting somewhere. So let me use silicon as an example. So silicon I show you has 10 times capacity. Silicon's a problem or also very similar to other battery materials problem. Silicon can take a lot of lithium coming in, 10 times higher capacity. However, volume expansion is four times you are facing two big problem. Number one is breaking. Number two is how do you be able to stable interface of silicon facing the liquid, organic liquid electrolyte? So this is the first of this project. I believe this is the batch of project Jesus ever funded on the batteries. So I joined the faculty 2005, starting roughly January 2007. This project coming in, I give a brief title is a nanowire batteries with me as PEI, a free princess code PI. I'll come back to the story later. So over the years, so we developed the tool to really understand how the battery charging work. This is in situ holder, now available at Stanford University. Have a lithium-cobal oxide cathode. You can put your anode right inserted into this ion in the liquid and then you use electron beam to watch what's happening. You can deposit particle and see what happened to particles as well. So I want to show you a video. This is video of silicon nanowires. During charging, the volume expand a lot. This is 200 nanometer scale bar. So it start from 200 and expand a lot, but it doesn't break. And the surface has a copper coating. The copper is broken. So this is really powerful process. This really give the idea how silicon can break the batteries. Now let me show you another video. This is a video with a nanowire and then a number of silicon particles right there. Some are small. There's one really big one. This is 800 nanometer in diameter. Once lithium coming in, volume expand a lot. Causing silicon interface shrinking create a more of a lithium silicon phase. Eventually the stress is too big. This particle cannot hold it. And it's going to be broken, but the small one is stable. So once this is broken, you lose the material. You lose electrical contact. The batteries die. The first time you charge your battery, you find out it's already close to be dead. That's the challenging back then. So using nanomaterials engineering, we find a way to make this to work, prevent a breaking. So that's this whole thing coming in. We have a generation of one design. That's nanowires. Diameters small enough they don't break. Then that's not enough. We want it to be more stable. We have a core shell design and a hollow design. After a couple of years, it seems like breaking problem is solved. But remember we have a number two problem. What about the interface? If something keeps expanding during lithium coming in and contract when lithium goes out during discharge, you don't have a stable interface. So our generation four, I consider as another big milestone after generation one is engineer a double wall hollow structure. The interior is hollow. This blue color is silicon outer layer. The red color is a mechanically strong coating. Once lithium coming in, charge up your batteries. And the volume expansion happened by expanding towards outside, inside, not towards outside. Outside has organic electrolyte. So you build a stable surface. So we adjust this interfacial stability issue. So indeed after this paper, there's a really numerous study in the literature mimicking our idea. Now can make all type of materials to work, known as silicon, germanium, tin, many others you can name it. It's trying to create a stable interface. We have been through 11 generation now. Let me highlight just two. Another generation is even it doesn't break. You have stable interface. However, nano structure will have too high surface area to react with organic electrolyte. Consume lithium, consume electrolyte. That's not good. So we took on this challenge to design a new generation. We call this as a permagriny like structure. Basically packing this nanoparticle into secondary structure and coated outside with another materials so prevent the electrolyte to come and effectively reduce the surface area and make it stable. So it's getting much better. And also along the way, this is another later recent years project just finished supported by GSAP, is what if you do have the breaking? Can you come up with an idea and let the breaking self heal? So together with Professor Junan Bobby you sell healing polymer. Self healing polymer has this functionality. It has this chemical hydrogen bonding. Hydrogen bonding is very easy to broken. However, this polymer chain can move around and find each other and reform the hydrogen bonding. Eventually self heal the whole electrode. So this self healing can be better seen. Coating it is self healing polymer on top of a balloon and make it visible by putting conducting carbon black. You can measure its conductivity. You blow out this balloon many times by still conducting the surface. And then you string it back the conductivity nearly recover. Showing you this self healing polymer is a very strong. It actually can make a micron particles to work. The micron silicon particle, they will be broken. They are different from nano particles. However, now using self healing polymer you can start to cycle this. So now let me come to the next thing for high energy materials such as silicon. After roughly 10 years of research we understand the material design principle. What about safety? The more energy you put into the battery if there's something bad happen you are going to have more energy to burn your batteries. So can we design something that really prevent the bad things happen? We have been thinking about this for a couple of years now. Let me share with you two ideas. So you have seen all kinds of accidents. I haven't had a chance to put in the Samsung cell phone yet but everybody knows about that. So let's see if you have a battery with cathode and anode. So what happened very likely is either due to internal reason or external reason. You have some sort of shorting. Once you have shorting happening well let me mention this reason the external reason can be accidents. You know outside there is a you know you bomb into your batteries inside can be you have manufacturing defect you have over charge or you charge it in the cold weather. Once you have shorting next thing happened is very fast release of electricity through this short. And it's going to heat up the batteries to roughly about 100 degrees C. Starting at this temperature you are going to have exothermia reaction of solid electrolyte interface the interfacial coating. During the battery charging you have decomposition of the electrolyte. This release heat once it goes roughly about 180 degrees C something really bad happening. The metal oxide on the cathode will react with organic electrolyte. This will just automatically go cold thermal runover. At this moment the battery was you cannot control it will start to catch fire very quickly and or even cause explosion. So in order to make your battery safe internally you don't want your battery to short. Can you really detect the shorting? You know you have manufacturing defect you might have over charge. So we come up with an idea in the battery separator right here this is cathode this is anode. We embed it over porous very thin roughly 50 to 100 nanometer metallic coating. Once the shorting happened halfway for example danger formation you measure the potential of this intermediate layer versus your anode you can see the voltage drop to zero. At this moment you know your battery is in danger you need to stop charging. So you're still safe you still have halfway to go. So your laptop if it's your laptop it just tells you shut down your batteries within a minute if you don't do that it will just automatically do it for you so to make it safe. So we actually show this idea work. Now the question is what about external extra? You don't have a control you know your car bomb into something and then do something you know really penetrate into your batteries. So what we want to control right here is prevent very fast release of the electricity through the short. So together with Professor Junan Bao we come up with an idea is on the metallic foil this current collector we coated with very thin layer of polymer. This polymer inside embedded with this nickel nano spikes at the room temperature they are connected electronically they are conducting as a metal but once this temperature heat up due to shorting and this polymer thermally expand and once the expansion happen it's going to pull this metallic nano particle open and electrons cannot move between these particles anymore. So the reason we need these nano spikes is when this spike touches with each other electron can go through tunneling you know very close distance. Once you pull this open by an extra or two the tunneling current scare with the exponentially with the distance one over distance. So and your color will go very small so you can have very very sensitive switching phenomenon. So we take these nano spikes of nickel coated with graphene graphene is to make a nickel stable chemically and you are embedded into polyethylene or poly problem it's in the polymer very low cost. And now you test out the resistance versus the temperature. Let's look at this green curve right here with about 30% of a nano spike of nickel in there you heat up this polymer film you see at the beginning is metallic conduction at about 90, 95 degrees C this sharp transition become a complete insulator the resistance change is eight orders of magnitude become a complete insulator. Now the battery material coated onto this polymer there's no way you can transfer electron very fast to your metallic foil current collector. So you prevent a very fast release of the energy due to shorting even you do have a short but it's not shorting through your metallic current collector your current density is reduced significantly so you don't heat up your batteries. So I want to show you this is the battery we built and the room temperature you can see you can get the regular capacity out we engineer this polymer film at 70 degrees C it start to respond and it becomes this battery become basically dead there's no capacity coming out but once it come back to the room temperature the capacity recover so this is a reversible switch and the small cell this is working very well we want to do the big cell testing and your realistic cell phone batteries and also bigger one for the car and see how this idea will work. So let me come back to the GCEP story. I want to mention something about GCEP funding impact. Some of you might know many of you probably don't. Roughly about 2005 to 2008 adjoining faculty 2005 and 2005 and 2008 roughly within this range there's very little federal funding available. DOEBS just cut their electrochemistry and battery program no funding EI still have some basically nobody is funding the battery research. So for about these three years when I write proposal to outside it doesn't get funded we were very badly because nobody believe batteries problem is important. But GCEP recognized very early. So the first GCEP coming in is this number. I still remember I was so happy to receive this funding otherwise you know your tiny case will be toasted completely without funding. So I want to also show you after that what happened and about 2008 also there's a big explosion of research recognizing the importance of the battery research for the electrical car and then later for the grid. So 2008 I co-founded Amphilus I co-founded sitting in the audience right there. Up to today roughly we have 60 acts of follow on funding compared to this initial one from GCEP and then there's a huge number of battery program coming into Stanford. Roughly there's about 100 acts of return including the battery hub and also the battery 500 consortium and also these also spin out new idea. How do you make catalyst to do a few generation borrow the idea from batteries to make a new catalyst. Let me mention a few things right here. Number one is Amphilus. So after initial nanowire publication so you have went through a multiple run of funding now there's several million batteries on the market. So now we learn how to use silicon for the high energy battery. So this is happening. Just to share with you a few pictures right here. This is Amphilus Wuxi plan. This is our board of directors taking pictures roughly about a month, two months ago and Wuxi China to celebrate this manufacturing plan. I also like to mention battery 500. This is a new consortium announced several months ago by White House. Battery 500 here means 500 watt-hour per kilogram of the batteries for transportation. DOE really like to double to triple 500 watt-hour per kilogram also roughly get to if you have a Tesla for the same size of battery pack it will go to about 500 to 600 miles if you can get to this 500 watt-hour per kilogram of batteries. So after the initial investment now we have very strong team here at Stanford. Indeed we are the co-lead with PNNL National Lab to lead this $50 million consumption. We have four national lab, five universities involved. We have a Tesla IBM to serve as our advisory board or industrial advisory board. So these are some of the key people you see right there. Some familiar face, so Jernan is from here. You see John Goody now, you see Stan Wittingham. Stan a long time ago used to be a post star with Bob Harkings. So you see a number of people right here. We identify the key chemistry. How do we produce 500 watt-hour per kilogram? How much time do I have? Probably getting close. So I will skip this then. I will come to the last slide. The future of the batteries. If you look at high energy for transportation. The current graphite, nickel manganese cobalt oxide, these combination anion cathode, roughly get to 300 watt-hour per kilo. That should be okay. Silicon combination with MMC can get to 400. And lithium metal and high nickel MMC get to 500. I think this pathway is highly promising. In the next five years, we could get to somewhere 350 to 400. And 500 range is very promising, lithium metal and sulfur combination. So we have a pathway to get there. But for grid scale, the story might change slightly. Lithium ion battery cells might still play very important role in the grid scale storage. But these are the ideas coming in possible. Aqueous solution batteries, low cost and very safe and also flow batteries. There's multiple choices for the grid scale. Well, let me end my talk by thanking my whole research group and GCEP support and also DOE support over the years. And a number of collaborators working together at Stanford campus right here. The culture is fantastic. The collaboration is just very active. Thank you for your attention. And for all the great work you've done over the years, we are getting close to the time to finish, but if there are one or two very quick questions, we'll take them. So I see there's one over here. So if there is a different technology, different chemistry, different physics that can go past the 500 and possibly sooner, is that something that you would consider? I will, yes, I will. However, getting to energy density, to get the batteries to work, there's multiple parameters to consider. There's some other things, maybe one parameter is so outstanding, but the others are so bad. So it needs to be a balance. You didn't mention anything about super caps in relation to this and especially the new work in graphing super caps. Maybe you can say a little bit about that. Sure. How they compare. So super cap has very high power, but energy density is very low. Roughly, you consider that super cap is about, it's better to get to a 200 watt per kilo super cap is in order of 30. So seven times difference. So I think for very high power needs, super cap can play important role. For the usual needs, if you do transportation, super caps role is not that big. Because the cost will be very high. You have seven times reduction of energy density. So the cost per kilowatt hour will go very high. But we start saying I think this certain special application, for example, if you do a bus, bus run usually, you know, 50 miles perhaps, you can do fast charging. And so every time you go to the end, you do fast charging and then come back. It might work for the case of the bus, but for the regular car, super caps role will be not that big. Even with graphene, it doesn't increase that much. I think graphene is oversold for many cases. It's over, it's oversold. Okay, well, I'm afraid we have to stop there. So let's thank you one more time. Thanks you. Thank you. Thank you.