 The most obvious way to extract the energy from the Sun is to use the solar radiation. And this is a huge energy input on the order of about 10 to the 5 kW striking the Earth's surface. About less than a thousand is usable due to mountains, oceans, and so forth. And we're only using about a fraction of that even by 2050. So this is a very abundant energy source. However, the Sun is not without problems. It only shines about half the time and it doesn't shine everywhere. Even if you think about a statewide Texas, for example, how many people call Texas? Do you see Sun all the time? Maybe not into someone when they're under storm and so forth. So when you have loss of solar coverage, then you have the problem of solar electricity ending very quickly. If you have cloud coverage, your power can drop by 80% in a matter of few seconds. So this is one big driving force for establishing a way to store the electricity that we generate for renewable sources like the Sun. Wind has exactly the same problem. If the wind doesn't blow all the time, if the wind stops blowing, then you have to have other, what we call peakers, coal-fired power plants in order to make up for the demand. So if you look at some of these control centers for power generation, they're actually watching weather patterns. And they say, oh, the thick of gush wind is coming from the Rocky Mountains. So we need to cut down generation for coal-fired, and there's no wind, we have to boost up the coal-fired. So the ability to store electricity even briefly, maybe on the order of a day, will make a huge difference when it comes to realizing carbon neutrality. The second driver is human mobility, more specifically sustaining it. So if you look at a lot of the developing countries, they're striving to be more mobile. You think China, India, and so forth. But of course, kind of a funny thing, if you look at a lot of the developed countries, we're becoming a little bit less mobile. So this is the contradiction in its own box. Everybody wants to be mobile, and we need to have a way to be mobile without causing damage to the environment. So obviously in this sense, using electricity is one of the best ways to pass no pipeline. We don't make CO2, we don't make water that as a point of use. So there's a way to pipe renewable electricity from the sun or from other sources to mobile applications like electric vehicles that we can have human mobility without all the environmental impact. So I would say that these two factors, realizing carbon neutral energy and sustaining human mobility, is what's really mobilizing us to develop storage technology to take something from the sun all the way to mobility, all the way to energy. So let me go into a few specific cases, a technical aspect to why this is a problem. This is a grid load balancing application. This is a power generation curve in California one day in 2013. 2013. And this shows basically as a function of the time of the day how much generation from the grid do we need. So this does not include the renewables. So this is from nuclear, from coal fire, other sources. And what you can see is if you follow the curve for 2013, it's pretty flat, but it goes up around 6 p.m. And the reason is because we come home at 6 p.m. and in March you may be turning on appliances. If you go to August you'd be turning on ACs and so forth. And it causes a huge spike. It's on the order of about 10,000 megawatt. More additional power is needed across the three-hour period. Then you go to nighttime, this is midnight and then it goes down. But what's really interesting is to be forecast down into the future. So from 2013 to 2020, that's only say four years from now, then there's a huge dip, and this is known as a dip, that there's so much renewable by 2020, mostly solar panels in California, that you actually don't need to generate that much electricity anymore. So this sounds like a good thing, but actually a really bad thing. I don't know if there's already a talk mark on grid, or will be a talk on grid. This is a big problem, you're suddenly cutting in half of the demand for electricity. Electricity is conserved, so you have to generate exactly how much you need. So when you have a boost of renewable coming out, then you have to dip your non-renewable sources in order to get the grid. This is the problem of storage. To have some way to move all of these excess electricity to the nighttime, for example. So the nighttime, the generation is quite low, but what if I'm able to reduce the generation required overnight? What if I'm able to basically make up for this giant requirement using something like a battery in order to provide transient electricity without having to turn on a coal-fired power plant or natural gas power plant? This is one of the biggest driving force right now, and I will show here in the next slide how to make the problem. This is a legislation passed two years ago by the Public Utility Commission in California that mandates the installation of electrical energy storage. It requires non-hydro, so the classical way of storing electricity is to pump water up an elevation, and then you flow it down to generate electricity when you need it. This is the classical way, but of course, if you calculate the potential energy in MGH, it's not much, you need a huge amount of space and you need water. So the legislation requires a 2 gigawatt of power installed by 2020. 2 gigawatts sounds like a lot, it's not a lot. Remember the figure I showed earlier on the slide is in terrorists. So we're still off by three orders of magnitude, but this is the start. To show you how big of an improvement this would be, this is where we were at in 2015. We're on the order of just a few tens of megawatt installed, and the legislation requires us to go to 2 gigawatt in 2020 in three to four years. So this is one of the integration with policy trying to get us to look at storage more closely. Motivation, as I mentioned, is mobility, and because we're something valid, we always make an example of the Tesla. So what is the question I'm going to get at in a bit? So you can cycle by our three characteristics of any battery system, which translated to electric vehicle energy density means range. How far can you drive? In today's Tesla, if you don't use your AC and you're in relatively warm area, you're looking at about maybe 250 miles of range. But if you are in Detroit, for example, then you're looking at 150 miles because you have to keep it at reward and you have to turn off the heater in the car. Power density more specifically talks about the recharging time. So if you don't have the right every single parking move, that would be great. But the recharging time, typically for an electric vehicle, is an order of many hours. If you come from China, for example, it would be hard to find a parking spot to park for that long unless it's overnight. And even if it's overnight, so the ability to recharge your car asks you to fill out the car just like you would for an internal combustion engine when you put the gasoline, because this is the most important one, it is the cycle line. For an internal combustion engine, recharge is not a problem. You can essentially fill on an issue. But for batteries that recharge in cycle lines, how many of you change iPhones or smartphones? So this is one of the biggest problems of batteries. For a portable electronic, it's made about 1,000 cycles if you recharge 1,000 times. What I want to show you now is a comparison to where we should be. These two plugs show you the power purchasing agreement for solar and wind electricity, so these are renewable sources. And it basically shows nowadays, so this is in 2015, we're approaching around 20 cents per kilowatt hour for solar. Let's do some simple math. What does this translate to for the cost of storage? So if I am generating at 20 cents, in a few years we'll be generating at 10 cents and in a few more years we'll be generating probably below 10 cents. So the storage should cost no more than the generation that possible. So that's what a metric here, I would like to store at the same price as generation, 10 cents per 250 mile range of electricity. I like to charge my car if you're driving 300 miles each trip times 500, that would be the 150,000 miles. That's roughly above the warranty of the car market. 100,000 miles for warranty or something like that. Mark Mark is the best one. So if you take between 500 cycles at 1,000 kilowatt hour per cycle, energy storage costs will be commensurate with the generation cost. What are they going to charge for 10 cents? Basically you don't have to make more batteries when you sell a car, the batteries do. But they're not really battery charged, but what we really need is something operating on the order of computer chips. They can perform calculations billions of times without data. So one of the big challenges we have with battery is getting the cycle life up because you amortize the cost of the battery over the number of times you can recharge it. So if we can get the cycle life up, we can make a huge difference even without changing the energy density more than the power density. And are you today that energy density is not quite there but when you're getting close. But the cycle life is nowhere close to where we should be at. And I would say in the next 20 years or so we will see battery back in cycle for tens of thousands of times rather than just thousands of times. And this will really change the business model because then when you sell a car, the batteries do go. Let's start thinking. Tesla probably won't be very excited about that, So what is a battery? Why do we need a battery? So I keep it at a very high level. A battery has a good reverse. That's how you store a battery. Energy is a battery. So if you have a low energy reservoir, energy reservoir, and you have a low energy reservoir, for example, for a winner. If we are not able to separate the traffic of ions and electrons, it will just be combined. It will just make heat. It will store energy. And what we have to do here is to have device material that is selective to the transport of ions and the transport of electrons. So in this case in blue, this material is selective to ions or selective to only electrons. So what happens is the ions flow inside. This is like water flowing down and the elevation difference. And then you have the electrons flowing outside. So this is like the generating turning and then you have the electron flowing through the circuit. If you keep this tuned traffic of ions or electrons apart, you will be able to store energy. So what is some of the key design trade-offs? It's always good to look at this because we know what we're trying to get to and we'll react. As I already mentioned, three of the main factors are the energy density, how many times you can cycle, and then also universally related, meaning if you have a battery that can cycle many times the cost, we'll go down, especially the capital cost, I'll make you do that. So let's take a look at the energy density of fuel, such as hydrogen, gasoline, and so forth. And you have incredibly user-fueled without damaging and insertion reaction. The mobiles we see often lithium in inertial batteries today were basically looping lithium between two reservoirs. So one way to do it is to stop the lithium in something else. In your cell phone, and these materials take off the lithium into the material without massively changing its shape, size, and so forth, the problem is that you have to carry a lot of dead weight because you have to have a host to store those ions, and that puts a specific energy density quite low, but because it's a very simple we're able to do this in and out many thousands of times, and this is what we're using today. Now, if we want to boost the energy density, what we can do is, well, why do we allow the material to change massively? Rather than just putting a little bit of lithium for a big host, wouldn't we put a lot of lithium? So for insertion reaction, we're often only putting one lithium, say, for one molecular formula weight of the host. But why don't we put five lithium? It's a huge reaction where we are reconstituting the battery every time we charge and discharge it. So we're making a new material, and then we're taking it away each time. However, the specific energy goes up, but the cycle of energy goes down because we're massively changing the morphology, the shape, the size of the material. It is basically a nightmare when it comes to a mechanical engineering of the material. Then we can go to transition growth reaction. What if we don't have a host at the host? That means you're getting every bit of energy without anything when you grow things like lithium and often grow like dendrites. And these dendrites are very bad for safety and also for the performance of the battery. So why do we get extremely high energy density getting up to that of fuels? But the cycle of ability is extremely poor. And when the cycle of ability is poor, the cost is very high. So how do we tackle some of these challenges? One of the ways we can deal with this is to up. A challenge that we need to address here is basically to tackle the transport of ions and all-iternal batteries over a diverse range of length scales. In a battery, you have about nine orders of magnitude of length scale, all the way from centimeter in a double-sized battery all the way down to M-strom in the individual atoms of electrons and everything in between. You are straddling device macro-meso-atomic. There's virtually no other chemical technology that stands this range of length scale. And one of the things that we're doing at Stanford, myself included, is to borrow a page from... Indigo imaging has transformed medicine because it allows us to study the human body from here alive. How we've been able to understand these degradation mechanisms is actually using the port's motive approach. We let the battery die and then we see what happens to it. So this is the fundamental viewing point of a battery and you can see it's actually really small. It's only on the order of a few microns. And here, it happens, we can understand how to improve the battery. I'll show you just a few cool videos of what we have done in the lab. This is a 2D image. The color tells you whether the battery is charged or discharged. So if you look at red, that means it's not charged and green is charged. And you can see in a battery it is extremely homogeneous. At the boundary of red and green, you have a pressure that's equivalent to about 1,500 PSI. So imagine we applied 1,500 PSI. So one of the major theme in battery engineering is how to illuminate the mechanical strength, especially when it comes to insertion reactions. And this is just a movie. So this is what we see in a post-mortar image. But this is what we see in a dynamic image. We now have the technology of working with folks at Slide National Lab to actually image battery one particle at a time. So this is really bringing the length scale challenge. We're now at the length scale of some microns. So this is a mesoscale range I was talking about. And by understanding how the gradation is happening, we have now the ability to rationally design batteries that go from 3,000 cycles to 30 working on alternative ways. For example, Professor Ben Amali in chemical engineering has been looking at cell healing polymers to basically suppress the effect of the particles. Bringing the battery into the part. If you can hold it together, that will still be fine. And if you surround all of your active material with the cell healing polymer, you have an opportunity to do that. So we can also use nanotechnology. This is an example from a professor in my department where he is looking at how to use pure lithium. As I mentioned, pure lithium is great, but it grows like dendrites sharp pins that will penetrate the battery. But what do we put the lithium metal in a thin container? Not. And we'll insert lithium into the layers of graphene sheets. And he showed that it's possible to completely eliminate the growth of dendrite having this very low volume post to hold the battery. And you have the opportunity to increase the energy density further by another factor of 10. So this will really help you to move up this particular trade-off this way. Okay? So with understanding the deformation process and putting cell healing materials, we can push up the cyclability with nanotechnology. We can also increase the specific energy density. But how do we get off the block? Okay? This is probably one of the most interesting questions that we get away from this. One of the common things about me describing to you is all battery material involves some sort of solid. You're inserting lithium into a solid. You're growing lithium dendrite, which is a solid. Okay? And solid is up to work with. That's why you have mechanical stress. What type of matter doesn't happen in it? And that's what we've been looking at as well. There are several groups that stand for looking at this for many, many years. In the 70s, GE commercialized a high-temperature liquid metal battery. So imagine solder. Okay? Solder is a liquid metal. You melt it at about, say, 200 or so Celsius. And then it becomes fluid. If it's fluid, it means it can regenerate itself with a pump. Okay? So if it was out of whack, you pump a little bit of it, things come back to normal. But the problem that they include is that it doesn't make it as efficient. On the other hand, we can also use a water-based flow battery. This is great because we can go to room temperature, but one tricky thing with water is only stable up to a voltage of 1.23 Volts. I think Thomas already mentioned this as well. So you can have a battery that is more than 1.23 Volts. A living amount of battery today is close to four Volts. Okay? So automatically, you have a 3x energy penalty. You develop a room-temperature liquid metal, so it is a liquid at room temperature to be able to have the flexibility and a high-voltage of a liquid metal without having the issues of a flow battery based on water. To take a look here, these four readouts couples show you the voltage of a flow battery with water to about 1 Volts or so. But if you go to a liquid metal battery, you go to 4 Volts. So automatically, you have a liquid metal at room temperature. And this is a picture of the liquid metal working in a battery. So it's nothing like what you would imagine for a battery. A battery, you think of as a solid state device. But actually, if you take away that constraint and go to liquid and you can solve many of the degradation problems. So these are just a few examples of the way of working on a scamper and swag. Hopefully, we give you a bit of an introduction. E, I think, and then you circle back to the original motivation. Energy, storage can really enable human-only energy scheme. And this actually are two of some of the biggest problems we have in the world today. So with that, I'm going to thank you for your attention and I'll be happy to answer any questions. Why don't we talk about energy through but instead of capacity? We can answer. So the question is there's a lot of talk about how the energy, how the cost is amortized into the technology. So it is indirectly included. Increasing the energy density by a 30% percent. And how does the mind close the fluid as well as energy? How competitive is it like cost-wise? How competitive is cost-wise? And myself is is it going to be we don't know, we can do the calculations we want, but until we make a fairly large order of a few kilowatt hours we won't know the actual cost. And fundamental mechanism whether there's any show software we can make techno-technology analysis but we won't know. So I'll be honest to you. High because present technologies are rather expensive. And innovation beyond the university to see where it should impact my DNA but we at the university need to be aware of what may be down the road. The manufacturing is high and power density of super capacity the second life of fuel cell is even unlimited. So is it possible to combine all these technologies in one single electric vehicle? Right. So the question is electrical devices like a capacitor can deliver a huge amount of energy quickly but because the transport of ions and electrons in those devices are extremely important. So one thing that you can try to do is the decrease in a catalysis perspective. How do you get lithium to build a liquid to a solid more quickly? So it is possible today to realize battery technology that approaches a power density of capacity within the battery. Do we need to recharge a car in 10 seconds? Probably not. Do we need to discharge our car in 10 seconds? Not quickly discharge a car over several hours or even days. So there is a limit but I think we are discharging for using the batteries good enough. Power density of the battery.