Yi Cui – Materials design for grid-scale energy storage | GCEP Symposium 2012

okay for our next speaker let me introduce professor each way he is an associate professor of material science and engineering here at Stanford University he also has a joint appointment at SLAC National Accelerator Laboratory and he has been a jiseop researcher for quite some time been very prolific leads a vibrant research group in addition to his research roles and teaching roles here on campus he’s also an associate editor of nano letters and a co-director of the Bay Area photovoltaics consortium which is funded by the US do II and he’s won numerous awards during his time year here I’d like to just point out a few the Wilson prize which the award was awarded last year the david filo and jerry yang faculty scholar 2010 the sloan research fellowship in 2010 and was a global climate energy project distinguished lecturer in 2009 so please join me in welcoming each way thank you Tom for the nice introduction so this is my starting idea at Stanford right here let me tell you my personal experience on G sub so I joining the faculty in 2005 and in 2007 my group has a power 10 graduate students in post-op you know how the things work in turn your check you start a package going down like crazy and I literally had the nightmare look waking up in the middle of the night thinking about how do I feed this a 10 why’d you steal him post up so my food is a big funding convenient that’s the Jesus I found funding that really saved my group allow us to continue doing research and the air wheels were interested in so now let me use next 20 minutes or so to give you an oval over where what’s happening and Annie love under the GZ project on the materials design for Greece scale storage this project was started in collaboration together with Professor hugging’s Istanbul right here initiated by a graduate student calling vessels he graduated and a few students joining in mara richer and Hume work they are doing a lot of amazing research on this topic so let’s look at the what do we really need right here we are talking about energy storage for Greece scale this is because we will really like to integrate when and soul into the grid these are the renewable sources they are intermittent so did they create problems to the grid if you would like to have a lot of them into the grid so you would you you see it is a minute second two minutes minutes to hours fluctuation in the power so you need to smooth out that a fluctuation and also when you are generating electricity at a time that might not be the time you use all of them you know you need to start that we need to have a so-called picture shifting from hours to days with this in mind now let’s look at what’s really needed for gryska energy storage compared to the energy storage such as bare as we know about and what’s the difference right here first of all you needed to be low cost because the scale problem you’re talking about is too big and if the cost doesn’t doesn’t meet the requirement there’s no way you can implement this in a larger scale so a good number to look at an issue cause from do e is a hundred dollars per kilowatt hour so you might be wondering this is too expensive our electricity only cause you know maybe ten cents also per kilowatt hour why this is hundred dollars that’s because you have a baby with a capacity of kilowatt hour you can use it for many many cycles at the end it’s the second parameter you should look at if you can get down to 2.5 cents per kilowatt hour per cycle that this might be become economic and let’s look at the scale also you know we are talking about each of this energy storage plan you want to have a megawatt to a gigawatt scare eventual adding together you want to reach terawatt in order to you know stolen in our electricity to regulate the grid so that’s a big big scale I’ll give you an exam for how big that is in the next slide lifetime 20 to 30 years right our babies and for yourself on your laptop you’re talking about three years now you’re ten house more life because your power plant is 20 to 30 years at least so you need very long life it is a gap right there and cycle life you want to charge discharge five thousand cycles what this is by assuming a roughly every day or every two days

you you charge in this chart we watch you calculate that you need at least five thousand cycles for doing that I’ll bet you can only run for 500 cycles for most of them you need high energy efficiency also too so much electricity pumping in and pumping out of these storage devices if your energy storage is not efficient then you really waste a lot of power so these are this requirement are different from what we are familiar with for the grid for the portable electronics way for the probably electron is the first thing you probably think about is energy density if i charge you have my cell phone how long can i use it before he looks like energy density is not the most important parameters by t is still an important in a sense you can affect the cost profile eventually so let’s look at existing technology for the grid scale they are not adequate the biggest pie right here the biggest one is a palm hyjal you know the area right here really miss Pam hydro are dominating all other technologies just you know consider you can neglect them you know the contribution at this moment so let’s look at the palm hydrous the problem ism let’s say the good things first it’s capable it’s a low cost it’s a long life that’s really amazing you look at this big damn right here where you just keep pumping water that’s very very simple but you look at the what’s the downside is a location dependence you probably don’t want this in your backyard and it’s a low energy density is to load that’s why you need is big bigger you know like Tae Bo structure right there and it’s a energy features it needs to be improved as well well let’s look at the bear face and see what about biophys beth is just so far away for meeting the requirement we mentioned in previous slide if you look at the cause it’s too expensive this is a course in the system level this is not individual in a system level what this is a horizontal line right here is the boe target you know all this battery technology are just too all these storage technology is just too expensive and looking at the barre just know enough cycle like we calculate the cost per per cycle energy and this is the target you know palm height you can be the target but other things are really a to too far away so if you look at this you say if I can improve my bed it’s like alive that certainly has an impact to the right hand side parameters the cost per cycle and we also need to think about a way to reduce the initial calls as well now let’s talk about this problem of scale last night I was preparing my sly I sit down I do a simple calculation based on how much your little my baby will produce each year let’s look at the recent data based on the reason they’d had a yearly production of lithium-ion batteries and can power the grid giving the power only for the gigawatt for one hour so if you want to look nervous the whole was greedy only can only store the electricity for you know probably ten minutes also that said you know we spent a whole year and everybody can know you cell phone your laptop’s all contribute their bare fees and doing the grid scale storage so that that’s the problem the scale is really really tough if you think about to get to tell what scale you need 30 years of production of lithium and battery to get there so the weenie really fundamental way to think about how do we do electrochemical energy storage the good thing about electrochemical energy storage is energy fishes is very high and if you can solve those a cost problem and is a hope we can use that for the grayscale looking at a bare face you know this is a basic structure let me use lithium-ion batteries and example you have to electro negative electrode positive will actual cutting onto a metallic foil you have a lecturer in between look at this fundamental structure the materials course is important how you may bear is important so we need to think about ways to change those things fundamentally so looking at the what the things you need to do this is not a comprehensive list but certainly you need to look at very low cost abundant materials to meet the course and also the scale problem from the for the electoral I for the separators for the election materials you need to go for example 500 cycles to 50,000 cycle we need hundred times longer cycle life and we need to think about different processing but has even be even device structure to my bare face now let me mention let me talk about the the work funded by G sub you know we jump start this project using our funding but now’s transition into the district funding this is the materials we work on cooperation blue is a material known 400 years is a material you used to dine your blue gene into blue color it’s very low cost Prussian

blue has a iron iron son I grew right there potassium you see this is all abundant you might be concerned about cyanide group right here it does not release or cyanide gas until you go to extremely strong as it lights a ph 0 type of acidity he has attractive feature it has this open frame of structure if you look at the basic structure right here these are the side of us are in peace I these are the metal I and iron and iron right they’re linked together by the CM Punk having this open frame of structure it has this an open channel this pink color that’s the location you can stop you can stop potassium you can start show them you can store other ions if you look at take one of these a cross-section a image and look at this this is the site you’re talking about look at looking into the detail this has the radius roughly about 16 angstrom you look at these metal ions are commonly used for battery lithium sodium and potassium they fit in there very easily so this open framework structure really welcomed ions to be inserted this potentially can create a really good materials for high-power fast applications and you can also tune this materials chemistry well you can use up you know p sy and answer can be iron and iron you can put in kabul you know low-cost you can put a nickel as well you can put in other metal ions to tune is a electrochemical potential so this really allow you to have a lot of a knob you can tune this material 2 2day degree you really like them to be so you we use are really simple synthesis to make this material in temperature precipitation very fast I for example if you want to make a Prussian blue you put in potassium iron you know I iron 3 plus 2 plus we sign a group precipitate out you will produce these up the nanoparticles of these materials right away really simple you can replace eye on these others you can control what a that’s potassium or sodium you can control p or are so forming these on the analytical structure if you look at its performance something amazing it’s really there you measure how you charge up this theory let’s use a potassium aqueous electrolyte so aqueous electrolytes feel very low cost but has a very low cost we know organic lecture I use a little man bear it caused a lot of money now you cannot start to charge up your batteries and it is charge your battery now you see one thing that’s the difference between the charging and discharging voltage is so small this tells you this materials extremely fast we are using this roughly about one hour charging in this charging speed the wattage hysteresis is done to a below 10 milli watt also it’s very very hard to find electrochemical system visa such low hysteresis if you speed up the way and go form all the way down to 83 c 83 c means once is run our 83 c means 83 times faster so you’re talking about roughly what 40 cycles are charging discharging for very big lecture we are making this is very fast hysteresis is actually still consider very very small for this so fast charging way but you notice one thing right here the charge storage capacity you can put in a 60 this is only one third of the lithium-ion battery material such a lithium cobalt oxide we be used so in terms of energy density this is not the materials for high energy density but we are trying to develop these low-cost material high-performance materials for the grid scale that’s a different argument right now and also looking at this structure this is about roughly at the 1.2 or 11 also vs. or standard hydrogen so this is for the cathode and aqueous solution now let’s look at further so this is also working quite well for sodium it has this open channel now we change the electrolyte to sodium nitrate by molar also during these amazing fast kinetics so if you will compare this with well-known lithium-ion battery material very fast material such as lithium ion phosphate and leave them titanium oxide the capacity retention at the top right here for our materials is comparable or better so these these fast kinetics we ever do do we have done be a long term cycling here’s the psycho performance right there no after the 40k 40,000 cycle you still have more than eighty percent of the capacity retention so you are talking about you have a hundred times longer than the usual berries you’re familiar with this is exactly

what you need for the grease scale applications this can impact the cost profile Oh quite a bit now let’s look at why these materials are so good and if you are you charge in this charge of your bare feet depending on state of the charge right here you measure the x-ray diffraction and see this diffraction peg shift and the degree of shooting is very very small we are zooming into this diffraction pick a lot if you calculate the lattice constant versus theater the degree of charging idea and you calculate the strain is only pawn one per set it’s a nearly zero strain right there that tells you when sodium or potassium ions at this case is sodium go into the crystal structure it just fits so nicely it does not need to expand the crystal structure construct a crystal structure that much if you look at the common lithium-ion battery material you use you know a carbon or lithium cooker bauxite you have up to about eight percent of the volume expansion right here now you’re having it is merely zero strain materials that’s the way to guarantee you have very long cycle life without doing other fancy you know materials tuning so as I say this material has very rich chemistry right there depending on you choose copper to go in together work with iron or the nickel you can tune its potential you can do the error you can tune its potential to be a position you would like you can also even choose once the cat eye and you are going to insert we we talk about just talked about sodium and potassium you can you can look at the ammonium you can look at lithium they have slightly charging potential when you look at Liam the charging potential right here is lower that’s because lithium has really strong solution in fact is water and that who is the strip of the solvation shell it takes some energy actually it make it harder for these crystals to take medium so the potential needs to be lower so what about the negative electrodes we were talking about the positive in order to make a food sales working foods you need to write negative electrodes ideally you want the negative electrode potential is close to zero what was the standard hydrogen electrode because you want to max out the battery voltage you wonder you’re positive lecture as high as possible the neck theory lecture as low as possible without splitting the water it turn out and you look at the development of a crisp air is so hard to find i equals negative electrode it’s just very very hard there are not many materials available even though available date their performance just so bad so we started to tackle this problem and this is a PP biopolymer it can actually you can do oxidation and reduction on this polymer when you turn now when you do the this solid curve right can you do the reduction right here you see this a pig close to zero world versus the standard hydrogen electrode but this material is not good to be so negative a lecture by itself because when you want to do oxidation this this plateau right there on the top and then this voltage just spent two very large range eventually you look at our oppression blue family material right here and when you do the discharge right here you will lose a lot of voltage so the cycle of is very bad people try to do study on that now let’s look at what’s the thinking right here we also know supercapacitors where you can do you can take activated carbon and to make a super capacitors activated carbon usually stuff is a potential power point 25 watt also due to the surface charges stay and then when you charge in this charge of these are super capacitors and the wattage will change and it actually has amazing cycle life but if you super capacitors activated carbon mechanism to combine with our a Prussian blue family the voltage you’re going to lose by a lot and I be idea is become what we come up this idea is to use a PP by as an additive to tune the electrochemical potential of the activated carbon now pull down this whole curve down so at the end you made official you can gain a lot of voltage for your bare fees we actually make it this is a reason they have very new to show you now we can use a chick using sodium borohydride to actually pull down these a curve two minus a point 25 watts and then we make a food sales right here this is the food cell voltage now you have a 1.21 a crisp air is a very fast potentially long cycle live we actually prove that we can do a long cycle like 1000 cycle without EK so this is still going and also you might ask the question let me use one last minutes to talk about this why don’t we also designed the open

framework structure having the right potential for the negative electrode we actually took this challenge and a poster in the group model past a pastor come up this idea really great and now replace iron with some manganese certainly play with the synthesis now we we find out when you do the CVR measurement you now to see if you use up station state of two to three now you can have materials with potential that 0 was a standard hydrogen or you turned out this one also has very fast kinetics when you do charging and discharging virtually no what is hysteresis very fast materials we are making food sale right now to testing out the performance of the of the whole berries and I also like to extend it further this open frame was chuck is ready to take ions then why don’t we think about this you know instead of just working on mono valent ions such as sodium potassium Elysium what about we do thi Valen that Waylon has the benefit one died Waylon iron goes in can you can charge two electrons so one atom two electrons that’s a really good deal right there you don’t need so many ions to go in so it turned out this open for infrastructure is amazing for Darrell and I on insertion as well we can do now magnesium we can do calcium we can do strong than we can do barium now we have all kind of a bare face now coming up by whele we are thinking about trivalent as well in the future now let me end my talk by summarizing we are developing a quiz perry’s new materials locals a very fast high energy vision to do a you know single waylon and i will embarrass i would like to stop right here thank you for your attention questions I professor trees thank you for your talk that’s very exciting research I’m wondering if you could talk a little more about tuning the the metals in the mineral organic framework you showed that pretty quickly and it looks like you can increase the voltage that way but what are the trade-offs and how far can you get with that approach the question was using these metal ions inside the open plain what structure how far we can tune the voltage range and principle we can tune the wattage strange very big because all this metal ions redox potential is the document in the in the in the database you can use that as a reference certainly when you put into the open flame but the potential will ship a little bit and this is a you can turn it a above water oxidation potential below the water reduction potential it’s all possible right there the actually the problem becomes whether they’re going to be stable in water they’re going to split water or not that’s the key requirement right here so far the highest one we can get is roughly 11 to 1 point2 wat and the highest the lowest one is about minus point to what versus a standard hydrogen beyond that you start to spread water so you are expecting bump on to what you know to 1.5 volt battery is also russian blue also is electrochromic that’s why I own for some years so there’s other energy applications for your research in controlling windows for instance absolutely a you are absolutely right the Passion blue when you this family when you are you know let’s look at this to picture I they have two different colors right there you can tune the colors yeah I just wrote a small proposal and using a Prussian blue for the electrochromic window see you know this is we didn’t wait out right now yeah thank you looking at the future if you have to store gigawatts hours at what voltage would you store I mean would you have a huge number of of these devices in series without striker the voltage and and and store stuff at at the vault scale how would you do that so this is 1.2 what you will absolutely need multiple bits perfect connecting in serials as well as Imperial parallel to build up a current series to build build build out the voltage certainly battery management system becomes important later Oh in the like in the tesla car yeah certainly in the tesla car is more

challenging my dear because you’re carrying the bear is following a bower bumping into other things to have aston and this is for stationary storage i will see the challenge will be less and managing the battery system charlie hey this is really exciting I had a question about calendar life I don’t know if they still die blue jeans with a Prussian blue but go up to san francisco owned by a really hip hair that’s really dark then three years later I look like dad from 1990 for how long I’m always worried about somebody asked me what about calendar life and look at that I say my glad you stealing only they stay here for five years yeah I so I don’t know that’s a good question Canada life is so important yeah we need to study that I by I also tell you the Challenger by five years squad you’re still in life so maybe I’m the only guy you’re sitting right there for 20 30 years old so it’s other corrosion mechanism though I I agree you know this many bad things can happen you have a lecture liar you have a lecture you have all the packages you need to take care of that it’s not trivial for 20 30 years any other questions well I have one last question for you e so you showed the 60 milliamp hours per gram I was curious how close that is to the theoretical maximum for this type of material Oh for these materials the theoretical maximum is somewhere around 60 to 70 so it’s already close to that so something I didn’t mention is in the crystal structure of this material is you can check water inside so that make our accurate calculation of the capacity becomes challenging desert a little bit aerobar doesn’t affect any conclusion right here so we don’t know for example is 65 or 62 because of the variation of the water content okay great and if there are no more questions let’s thank you one more time thank you