Linda Nazar: Energy Materials & Climate Change

good afternoon my name is Ashok Gadgil I’m the division director for environmental energy technologies division and it’s my great pleasure to introduce professor Linda nazar she’s a professor of chemistry physics and electrical engineering at the University of Waterloo in Ontario Canada and as as the three disciplinary core overlap indicates she has actually fingers in many areas all of them focusing on materials for energy storage and conversion with research that spans lithium ions sodium ion batteries lithium sulfur batteries lithium air batteries and fuel cell catalysts professors NASA has a number of awards and honors and is and holds a senior Canada Research Chair in solid-state Energy Materials she was in 2009 they received the electrochemistry society’s battery research award last year she was the more distinguished scholar at Caltech and this year she received two awards the International battery Association award and IUPAC distinguished women in chemistry and chemical engineering award and she also has been elected to fellow of the Royal Society of Canada so without further ado professor Messer I think I I think I’m wired up well thank you very much for that kind introduction that was some very very nice it is really my pleasure to be here and I thank you for the invitation to come and visit and I’m sorry about not being able to visit in January I’m somewhat better now and so today I’m going to talk about energy storage the the part about the climate change is really just to move people into the talk I’m not really a climatologist and I’m not going to be talking too much about it this pretty picture is not of Waterloo I’m sorry to say it’s of Toronto but it’s our Airport which is about 45 minutes from my house if I Drive quickly and at Waterloo we have an Institute for nanotechnology we also have Waterloo Institute for sustainable energy and Institute for climate considerations and also quantum computing and all sorts of other things and I am primitive dominantly affiliated with these two institutions so we don’t have a material science department which kind of makes me be localized someone so I will address a few issues of climate change but if it said very very briefly and just to bring up a couple of points this is in May I watched the BBC news and a regular basis and read their website and so on the 30th of May they pointed out that the global carbon emissions have reached a record according to the IEA and this is of course another wake-up call pointing out that we’re not really going to reach the level of emissions that we should not be reaching until 2020 if this so-called 2c target is going to be attained that is the target for the reduction of emissions and whatever one can say about global climate disruption certainly carbon emissions are a part of it I think that most of us would agree so it turns out that the day just before that on the 29th of May the BBC happened to report that the Saudi prince bin Talal said that he wants oil prices to drop so that the United States and Europe don’t accelerate efforts to wean themselves off the country’s supply and he clearly actually said this I saw the interview we want we don’t want the West to go find alternatives because the higher the price of oil goes the more they have incentives to go do that so of course the economy seems to be fixing the problem right now because the price of oil is down but that’s of course not what we really want to have happen and in Canada we particularly see the effects of this with the reduction in the sea ice which is this data is from the Copenhagen climate science report in 2009 and this has only become more significant with the DES munition of the sea ice and of course for Canada this means that now a lot of water can be opened up in all kinds of economic advantages that prove it of course is an enormous amount of economic disadvantages are all crew from this so I think it was safely safe to say that this is a threat to humanity in the 21st century if it continues unabated and it is certainly a scientific challenge of our time so energy storage is certainly not going to fix the problem there’s many solutions or many ways that one can

approach this and that includes investing in catalytic chemistry more energy efficient reactions and processes co2 sequestration for example photovoltaics most certainly perhaps hydrogen generation and energy storage is on the bottom of the screen and if not I don’t know if it’s possible to fix that the people that are running the video but anyway there’s some energy storage at the bottom so as I said energy storage is only part of it it’s really what we’re looking at is trying to manage energy demand through energy storage and this is obvious I think to many people in the audience solar wind and wave are intermittent supplies that have to be coupled to energy storage devices in order to effectively make them dispatchable sources of power and of course there’s also the electrification of transport and this is very much a commercial concern at the moment and to a certain extent a government concern and this can lead to of course to reduction in co2 emissions and the conservation of oil so energy storage is critical for both of these it would be safe to say that lithium-ion batteries are probably not going to be economically viable for this purpose although the jury is still out on that the moment there’s probably other forms of energy storage that are necessary and certainly lithium-ion technology or lithium technology is going to be an important point of the electrification of transport so I stole this slide from the do II and this is what in electrochemistry we call a Virgo nee plot which is just a plot of energy versus power and so this axis for the non electric chemists in the audience basically refers to the range and this axis effectively refers to the acceleration of one’s vehicle and so this is good old wet acid and this is nickel metal hydride that is still driving the Toyota Priuses and this is a lithium-ion cell as it currently stands today so if we take a look at where we need to go in these systems Phe phe’s are for at least short-range transport in the vicinity of 20 to 40 miles pretty much adequate if that’s all that you want to go the problem is that of course people don’t want to drive for just that distance there are various targets that have been suggested arranging from a hundred miles to 500 miles but perhaps 150 is appropriate and and certainly none of the batteries really match the easy goal at present so of not surprising we need to find better G energy storage systems so it turns out that electro chemists are first of all in short supply in the u.s. perhaps not everywhere in the world they are also insert supply in Canada but it’s um it’s a well worn technique some people think that Volta was responsible but it appears that in fact it was developed in well a long time ago so these devices which are quaintly called the Baghdad battery have been unearthed they exist in the Baghdad Museum and they are excavated by from Sumerian sites dating back to 2500 BCE and it’s effectively just a good old electrochemical couple between there’s a copper and iron rod that sits in the center copper sheet on the outside and it has thought that the electrolyte was something like citric acid or vinegar or other type acidic solutions and when these things are coupled up they actually produce a current um woman who was at the naval research labs Deborah Rawlinson and I wrote a where we edited rather an Mrs Bulletin Edition which has just come out and Deborah actually has built these in her lab just for the sheer fun of it to prove that they work and so you can see a picture of one of them not from the Baghdad or not the museum but the actual battery so it would appear that the Parthians inherited their batteries from one of the earliest known civilization and of course this is just the general concept of energy from chemicals using the well known Gibbs equation so then if we consider what electrochemistry can do from us for us well there is electroplating which is possibly what they did in with the ancient Sumerians so this is just chemicals to other chemicals through electrochemistry there’s also chemical energy to electrical energy and work which is the normal discharge of a battery for example the functioning of your primary cell that you use in all the time and then finally there is the you know the real frontier which is converting the electrical energy to chemical engineer energy which is the battery recharging and this is the of course the storage problem that challenges pH EVs and the renewable energy technology so if we go back moving on from the Sumerians a little bit forward in time to the lead acid battery developed by plant a in 1859 this is a pretty simple reaction and the point here is that this reaction works in the bulk and it’s actually pretty amazing that it works at all so there’s two electrodes and effectively it’s the

conversion of lead to lead sulfide sulfate or let oxide to lead sulfate which drives this the EMF difference between the which drives the reaction and of course sulfuric acid is an electrolyte and that has a pretty poor gravimetric energy density and some of these numbers will come up again so perhaps keep them in mind so this is 30 watt hours per kilogram with a volumetric energy density of 60 watt hours per liter and it has problems of durability and self discharge but they still work pretty well and in general one I think can say that the success really depends on the kinetics of the electrolyte and the electrode processes and so the persistence that the technology is really around despite this unfavorable mass because of that kinetic factor unless you think that everything is known about loud acid batteries and that they’re really boring my other favourite magazine is economist and there’s an article here on Einstein in car batteries so I won’t tell you all about the article because it’s rather long but the point is that apparently the batteries in your car actually have something to do with Einstein and there’s even a fez reflect article just out this year all about that so there’s always more Under the Sun than we think ok back to a step next in history so from 1859 we now moved to sort of the 1980s and this technology was being developed at a lot of places around the world at Exxon corporate research where I started my postdoc but even prior to that at the University of British Columbia where I did my undergrad and this physicist professor Rudy Haring he didn’t really patent the first rechargeable lithium battery but they were the first to commercialize it so this slide is slightly an error because it was the first as I said successful system and he was based on a reversible reaction in the cell which involves a nice cheap material molybdenum sulfide and one can then use this as the positive electrode it will react with lithium and when it is deleted it forms a highly conductive form of the love to them sulfide so the beginning the starting material is actually a semiconductor and this reaction changes the phase and turns it into a great conductor by processes that are understood by solid-state chemistry and it was this thenn that really started this company they use metallic lithium as a negative electrode because they didn’t know any better in 1979 and ultimately the technology failed due to problems with dendritic growth of that metallic lithium their negative electrode now back in those days they were using electrolytes that we we we’ve gone far beyond that separators weren’t developed all sorts of things weren’t developed but it did sort of jumpstart a lot of interest in the area and so today we have lithium ion cells and this little cartoon shows how they work so in short while you watch this thing cycle around I will explain that no longer at least in a commercial lithium ions do we have carbon so this is the negative electrode which is carbonaceous electrode which replaces the lithium and a positive electrode which intercalates or incorporates lithium at appalled at high redox potential versus the negative electrode so during the process have discharged the lithium ions enter the positive electrode and during the process of charge they entered the neck they enter the negative electrode and so the lithium ions travel back and forth between these two host materials now this cartoon is carefully designed such that it marries this process so this is the process on discharge and the voltage goes down in the cell and then we recharge the battery we shove the lithium back into the carbon and it comes back up again and you can see that the energy difference between those two electrodes actually changes as a consequence of the fact that the Fermi energy of this positive electrode is changing so we call this a double intercalation cell sometimes it’s called on the Japanese used to call it a shuttlecock cell but the point is that it works very well and it’s also limited by both of the capacity of the positive and negative electrode to accept the electrons because in this process the electrons are transmitted through this external circuit so the positive ions effectively are just there in order to account for charge balance that’s all they’re really doing now they do however have to be hosted in either the positive or negative electrode so the limitations in the system are due to the fact that either mass this either one of the electrodes can’t accommodate that much lithium or electronics they can’t accommodate that mention electrons without undergoing a phase change so I’ll be talking later about systems that do not operate on this principle that that get around some of these issues so in the last since 1991 we’ve gone from small-scale to needs of large scale systems and we’ve also gone from micro to nano systems so back here in 1991 lithium cobalt oxide was the positive electrode in that cell that I just described and we’ve gone through some developments I’ll tell you a little about lithium iron phosphate today and

here we are well this is a little out of date but effectively we’re still here in 2010 with some other oxide material and lithium iron phosphate has now been implemented for those of you who are again not electrochemist we have a figure of Merit which is a gravimetric capacity which is calculated quite easily by the number of electrons transferred in the reaction Faraday’s constant and the molecular weight so obviously for a high capacity cell one wants a large number of electrons to be trained to be carried to are either a large amount of electrons in the reaction and one wants a low molecular weight and limitations to intercalation systems then rely on this value in because it’s very hard to transmit a lot of electrons in the reaction whereas the integration reactions and is somewhat less limited so the intercalation systems work really well for driving small portable devices and laptops and things of the sort but these days we’re trying to find car batteries for example to hit this is a picture of the Tesla because it comes from California and also we’re trying to get to small-scale systems for medical applications this just shows an implantable battery so along with that we have as I said gone from micron to nanoscale systems in order to achieve some of these the acquirements of these demands and this is just a cartoon that illustrates the advantage of going from a bulk material where you have a lot of mass transport limitations to a nano size material where of course the path link for transport is is far diminished and one can also then imagine going to nanostructured materials where they convey all of the benefits of nanoparticles but in a system that is somewhat larger in spans microns so in this case they materials will have a high tap density which refers to the actual density that you can get in the electrode but there can be low utilization due to these diffusion limits because of mass transport here we have enhanced kinetics these materials can be hard to fabricate in this case the as I said they convey the advantages of a nano size whereas they’re much easier to handle so just to remind you of those limitations of mass transport the time in any system for intercalation can be described as the diffusion square of the diffusion distance divided by the diffusion coefficient so you either minimize the diffusion coefficient or you change your diffusion distance pretty simple so the other than consequence of moving to nanomaterials nano structured or nano particulate systems is the fact that the surface reactivity is always a concern in any electrode so I mean in any battery so at the positive side of the cell there is electrolyte instability to oxidation and the negative side there is electrolyte instability to reduction so the smaller these particles get the more the surface area obviously is increased and the more destructive these side reactions in the surface can be so we always have to work within the thermodynamic range of the electrolyte and that’s what really limits the energy of the stone so we have our gravimetric capacity that I just described which is a combination of that at the positive and the negative electrode and then the energy of the cell gives us the voltage of the cell gives us the total energy so the capacity times the voltage is equivalent to the energy density so the more we can separate these two the higher the energy we can get but we’re ultimately limited by the electrolyte itself so I’ll be talking about positive electrode materials in in today’s lecture exclusively although there’s certainly a lot of work to be done on a negative electrode and in this case there is a particularly a problem with as I said reactivity with the electrolyte for with respect to oxidation so surface coatings and other processes have to be implemented so the first portion of the talk will basically be discussing poly anions and the first of these is lithium iron phosphate so it’s structure is showing here it is I often like to say that it is the along the lines of Helen of Troy who launched a thousand ships this structure has launched a thousand papers probably fifteen hundred at this point if not more and these are sometimes called generation 2 batteries although they’re almost generation 1 at this point because they are being implemented as we speak and here is a plant which is using these devices for energy storage in the 100 megawatt hour range this is a company down in Texas so this material called trip light with new iron phosphate is cheap well low cost it’s resorts unlimited it has high safety it has all of the right characteristics except for poor conductivity and it was reported by good enough in 1997 and it has been as I said the subject of much work since so it has octahedra of iron oxide with tetrahedra phosphates lithium ions running down these tunnels and the reaction can be described as this all of them looking effectively can be removed from the material at a potential of about three

and a half volts and it gives rise to a capacity a theoretical capacity of 170 which can effectively be realized and almost realized in practice our work on this started in 2000 well 2001 is when we actually published this and Michelle our mall at the time was I was a step ahead of us they were looking at carbon nano painting as they called it this is our work where we actually we took carbon precursors milled them up with I mean the lithium iron phosphate precursors milled them up with carbon and effectively use these as a kind of a shrink wrap to constrain the size of the particles so we ended up with composite materials that were about 15 to 15 percent carbon in which we constrain the growth of the lithium iron phosphate and then moving along to know 2004 we used our we considered a different sort of process which is effectively to use carbon thermal reduction to convert the carbon to a combination of Fe P and a metallic phosphorus iron phosphorus are been glass this is a metallic glass which has a conductivity similar to that of iron phosphide so this is a cartoon of an actual TEM image where we did careful analysis of the grain boundaries and learned this is the iron phosphate particle itself shown in yellow and the green boundaries shown in grey and blue contain this mixture of substances and it’s kind of complicated to go through in this talk so I refer you to our paper but effectively the process relied on the fact that if you take them at this iron phosphate which is slightly stumped stoichiometric and lithium and you heat treat it gently you phase separate out and iron phosphide pyrophosphate and material is quite subject to refer the reduction to fep so on the old Ellingham on the carbothermal reduction diagram this is more prone to reduction than is this so it reduces and this does not and it provides a basically casts out this material on the surface of the material on the iron phosphate and more recently so this led to what we consider to be a new area of surface transport of electrons and more recently Garrett cedar has used a rather similar technique to cast out a lithium ion conductor on the surface which is lithium pyrophosphate so this is just the same sky addict I just used the same cartoon showing now the growth of Li 4 P 2 O 7 on the surface of these particles and this is reputed to be a fast ion or at least a fairly good ion conductive material so instead of forcing out the formation of the fe p 207 he forced out the formation of the lithium iron by simply going sub stoichiometric iron so the same sort of idea and the other concept here is that this glassy coating of this iron phosphide iron pyrophosphate rather seems to control growth so they are able to produce defect free materials with a Zion conductive coating this material has enabled very rapid discharge rates with good capacity and is sort of a bit controversial so I won’t get into it so where do we go from up from there well we have new positive electrode materials this is just shows a whole lot of them most of which are theoretical so these are the dark little splotches here these are the ones that we already know about and the take-home message from this slide is that we can in principle get to capacity I mean energies of a thousand watt hours per kilogram by accessing some of these new materials and accessing those new materials is rather difficult so we have to rely on techniques that are not perhaps common to the solid state chemist ones in which we deviate from these classic ceramic methods and move down the scheme although the area goes going up into these other sorts of processes where we carefully control the chemistry so that we can access metastable materials ones that would literally decompose at the higher temperature and also possibly access nano dimension materials because ultimately high temperatures lead to large aggregation and large crystal lights so the ones that we know the ones that were particularly familiar with are always the ones that mother nature makes thermodynamically stable ones of both dimensions now Garrett cedar is calculated well well over a thousand new materials and so the hunt is on and the question is how to actually get them so as I said the discovery of our materials is limited by our our toolbox so this is not a picture of a politician in Congress getting really upset with the way things are going but this is a picture of a little hydrothermal worm they’re apparently Foreman hydrothermal vents in the ocean and we often wonder if we’re missing these and our hydro thermal reactors because the cat will control of the chemistry in these hydrothermal bombs it’s really very much a kind of an art and there’s magic involved and you throw things in and sometimes it happens and sometimes it doesn’t so as it said maybe maybe these

little guys are necessary but if you’re if you’re lucky you can control the chemistry so this just shows an example of the synthesis of a new class of materials these are looking in fluro sulfates and this material you can easily get 90% of the theoretical capacity at three point six volts a little higher than the all of u n– I just mentioned this requires virtually no fussy complicated carbon coding and it’s because it has a different structure which is called a tab right which has three-dimensional ion diffusion rather than one-dimensional such as the olivine and in this particular case we form the material using just a glycol and as our solve ol thermal reaction medium and the reaction is probably easier seen here and it’s what we call a top atactic reaction so we just take iron sulfate and effect this is the heptahydrate we convert it to a porous material by blowing off some of the water and then this material can be seen as an H Oh H so the H is exchanged for lithium and the O H is exchanged for fluorine and we end up with Li Fe s o for F and that little diagram is showing here so this produces a porous light porous material that allows really good electrolyte access which allows us to achieve fairly good energy densities at reasonable rate capabilities for the material we’ve also done modeling studies which are reported in this chemistry materials paper together with cycle islem at bath and we’ve learned that the activation energy for ion transport is about half of that of the olivine so not only does it have three-dimensional transport it also has a much lower activation energy for energy hopping and the point is that you can’t make this material under normal circumstances it decomposes before the kinetics will actually allow it to form so another material that falls in the lines of this are is a sodium ion material I’ll just briefly mentioned sodium ions cells are thought of as being a possibility for coupling to renewable energy they’re not going to be in automobiles because they don’t have the gravimetric capacity nor the volumetric but the fact is that sodium is abundant it’s less costly than lithium and the demand for lithium was certainly going to rise as auto manufacturers we hope start to incorporate these in hybrid electric cars so there is certainly a reason to look at sodium ion materials whether or not you think the reserves are limited they’re all in the high endian planes so we developed a material which is again looks a little similar to that lithium iron phosphate this is a completely different structure the material forms as this sodium compound the sodium can be exchanged readily for lithium so you can also use it for lithium purposes but the point is that there’s two sodium ions in the structure the pink and the green so to speak and on oxidation just as in the case of lithium iron phosphate you pull out one of those ions and so it turns out the pink ones disappear and the green ones remain in place and they effectively hold the structure together and they they stabilize it they pillar it and so that allows you to get away with a really low volume change of only three and a half three point seven percent between the reduced and the oxidized form and if you contrast this with lithium iron phosphate the olivine that’s about six point seven percent so it has half of that volume change which allows it to cycle much better we call this a low strain material and we reported this in 2007 this material also is metastable it falls apart at less than 650 so we have to make it either by hydrothermal or civil thermal processes because solid state reactions despite what the patent bivalent says don’t get you there because it it can’t be made at 700 under ceramic conditions so we were hoping that this would be a new family of sodium ion conductors it turns out that getting sodium ion conductors to work well is very difficult because sodium is just a little larger than lithium and so in most cases the volume expansion on taking out the sodium or vice-versa putting it in is enough to cause enough strain in the material to result in really poor kinetics so we were kind of lucky with this material because of this process of the so-called pillar ring and we’re hoping to find other materials that behave similarly so if I was to summarize at this point I would say that the way in the future is to go from nano crystallites perhaps to nano structures which is what I’ll be talking about in the second half to interphase organic polymer and inorganic material science which is difficult for any one lab to do because typically one tends to know one thing and often not a lot of others and this is to define materials with tailored properties to control and design grain boundaries at the inter particle and the inter electrode and inter battery level which should have a big star beside it because it’s incredibly important and then finally to turn back to the electric chemical reactivity of bulk solids like I was discussing with the old lead acid battery so I sometimes refer to this as research and that’s where you go back and you sort of did what you did before only you do it with a different set of

eyes that was actually from Gary McVicker who was a colleague of mine at Exxon he was the one invented the term so if we want to then think about going beyond limits of intercalation chemistry to higher energy cells where do we have to where what is beyond there so this is the NMC material that i this is nickel manganese cobalt oxide one of those layered oxides I described this is the this is all gravimetric energy density this is the same material coupled to a silicon negative electrode this is the spinel which is used in the Chevy Volt this is lithium iron phosphate which is actually made its way into the BYD car and you can see that there is you know respectable but not high enough gravimetric energy densities for these materials if you look at volumetric energy density now this material is looking rather good but then there’s these other gleams and folks eyes lithium sulfur and lithium air which obviously offer even from a volumetric perspective much higher energy density values and so we’ve effectively I think gone from way down here somewhere in the floor which is the development of the Sony battery through careful engineering to get us to these materials which have now been expanded with a range of new materials formed by a series of complicated reactions to now as I said what we call beyond intercalation so that’s what I’ll be describing next so this is just a comparison of batteries and fuel cells that kind of contrast the sulfur the lithium air and the ND and the old fuel cell itself so I’m going to start from the right and move this way and if you will cell oxygen and fuel both off-board and the product is eliminated go low of water so that’s a great cell except we know that there’s all sorts of problems with it so the lithium air cell is has oxygen as a not on board but the lithium is on board and the product is also starred on board which is ostensibly li 202 and the sulfur battery is rather like a typical battery but it has compared it it has something in common with this which is that the sulfur is also stored in a carbon cathode as the peroxide is stored in a carbon cathode here and again it has metallic lithium at least at this point as the negative electrode and so I’ll be describing these two cells nothing about fuel cells today so if we compare the reaction chemistry for these two systems are actually rather similar this just shows the voltage profile for the lithium sulfur cell and the reactions that occur and this shows the reactions for the lithium air cell so in both cases we have elemental either oxygen or sulfur starting off the protests we reduce the material to either form ultimately this soluble poly sulfide or in this case this highly unstable reactive intermediate lithium superoxide in the case of the sulfur cell all of the polysulfides is a cascade or a sequence of them here form a series of their soluble species and at this point this system becomes insoluble as you move through this part of the of the regime and in the case again comparing things finally at Li to us – we have an insoluble product as we do in the case of the peroxide so the oxygen mechanism has been elucidated by Abraham at Boston so in both cases the question then is how to reverse to actually make this cell reversible because especially in the case of lithium air the primary cells have been known for a while in a case of the sulfur the reversibility is actually pretty easy in the case of the oxygen cell it’s particularly difficult and there are different problems with each cell which are have prevented either one of them being commercially implemented in the case of the sulfur it is the soluble polysulfides that are the problem in the case of the lithium air cell it is the reactivity of the off of the superoxide and the in solubility of the peroxide that are effectively some of the problems so we have been addressing this problem through the design of porous carbons because those are required for the host and they’re required to electronically wire the system so this just shows a typical sort of cooperative assembly using surfactants to assemble a silica which is then converted to a carbon this as well established by many others in the field including Galen sticky at Santa Barbara and ryeong Rio and in Korea so if you control nucleation of these systems well you can end up with nano versions of these mesoporous materials you can institute Carrboro carbonized the surfactant rather than having to extract it and stuff in carbon afterwards to obtain relatively highly ordered materials this is a catalytic carbonization process and we’ve also used things such as metal foil Salonen to produce fairly graphic versions of these materials where you can see the graphic peak in the x-ray diffraction

pattern compared that of using sucrose for example for stuffing so these are just some of our initial attempts we’ve we’ve gone a long way in looking at different sorts of carbons which is led us into interesting avenues so I’ll briefly describe these then for the sulfur stone so I already explained that you start off with elemental zap sorry elemental sulfur in this salt and push the wrong button so this is just described by this block of sulfur and if we then start reducing we go through our cascade of polysulfides including s8 – – and the s4 – – species and in principle these are then contained within this carbon host which is not shown in the diagram if you continue to reduce you end up rather than sometimes reduction in forcing what’s called a shuttle mechanism because of the solubility of these sulfides they diffuse out into the electrolyte and they will engage in a sequence of redox reactions where the more ox were more the reduced species is sequentially reduced and oxidized within the electrolytes so effectively this transports over the negative electrode and back to the positive electrode and this little shuttle mechanism operates within the electrolyte it actually serves as a safety mechanism but it also reduces the efficiency of the cell dramatically so this is to be at all cases under all situations avoided or it is to be avoided at all costs so if you can’t avoid it at all cost that means containing the soluble polysulfides in the positive electrode and that then allows one to get down to deeper capacities there’s a lot of reports in the literature should anybody review be reviewing lithium sulfur cells that have present capacities about 420 or 500 and the fact is that cells that usually cycle about 500 milliamp hours per gram usually just work in this soluble poly sulphide regime so in order to get down to the deeper capacities where lithium sulphide is formed in the positive electrode one is its it has to contain these and at that point then just at the bottom the lithium does not have carbon this is carbon free so the carbon the lithium the sulfur and the products are contained in the carbon host at the positive side in the long run so that’s a big question perhaps we can save that for the question period currently the cells that we were exploring how putalik lithium is a negative electrode now in 1979 it was not considered safe we’ve come a long way since then we didn’t even have laptops in 79 so ways to protect the negative electrode which are on the next slide and a variety of other things can possibly get us there but really the aim is to go to different negative electrodes such as silicon ok so that’s a short answer to your question yeah but carbon just here so this rather high capacity of 1675 results from the fact that we don’t have an intercalation chemistry reaction we just have a simple reaction of lithium plus sulfur to give Li 2’s so I call as I said this an integration process so the good thing about sulfur is that it’s really abundant if you have to go to Jupiter there’s tons of it there this is a volcanic moon Isle which is absolutely covered but it’s only 150 tons on earth so we don’t have to go to Jupiter anytime soon and it’s resource unlimited even more so than L IFE po4 it’s all over the place petroleum tar sands all over the place and the resources between Canada and the US account for two-fifths of the world’s total so we don’t have a problem there this is from the US government database so there’s two approaches to solve for batteries and one of them is what I call or what most people call a cath light cell so that’s where you just operate in that limited regime of the soluble polysulfides typically speaking so these ultimate Lea would deposit on the metallic lithium negative electrode and cause all sorts of problems insoluble products so such cells have to have a protective coating which is an ion conductive glass so that insulates those polysulfides from reduction and therefore they do not reduce to form the insoluble sir precipitates and the cell can can carry on so these are absolutely required these were commercialized by silent our working together with poly plus and these tools have some advantages they have high power because they’re effectively a liquid electrolyte system the product is stored in the electrolyte but they have rather poor volumetric capacity and that includes well typical values are 100 to 150 watt hours per liter now in the case of the cells that we are examining we’re trying to develop we call these contain cathode systems so we try to constrain the

polysulfides to our carbon host this gives rise to much higher volumetric capacities I’ll show you one for just our cathode around 2000 watt hours per liter so substantially higher that you do have somewhat lower power characteristics as a result of this so I’ll just point out that all of our studies are conducted with an unprotected metallic lithium negative electrode which we consider to be the worst case scenario and once one starts to do fancy things with passivating that negative electrode well it gets a lot better but we’re trying to really see what our positive electrodes can do so we just said leave that unprotected so for those of you who are not familiar with this this is work we published in 2009 which is what we called proof of concept so we organized that carbon those little micron sized arrays I showed you about three slides previous we organized these they have carbon fibers between them that effectively span organized space which is comprises pores of about 3 to 4 nanometers so this is a very easy fairly easily made me so porous carbon which goes by the name of CM k3 and it has conductivity of about 0.25 Siemens per centimeter and when we stuff it with carbon the conductivity is effectively the same so it’s like inserting a whole array of conductive posts in the sulfur mass and the sulfur fills this space when we melt it inside it imbibes by capillary action and we tune the space to accommodate the swelling of the sulfur to form the li 2s so we actually leave space here for the electrolyte ingress and also as I said ultimately they this is on solidification ultimately when the Li 2’s is formed it completely fills the space and this enables the back reaction to occur in other words the reaction of Li to s back to elemental sulfur because both elemental sulfur and Li to s are extremely good insulators so that level of contact is necessary in order to actually activate the electrochemical reactions the other important point is that that array of carbon wires which are slightly porous also are somewhat efficient for encapsulating the sulfur they’re not perfectly efficient so this shows the surface of the electrodes before cycling on the 30th charge 4cm k3 sulfur so this is just sulfur incorporated in the carbon and you can see in a 30th charge that we do have a sort of an unpleasant-looking coating on the surface which are those insoluble polysulfides on poly sulfide which have now reduced to form insoluble sulfides on the on the surface and the problem with that is that these are fairly impermeable to live human diffusion and therefore they they inhibit the reaction but still we’re doing pretty well compared to some other systems we have polymer coated visas these materials with polyethylene glycol and we find that this effectively completely inhibits this deposition of the glassy sulfides in the surface because they are well for a variety reasons one is that they’re insulating and the other is that they act as a hydrophilic gradient if we divide this by half which is a figure which is a typical fudge factor we’d be looking at about 600 to 800 watt hours per for a full cell and the question then is how to get this capacity cycling for a long period of time and the other question is how to get cheap carbons to do this so we’ve made porous carbon replicas from a colloidal silica this is just luda can see size controlled blue docks and these have a pore dimensions than the city about 12 nanometers with a suitably high pore volume and surface area such that we can now hire them out of sea mk3 so that we can now get 85% sulfur into these materials and this just shows the the sulfur filled material now we don’t actually go to that fill ratio because it doesn’t allow us the space for access we can get pretty close to that and what we learned when we did this is that we had quite a problem with poly sulphide diffusion because now we have these twelve not only your pores and a lot of the sulfur on the outside was escaping so we borrowed a concept from the drug delivery folks the pharmaceutical people who use this material SB a 15 which coincidentally is the silica that we actually use to cast the carbon in the previous studies I mentioned so these are Mesa porous silica is that have poor dimensions that are somewhat along the same dimensions as the poly sulfide anions so we incorporate these to the tune of five through ten to eight percent in our carbon matrix and on discharge to the point where we have a lot of polysulfides a maximum poly sulfide concentration these incorporate their adsorbed into these little silica particles but because this silica is insulating they don’t become reduced they just simply diffuse out such that at full discharge we now have the lithium sulfide or Li to s2 incorporated in the carbon matrix so they sort of act as a little reservoir to contain the polysulfides just where when we need them but they are then controlled they

are released in a controlled fashion on full reduction so the concept then is showing here if we assume a desorption process we can we can calculate on the back of the end of lok that about 20 to 30 percent of them can be contained and if we assume an adsorption absorption process up to 96 percent can actually be contained so this paper is actually cut out by now so the system actually the concept works proof of it is given from careful physical analysis I’ll show you the electrochemistry in a minute what we did to prove this was we carried out a SEM and EDX measurements of the materials either at this point which is the maximum poly sulfide concentration or at this point at full discharge so at this point the results shown here where we analyzed one of these SB a 15 particles we see that we have a high sulfur concentration and we reference to phosphorous because our electrolyte is Li pf6 and it’s everywhere so it’s kind of a normalizing factor so at this point we have sulfur to fosters ratio of 3.4 whereas the depth of discharge a full discharge it is only 0.2 this just shows the e-tax data and we are analyzing again an SD a 15 particle and we didn’t do this obviously for one particle we did them for 100 particles and averaged and what we learn is that 94% of the sulfur that’s absorbed at this point is actually released on full discharge and the other thing you’ll notice here is that there’s no glassy sulphide particles whatsoever in under 40 cycles so then to summarize the sulfur part we’ve taken this is the this is data that shows the sulfur in the electrolyte this is a careful chemical analysis of the sulfur concentration and if you just mix up carbon and sulfur within 30 cycles you have almost 90 percent dissolved in solution the C mk3 manages to hold on to most of its polysulfides for a reasonable amount of cycles so we were getting about 40% dissolution but if we if we modified with the polyethylene glycol we’re down to just above 20% if we go to a larger pore system we now have a fair amount of dissolution whereas if we incorporate these poly sulfide reservoirs we again diminish that poly sulphide in solution down to about 20% so as we’ve gone from a system where we have lower discharge rates and sort of unacceptable dissolution now too fast discharge rates at about a C rate with very little dis well 20% dissolution so this shows the most recent data that we’ve had which does not incorporate SD a 15 but it’s the same sort of principle just a different porous material and so we’re able to sustain about 850 milliamp hours per gram over a hundred cycles and that accounts for about 77% percent retention over that value so we’ve also been looking at bimodal systems and that’s reported in this energy and environmental science paper that I don’t have time to describe so this is data as a see rate and these the difference between the red and the blue are just different porous materials and so it’s kind of proof of a concept and this is without the additive the in black so you can see the effect fairly clearly okay and the remaining roughly four minutes of my talk I will tell you about the problems of lithium-air and so I want to say at the outset that although we are studying this we are very cognizant of the problems with this system and I certainly do not intend to overhype this by any stretch in fact what I’ll probably do is show you all of the problems more so than the benefits but there’s a lot of exciting stuff to be done so just like in the sulphur case you need a porous carbon as it said in this case to host the Li 2 O 2 which is the reduction product this has to allow all of the reactants to the active catalyst sites catalyst you say what’s that for well it turns out that oxygen unlike the sulfur system which cheerfully reduces and oxidizes oxygen needs a catalyst both for oxygen reduction and oxygen evolution and especially for oxygen evolution this is an issue so the poorest carbon system has to host the products it has to maintain electronic conductivity it is the worst aspect of the fuel cell because you need a triple phase boundary and the carbon surface area is particularly important and this has been studied by a host of people over the last couple of years and what we have learned is that as I said there are many problems so the first is the electrolyte problem this superoxide species just like it is in your own bodies where it is causes Aging in this case reacts quite vigorously with the electrolyte if it’s an alkyl carbonate and produces this whole host of materials which then engage amazingly in oxidation but this is not a lithium air cell and it’s rather difficult to get these materials to actually oxidize on the cycle and this was first shown by at least in the literature by Doron Orbach in 2000 Mizuno really clearly showed this electrolyte decomposition pierre bruce

has elucidated how it decomposes and this group at IBM has determined what the stable solvents are we have been looking at some of the salts and we’ve discovered as I said that this tears up almost anything so even things like lye Bob and Li df4 are chewed up by the peroxide in this case for example to form lithium oxalate so we have a big problem the first intermediate is a real issue so we don’t use L kill carbonate or lye Bob or Li bf4 we’ve been rather slow to because we’ve been trying to get things as my colleague yang shellhorn said it’s good to get the systems worked out so that no one has to then publish showing how you were wronged so if we were trying to get there before other people show we’re wrong so the IBM group has discovered or is claimed that these are stable electrolytes we have learned that these seem to be reasonably stable so we’re now looking at efforts of catalysts and trying to understand that process better and the role of surface defects I won’t have a lot of time to tell you about this but I’ll just show you some nice pictures so this is a nun catalyzed system and this is just carefully designed carbon that’s all there is to it so you see a nice flat discharge potential exactly where it’s supposed to be no slopes that indicate all kinds of electrolyte messes this is characteristic of a two-phase reaction as as one would hope switch pointers here and is characteristic of peroxide formation and we see this in the XR d it’s quite abundant and a little hard to see here from this number but this gets to about seventy nine hundred million powers per gram for carbon because it’s a nicely defined carbon and you can also see that the peroxide has nucleated to form these interesting little toroidal aggregates so the this is at 2500 milliamp hours per gram along the curve and it’s a little hard to see here but those are showing here these are sort of pre toroid shown by these little yellow dotted lines and you can also see real towards that are starting to form with these white circles and this is characteristic of peroxide formation and the important thing here is it’s it’s not filling the pore space the way we think it’s actually growing right out of the surface of the carbon and this is further along the curve or you can see quite a field of these toroidal aggregates again pulling or literally coming out of the carbon surface and these don’t really grow in size they start off at about 500 nanometers an even part way through there’s still only at about 700 nanometers so it’s the number that is increasing as opposed to the actual growth of the particles and you can see these form as I said these interesting torrents and these are actually agglomerative tiny crystallites that are about 15m in size based on sure analysis of the diffraction patterns so it’s particularly interesting to consider how these form because there must be some sort of dissolution and precipitation that allows these to aggregate in this form and I’ll let that open for the question for the question period at the very end of discharge again same thing more towards no particular changes and the problem is that carbon does not engage in oxygen evolution reactions that you have to take it to four and a half volts we cycle between up to four point three because we don’t want our electrolyte to undergo decomposition so the reversible capacity drops like a rock for a carbon type system and so there’s a real need for a bi-functional catalyst to decrease this potential on overcharge it’s not so much for this reaction but it’s for this reaction where you really need to minimize or bring down that potential in order to create reversible systems now if you of these have been mentioned in the literature alpha mno2 reported by bruce platinum gold particles reported by yangshou horns group unfortunately all of these studies were done in what we call B PC which is before propylene carbonate or before propylene carbonate was discovered as having problems and so there is carried out in a propylene carbonate electrolyte where the oxygen decomposes the super oxide decomposes a solvent so the reversibility in these systems is not li 202 being charged it’s all those nasty those nasty products so we still don’t really know a lot about oh we are that’s kind of the bottom line today we have looked at a new catalyst it’s a metal oxide we’ve made it porous and nano crystalline it’s based on a pyrochlore structure this shape data shows this is our favorite electrolyte right now and this is just shows the data for it alpha amino to either in PC or integ DME so the take-home message here is that just by changing the solvent you can increase the capacity even for a poorly functioning system and it’s poorly functioning because as you can see although it oxidizes it oxidizes those nasty decomposition products in PC it really doesn’t do it very well integrity about Tegh DME at all you have to go to high potentials to get that Rockside to oxidize so our system as I

said we have a me so porous version and a nano crystalline version this is the second cycle of the cell where the carbon has already kind of died off of it so this just shows the carb pure carbon material this shows alpha mo no.2 in the second cycle very high potential for oxygen evolution and this shows the catalysts in a nano crystalline form and the catalysts in its Meisel porous form so we’re able to get to fairly high capacities in these system approaching 10,000 milliamp hours program and we’ve lowered the charge potential 3.8 5 volts and we know we form peroxide because the XRD shows that clearly and the most hopeful side of this is that we can cycle the cell now this is only four cycles and I appreciate that this is only a tiny step but compared to alpha M&O 2 or no catalyst we are at least able to sustain this even increase the capacity a little bit on the third cycle to you know as I said reasonable levels what we’ve actually done is taken the cell apart and discovered that because these cells go slowly the rate capabilities for living there are pretty pathetic we have a totally fuzzy white metallic lithium negative electrode at the end and so what we’ve done is we’ve taken the lithium out and we put fresh lithium in and we can keep going for another few cycles so the problem really is protecting the negative electrode so when we count up the binder and the catalyst and the carbon we’re looking at about 1090 milliamp hours per gram capacity over those few cycles compared to lithium cobalt oxide so I’ll just remind you that the voltage is higher than lithium sulfur so that gives rise to a higher energy density so we can easily get 2000 watt hours per kilogram but really with poor efficiency because it’s still of that that over potential so if anyone had to ask me where I put my money it would it would definitely be lithium sulfur but the fact is that there’s a lot of intriguing work to be done in lithium oxygen I think as we start to the understanding really expands so this is just a summary I’m sort of not quite on time but this just shows where we’ve gone from lead acid all the way through the battery regimes of nickel blah blah blah the way to lithium-ion so I told you a little bit about sulfur and oxygen theoretical capacities that are achievable in full cells and sometimes people ask heavens knows why do you ever think that batteries will approach gasoline and so the answer is no they won’t now we don’t have the problems of the inefficiencies of the internal combustion energy to deal with when we’re talking about automotive transport either so there is advantages and I guess the other take-home message is that we’ve gone from bulk intercalation to nano intercalation to what I call nano integration and we’re getting more and more complex as the demands become more complex so for those who say that batteries haven’t come away a long way in the last 10 or 20 years well they’re just plain old wrong so I’d like to thank all of my graduate students who did the work they’re all listed here in most of them have contributed to this one couple of new students here david contribute a lot to the lithium sulfur work really did some excellent studies and he’s now at santa barbara it’s a postdoc working with Galen Snooki and my postdocs past and present and some of my collaborators and these people for funding and I thank you for your attention we have a full house even going into a time a little bit I’m sorry about that yeah I knew that would happen it’s just so much interest so I want to formally announce this close to those of you who have other commitments nearly but I would like to request five more minutes of questions I think we actually have 15 so we have five you say we have five we have five okay fire away yes can you turn on your mic please not working it’s okay I can hear you just no one else can I think also coming on the lithium there were the lithium sulfur depends very much as you pointed out on the on the molarity of the sulfur in the electrolyte so I wonder when you examine to yourselves with nanotubes and so forth do you actually determine the sulfur molarity of the electrolyte yes in effect so those results that I briefly showed on the concentration of sulfur in the electrolyte were taken from large cells where we ran them to certain number of cycles as I showed and we carefully disassembled the cell of

each point extracted the electrolyte and sent it off to Galbraith for sulfur analysis so we’re determining the molarity which is basically the amount of sulfur in the electrolyte and we’re trying to minimize that in our sorts of cells so but then in that case also it’s also the sulfur molarity in the cell so that is in the cell well how much excess what was the ratio if you take yourself and you calculate out what the molarity of the sulfur would be assuming that all the sulfur were in the amount of the electrolyte that you have you what was the number so that’s the percentage that I reported so when we measured that sulfur content those values that I showed you are the percentage of sulfur that is dissolved I don’t remember the molarity values because probably my graduate student work them out but those that is what is I think that’s what you’re trying to ask what you’re trying to get at which is what is the fraction that’s being dissolved is that what you’re is that sort of where you’re after no if you have a large excess of electrolyte on you oh I see what you’re trying you can end up with a low molarity in there but you’re still dissolving yourself away from that other part okay so this is um so I can get would have to get my calculator out because I don’t have those numbers at hand but these are low electrolyte cells so we don’t have a vast excess you don’t have a large off electrolyte so that the point here is we I mean really there are sort of nine drops of electrolyte in a small cell so it is possible in these cells to dissolve all of the sulfur but the point here is that we’re retaining eighty percent of it and that’s really to take home message now your question on I mean I’d have to work out the we’re certainly within we’re certainly operating in a regime where if we could dissolve dissolve all of the sulphur it would happen I think that’s the best answer to the question the question also with respect to the lithium air cell none of your tests were done with dry oxygen or with our experiments are done with dry on is about the viability yes the other question is that you also have to have a solvent that is stable lithium and you know this is some of the so-called stable solvents – Li – Oh Cal I – it – may not be stable well that’s why we are guided by the work of the group at IBM who are our collaborators and they are surest that these are suitable solvents now none of these are really stable with respect to superoxide formation but they’re somewhat stable so there are Saudis not okay with barrel Asia you’re dying for under question yeah I’m sorry can’t talk that’s two oh yeah I see what you mean yes no that is not for that is not okay for living and that’s just okay for the for the superoxide so we’re using tech via me and other groups at MIT are using DNA so we’re we’re using solvents that are at least extensively stable and DME integ DME are fine with metallic lithium so what is the purpose of introducing the flooring into the battery I saw the fluorine is toxic so we’re trying to get rid of it your first part you you have a F into like the phosphate and you’re talking about the poly anion materials yes yes question is why are we introducing the fluorine yeah the fluorine is strongly bound to the system and it is not accessible for reactivity it is as strongly bound as the oxygen in its it’s part of the metal coordination environment so it’s not a toxic substance such as the electrolyte salt and the point of introducing is is that effectively it alters the structure of the material so we can change the connectivity with that incorporation of a fluorine to make new structures that we couldn’t for example we want to move on beyond lithium iron phosphate so if you take a look at the same material instead of sulphur with a phosphorus it forms this same tabloid structure so that the fluorine is not so much important because it’s fluorine it’s because it is a negative charged anion that changes the whole structure and charge distribution in the material and it also increases the potential by a small amount but that’s kind of incidental so Linda Jordan asked a question first

just showed it up well.alright know where the mic is there is so in your opinion how much sulfur dissolution can we live with how much sulfur is what how much software dissolution can we live with and still get the battery to what we wanted it to be well is it zero or the week and we live with something that’s a hard question to answer and I’ll be able to tell you in two years from now I’m guessing you know the fact is that we still know for example if we take a look at this data where did it go you know we have we have probably this is a slightly better me Zipporah system but I would guess that we have retained about eighty to ninety percent in this system as well so I think it’s not a combinated seems to you know level out so I don’t know is this is this acceptable over 100 cycles well not if you want a five thousand cycle battery no it’s not but the fact is that this rate of decay really drops off dramatically as you keep cycling because now you start to precipitate things on the positive and negative electrodes so that capacity fee becomes almost self-limiting after a certain point it stops because once those layers build up they don’t form they don’t build up anymore so the problem can be addressed by interfaces at the positive and negative electrode and that’s what we are working on now which is why I would say that I’ll tell you in two years yep but I think what you you’re not realizing is that we’re not working in a Cathal 8 cell so this this little picture that I showed here this is not our design we don’t have sulfur in solution we’re trying to contain it in the in the electrode itself so your questions about molarity are a little hard for me to answer because we’re not we’re not putting it in the electrolyte we’re trying to keep it out if you can’t stay below about 3 molar sulfur in your electrolyte however much there is your cycle ability will be much higher well Jody’s question really relates to how much sulfur can you precipitate on the positive and electronegative electrodes and still maintain ion conductivity such that the system works actually what it comes down to because if that if you can build up only thin layers in the cell and we don’t know what those layers are that’s why I can’t really answer the question that’s one of the things we’d like to know but at this point for example in the cell if we have ultra thin layers of sulfide insoluble sulfides in both the negative and positive electrode and the sulfur that’s in the cell has effectively been precipitated and it’s just sitting there then we’re good to go the question is you know where are we in that limit so all I can say is that this is probably about 10 to 20 percent sulfur and for a typical cell that’s that’ll probably produce a coating that as long as it’s not too thick will carry on for a while if that answers your question