Sri Narayan - Advances in Suppressing the Polysulfide Shuttle in Lithium-Sulfur Batteries

good afternoon thank you George and Tom clear for having me here so as George said trying to share with you some of our work on how to suppress the Poli Sci Fi Channel also found that the Green pointer actually works ok so maybe next time so the work that was done my graduate student Dirk Moy he did all the experimental work and this was funded by the University of Southern California local hydrocarbon Research Institute ok so I think you've heard enough about what have you doing let himself Oh batteries but so one last time I suppose up so we we all are in the mode of trying to build high energy density systems and as you know lithium sulfur is not like huge market share of you which hard to break in battery markets so if you can get to those those metrics of a thousand milli ampere per gram and maybe over a thousand cycles maybe are something that we can do with it over a larger application space but then the challenges as we have heard many many times already is that we have a short cycle life we we have relatively low utilization of contact materials of the cycle life of the of the lifetime of the Maccabi of the battery and we have premature failure by electronic shorting and I think the elephant in the room is also the recharge ability of the anode which hasn't been talked about as much but I think it's slowly talking to get attention so so what have we been doing about this so when we let's let's take a look at the the polysulfides shuttle again I call it the bane of the polysulfides Cheryl because it's been amply blamed for all the problems that that we have and and so we have self discharge and go charging efficiency we also have a reversible capacity losses and that is our version of the cartoon so I don't think you can learn anything more from that cartoon so I'm going to skip that power there have been a lot of efforts as you as you know here up with trying to address Poli Sci Fi Channel and the first one being the electrolyte additives it's been pretty successful as a baseline with lithium nitrate a consumable additive than all these interesting encapsulation techniques that you've seen with carbon nano structures and also porous materials that will pretty much slow down the transport of polysulfides and of course the Holy Grail is to have a solid electrolyte that will completely exclude the kala sulfides and just let the lithium ions go through so when we started to work on this problem this is back into 2013 as trying to figure out how we can approach this and so before we could embark on trying to find some solutions I wanted to understand the scope of the problem and so how to quantitate really this polysulfides chattel and try to get a measure of how much of this shuttling is happening so what we set out to do was to measure the shuttling process to scope out the problem and also if we were trying to make some improvements then I should be able to figure out what exactly these improvements are doing to the shuttle itself so so essentially we wanted to measure the rate of shuttling or the shuttle core and so measurement of the polysulfides shuttle current would be a very useful thing to do and so when we when we started to do this how did we go about it so this is the description of how we measure the shuttle current so we take a cell a typical cell I'll show you in a second what what these cells are so we would start with cycling the cell about three times so that we can get all the electrodes formed and then we would discharge it to the targeted value of a depth of discharge so let us say 10% depth of discharge and then you would measure the open circuit cell voltage allow that cell voltage to equip it for a little while and then hold the cell voltage at that open circuit value for with a potential stat and measure the steady-state current so when you try to do that this is this is what you will you will match you see of course a very transient rights of corn and then finally you will see a steady-state value and that turns out to be the steady-state shuttle current so now if you were to have an additive like lithium nitrate you can see well without the additive the shuttle current is significant whereas with new tiem nitrate 0.25 molar of that your shuttle corn is gonna so this seems like a good method of assessing whether you are having polysulfides shuttling or not okay and our our cell design is very typical so carbon 60 percent and 35 I mean also sixty percent and about thirty percent of carbon and 10% of PVDF and we use TFSI electrolyte and we have ability in my lab okay so we all we all also are quite familiar that unless you are you are operating the soluble region the polysulfides Shadle is not an issue so if you were to just look at our current voltage so the voltage timing curve that's a discharge curve you find that we are this is the soluble region and this is the only place where you will actually see the shuttlecraft so if you measure the shuttle current of the function depth of discharge you find that the shuttle current is fairly high when you start with but as you go down to the insoluble region the shuttle cord actually false and so if you have lithium nitrate you should not see anything in that region so it's kind of corroborates the fact that you are indeed measuring the shuttle current and not some something else okay so so we therefore confirm that the shuttle current is actually only present only when the soluble salt polysulfides are present and there is a direct evidence that lithium nitrate actually affects this process so these are sort of sanity checks that you have to make as you as you go along okay so now now let us okay so if you're really interested in the shuttle current measurement you could you could go and look up this article that was published in 2015 and I also understand that axis is trying to use this as a method for rapid screening of various electrolyte kompis and I'm happy to hear that because it can be pretty easy to set up and run these types of experiments okay so what what have we done about deepali so by Channel so what we have done is we've taken something called the mixed conduction membrane and have that selectively transports lithium ions and introduced it into the cell so the way we put it in here is we have this mixed conduction member in moment I'll tell you what that is it is sandwiched between two layers or separators of the electrolyte on this liquid electrolyte on this side and liquid electrolyte on this side but then you have a layer called the mixed conduction membrane so the mixed conduction membrane is a non per first of all has to satisfy the requirement of being non-porous so that doesn't let any call itself icicles which it impervious to any soluble polysulfides species it should have excellent ionic conductivity so fast ion transport for lithium and in this particular case we are not using a solid conventional solid electrolyte what we are trying to do is take advantage of lithium ion transport using an electrochemical mechanism much much in the same way as lithium ions are transported in in the cathode and the anode of a lithium ion battery okay oh I didn't actually do that okay the the other requirement for such a system is that this particular membrane has to also be electronically conducting otherwise it will not sustain the electrochemical processes further if this material has to be stable in contact with the polar sulfites are not electro chemically react with them so it should have a an electrode potential of something greater than 2 point 7 volts so that doesn't actively react with D with a poly sulphide ok so what are the examples of these materials are these well known cathode materials that are used in Lukie amount batteries lithium cobalt oxide lithium nickel cobalt oxide and the combining these oxide so on so forth so we can use any of these materials and the moment I'll show you how exactly this is done so you take this mixed conduction membrane which means it conducts lithium ions and also electrons and here you see that it's deployed here between the two layers of separator there is no reaction of the polysulfides with this layer at all there is no reaction per se however the injection of lithium ions and the removal of lithium ions across this layer occurs by the same mechanism as I said of the charge and discharge of a lithium ion battery so you have electron flow and ion to a couple together as a result of which you can actually transport these with amounts fairly rapidly and we can discharge these lithium ion cells at very high rates I mean to see rate is pretty common these days so lithium ion transport through these structures should be fairly fast and indeed we do find that this this is actually true okay so the other way of looking at this device is sort of like a back-to-back splits out and then if you would think about it just to make sense out of it it's just two back-to-back lithium ion electrodes put together and you is pretty much shorted across this and that's how is it the amounts are getting conductive okay so so given the fact that this this particular membrane is electronically conducting you do need you do need the two layers of insulating separator on either side of the membrane so you can you assemble the cell you have to sandwich it between two layers okay so let's see whether this particular device that accomplishes the purpose that that we set out to do so let me let me just describe to you how we actually make this membrane we we start with we take a look in cobalt oxide as a good example of how how how we might actually deploy this membrane should be about commercial ad aided cobalt oxide ball mill it with with the binder and coated on an aluminium substrate dry it and then hot press it and you can then cut it into disks so this is this is how it actually looks once you once you try to fabricate these membranes so and it's also very flexible and you can get freestanding membranes about 50 micron in thickness which are pretty dense so when I say pretty dense I always want to know what what I mean by that so we did a permeability test and the permeability test is very simple you just build a little permeation cell where you house the the circular membrane in between and on one side you have colored substance in this particular cases and anthraquinone dyes salt on it so that you can watch the permeation of anthraquinone dyes all connect to the other side using spectrophotometry and you can see that this particular picture was taken after 72 hours and you you can you can believe me that there is no change in color there and it was established by selective or comment photometry that there was no detectable amount of material crossing over so this was a good proof of concept for for initial studies and so we went on to start deploying these in actual coin cells and trying to cycle them okay so the first thing that we did was after all we are armed with this tool of trying to measure the shuttling process using the shuttle current as a measurement technique so we said haha let's apply this to the to a new to be mixed conduction membrane cell and lo and behold we find that the shuttle current would be with the lco membrane between the two electrodes is actually much much lower than with the cell without the membrane and in fact it is about comparable to that with lithium nitrate okay so the shuttle current can be significantly reduced by using the mixed conduction membrane now if you were to if you were to believe this then we would go next step is say okay let's say we can start charge and discharge the cell and you find that you can actually charge and discharge the cell and the voltages are very comparable and also you find that your discharge capacity is much much higher than without the membrane so in which case you're looking at the charge to discharge ratio here the coulombic efficiency is actually pretty good so we continue to actually cycle these cells and so we're looking at now cycle life and columbic efficiency sorry for the all these lines in the chart but if you were to look at the kuhlenbeck efficiency which is the top two lines with and without the membrane you find that with the membrane you can get coolant with efficiencies as high as 96 97 percent whereas would be without the membrane you are down in the AES and if you were to cycle these cells for over 200 cycles you'll find that there is a big difference in the fade rate now this is not sufficient because although you have a big difference in the side rate it only proves the point that stopping the shuttle is not the answer to all our problems and and I say there's only because when I started out I thought the shuttle was the main culprit for for for destroying this cycle I to the battery but here we apparently have stopped the shuttle but there are still major issues with the lithium sulfur battery that bring down the cycle life however since to keep the tempo of this stop going I said okay let us let's let's look at what what the actual what what is the benefit of using the shuttle now we already know that the the polysulfides shuttle happens only in the soluble region of the of the curve so so here is our soluble region and so let us say if we were to cycle the cell only in the soluble region we should be able to expose the benefits of stopping the polysulfides shadow to to the maximum possible so how do you just cycle in the soluble region so we we figured that we if we were to cycle the cell at very high weight charge and discharge damage for example see over to rate you are pretty much operating only in the soluble region okay so if so see over to with a cut-off of about 1.8 volts for the cell would pretty much keep you only in the soluble region and will allow you to expose the effect of of the stoppage of the polysulfides out so we we continue to test these cells now cycling @co were to charge and see over to discharge and here you find that if you would look at the the capacity retention as a function of cycle I'd be the the cells which had the LCL membrane or the mixed conduction membrane seem to lose almost no capacity at all and on the right if you were to look at the cell voltage versus capacity retention curve after 170 cycles this thing has that in capacity retention as as you started in cycle one whereas a cell without the membrane seems to rapidly degrade so if we were to work just in the soluble region yes we have an answer to the problem but if you were to go into the insoluble region and and you get all this precipitation of lithium sulfide and the segregation of material in the redistribution material this story is quite different okay so so the capacity fade definitely is lower with the would be looking cobalt oxide membrane but the effect of the effect of shuttling on cycle life can be can be mitigated to a large extent but not completely you've been looking at these charge/discharge curves and there hasn't been anything different about using the lithium cobalt oxide membrane in terms of the voltages so if you were to run an impedance curve on this you find that the resistance is are very comparable the high-frequency resistances are very comparable within the very abilities of making coin cells I would say they are pretty much the same I don't find a huge difference in this so so now after few cycles this cell for about 200 cycles we decided to open up these cells and see what the status of the membranes were okay so we take a look at the lco membrane and it looks very similar to what what we had seen where before we put that in and we also looked for things that had that could have been incorporated in that we could not find any software we could not see any change to the composition of the litigated cobalt oxide member and that's good news because if there was any reactivity of the polysulfides with the Lithia did cobalt oxide it would have shown up in an X Rd so we didn't find any other any other places other than thirty-eight cobalt oxide that we started with and when you look at the lithium electrode you find the dursa there's a significant difference in the morphology of the lithium electrode I do not fully understand this but at this moment it looks like the litigated cobalt oxide seems to give you a more smooth morphology compared to the one where we didn't have the membrane and on a closer look at these surfaces you find that the surface of these mum of the Librium electrode after 200 cycles seems to be fairly uniform whereas the one without the MCM membrane or the mixed conduction membrane seem to be patchy with with insulating material so now you get curious as to what this material is and I'm hoping that it's all for because it will prove the point that we haven't that we have that we have actually stopped the sulphur from migrating to the to the lithium electrode so so started to do some EDX analysis and look at sulphur okay so when you when you are doing EDX analysis looking for sulfur we are using lithium TFSI as a salt so there's another source of sulfur that that can actually appear at the anode which is material coming from the created Kiev TFSI so Howard litigated TFSI is also associated with chlorine okay budget by the nature of the molecule so the sulfur to fluorine ratio in litigated TFSI is is a fixed number so so we decided to compare the sulfur to fluorine ratios for both these surfaces and so here are those numbers you just focus on on just these numbers so on for litigated TFSI you have a rage sulfur the fluorine ratio of three point six three whereas for the litigated lithium electrode which we did not have a stopping membrane we had a very large number of sulfur to fluorine which means that the excess sulfur and with the case of the litigated cobalt oxide member and we had a number that was much smaller so I would I would argue that maybe we were able to stop about eighty percent or so of the of the transport because we still have a little bit more sulfur than we would like to see and so and of course we have a coin cell design which means that there is electrolyte that can go around the edges of the electrode and so on and so forth so so I know I've rushed through this quite a bit but I think we have a few conclusions here first of all the mixed conduction membrane does actually block and stop the poly sulfide from moving around it does improve the utilization of sulfur because you can keep the sulfur on the cathode side it protects the lithium electrode three from being degraded by self compounds we don't know whether that's good thing or bad thing but definitely we find a lot less sulphur when we use the litigated cobalt oxide membrane I think we we have we have a material that you don't have to go and invent you already have these materials around and some of them are fairly easy to obtain so these are whirlpool and cathode materials you can incorporate them in these cells and they're fairly robust recycling so some of the unproven benefits that we think might come out of this is that you could you could prevent premature failure of these cells by electronic shorting because this membrane is fairly dense and strong and it's made out of literally a ceramic type of material which is created cobalt oxide we also have the possibility of isolating the cathode side from the anode side and you we've been talking about different solvents for the cathode and the anode and here is here's a possibility of doing that because you don't need liquid electrolyte to move around in the south so so that that that is essentially the all that I really wanted to say and if you wanted to get more information on this we just just published a paper very recently which has a lot more details on what you've heard okay and thank you for your attention okay beautiful talk are there questions yes there's one and another there where in life talk I just want to ask you what's the mechanism when you use using proper oxide membrane to prevent a subtle effect it is a physical prevention or chemical of knowledge option you basically inject lithium ions on one side of the membrane by the electrochemical mechanism just like you do the integration integration they're not in a cathode of a lithium ion battery and so lithium goes in and lithium comes out so this intercalation if you want to think about is indication on the front face and the indication on the other space and then you have electron transport occurring visit an electrochemical field driven process so you can actually transport with the amounts from one side to the other yeah okay thank you yeah so and what is the velocity of your membrane okay to be honest I have been asked that question and I have not measured it but the the it's it's under an SCM it is very very dense okay and this permeability test is the only one I have to to to sort of support the fact that it is non poor it doesn't have to reinforce that's that's one way of putting it but it is it is like a lithium battery electrode I mean if you press it hard enough and and it's not dense as in a sintered form because it has got a binder but it looked like a lithium battery electrode thank you more questions I see several heads did you put a carbon with LC oh no it's it's sufficiently conducting to do the job and I am sure carbon will have a beneficial effect on for certain types of cathode materials when we like lithium and phosphate or another cathode material but most of these are have some conductivity and so it seems to be working fine that's free how much weight is the LC Oh at the system is what your energy density look like I knew somebody would ask this question yes so well so clearly we have a 50 micron thick membrane okay and there is no need for it to be that thick if it it was non-porous yeah so we didn't want to have a porous membrane when you start working on this concept so we have to start somewhere so I'm thinking that we could probably get it down to about 10 microns by the method that we are using and then of course there are all these interesting ways of doing putting coatings on which like PLD on a separator you could fill it up so you can in principle reduce this to an order of one or two microns I think I don't have a proof for that but that's just my guess yeah I have two questions do you know about the ionic conductivity in your membrane did you murder of it it's very hard to measure ionic conductivity where it's mixed with electronic connectivity so it's extremely hard you have to use blocking electrodes and try to get some sense out of it but I showed you one impedance plot that essentially is that of the lithium sulfur cell with a number in between and it's at least as conducting us the as the electrolyte liquid electrolyte that you have in there it's hard to believe first of all it's not so hard to believe if you believe that the lithium ion battery can be discharged at to see rate which means stuff just goes into the electrode and comes out look around the moving at that way so then of course comes the question as to what at what weight will will this max out okay so we can you discharge it at a 100 C way that somebody was trying to propose I don't know I don't know that spheric probably will stop I mean it's probably 2 or 3 C is probably okay but I don't know how far you can push it and depending on how thick it is I'm sure so actually my second question vanished with your answer so thank you okay thank you there's one more question and maybe another you have on the description of how do you matter this shuttle current so look like a W just matter SEP discharge maybe as per time over you know so in this case the you assume that in the chateaux curve is the main cause for the Savi this chart right yeah so I'm really not assuming anything when we I think we've been talking about it all of yesterday we we said this is not a thermodynamic system it is the things are happening in there right so if it was a true thermodynamic system at equilibrium then you if you have an open circuit voltage and you match that open circuit voltage on your potential set there should be no flow of current that would be the definition of a system at equilibrium but when you try to do this and there's flow of current that means there are processes happening in here so I'm you're absolutely right and attribute all of that to the to the transport of the shuttle but I can't think of anything else at the moment that would contribute to it certain charge that causes maybe you know complicated from different days back so if you here you are only considering the current part in the liquid side so if for example just sulfur diffused away from the from the surface of the electrode that would not show up as corn that's true so we are measuring only those processes that are coupled from one side to the other another question I actually have a question so you mentioned some other cathodes and you can think of all the ones that have been used are you do think there are other cathodes it might be better than cobalt oxide yes I think I think we just picked cobalt oxide because it was convenient but we just have to make sure that the electrode potential is about that of the sulfur record otherwise it would start participating in reactions and that would complicate life so so the cathodes I listed specifically are in that region of potential as much as I said lithium ion phosphate might work we we have to make sure that it's it's in the fully charged a so that it's at the right voltage so I don't know whether I answered a question but that the criteria would be to have the potential that the reduction potential of that electrode fairly positive about that at the sulfur couples No yeah beautiful thank you so thank you three that was a wonderful car thank you so this concludes the morning session time for lunch which is in the usual place so see you in the afternoon

Channel: Energy Futures Lab

Views: 4,572

Rating: 5 out of 5

Keywords: Energy Futures Lab, Imperial College London, Energy

Id: n9-Rz4sH6yk

Length: 28min 47sec (1727 seconds)

Published: Wed Jun 07 2017