Battery 2030+ Excellence Seminar, Yi-Chun Lu, Energy Storage Technologies, Jan 23rd
please thank you Eric um hi everyone uh my name is ichin l and I'm thankful for this opportunity to share our work uh today I will be mainly talking in in the area of aus batteries uh I thought it would be good to uh start uh introduce the overview uh of my research activities uh as Eric mentioned we have uh we have activ ities in metal air metal sulfur batteries um looking at the fundamental understanding of Electro electrolyte interfaces um and then aquous batteries and Redux flow battery and today I will be focusing on mainly on the aquous battery and the Redux flow batteries uh we actually built uh several Institute in-house uh characterization systems to understand battery chemistry uh while we are operating uh the the batteries we would like to know what kind of side reactions what kind of uh degreg mechanisms ongoing so that's really very much dear to our heart um uh in addition to fundamental science we also uh dedicate a lot into Tech transfer and so uh we founded lucose energy uh looking into how to we commercialize large scale energy storage uh for uh renewable energy integration and so on so uh to begin i' like to start with the motivation for uh first part of my talk Aquis battery so uh we all know that non aquous so of non aquous battery they have high energy density however they are flammable and so in a lot of application it may be it may be a risk uh involved for instance EV or of course most importantly large scale energy storage uh the larger the scale the larger the risk could be so if we looking at uh chemistries that Beyond uh non aquous we looking at aquous battery however aquous battery uh suffer from narrow voltage window limited by water stability so from this plot you can see that the voltage window uh that operated for Aquis battery as limited by the oxygen Evolution and the hydrogen evolution so we we have to exclude a lot of material uh that's beyond this uh stability window uh recently there are a lot of efforts uh in the area of highly concentrated electrolyte trying to use highly concentrated salt to stabilize water um so for instance watering salt uh a hydr melt a lot of gray approach has shown much improved stability uh in the aqua battery and expanded voltage window however these approaches uh involve highly concentrated salt and they are usually involving high cost and potential toxicity issue so therefore uh when we started we try to try to find a approach that can stabilize water without involving highly concentrated salts so there is a phenomenon in living cell called molecular clouding it's actually talking about the property of the solvent uh or the liquid liquid could be modified when there is a large amount of large molecule such as protein U around these solvents and so when they reach high concentration they can actually modify the activity of the water molecule therefore we ask can we utilize this phenomenon and to apply into the aquous electroly and see we can reduce water activity by so-called crowding agent and these crowding agents should have strong interactions with water and so we come up with a a series of uh water miable polymers they are low cost and eco-friendly so our idea is to improve to increase the water a agent interaction so that we can discourage the the water water interaction and so the weakened hydrogen bond network from the water water interaction will help to strengthen the O calent Bond because now the water are surrounded by the crowding agent instead of the water and so therefore the hydrogen bond strength between the crowding agent and the water is weaker than the water water uh hydrogen Network so by doing such replacement uh we hope to strengthen the O calent bond strength therefore discourage water splitting so to start to test this idea we started the a very common crowding agent in biology uh polyethylene glycol and this uh polymer actually you you can mix with water in any ratio so uh we use in this case uh various of water concentration various of phg concentration to try to understand how the electroly property will affect will be changed as we change the pH content and of course as I mentioned earlier these type of material are non-toxic and very low cost compared to the lithium salts so first we look at the how the voltage window change as we increase the peg concentration so as you we increase from 0% uh so that's would be the low concentration two two Mo uh two two more lithium tfsi with no Peg all the way to 70 and even 9 to 94% you can see the hydrogen Evolution potential has been delayed significantly as we increase the PG content so this supports the idea that the water splitting are suppressed by the presence of PG and so we further look into some of the uh reasons why this will happen so as we increase the phg content from uh proton NMR we see that the as we increase phg content uh the the nucleus is more shielded and indicating a weaker hydrogen bond Network which is as we mentioned earlier as we replace water molecule with crowding agent the strong hydrogen bond Network between water water has been broken and uh at the same time we also observe a improved a stronger o calent Bond blue shifted so these two uh evidence are supporting our hypothesis and therefore we use this this electrolyte with 94% PG uh 6% water to uh serving as a electrolyte and to test the LMO lithium manganes oxide and uh lto as an anal as a full cell so first thing we thought it's important to check is whether uh in this even though we're looking at voltage window are suitable for this material we still need to check whether in the real battery there is side reaction happening so we use the online electrochemical Mass Petry to detect what type of gases are coming out from uh the battery operation so you can see from uh first and 10th cycle you can see uh both charging and discharge uh the there is is no obvious hydrogen Evolution or oxygen Evolution happening in this type of cell you sing the molecular clouding electrolyte um and so this type of uh cell then can be operate more than 300 Cycles U at more than 80% capacity retention so we thought it would be interesting to compare this low concentration uh low lithium concentration electrolyte with other highly concentrated electrolyte and we thought we found that it actually provide the one one of the widest uh potential window for water aquous electrolyte another thing that was very interesting during our uh investigation is that some of the electroly even though the voltage window seems to be suitable for LTL LMO sometimes you actually if you use the mass Petry to detect the water uh splitting you can see a lot of hydrogen Evolution for instance uh the 21 mo mo lithium tfsi has a lot of hydrogen Evolution uh in operation also other type of Electro highly concentrated electrolyte also see some trace of hydrogen Evolution but the molecular clouding electrolyte uh does not evolve hydrogen uh in this in this battery operation of of course uh the motivation of using aquous electrolyte is for safety so uh since now we reduce uh large quantity of water so we want to also check the safety um and so in this case we can see that uh the in addition to adding water into the P does help the reducing the flammability of this electrolyte um so this is a sort of a a first attempt for this type of molecular crowding electroly using for aquous battery but of course uh one of the drawback uh in the our first attempt was that the PG electroly uh with water actually has very high viscosity and that give rise a consequence uh in the low ionic conductivity so in this case we were only able to obtain a 8 mement per centimet but compared to say water in salt uh we is significantly lower right so but of course other type of highly concentrated electrolyte also in a range of less than one m per ctim so we do want to improve this ionic conductivity and the thought it's uh simple that you know if we can use other type of water missable polymer but with lower viscosity then we can significantly reduce the overall viscosity and improve the ionic conductivity so PG actually have a very strong intermolecular hydrogen bond interaction between uh the molecules so uh if we can remove such interaction and replace with u pgdm so in this case the end group uh does not have the hydrogen bond interaction so this can significantly reduce the viscosity um and so you can see from this U measure that the PG DME has a much lower viscosity compared to the pg and of course using this type of um crowding agent pgme you can make a even the same percentage of a polymer you can have much improved ionic conductivity and correspondingly you will have improved uh polarization much reduced polarization compared to the p in this case um at 1C you have much reduced um hit uh voltage hit teristic and of course uh if you check the mass spectrometry you also see no hydrogen Evolution or oxygen Evolution for the pgdm molecular clouding electrolytes so this uh is really a a beginning of this type of uh concept applying into aquous electrolyte and we do see a lot of papers coming out uh really applying molecular clouding Electro in various of uh battery and uh different like zinc batteries or even super capacitors so really to looking into um the mechanisms and also applying in different U devices um the next part uh I'm going to switch gear to flow batteries and the motivation uh to investigate flow battery is really to looking into large scale long duration ation energy storage um so for long duration we want to be able to freely uh extend the uh discharge duration and flow battery actually offer a exactly that kind of purpose so for flow battery we store energy content uh in the external tank uh dissolving in the aquous solution and uh positive electroly negative electroly essentially dissolving uh the active material at the positive potential and negative potential then we can circulate the liquid electrolyte into the stack um and separate by ionic conductive membrane uh and then we can convert the energy convert the uh chemical energy into the electricity through this device so there are several um features uh and also advantage of this type of flow flow configuration for for for instance in this case power and energy are decoupled meaning that you can freely scale the energy without having to change the power stack so if you want going from 2 hour to 10 hour duration you can just uh scale up the tank without changing the design of the stack so that's very easy to scale up um of course uh if you use aquous electroly then it will be nonflammable um and uh also it will be uh very easy to control the self- discharge because you if if the device going to be idle for a long time then the pump can be uh just disabled then you will have no self- discharge issue over a long period of time right so there's a lot of uh configurational Advantage for flow battery um and like I mentioned if you want to double the duration in in the in the case of lithum battery you will just double the cost uh to buy two sets but for long duration uh using flow battery you can have added small cost using the uh electrolyte tank um and so right now the uh challenge for flow battery uh several fold essentially uh we are looking at expensive electroly needs to be replaced so vadium is the most commercially available and also most mature flow battery chemistry um Vanadium 54 and vadium 23 uh on the positive and the negative side um and this battery system is quite mature but it's limited by the high cost as well as the uh limited energy density right right now it's about 25 to 30 wat wat hour per liter which is much lower than um most of the other rechargeable battery system so we decided to look into something more Earth abundant and something intrinsically low cost uh intrinsically safe uh so uh we're looking to sulfur poly sulfi in Aquis uh could be a potentially much easier to scale up since is very cheap so the cost for storage compared to uh Vanadium it's uh 1,000 times cheaper um and so globally it's also uh the annual eeld uh is actually also significantly higher than Vanadium and so with this intrinsic Advantage um back in 2016 we put together a a poly Sufi iodi flow battery just try to find some couple that we can demonstrate this uh this type of uh system and at the time we were able to uh look into some of the different concentrations so both iodide and polysulfide uh has very high solubility so potentially it can actually provide uh theoretically 80 W hour per liter and at the time we we actually achieve around um 50 hour poo um but one of the most critical challenge for poly Sufi flow battery is the lifetime so you can see uh back in 2016 uh we using uh the commercial napan membrane we were able to only cycle for 50 cycles and um uh this is really due to the crossover uh of poly Sufi as well as the uh water migration and so we decided to look into this issue uh by modifying the nafal membrane and so there are several issue we need to address right so we want to reduce polysulfide crossover and also polyiodide crossover we also want to reduce water migration and the water usually the polysulfide actually sometimes most sometimes uh come with water together migrating through the membrane so to stop this um crossover we we decided to have a coating that consist of a hydrophobic polymer binded uh carbon and the the the role of the hydrophobic polymer pbdf in this case is to uh reduce the water migration and really to reduce the water affinity and the the carbon uh on the both each side is actually try to absorb poly sulfide and polyiodide and so that the when they are immersed into these uh anion uh they will be charged and to further prevent further crossing over through this layer so we call them charge reinforced because in our design that we hoped that they can reduce uh the same charged ion from further crossing over so that's the idea of this type of uh uh coating uh apply on the nail membrane so while this is a very uh I would say simple idea actually is surprisingly effective so you can see um the the polysulfide flow battery using napon membrane from one to maybe 50 cycle you almost uh Decay uh one than half and after 300 cycle all the capacity are mostly gone and this is exactly the issue uh that all the previous poly suf flow battery are facing so with this PR membrane this membrane modification you can immediately see a huge change in the charge discharge profile um and again this is reoperating at the 100% s so uh looking at very deep uh uh deep DOD with this type of test we can cycle this uh polysulfide flow cell for more than three months without degradation and using commercial membrane either one piece or two pieces you can see that decay significantly over time and so the in the flow battery test uh we operate for uh 500 Cycles more than 2,000 hours and then after that we did a uh check and realized that they still have more than 99% capacity left after three months so this is the really the first time that we can see uh stably utilizing poly sulfi in in the flow battery so why this is so effective so we try to understand more uh so we use the small angle x-ray to understand the water cluster size uh so as you can see using creas membrane we actually are shrinking the water channels uh from 3.7 nomer to 2.7 nomer so this is a significant difference which can explain the reduced water uptake for tras membrane and the reduced swelling as well and we also use the Institute ftir try to understand how fast or how slow this water molecule can go in and out through this membrane so we uh we lock down the water uh water Peaks and observe the the evolution over time as we hydrate and dehydrate the uh the membrane so you can see with napon membrane uh you can see the water uh migration or water movements very very much faster compared to uh the Pras membrane which again which a much restrained water water uh activity water mobility and so the next uh challenge we need to address for poly Sufi um is the slugish kinetics right so U most cases uh you need Catalyst but then even using the reported Catalyst the Round Tree efficiency is roughly is almost less than 50 or 60% and so these are from also from literature uh this is very Universal uh for polysulfide type of flow battery and the Really the the reason is is this uh poly sulfide reduction from s42 minus to s22 minus a breaking sulfur sulfur Bond actually is very energetic intense process and so most of the reported methods uh is is not very effective because for one it the Catalyst the may may not very active second uh the Catalyst need to be very well dispersed and they still have very limited active site so we decided to go a slightly different route from uh I guess solid Catalyst into looking to a soluble catalyst so a molecular Catalyst um so essentially uh instead of directly electrochemically reduce polysulfide we reduce uh a a Rous molecule uh itself in this case AI ribo flaven uh sodium phosphate fmna uh this molecule can be reduced uh with much lower over potential so much faster kinetics and once it's reduced it can react chemically with s42 minus and to reduce s42 minus to S2 minus chemically and itself can be reoxidized back to the oxidized phase and so through this cycle then the oxidized FNA can go back continue to to receive electrons from the electrodes and then to reduce chemically s42 minus so this is a very simple scheme uh to essentially convert the slugish electrochemical reaction to a very fast chemical reaction with a fast Redux um relux molecular Catalyst and the first thing you need to check whether this approach work or not is this whether or not this reaction actually exist right so we assume that once we uh reduce fmn we can chemically reduce s42 minus so we actually can use a very simple xit to UV so you can see this gray line is the um solution for S4 2 minus and the the blue line is the reduced form of the fmn so once you chemically mix them you can see immediately uh we will form as fm3 plus S three minus so to indicate the charge transfer between the five minus of and then the four S4 2 minor to form fmn 3 minus okay so this is a uh indication uh that this reaction actually exist and then we go ahead to start putting uh I think in this case uh five to 10 uh Mill moo of the fmn into the poly Sufi electrolyte and so you can see without fmn just the pure polysulfide is this light blue at a 40 milliamp square cm you can see a huge polarization um and with fmn molecular Catalyst you can have much reduced um over potential and much higher capacity utilization right because now we can um the the cut of voltage uh will be far away from the actual plateau and also what's Inc encouraging is that this Catalyst effect uh is persistent throughout the Cycles right so that's also very important um that it has to be stay stable over Cycles uh so you can see that not only it will improve the rate capability but also it improve the cyclability because now we can access most of the capacity in the electroly without reaching uh the cut of voltage right because the kinetics is significantly improved and so we also looking to more evidence that we actually are directly improving the over potential of the poly Sufi side so then we operate this four electral system so you can uh decouple all of the potential using different the reference electrod so the red profile show the so this Gray Line on charge and discharge that's the overall cell voltage and you can separate them into the positive electrolyte poly and the membrane over potential and then negly potential so the cell with fmna and without FNA you can see they have a huge difference on the over potential at the negative electrolyte side and so this is further um indicate and support that is this molecular Catalyst uh indeed reduce the over potential at the poly Sufi side and so we're looking into the the rate capability of this cell because some very often as you increase the rate capability uh the the power and the current density you will reduce the capacity in this case we do see increase of over potential uh most likely because of the membrane but you can see we still can access the 100% of the capacity even at 80 million per square centimeter so we are curious whether this over potential WEA is uh mostly from the membrane so we we did a simple I correction just to look at uh if what if we remove the U the dominant membrane uh omic drop what would be the profile look like you can see they have they are overlapping perfectly even from 10 to 80 milliamp Square centimeter so this shows the the potential of the PO this Catalyst for polysulfide flow battery and so we also looking to U op parental uvv spectroscopy try to understand the interaction between the Catalyst and the polyware so uh so essentially you you're looking at the charge discharge profile uh as we operate the uvv uh and what you're looking at the first uh on the on the left side a panel looking at only fmn so as we uh reduce fmn you can see the fmn 3 minus uh forming to FM 5 minus and then forming back to fmn 3 minus so this is the changes that you will see with fmn alone okay and with poly Sufi uh in the range that we are looking at there is not much change during this uh charge and discharge and so once you put them together once you put the fmn into the poly Sufi electrolyte that's something interesting you can see so first of all um we are still as we as we uh charge the cell so initially you will see fmma uh to reduce from three minus to five minus and you will see the color are reducing but in the case if you have four s42 minus you will see that the fmna uh 3 minus will not Decay as fast as the one just with the catalyst so essentially is indicating that the FMA 3 minus has been regenerated in the presence of s42 minus right so otherwise it will be dropping and forming uh five minus but in this case we are continuous regenerating the 3 minus through this chemical reaction with polysulfur and so this you can see that uh is another indication of the the mediation uh process that we uh designed through the molecular catalyst so uh looking at the stability of this type of uh cell so you can see the cell without fmn you can have uh St stable with the creas membrane but the capacity is quite low but uh if once you add fmn you can have much reduced uh over potential and much improved Energy Efficiency we also uh do another test where where for the cell without fmn as we show in the light blue here uh after around 900 cycle we add F to the exact cell and then you can see the capacity just jumped um significantly from around 60 to 120 you can see this is a direct proof of the effect of the mod catalyst so uh in the long long-term cycling we uh cycle for more than uh 2,000 cycle and more than 1,2 uh 200 hours and to extrapolate the Decay rates uh so this is uh again not only very low Decay rate but also we can operate at much higher current density than before we also further scale up this system and looking at um amp level of the of type of cell so it's 100 um milliamp Square centimeter tested in the 100 cmet square cell so you can see um still operate quite stable uh at scale of electrolytes uh the last work I want to share in flow battery uh is really targeting for uh low temperature uh application so um extreme cold weather reduced the range of uh flow battery application and so of course also for Ev but for a greas scale storage we need to really withstand uh the Cod um so we decided looking to what type of material would be more suitable for extreme low temperature application and we decided there are several thing that are important first at the low temperature the kinetics will be sign ific reduced so the material in initially should have very high kinetics uh second we need to have low freezing point right so the electrolyte usually I mean water solu water medium electrolyte freezing point needs to be lower than much lower than uh water um and we need to have high conductivity because again at a low temperature you will have redu conductivity and also we also want to have high concentration of the active material to preserve the uh energy density so we look into literature and this polyoxy metalate type of uh material call our attention U because again it has a very high kinetics itself and also it has uh six electrons um active uh that can be stored in this P um there are some literature showing uh up to 12 elron storage but in our case uh we see that a lot more electron at the end are mostly hydrogen Evolution so we decided only stick to six electrons and which is already uh pretty good and um and that's also very important to avoid side reaction of hydrogen Evolution so then uh in the literature uh the Le lithium cat is also what's used uh in the flow battery but we decided to go for proton uh so this uh proton type of hpom uh we believe that will be more important uh for the low temperature application and here are the the reasons so we look into the cogram to understand the kinetics uh as we change the temperature so you can see the hpom uh at the 25 deg celsus and the negative 20° celsus you you see that the cogram does not change significantly over a very wide temperature range but the in contrast the vadium 2 2 plus 3 plus which is the neite of the old vadium flow battery has a significant impact on the kinetics uh as we decrease the temperature so this indic indicates that hpom will survive much better uh at the low temperature environment and then we also look into uh the differences between different cation in the concentration of pom right because we need to maximize the pom concentration so uh the proton form will have a much higher concentration compared to other cation lithium sodium potassium and this is because the hydration shell of the proton total hydration um cation actually will have much weaker binding with P because the hydration shell is um is smaller so you can see in this case the weaker binding will provide higher uh concentration higher solubility of the P so these also uh are supported in the concentration measurements and then next of course is the freezing points uh the hpom we we detect uh down to negative 30 degrees 35 degrees C at 05 mole is still maintained as liquid form but other type of cion already uh Frozen uh at much higher temperature and so this again further indicate the proton p is a very critical Factor if we want to maintain the low freezing point then is the conductivity um so uh benefit from the special grossers uh mechanism ion conduction hopping mechanism this hpom has much higher ionic conductivity at uh low temperature compared to other ion ionic conduction mechanism operated by lithium sodium and potassium uh at the negative -20 degrees Celsius it can even reach more than 70 mement per centimeter so uh looking at the full cell we couple this hpom with vanadian 45 and to couple to see the full cell operation at the room temperature uh this hpom can operate at extremely high power density and high current density so we can run from 100 million per square centimeter to all the way 500 million per square centimeter uh and the the stability is also very reasonable of course the most important thing uh for us at the time was to understand the low temperature performance so we bring the temperature down to negative 20th degrees Celsius and looking at the rate capability from 100 to 240 million per square cimeter uh and the the cell can still provide more than uh around 300 millatt per square cimeter and the stability can sustain uh for more than one 1,200 hours uh more than 1,000 Cycles atga -20 degrees at 160 million per square centimeter right it's a very high current density for a low temperature operation and even after uh cycling we can observe no structural changes uh of the this type of material and so to put them into a context right so low temperature operation for flow batter is very challenging can be evidence from say if you use vadium or vadium V V2 V3 V4 V5 uh even at 5 degrees Celsius 40 milliamp you can already see huge Decay uh not to mention at uh -20 degrees celus 160 million but for hpom um you can operate a very high current density at the low uh negative 20 degre cus so this is the comparison also with other type of zinc iodi for instance also cannot sustain U at the low temperature so uh this is I believe is my last slide uh think my time is also up uh we also uh wrote a perspective uh uh to to share about uh the assessment method and performance metrix for flow batteries and looking into to what type of uh cell configuration uh for long-term stability test what type of analysis uh will be important for different type of flow battery right so today I only discuss the the all liquid type but a lot of time you you will involve zinc uh zinc metal as one side and so you may look into D drve formation and so on and so uh I believe that this uh is uh I guess little bit on the educational side uh will provide some of the insight into uh how to test and how to find out why the battery is decaying and uh what should we pay attention to so with that I'd like to uh thank the funding agents uh and also uh students and postart for their work and I'm happy to take questions thank you very much Eun uh for very nice talk
2024-02-20 22:11
