Optimization of electrolytes and electrolyte additives for Si anode applications 

Surya Moganty – CTO NOHMS

0:00:00   Yeah. Good morning. Good afternoon. Good evening. Depending on where you are. This is Surya 
0:00:07   Moganty from NOHMs Technologies. I would like to first time to organizers for giving them this
0:00:13   opportunity toe talk about the electrolyte products that company has been developing for silicon and
0:00:19   or applications. The structure off this presentation is as follows. We're gonna talk about the
0:00:27   requirements of electrolytes, especially for silicon containing an or systems on. What are the
0:00:33   challenges associated with state of the art Director lights. Then? Then I'll show you some off the
0:00:40   an order details. And Catherine reduce we've been developing. Thing will present some off the data
0:00:46   in terms off case studies, both on the new additives for silicon and our systems, as well as some
0:00:52   off the safety challenges that I know high energy systems will face eventually. So I'm going to use
0:00:58   a proxy Cylindrical cells, uh, toe provide some insights and to temple provocation on how to
0:01:06   mitigate those challenges.
0:01:10   The challenges with silicon and nerds are quite well known. I'm not going to spend time on this one.
0:01:17   Uh, but from an electrolyte perspective, it certainly need on electoral design that conform a very
0:01:27   robust and efficient C I, which is critical. Eso you don't consume too many little Myint in
0:01:35   repairing that because, as shown in the schematic about during the litigation process of silicon,
0:01:42   the silicon expanse, the expansion causes stress within the accord on the stress, if not managed
0:01:49   with the electoral architectures, even if you manage without a dark picture that would break the C I.
0:01:55   Hence, the FBI needs to be redesigned in a way that has a little bit off. Elastic are Paul America
0:02:03   nature so it can absorb these kind of stresses during the long litigation debilitation, our long
0:02:11   cycling processes. So, um, the literature that has been a lot of studies, um, flooring nation is key,
0:02:21   I would say must have a silicon dominant then or systems. So here is a schematic that shows two
0:02:31   different reaction paths. Uh, that's containing only slightly carbon it like easy what happens and
0:02:39   then when you have ah, little bit of a fizzy in the system. So just let's look at the top part off
0:02:46   the schematic where easy reduces toe form, part American films. These films are actually are not
0:02:54   stable because off the expansion right on daybreak, Andi when they break, they get the formed hands.
0:03:04   You can see the thickness of their CIA grows. And that's where consumption off lithium on
0:03:10   effectively. Um, you're losing capacity. Eat cycle. Well, if we have enough Easy, there are reaction
0:03:19   path that shows that it actually forms to different kinds off a C. I one is lithium fluoride rich
0:03:27   that helps to protect and improve the Lydian penetration through the CIA and other is ah Pol
0:03:34   American Nature Upali VC type where it helps toe contained. The expansion, however, have fizzy um,
0:03:43   this kind of f e c derived CIA has disadvantages especially, you know, in your cycling and higher
0:03:50   temperatures. Um, you know, if if he can have gassing challenges and as well s the CIA consumption
0:03:58   reconstruction depends on how much, uh, 50 you have in the system. The higher the amount. You know,
0:04:04   you get along like life, but at the same time, uh, it can also affect the performance in a high
0:04:11   temperatures in a negative way. And, you know, how do we find the replacement for FBC? Actually, can
0:04:20   we find that? So in order to do that, what we have when we come up with a few simple steps. One is,
0:04:25   you know, let's identify Prototype sells. You know that design these cells then define and
0:04:33   accelerated testing protocols that identify a base electrolyte composition, then create the negative,
0:04:44   make them test them. So based on this kind off approach, the goal is Can we identify? Edit is that
0:04:53   can either complement are completely replace the Floren ated system. So here is the prototype cell
0:05:00   design that we've bean working with. So this is about one point example grows in size. Hey, it is a
0:05:09   physical picture off the cell itself. The cattle is 811 cattle singles I loading is close toe 28.9
0:05:17   million proper milligrams per centimeter square. Likewise, on all these combination off graphite and
0:05:24   silicon carbon composites, the loading is is given here It's close to 15 mg per centimeter square on
0:05:32   the silicon we obtained from a company PTR sells for fabricated. That Korean cell manufacturer uh
0:05:40   811 is also coming from a Korean supplier. Here is the design of experiment that we have done using
0:05:49   these cells. So the base compositional Accolate, which was was very simple. Easy e m c 3 to 7
0:05:55   containing one more appear six. Then we looked at 20 different new negatives we made them on That
0:06:03   put them at the concentration about 1%. So there will be three sells for each little group that we
0:06:10   made on. Then we tested them using one see the cycle Life studies just to understand, you know, what
0:06:17   is the effect off the sedatives on the performance of the self? So totally. We designed close to 26
0:06:24   electrolytes on the next life shows some off the initial data. Right. So what are you looking here
0:06:33   is electrolyte number versus first cycle efficiency. We use this particular parameter as a ZA means
0:06:40   to understand how much loss of your team is happening. So give or take the cells with the reference
0:06:47   electrolyte containing FCC on GC kind of fairy tales, which is baseline. You get about 80 around 80%
0:06:56   Kulenovic efficiency, meaning 20% capacity laws. So among the 26 different dialect lives that we
0:07:04   tested, we way see a few of the electrolytes. I did not meet the requirement, but most of them are
0:07:11   meetings. That means that the design elements are at least met, right then we went to the on here.
0:07:19   Are you know, I'm not going to show you all the results. Some off them perform girls. Some of them
0:07:23   did not. But here is the best performing at a collect composition. So let me just walk you through
0:07:30   the cycle life, right? Then I'll talk about some off the S C I using the d Q T V analysis, which is
0:07:37   the very first formation psychic. So here is a capacity with the cycle number A 21 spirit 4.2 to 33
0:07:45   wars at room temperature. All the electrolyte based components are, as I told you, easy M C 3 to 7.
0:07:52   It's one more larvae of sex in it. So without any details with beyond 100 cycles, I think this cells
0:07:59   died completely. I could see the red one knowing that it was just count for then adding 1% we see
0:08:05   slightly improved, but not much. And adding, you know, 5% deficit, which is the industry standard
0:08:10   for 10% containing silicones. You can see, uh, it enhanced the performance compared to V C about 400
0:08:19   cycles. And after the slow leaves, you're seeing a capacity fit things has. It has to do with the
0:08:26   consumption off the FCC. As I was explaining in the previous slides. Then, with our two formulations
0:08:33   among the you know, different 20 different entities that we tested, a couple of them are showing
0:08:38   promising very promising results where the great force and is close toe one right, They sedatives
0:08:44   improve the performance compared toa v c r f e c. Initially there is a little slide, uh, capacity
0:08:53   fear. But overall, after 500 close to 700 cycles eso these two formulations of showing much better
0:09:01   performance compared to a fizzy. If you look less, then understand why this is happening. If you
0:09:07   look at the Deputy V plot, here is the control that does not have any edited on the easy is the
0:09:15   predominant. See a farmer on the animal so you can clearly see easy peak. Then when you are VC, VC
0:09:21   dominates and then the easy pick surfaces. Then you are f easy to react so early and then you can
0:09:28   see the reactions off Easier surplus, so similarly with norms, formulations, um thes are actually
0:09:37   additive called names, which is a electrolyte name is 23 24 you can see the both green and blue
0:09:44   curves. Also, did the similar, um also performed similar towards the FCC, which showing like where
0:09:53   it can mask the easy reactions Because these elected lives contain 30% off easy on def. They do
0:10:01   needs to be effective. It needs to form an A C. I react before the easy so that you know, you're not
0:10:09   doing you're not continuing the reduction reactions on that or surface has just like so overall, Um,
0:10:17   um, like a someone I used in the bullet points here, the control electrolytes sauce, the easy peak.
0:10:24   But adding FC definitely, um, overcomes the easy reduction. Likewise, the norms that it is also act
0:10:35   in a similar fashion. But what we recommend from these, uh, Italy Discovery study is what we found
0:10:44   is that there are actually alternatives that we can either replace Fizzy completely in the system.
0:10:51   Uh, you know, reduce their physique content depending on the annual architectures, cell design and
0:10:57   psyching life requirements. Then what are these video right? Eso here is in a fizzy again. It's a
0:11:06   visit derived. Rossiya comes from this kind of problematic reactions and also lithium fluoride
0:11:11   formation because it has the property of ring opening, uh, drink opens on, reacts with lithium ions
0:11:20   and electrons to form things. Kind of problematic films. So, Norm sedatives some off the ones that I
0:11:29   showed you one of the factories like this without going to details thes have three active groups
0:11:37   that can do both polarized from pulverized products on the surface of an or as well as, uh, the
0:11:44   reactive fast food court can create much Robert S C I. So the GDP plots that I showed in the
0:11:51   previous lives certainly speak to that. Then we slip this details, tow our customers partners they
0:12:00   tested the Here are some off the results, right. So here is a cell that was designed with 811 In
0:12:07   this case, they used apartment used 5% silicon oxide. The reference that actually this carbonates on
0:12:14   B. C and officiated him on. Then they added, actually, our editor toe there the difference. Uh, Onda,
0:12:23   let's just look at the D. Q t v. Like I showed you our editor who has reaction that say they're
0:12:31   closely official before, so you can clearly see those reactions are happening in conjunction with
0:12:37   FBC The oranges norms edited containing acolytes Andi in this particular case. But you can witness
0:12:45   both physically and we see reactions because the and or design is different than the ones who tested.
0:12:50   But when you start looking at the cycle, life here is a cell. It's a close toe. Three. Empower 2.8
0:12:57   empowers precisely a T one c room temperature cycling more than 1000 cycles about 92%. They're
0:13:04   expected to reach the design close to 3000 cycles. We are that we're on par with those expectations.
0:13:13   What is interesting is that every 200 cycles measured impedance the resistance clothes using this
0:13:20   year measurement. So here is the data the impedance growth for reference and Norm Select light where
0:13:27   you can see, uh, addition off Norm Sedative certainly helped to lower the resistance even though
0:13:35   we're seeing similar performance. The additive in conjunction with deficit is certainly helping toe
0:13:41   maintain the resistance to a lower level. Um, things has used impact when it comes to, uh, cycling
0:13:49   at higher temperatures are fast for aspect for cycling. These low resistance Certainly there is a
0:13:57   big here is another partner they made 3.7. And parcels. These cells have energy density close to the
0:14:04   human. What else? For K G. Silicon oxide? 811 you know, black and gray are the partners modified
0:14:13   electrolytes on? They are. They're already thio at a very low concentrations. Um, these are cycled
0:14:19   at 45 degrees. You can see the performance is a lot better compared to the best 23 electrolyte. They
0:14:27   also measured the gas generation. You can see we have after close to 430 50 cycles. Very, very low
0:14:37   gas generation on the better. Citing absorbed so overall the and entered it. It is that we have
0:14:46   developed not just so performance in in the in house testing, but also with the partners then moving
0:14:55   on to the next part of the story is the cattle, right? So here is so why we need to worry about the
0:15:02   stability and safety, especially silicon containing animals. What I've done is I've given you a
0:15:08   cross sectional off the battery. Um, the graph. I'd dominant and or dominant an organ silicon,
0:15:13   replacing that with the silicon. So as we know, silicon has higher capacity than graphite right. So
0:15:20   if the right cell designs right and our architectures are available, you can replace gratified with
0:15:27   silicon. Then, in the same cell volume, you can have more cattle. That's how you increased energy
0:15:33   density. But but having higher amount of cattle means safety challenges. Here is a plot showing
0:15:42   coming from military to study. Where they looked at the oxygen really is as a function off different
0:15:48   nickel content. Off course. The 811 pairing with silicon has the potential to increase the energy
0:15:55   density, and you can see at lower temperatures you're releasing oxygen, and also the safety off cell
0:16:03   heavily depends on on the content off nickel in the system. So overall, going with silicon means
0:16:11   putting more graph mawr NMC in the cattle means terminals challenges. So one of the state of the art
0:16:19   Katherine, our attitudes would look like right with Here is a summary for coming from a recent
0:16:26   article where they looked at the multiple types off reactive additives toe help improve the cycle
0:16:34   life off, you know, nickel dominant systems. Most of them use the border on our hospitals are some
0:16:43   other type of materials, all of them none of them actually, uh, after commercially available are
0:16:50   even are Indeed it is that looking at the Catholic side of things, address the thermal stability. So
0:16:56   we have created, um, using similar kind off court chemistry, multiple cattle relatives heavily
0:17:06   focusing on fast for a score. Our idea is, can we engineer the electrolyte er of the cattle relative,
0:17:13   so to speak, that can reduce the heat coming from the electoral electorate? Reactions also, can we
0:17:21   shift these reactions toe higher temperatures? I'm finally not completely mitigate but delay this
0:17:29   thermal runaway reaction so that you know you can. You can increase to say, thermal stability off
0:17:34   this high energy cells. Here's the D SC data, which is on a material level. So we really charged to
0:17:41   sell 811 graphics and in this case, then took the Graham 811 in a fully start stay and then Brana
0:17:48   DSC Measurement differential scanning Kalorama tree toe. Understand the reactions and you can see
0:17:53   with the reference a life flight. The reaction happen around, you know, starts that are 1 80 then
0:17:58   peek around 2 10 on the area. Under the curve is the heat released 97 jewels for Graham with non
0:18:07   selective light, you can see these reactions are suppressed and also happens at higher temperatures.
0:18:13   So bye bye, Custom designing electrolyte compositions. Certainly there is a way toe believe this
0:18:21   symbol reaction. That's right. That's on a material level. Show you next slide on words. Some of the
0:18:27   case studies that were conducted on cylindrical cells. It is a high nickel 18 6 of the cell we
0:18:32   conducted our convention mints, uh, accelerated rate Kalorama tree measurements on these cells. A
0:18:39   two results coming from that. So we do again. The CMC 3 to 7 Assauer different selected with PC,
0:18:47   attitude and norms. Cattle ready to containing electrolyte When we ran the Ark measurements, I'll go
0:18:56   through some off the data here where it's a self feeding the versus temperature. That's most
0:19:01   important. One A Z. You can see there are multiple cells for each electrolyte, consistently norms
0:19:08   containing a little delayed. The thermal runaway reactions when you look at closely thermal runaways
0:19:14   defined as intending this per minute right, if you look at that around with our norms, electrolytes,
0:19:22   we can certainly delay this reactions to higher temperatures. Thing we connected. Some thermal
0:19:28   simulations are using this data hair is a, uh, a pack that we built with cylindrical cells on you
0:19:36   can see with, uh, norms on then traditional electorate. What happened? So we deliberately induced
0:19:44   Tamerlan away in the cell in the middle. Then observe what happens with the traditional accolades.
0:19:48   You can see the full pack goes into Campbell on away at the same time. You know, it norms containing
0:19:54   regulate these cells, even if you deliberately make one cell going toe Chamberlain away. Given this,
0:20:01   cells have higher the tolerance towards thes reactions. So we're able to mitigate these, uh,
0:20:08   propagation, uh, in in a in a simulation based on the ar commissions. So then move it
0:20:51   next slide. What we've done is what does this mean to real World Bank design? Right. So we can lower
0:20:58   the This is a pack design, then conducted some simulations on this. So currently, the pack would
0:21:04   require three mm gap with our electrolytes. Weaken goto 30.5. Mm. That means that more cells So
0:21:10   which, which means at the end, we can we can improve the energy density by 11 person, just very
0:21:17   reducing the gap between the cells. Then the next question is the performance? Of course. This sells,
0:21:23   uh 18 60 cells with our electrolyte perform similar toe for himself. Flight Um after five seeded no
0:21:30   problem on their expected to reach 1000 cycles. See these three cells with homes left like that
0:21:36   being tested as we speak on the performance is quite similar. Onda hopefully in somebody What I've
0:21:45   shown you is the silicon dominant and or architectures requiring you calculate design Onda as well.
0:21:51   As you know, captured executives are also required So currently norms is offering these products for
0:21:57   both Catherine and Nano eso Please feel free to contact us A to the information given below thank
0:22:05   you for your time and attention.