Solid State Batteries Get the Hype. Sodium Ion May Arrive First | Prof Nikhil Koratkar | The SP Show の全文書き起こし

engineering systems they use nanomaterials at various levels but it's not at the forefront. So people maybe don't realize as much as is being used. For example in batteries, carbon nanotubes Sri Popuri, Ph.D: Mm-hmm. NIKHIL KORATKAR: are used to make the electrode more conductive, right? so again, it's not the principal element of the battery, but it's an important enabling element. similarly we all know about sunscreens and TIO2 nanoparticles are used there. So them I think across the board, you hear of graphene paint that's used on cars, to keep the cars clean and all of that. So so th there's a lot of use on nanomaterials, but it's just embedded rather than right on the surface. So people don't realize you know how much it's being used. Sri Popuri, Ph.D: But particularly, think if you're talking about nanomaterials, right? as you already pointed out, carbon nanotubes, graphene, These are very buzzwords, keywords people use everywhere. And of course, the properties will enormously change, When you're moving the material from the bulk to the like micro level to the nano level, for example, And and you have a lot of excitement that coming out of the laboratory results. But Based on your experience, Professor, that why it is becoming difficult, to commercialize these kind of nanomaterials at a scale, What is the kind of few bottlenecks we have seen in your experience? NIKHIL KORATKAR: I think most of the time it's cost. you know, nanomaterials are expensive. And you know, when you do academic research, you don't care about cost, right? You just want performance. But when you Sri Popuri, Ph.D: Yeah. NIKHIL KORATKAR: talk to industry, cost is equally important as performance. In fact, maybe even more important. also ease of manufacturing, repeatability, scalability, all of these things become very important when you talk to industry. even in the you know, in the battery space, if you try to the active material that's used for the anode and cathode, if you try to scale it down to nanoscale, the big challenge is again cost, because we all want to reduce battery cost. And you know, adding expensive elements into the battery is sort of really hard to do. and also again just repeatability, right? Because nanoparticles tend to aggregate, they tend to oxidize, their properties change over time much more easily because they're very surface dominated. Sri Popuri, Ph.D: Mm-hmm. NIKHIL KORATKAR: And so achieving repeatable, reliable performance over time, all of these issues become important. And with nanomaterials it is a challenge, in all of these Sri Popuri, Ph.D: Mm-hmm. NIKHIL KORATKAR: issues. Sri Popuri, Ph.D: So having touched upon the batteries, right? So what kind of problems you are trying to solve in your research group at the moment? NIKHIL KORATKAR: So we're looking at lithium and also beyond lithium. so in the lithium context, I would say a big push is to reduce the cobalt content that's used in the battery. so if you think of the NMC technology, the lithium nickel manganese cobalt oxide technology, industry is pushing for NMC eight one one, which is the the ratio of manganese to nickel to cobalt is eight to one to one, right? so trying to push in that. Direction industry is also interested in 955, where you have 90% nickel, five percent manganese, just five percent cobalt. it's very difficult to do this because when you reduce the cobalt content in a layered structure, the tendency of a structure is to sort of the layers is to sort of fall on each other and collapse. also it becomes less conductive when you remove the cobalt, you get lithium nickel intermixing. There are a lot of challenges in doing that. but Our approach is really to take these particles and to coat them to prevent side reactions with the electrolyte. So we are using magnetite as a coating. There's also a big push towards spinal structures which don't use cobalt at all. So you have lithium, nickel, manganese oxide, which is a spinal structure, not a layered structure. It doesn't contain cobalt at all. I think that's one area of research which is very important, which is to lower the cobalt content. the other area of research that I think is very important is the flammability of lithium ion batteries, right? That's one of the big problems why you know lithium ion batteries have concerns when it comes to grid storage where you have a lot of lithium in one place. And if you have the risk of fire, then you have insurance cost. And insurance cost is not something that you pay one time, right? You have to pay every day. So Sri Popuri, Ph.D: Mm-hmm. NIKHIL KORATKAR: So that really drives the cost up. And so flammability issue, the obvious target is aqueous batteries, but the problem with the aqueous battery is the operating voltage goes down a lot because of the danger of water electrolysis and things like that. The other obvious solution is to go with a solid state battery. and with a solid state battery, clearly you know, you can't use a graphite anode to get gain the advantages of energy density and power density. so lithium plating stripping, so using storing lithium in its metallic form, sort of a lithium metal solid state battery, that's the way to go. but with lithium metal, again, the problem is the dendrites, right? So when you have plating and stripping, that process is uneven, it's non uniform, and it creates these sharp pointy structures that grow from from the anode side, which ultimately shorts the battery and causes fire hazard. so we've been exploring techniques like using heat, using temperature as a handle to try and increase surface diffusion. So when you have high fast surface diffusion, the dendrites tend to planarize. So we explored using high Current pulses to try and create local heat at the interface between the dendrite and the electrolyte to try and planarize dendrites. And we had a we have a paper in science on that that came out around 2018. but controlling the dendrite problem is the

key in solid state lithium metal batteries. And so we're exploring seeding layers. So the idea is that you create a lithiophilic layer on on the anode surface, which because lithium loves the surface so much, it spreads out beautifully to form a very even layer. So the plating stripping process becomes much more uniform. You can also create alloys. So for example, lithium magnesium alloys is something we are exploring with the help of a group in China where you know the when you have the alloy reaction it it's less likely to create a dendrite. So Because the process changes from plating stripping to alloying dealloying. And that fundamentally changes the nature of the material to create dendrite. So I think these are the two main areas in the lithium space, which is cobalt elimination, and the other one is going to a non-flammable lithium system, which would be a solid-state electrolyte. Now you think of beyond lithium, that is a very interesting space. so sodium is the big, big target. a lot of interest in sodium iron batteries. We are also working in that area. What we are trying to do is not so much on the cathode side. our focus is more on the anode side, where we are trying to create anode-less sodium ion batteries. So the idea is that instead of Hard carbon. Hard carbon is typically the anode right now. Instead of using hard carbon, we would have again an anode free system where sodium would just plate and strip. So this is again taking off from our previous work with lithium, where we would use seeding layers that are that are sodophilic now. so sodium loves those seeding layers, it spreads out beautifully, and then you can uniformly plate and strip and not create dendrites. So that's one area. We are exploring a range of different seeding materials. What is the most sodiophilic? What is the easiest to deposit? These are all issues. For example, again, you know, we talked about industry likes things that are simple, things that are elegant, are the ones that work, right? If you do something very complex, the chances of industry adopting that go down, right? So instead of actually depositing that seeding layer before you assemble the battery, what if the seeding layer self-deposits? When you assemble the battery. So you you add an additive in the electrolyte, a simple additive, could be some type of a polymer. And that polymer self-layers onto the surface of the anode automatically when you assemble the battery. So no extra time, no extra cost spent in actually coating that anode surface. and if that seeding layer would work, then you know those are the kind of simple, elegant solutions that you need that can really move from lab to market. So those are the kind of things we're exploring. and so of course sodium ion batteries if you can make them non-flammable that's a big deal as well. and again there are various kinds of additives one can think of that you can add to the electrolyte that could you know make them less prone to to catching

fire. So non-flammability again, a very important angle in sodium ion batteries. And lastly is rate capability. So if you think about AI centers, right? They all need battery backup. one important property that you need for any kind of battery backup system is, of course, sustainability and cost, right? Should be low cost, should be non-toxic, safe, should be materials that are sustainable. sodium does a lot of that. but one thing that sodium doesn't do a great job is rate capability. So with these AI chips are so power-hungry, right, that they require very high rate operation. Charge discharge has to be at least for small time periods has to be at a very high rate. and in you when you think of a hard carbon anode, the diffusion of sodium into a hard carbon is a slow process. So it's very difficult to achieve charge discharge rates of like five C, six C, which you can get with l lithium ion phosphate technology. one C means you charge or discharge in one hour. N C means you charge or discharge in one nth of an hour. So if it's 5C, you're discharging in 12 minutes, right? if it's 10 C, you're like discharging in six minutes. so industry right now with LFP technology is able to get 5C, 6C, but for these AI center data, data backup, battery backup, you're looking at 10 C. You know, you want to push towards 10 C. And even lithium ion phosphate technology is not able to do that. But can we push sodium to something like 10 C? So again, if you're doing plating stripping of sodium, that's a faster process than if you're doing diffusion of sodium into a bulk system. So we are really trying to push the C rate. And of course, you want to maintain cycle life. I mean, with lithium Sri Popuri, Ph.D: Nope. NIKHIL KORATKAR: ion phosphate, one of the big advantages is you get 10,000 cycles, 5,000 to 10,000 cycles. Can you get anything close to that with sodium and technology? So again. These are open research questions that we are pursuing. beyond sodium, there is this whole Pandora's box, so to speak, of multivalent ion ions, right? Like calcium, zinc, aluminum. And aluminum, of course, is the holy grail in a sense, right? Because aluminum Sri Popuri, Ph.D: Yeah. NIKHIL KORATKAR: is the most abundant metal on planet Earth. And it's also pretty low cost. Zinc is also low cost compared to lithium. multivalent ions give you the advantage of multi multiple electron transfer per atom. So aluminum you get three electron transfer, zinc, you'll get two electron transfer, and so on. So you can get high capacities. you can sh store more charge per unit mass. The problem, of course, with multivalent ions is many of them are big. calcium is a really big ion, hard to squeeze calcium into tiny pores, right? The other problem with these multivalent ions is their charge density is very high. Like aluminum has plus three charge. So it has very high charge density. And so there's a lot of electrostatic repulsion when you try to push aluminum into a host which has cations of its own, right? So that electrostatic repulsion makes desolvation of the ion hard. Insertion of the ion very hard. And so also a big problem with many of these these ions like aluminum is they're so reactive that if you try to push them in the electrolyte, they react with various elements of the electrolyte, like various solvents present in the electrolyte can react with aluminum and it can form complexes which are very big, and those are really hard to insert. So, in general, very little is known with the multivariate ions as to What is the best anode? What is the best cathode? What is the best combination of anode and cathode to use together? What is the best electrolyte? Because again, we want we want to make sure that these cations don't react with the electrolyte. So a lot of design has to go into maintaining a contact ion pair where the aluminum remains with its counter ion together and doesn't react with elements in the electrolyte. So a lot a lot is unknown when it comes to Multivalent ion batteries. And I think that's really the frontier of if for fundamental research, Sri Popuri, Ph.D: And I just want to ask you that industry loves NMC, At the moment. So what is this magic combination, Why nickel manganese cobalt is so good, To give the best performance, I think one of the benchmarks of the industry, right? NMCs. So why it is so good actually? NIKHIL KORATKAR: Yeah, so when lithium ion batteries first came to the fore, the cathode was lithium cobalt oxide, which is the most simplest form. it's like a layered structure into which lithium can diffuse in and out. now when people started you know realizing some of the problems with cobalt, like it's available only in Congo, it's mined only in Congo, it makes a battery toxic, they started to think about can we replace cobalt with something else? So nickel and manganese were the dopants that people started exploring. what some of these dopants do is also they increase the capacity. So they allow you to extract more lithium out of the lithium cobalt oxide structure. so that's one advantage. and and then also just in terms of stability of the structure, in terms of outgassing, from many of these points of view, it's helpful to have these dopants. The main reason why industry loves you know these NMC systems is energy density. So you can get energy density of about 250 watt per kg and you can get you know the volumetric number is also important. That's around 600 watt r per liter and in terms of gravimetric as I said it's around 250 Watt hour per kg. So these are outstanding numbers. Like no other battery systems comes even close to this in terms of energy density. So you think about lithium ion phosphate, which is a completely different type of system, which is a sort of a 3D framework into which lithium can diffuse. in in in LFP, again it's cobalt-free, but its energy density is around 150 watt hour per kg and around 300 watt are per liter. So about half of what you get with NMC. So if you're

really looking at high performance systems, let's say laptops, cell phones, where you want the battery to last, you know, a long time, even EVs where you want to have maximize your range, a lot of where cost is not maybe you're willing to pay the price to get that performance. Sri Popuri, Ph.D: Mm-hmm. NIKHIL KORATKAR: you know NMC makes a lot of sense. It's the best performing System. However, for grid storage, the question is if you're stationary, You're not moving, does gravimetric energy density really matter? Because it's a stationary application. Volumetric still probably matters because you want your battery to be as compact as possible. if you're storing batteries in a built-up area like Boston or New York City, space is very expensive. So volumetric will still matter, but I think in general, a lot of these metrics can be compromised to some extent when you have stationary application. And that's where sustainability, cost, safety that takes over. And that's why LFP is starting to now displace NMC. and to a large extent where where it is doing this displacement, I think is in the stationary applications. but for the very high end high performance application, It still makes sense to use NM. Sri Popuri, Ph.D: So do you believe that the conventional lithium ion batteries has reached their limit at the moment? NIKHIL KORATKAR: I would say the conventional liquid electrolyte based has in terms of energy density, power density, cycle life, yes. But the solid state is i it could result in a significant quantum improvement in performance. because again the separator, right, which is going to be solid state electrolyte also serves as a separator. That can be very thin, so you can save in volume. again, if you use lithium metal as the anode, so if you do plating stripping. of lithium metal onto a like a substrate like you have a copper substrate you have aluminum substrate or something like that you just plate strip lithium onto it you have no anode basically it's an anode free kind of battery again you save a lot of weight you save a lot of volume cost well cost is a little bit tricky because you the the solid electrolyte right now is more expensive than the liquid electrolyte. so the manufacturing of the solid electrolyte is in general more expensive. Although there are a lot of people doing research on trying to make the solid electrolyte in a much cheaper way, using liquid-based processing, so and and then followed by heating and all of that. So there's a lot of research going on and trying to cheapen the the to lower the cost, but but cost is a little tricky right now because there's some trade-offs involved in in how expensive it is to make the electrolyte. Sri Popuri, Ph.D: If you, Professor, if you look at the solid state batteries, right, either it is anode less or with anode, lithium metal, for example. So we are almost sharing the solid state batteries for almost five years, right? If you look at Toyota or Honda's of the World, right, they are promising this technology almost like five years ago, right? At least in the niche applications. But it is still a a dream, right, in in large-scale applications. So why is it so? Because as far as my knowledge goes, Solid state being researched for almost like a more than a decade now, if not more. Right? But still NIKHIL KORATKAR: Yeah. Sri Popuri, Ph.D: commercially it is lagging, right? NIKHIL KORATKAR: Yeah, yeah. So I think a lot of the initial let's say ten years was spent on figuring out what are the solid electrolytes, stoichiometries, and compositions that will give you fast diffusion of lithium. Because initially lithium diffusion was too slow. But that problem has been solved now. The lithium diffusion kinetics in a solid state battery now is very comparable to a liquid electrolyte system. so a lot of the initial, time that was spent was spent on focusing totally on that problem. Now that that problem has been addressed, people are now looking at other problems that have emerged. So, for example, this whole dendrite issue, when you have an anode-less system, or if you have a lithium metal type of electrode system, how do you address the the whole because industry again is not going to adopt this technology if it's an expensive battery if after a thousand cycles it shorts. The whole cost issue, right? You build such an expensive Sri Popuri, Ph.D: Yeah. NIKHIL KORATKAR: battery in a thousand cycles, it's short. So industry is going to be very, very cautious about moving any technology to market until you really show that you can demonstrate that you can solve a dendrite problem in a very reliable way. Multiple cells have to cycle very long time, has to be very repeatable and reliable. That has not been achieved as yet. the other issue is the interface resistance between the electrolyte and the electrolyte. Now you have solid solid contacts. The beauty of a liquid electrolyte is that the liquid spreads, it goes into all these pores, it wets the electrode beautifully, right? But in a solid electrolyte system, you have solid, solid contacts, and these contacts they change over time, they evolve, the interfacial resistance can grow. You need a lot of stack pressure to, you know, sort of the way people try to reduce, yes, Sri Popuri, Ph.D: Keep them attached. NIKHIL KORATKAR: the way people try to ensure that interfacial resistance remains low is by applying huge stack pressure. And that stack pressure creates its own problems in terms of creep of materials, in terms of plastic deformation over time. and is it how practical is it in a real cell to apply that kind of stack pressure, so these are all the other issues that came up later. So initially the community was so focused on solving this lithium diffusivity challenge that they sort of ignored all these other. problems which are also real issues that need to be solved. And now the community is trying to attack those problems. so the next 10 years will be crucial, right? Like if

these problems are not solved in a very robust manner then obviously technology will not move forward. But if they are then I think in the next 10 years you could see I think you could really see the sodium ion technology moving and you could see the solid state lithium technology moving. These are two things that I see could make real gains and real advancements over the next ten years. Sri Popuri, Ph.D: Are you are you hopeful of this all state technology? NIKHIL KORATKAR: So I don't work a lot in that area. I'm just getting into it with these lithium, magnesium, alloying systems, which don't form dendrites because the whole mechanism of lithium insertion extraction changes. so I I'm just getting into it more again from the anode side, trying to solve this this lithium dendrite challenge. I'm too early into it. I don't know quite enough about it. Most of my prior work has been with liquid electrolyte systems. But technically, you know, there's no doubt that lithium, liquid electrolyte, has saturated with the NMC kind of giving you the best possible performance and LFP giving you the best cycle life and cost. so I think the only advancement in the lithium space that I see the Big advancement, quantum jump, can come from the solid electrolyte Sri Popuri, Ph.D: So it's NIKHIL KORATKAR: s system. There's no other silver bullet out there that I see. now can we solve Sri Popuri, Ph.D: Right. NIKHIL KORATKAR: the challenges I just mentioned to you? cost of the electrolyte, dendrite challenge, interfacial resistance, over over the next five, ten years? That that's still an open question. if you were to ask me to bet, is sodium ion technology gonna move with liquid electrolyte, is that gonna move forward first or is the solid state technology gonna move forward first? I would probably bet on the sodium ion with liquid electrolytes that are engineered to be non-flammable. I think in terms of cost sustainability, because sodium is so abundant on the planet, right? That you know, I I think there's a very strong market case to be made for sodium ion. in fact I have a company that I founded called Alsym Energy and we made a pivot to sodium ion about three three years ago and you know we really realized the the possibilities are are really really good with sodium ion. So so so so I think if you were to ask me to bet Which would come out sooner, which would be more likely to make commercial impact, I would probably say sodium ion with liquid electrolyte. But but the possibilities are are exciting with solid state. It's just that are the solutions simple enough? is the manufacturing is it feasible, is it scalable, is it is it cheap enough? And I think these are open questions. I I'm not enough of an expert on on solid state to really comment on that, but it's obviously more challenging to make a solid electrolyte do the kind of things that a liquid electrolyte can do. Sri Popuri, Ph.D: sodium ion, is more displacing the markets of LFPs, right, because of the energy densities, It it may not touch the so-called the shining star of the batteries, which is NMCs, for example, NIKHIL KORATKAR: yes, you're right about that. in terms of energy because this operating voltage of the cathodes that are used in sodium and batteries are lower. And the reason is that sodium is less electropositive than lithium. So the electrochemical driving force to push sodium to take it out of the cathode is less. So the operating voltage drops. So while lithium works at a higher voltage, it gets you more energy density. So Yes, you're right about that. that in terms of energy density, it might be difficult to match NMC. but I feel that the stationary, you know, if you think about where where the maximum growth is right now in the battery sector, it's probably in the stationary storage, right? Where you have these this explosion of AI data centers, you need massive battery backup all over the globe. so I think there is massive market opportunity for these kind of systems, especially if they can be made non-flammable. So I think with sodium, it is easier to make a sodium battery non-flammable than a lithium battery. problems with thermal issues are mitigated to a large extent because of the type of minerals that are used, because of the type of electrolytes that are used, and special additives can be used. All of this can be used to mitigate the fire hazard. So so I think if you think about yes, can they replace NMC in terms of performance? I think not. Just because fundamentally, the fact that sodium's electropositivity is not as good as lithium is gonna you know create a fundamental science barrier there. but the question also is can you improve capacity to counteract that? So so energy is voltage times capacity, right? So maybe you can improve capacity, which allows you to then get s elevate the energy density. And the way you could improve capacity is maybe again going to these anode less sodium systems, where you know you get rid of the anode entirely. So that entire weight is gone, that entire volume is gone. maybe the propensity to form dendrites in sodium can be different from lithium. you could maybe address that challenge as one way To improve energy density by targeting the capacity angle rather than the voltage angle. So sodium research is still not that advanced. It's been the last four, five years. So I think over the over the next 10 years, who knows? There might be breakthroughs that might result in Sri Popuri, Ph.D: Mm-hmm. NIKHIL KORATKAR: sodium pushing towards NMC. I doubt it can reach the level of NMC, but it can get closer, and that's the hope, Sri Popuri, Ph.D: So do you expect that any kind of solid state systems based on sodium ion batteries? NIKHIL KORATKAR: That is even further down the road, right? So right now all of the solid state work seems focused on lithium. I think sodium is a long journey. Again, people will have to come up with electrolytes where sodium ion diffusivity is good enough. just

like lithium, right? That whole process has to repeat. That in people have to develop sodium electrolytes that are solid state that give you fast sodium diffusion. Then all of these other problems that we talked about, like the dendrites and the contact resistance, all of those Sri Popuri, Ph.D: Yeah. NIKHIL KORATKAR: have to be solved. But the hope is if many of these problems are addressed with lithium, then those same ideas could be directly applied to sodium. So it might be a faster learning curve with sodium Sri Popuri, Ph.D: Yeah. Yeah. NIKHIL KORATKAR: solid state than with lithium solid state. But that that's again very far down the road, right? Right now people are focusing on liquid-based sodium. Sri Popuri, Ph.D: So as as I mentioned earlier, I think I read your beautiful review, right, in materials today, which is in front of me here, right? where you talked about a lot of multivalent systems, right, beyond standard NMCs like sulfur, selenium, and so on, right? And I mean fundamentally what I would like to understand is that an ion is carrying, I don't know, let us say two or three charges. sounds better compared to lithium, right? Which is carrying only one. NIKHIL KORATKAR: Yeah, so again as we just talked about that energy is voltage times charge. so you're right that if you're using a cation that has multiple electron redox process, then clearly it can store more charge for sure. but the voltage of the battery really depends upon How much a cation, how electropositive it is. Because ultimately the electrochemical driving force that you need to push that cation out from the cathode into the anode. That's the key. So lithium is the most electropositive element, right? And things like aluminum, zinc, sodium is not bad. It's a little bit off from lithium, but it's not bad at all. But if you think about aluminum, zinc, calcium. These are somewhere in the middle. So they are a lot less electropositive than lithium. So the cell voltage that you would get from an aluminum ion battery, for example, might only be one volt or 1.5 volt, while with lithium you could get 4.2 volts. That means that even though you're doing three times electron transfer, if your operating voltage is three times less, then the net energy density of your battery is still the same, right? So that's the beauty of lithium, right? Then even though it has one electron transfer. Because it's so electropositive, it gives you very nice cell voltage and very nice electrochemical driving force to during charge, the high voltage to get it out and during discharge, very strong tendency of lithium to then go back, right? so I think multivalent ion systems, to my mind, can they beat lithium in terms of energy density? I doubt it. But again, the the space they are operating in is again this stationary storage space. Because if you think about aluminum, if you think of a stationary storage battery, that would be a dream, right? Because aluminum is so safe, it is you know so abundant, the cost is low. it's available geographically everywhere, right? I mean, the w one problem with the a lot of the battery. systems is that you're looking for a material that's only mined in South America, only mined in Congo or in Australia. And so it becomes so difficult to build the kind of supply chains that we need to make batteries in the West, right? but aluminum everywhere, right? You can get aluminum in so many places on on planet Earth. So an aluminum battery with let's say an aqueous electrolyte, right? Which will never catch fire, will give you operating voltage of around one one point five volts maybe. but still for stationary storage, if you you consider all the advantages of cost, sustainability, supply chains, safety, aluminum is so easy to kind of dispose and to work with and to recycle also. I mean that would be an ideal battery, right? So so I think that to my mind, these a lot of these multivariaton systems like zinc aqueous batteries, aluminum aqueous batteries, these would be a great fit for that. Again, that stationary storage market. And maybe sodium might push more towards NMC. And as you know, as lithium on the planet gets used up, we're talking of you know 20, 30, 40 years down the road, lithium gets used up. Let's say we're not able to recycle it effectively. then sodium kind of does what lithium is doing right now in terms of the high-end applications. And because it's gonna keep on improving, right? The energy density of sodium batteries is gonna keep on getting better. and then this the stationary storage market again starts getting replaced by by these multivalent ion kind of systems with aluminum. So we are doing a lot of research on aluminum ion aqueous batteries, and one problem with them is that the aluminum metal forms an oxide, an aluminum oxide interface. And that interface is very impermeable to passage of aluminum ions. So you cannot push aluminum out of that oxide interface. and so we are working on aluminum alloys that where the bulk, the principal element is still aluminum, but we add alloys that that create a very thin oxide interface, a very permeable oxide interface, a mixed oxide of aluminum with other oxides. these are high-entropy alloy systems where you have five elements in the alloy, which of course dominated by aluminum, but we're able to Create very permeable, ion permeable oxide interfaces, and we're able to push out aluminum ions out of that. so that's the kind of fundamental research we are doing to try to try to enable aluminum ion aqueous batteries. and we're also looking at non-aqueous aluminum ion batteries. So aqueous, of course, the the beauty is there's no risk of fire. but there's always a danger of hydrogen evolution reactions because. To move aluminum ions in a stable manner in water, not to form aluminum hydroxide, for example, you need very low pH. And when you have low pH acidic electrolytes, you have excess of protons. And those protons tend to take part in the

reactions and form hydrogen. So we're also working at non-aqueous aluminum electrolytes, where the challenge is to prevent the aluminum cation from reacting with various. Solvents that you use in the electrolyte. Al3+ is so reactive that it just wants to kind of react with stuff, right? So we are designing electrolytes where we are able to maintain the aluminum and its counterion, anion and cation system. In the immediate solvation shell of aluminum, you have anions dominated by anions, which sort of keeps the aluminum stable and sort of prevents it from reacting and sort of move it along. So I think there's a lot of opportunity for basic fundamental research in these kind of systems. but where would they go? Is I think they would again target this stationary market where energy density isn't that critical because again the operating voltage of these systems with aluminum is not going to be high. Aluminum is not that electro positive compared to lithium. Neither is zinc, neither is calcium. and so you have to understand you know the limitations of these systems, but you also have to understand the advantages of them, right? Sri Popuri, Ph.D: So in terms of the you mentioned there, a couple of different systems, Like zinc, magnesium, calcium, aluminium, iron, maybe, which is coming up as well. So what is the research status of these systems? And when do you think they can be coming into the, for example, as you mentioned, energy storage applications? What is the timeline NIKHIL KORATKAR: Correct. So I think if you look at aluminum ion, zinc ion, calcium ion, magnesium ion, and iron now, I would say zinc ion is the most furthest, it's the one that's developed the most. the reasons for that are that zinc oxide interfaces are ion permeable. so with aluminum you have that huge challenge. You don't have that challenge with zinc. calcium, magnesium, all of these they react with water. So in any kind of aqueous system, if you have a calcium foil or like a magnesium foil, or if you have even iron, iron is also you know reacts with water. so zinc and aluminum they have advantage that they form a passivating oxide, which is very stable. So you can use them in aqueous type of systems. but aluminum has this problem that the interface is blocks the passage of aluminum cations. So zinc is really the one that has advanced the most. people have demonstrated zinc ion batteries that operate over thousands of cycles, you know, approaching four thousand, five thousand cycles of operation. I think the problem with zinc is it does form dendrites, which is not catastrophic problem because it won't catch fire in an aquous battery, but you know it does short the battery eventually and Again, industry doesn't like that, because you have to then you know build a entirely new battery, replace the entire system. so we've been doing a lot of research trying to address the zinc dendrite problem. but I think zinc is sort of the furthest advanced among all of these at this point.

to commercialization, I think the issue is that is it a better alternative than the incumbent, which is LFP, right? And at this point, it's not better than LFP because A, it forms forms dendrites. B if you think about zinc ion technology, you always have the small that the risk of hydrogen evolution, which you don't have in the case of LFP technology. Also the rate capability, right? So LFP, the beauty of LFP is that it can go to very high charge discharge rates of up to 5C, six C. which the zinc ion technology is nowhere close to those kind of rates. so because of all of and the cycle life. 10,000 LFP does 10,000 cycles, which is beautiful. I mean at least 5,000, but people are claiming like up to 10,000. which again, zinc ion will really struggle with the hydrogen evolution problem and the dendrite problem. So at this point, I would say it is not competitive with LFP. But in the future, could it be competitive? It could, because lithium costs could keep going up. And a lot of the challenges I mentioned with multivariant ions, with more fundamental research, they could be solved. But then the competitor also is sodium ion, right? So sodium ion is also Sri Popuri, Ph.D: No. NIKHIL KORATKAR: rising. And so these multivarient ion technologies have to also compete with sodium ion. So I think this is you know one thing we found when we formed our company, Alsym Energy. you start a company with certain vision, but you have to be very flexible and nimble. because technology is all the time evolving. And so what you think is a winner at a certain point in time, the battery industry is evolving so fast that even in five years, you might find that what you thought was a winner, there'll be another technology that comes which is better. And then you have to pivot. You have to be very flexible and nimble. and that's just the nature of the business. Sri Popuri, Ph.D: So as you mentioned, your startup, why you started one and what is your experience actually by translating this academic research into into industrialization, Almost. And also what I've seen that you raised more than hundred million, From the capital markets. So what's your journey around that company actually? NIKHIL KORATKAR: Yeah, so I think as an academic, life of an academic is the first I I don't know how it is in Europe and in India, but in the US, the first five years is about getting tenure, right? And that's Sri Popuri, Ph.D: Mm-hmm. NIKHIL KORATKAR: a very high stress point in the life of any new faculty member in the US is to get tenure. that time you're just trying to publish or perish, right? And you're trying to graduate students. you pretty much don't care what you work in as long as you're successful, you just wanna Show the university that you can be successful. and then after you make tenure, I think you start thinking about what do I really want to do now? Now that I have tenure, I have stability for the rest of my life in this place. What do I really want to do? Like what is

really exciting to me? What makes me want to get up in the morning and come to work every day, right? Those are the kind of questions you start asking yourself. And in my case, the first ten years of my life in RPI from 2001 to 2012, I was all over the place, right? So I was making all types of nanomaterials, carbon nanotubes, graphene, graphene oxide, moly disulfide, tungsten disulfide, phosphorine, tellurene, all kinds of 2D materials, studying their fundamental properties, and trying to study, you know, mechanical, thermal, optical, electrical, optoelectronic, everything. so all over the place and can we discover some new exciting property which is different from bulk and we did well we published one paper in nature we published two papers in nature materials so we were very successful but after that phase when I went through tenure and then I became a full professor, then I really started to think about what do I really want to do, right? And Then I felt that I want to do fundamental science and make discoveries and go after nature science. Definitely I want to do that. But I also want, the side of me that was an engineer wants to develop something that has societal impact. so these discoveries they make, besides creating knowledge, just for the sake of creating knowledge, I also wanted to make sure that they benefit society in some way. And that's the engineer in me coming out. And that's why I decided to focus on batteries because in batteries I saw this opportunity for doing both. fundamental science discoveries and coupling that with real engineering systems and devices, right? So around 2012, I made that decision to totally focus on the battery research. and it was a good fit to my nanomaterials work because a lot of nanomaterials can be deployed in batteries very effectively, right? so that that was a transition. and then after doing that for another 10 years, then you ask yourself the question that okay, I've got all these papers, all these publications, citations, Hirsch index now approaching hundred, you know, over forty thousand citations, you become fellow of all of these societies. and then what next, right? Like so then the next question comes that if I publish all this work, I'm depending on other people to then take that knowledge and then to convert it into real devices. And you don't want to rely on other people, right? you the question is: can you yourself do activities that result in technology transfer rather than depending on other people? someone will read my paper and then will take this idea and then use it in his or her battery system. instead of that, can you yourself be proactive in doing that, right? And that that's when you start thinking about can I start a company? now what happens is most academics we live in a different world, right? So for us, we don't have the training that it that you need to start a company. To be an entrepreneur is not easy, right? It's a completely different mindset. and certainly I didn't have that, right? so and most academics don't. There are some exceptions of of course, but most academics don't have that mindset. The skill set that you need to start a company to raise money from VCs and that kind of thing. I was very lucky that there was a gentleman by the name of Mukesh Chatter who is an RPI alumni. and he's founded companies in the past, he's been very successful, and you know he was very passionate about energy storage too. So he took the lead in all of that, right? So he really is CEO really of the company. and He hired one of my students as a CTO. That student has now moved on. So now there's a different team in place. but I'm more in an advisory role now. but the point is that you start thinking along those lines about technology transfer and trying to push the technology transfer, and that's when you really start thinking of the companies. And you know, one thing that makes me happy also Company is that we've hired so many people, right? We had at one point over 100 employees. Now we've downscaled a little bit, but we had over 100 employees. And the fact that you can create employment for so many people also makes you feel satisfied. Is over a 20 to 30 year period, your priorities change and you start thinking along with. Sri Popuri, Ph.D: So you mentioned about the starting a company, right? So based on your experience, at least, you know, what makes this kind of highlighted publications, and also you cited a couple of famous journals, at least from the H-Index perspective, Nature's Science of the World, But when you publish a paper, it is fantastic, right? As an academic, And you get a lot of citations and so on. And also, in order to publish in those journals, Professor, that it's very difficult, Because you need to have very unique, almost groundbreaking ideas and thoughts, But why such kind of publications are also struggling, To make them into real products and real applications, actually. NIKHIL KORATKAR: Yeah, I I think it goes down to the same discussion we had about simplicity. So a lot of the work that's kind of published in nature and science many times is very exotic. very exotic materials, very complex materials, right? to an academic that's exciting, right? the fact that something has complexity in it. but i in the academic world we don't think about manufacturability, scalability. all of these issues, cost, supply chains, availability, there's so many things, right? That we don't even think about. And so I think that's probably the reason why a lot of these breakthrough discoveries also keep in mind that that industry cares a lot about repeatability, right? So if your battery if you build five cells and four of them work. we will write an academic paper based on that, right? Industry that's not good enough. it has to be all near 100% reliability, repeatability, right? So Sri Popuri, Ph.D: hundreds of cells, yeah. NIKHIL KORATKAR: yes, so the standards, the requirements, because if you have a very

complex material, it's difficult to make it. Every time you make it has to be exactly the same. It's sometimes hard, right? And for industry that becomes becomes a concern. also simple things like so we published a paper in Nature Communications back in 2014, where we were showing dendrite free plating stripping of lithium on a three dimensional graphene network, right? So we were getting long cycles without forming dendrites. So we approached industry, we talked to some folks at, for example, Tesla, and they told us, you know, this is great, you're not getting you're not getting dendrites, and that's fantastic. probably because of this 3D structure of your electrode, which traps Sri Popuri, Ph.D: Mm-hmm. NIKHIL KORATKAR: the dendrites, and you form lithium metal inside nanoscale enclosures, it remains stable inside those enclosures, and fine, it's not forming dendrites. But what is your volumetric performance? Now, we are not even calculated, we are not even bothered to calculate the volumetric energy density and the volumetric capacity of this electrode. We were only looking at the gravimetric number, which is You n you normalize by mass. Volumetric means you normalize by volume, right? So because we had this porous 3D network, we had really good gravimetric numbers, no Sri Popuri, Ph.D: Yep, yeah. NIKHIL KORATKAR: question. But our volumetric numbers were actually bad because you know you have all this empty void space, right? and when you talk to industry, what industry tells you is that look, we it's good to have gravimetric good numbers, but we really care about volumetric. So a company like Tesla it really struggles to find the space to put all its batteries, right? So for example, Sri Popuri, Ph.D: Course. NIKHIL KORATKAR: the trunk of the battery, a lot of the trunk space is lost because you have battery there and and obviously if you're a driver, you want trunk space, right? So so Sri Popuri, Ph.D: Yeah. NIKHIL KORATKAR: volumetric is very important. And again, if you look at the 99% of academic publications in battery space will only quote gravometric number. Now slowly people have started to realize the importance of aerial capacity and the importance of volumetric capacity. But it's a slow process. People still in the academic community, they just brought gravimetric numbers, right? So think sometimes industry and academia are talking different languages. And there needs to be better communication so people really realize that what is it that really industry needs, right? And we didn't realize it back in twenty fourteen and twenty fifteen, we we did not realize that. And that's why that's the reason why I published a Article in Nature Reviews Materials, which came out in 2022, in which I discuss a lot of these issues. I talk about nano versus micro, why industry is hesitant to adopt nanoparticles as the active material and batteries. What do you gain when you use nano? But what are some of the fatal flaws when you use nano? And this

whole issue of volumetric performance and all of that. I bring a lot of this out in that review article. I'm very happy it's been cited now I think more than three hundred times now in the last three or four years. So people are starting to you know understand the real practical issues more and more now. Sri Popuri, Ph.D: Because why why I'm asking this question actually? Because I have seen those both sides of the world, both the academics as well as industry, Because sometimes people approach me saying that I discovered X and Y material. I want industry to scale it up. But the fundamental point is that they manage to get that material uniformly, for example, in a gram scale. NIKHIL KORATKAR: Yes, Sri Popuri, Ph.D: when I ask them, can you make it in let us say kg scale? Either they will say that they don't have the facilities or it is out of the scope for them. But the kind of conditions they use to make that gram-scale material is almost impossible, To replicate in any possible chemical environment. NIKHIL KORATKAR: Absolutely, absolutely. And your point is is it's a great point. just a basic question that the manufacturing process you use, for example, if you if you use sputtering, right? Sri Popuri, Ph.D: Yes. NIKHIL KORATKAR: So I I told you about the fact that there are these seeding layers that we need to deposit on the copper foil to kind of get lithium or sodium to plate uniformly. If you deposit the seeding layer by sputtering, how is that going to work? That requires high vacuum. It has to be a roll-to-roll process, right? So so again, for an academic paper, it's fine. You do you do the sputtering and you get great results, you publish it. But it's never gonna go anywhere because the you can't do that in the real world, right? So similarly, people talk about manufacturing in grand quantities. Many times they use processes that are batch processes, small, small quantities, Sri Popuri, Ph.D: Yeah, exactly. NIKHIL KORATKAR: right? you cannot scale that. So that is A very important point. but I think the way academics think is that let's demonstrate it with this process and then somebody else will come and will figure out how to scale. that's how academics think. But but it's hard, you know, unless you yourself are proactive in taking those steps, it's probably not Sri Popuri, Ph.D: Yeah. Because as as you mentioned, not only spottering, I have seen a lot of startups approaching me saying that they're using the ALD and PLD kind of techniques. Right? of course, I mean the the results are fascinating, definitely, Because you can get a almost like atomic layer like control over your product, right? And obviously it gives the excellent battery performance or the stability or whatever it is. but when I talk to them, And I need not 10 grams, I need 100 kg, right? Because I don't want to make five coin cells. I want to make, I don't know, hundreds of 21700s or 4680s, for example, right? In order to have a qualification process, right? Until and

unless we are thinking in the direction, it's quite challenging to translate those fascinating research into commercial products. In the similar lines, Professor, you talked about the applications of batteries in some healthcare I think a lot more cases what you are targeting is the a huge, batteries, right, in electric cars. We already discussed about the energy storage, for example, for data centers and also the electronics, like mobile phones, laptops, and so on, What kind of technologies you are involved with at the moment, or our academic research is going into about the medical devices, for example, because given this AI revolution, right, we are getting a lot of wearable devices and so on, particularly for tracking the health, which is a huge market, So, what's your take on that? Where the batteries will fit into this kind of application, which technologies are you working on now? NIKHIL KORATKAR: I think obviously flexible batteries is like extremely important because if you're gonna have wearable devices you need flexible batteries. flexible batteries again the issue is that the choice of materials changes a lot because you want things that can bend and you know not undergo damage. so the performance can go down because everything right now is optimized for performance. So when you change these materials, you can have loss of performance. So that's one issue that can be mitigate that loss of performance and yet have flexible batteries. So that's one thing. The other thing is use of any toxic material should be avoided in these kind of batteries, right? Like for example, cobalt, the use of cobalt. And you know, they should be very safe, right? Because if it's something on your person, then you cannot have any risk of fire and any of those things. so I think this is a new nascent area of of growth, I think. but these are some of the important criteria of flexibility, safety, and also you know performance loss being mitigated. because everything is right now so optimized for these these regular batteries that that when you go to flexible batteries, they don't perform as well. They lose energy density. Sri Popuri, Ph.D: when you're talking about the sodium, a lot of these things are derived from the lithium systems, NIKHIL KORATKAR: Mm-hmm. Sri Popuri, Ph.D: That means they might also contain some sort of elements such as cobalt and so on, Which might also involve if you want to push their energy densities. So what's your take on these NIKHIL KORATKAR: Yeah, so sodium, I think the beauty of sodium is that it's an intercalation chemistry, very similar to lithium. So a lot of the learning from lithium has got transferred to the sodium, right? So for example, you have LFP, lithium ion phosphate, and similarly you have NFP, sodium ion phosphate. So you don't have any cobalt. you have manufacturing of NFP is not that complex. it can be done, it can be scaled. And you know yes, graphite doesn't work. sodium doesn't go into graphite because again,

sodium being less electropositive than than lithium doesn't have that that driving force to go into into graphite. and also the ionic size is larger, so it doesn't go into graphite. But people have shown that hard carbon works really good. So hard carbon again is easy to make. If you think about hard carbon, it's basically a disordered form of graphite with a lot of defects. And it can be made from things like waste, you know, it can be made from all kinds of waste can be used to make hard carbons. And there's a lot of work on this, like how to make hard carbons for batteries. So the manufacturing of the anode is scalable, possibly even low cost if it's made from waste. the manufacturing of the cathode is you have to avoid cobalt-based. designs and I think go after things like NFP. they the the NFP will have to be doped to make it work better obviously. So you have sodium, ion, phosphate, pyrophosphate type of systems, and then you dope the iron with manganese, which expands the kind of network, right? Because manganese size is bigger than iron size. So it expands the network and it allows sodium to diffuse easily in and out. So So, yeah, you will you will have to do some doping. Similarly, oxygen is doped with fluorine because fluorine is more electronegative than oxygen. So that helps to stabilize the oxide better. So, you're going to have to do doping and things like that, but it's not too I would say cumbersome and onerous. So the the way I see it is that a lot of learning from lithium can be applied to sodium. We can figure out What are really the practical sodium systems that we can develop in the future? And again, the electrolytes, you know, people right now are using organic electrolytes. but there's been a lot of breakthroughs on the electrolyte side for sodium, which have not happened for lithium. So, for example, there are Chinese companies, C ATL and I think BYD, they have shown non-flamable Sodium ion electrolytes. So they have used certain additives that kind of suppress the ignition temperature of the electrolyte. So again, the possibilities seem very exciting with sodium. and I think the chemistries are such that they are less mineral intensive compared to lithium. So far from everything I've seen, sodium Looks promising and is already starting to displace lithium in in some areas. as I said, the Chinese again, the C ATL, BYD, they're manufacturing sodium and batteries, selling them. And they're slowly starting to, you know, where cost and safety is important, they are starting to display because LFP, despite the fact that it's less flammable than NMC, it is still flammable. And anything that's still flammable will always have that insurance cost, right? So we're starting to see displacement of L F P by sodium ion already starting to happen. so I'm quite bullish on sodium. Sri Popuri, Ph.D: So, what about the supply chain for sodium, for example, Because the argument is always comes back to whenever there is a lithium price is shooting up, people talk a lot about sodium, right? When lithium is coming down, everything's lithium-lithium, right? That's the kind of trend we are seeing. And also from the recycling perspective, actually, definitely lithium got the advantageous scenario, right? talking about the through systems and so on, right? Where sodium might not have. until and unless we're having any any precious metals there, nickels and Mn and so on. NIKHIL KORATKAR: Correct, correct. Sri Popuri, Ph.D: Right? So from the recycling perspective, I think sodium can also struggle in in that domain, isn't it? NIKHIL KORATKAR: Yeah, but I think for for sodium you may not need recycling because you have a lot of abundant Sri Popuri, Ph.D: Good. NIKHIL KORATKAR: sodium and you and hard carbon again very straightforward to make. so I think that the driving force for sodium recycling is not there. Driving force for lithium recycling is there. but again, cost is the ultimate decider of these issues. So As the price of lithium ion batteries goes up, the question is lithium ion batteries, you recycle them, is it more cost competitive than just going with sodium ion, right? So so I think that is the open question, which we'll come to know in the next 10 years, right? it's all a function of how fast lithium price rises and how fast cobalt price rises, how fast prices like things like manganese, nickel, how fast they rise. And and then the question is that recycling then becomes more attractive as the prices of all these things go up. But recycling has its own cost. So if Sri Popuri, Ph.D: Yeah. NIKHIL KORATKAR: you look at the cumulative cost then of lithium mined battery plus recycling, how does that compare to sodium? And if sodium is cost competitive, then then it will you know gain. But also, I think the flammability issue is very important because if sodium ion batteries are shown to be reliably non flammable. That is a game changer because that then gives it an advantage over lithium in many applications, right, where maybe performance is not that critical, right? in those kind of systems, having that non-flammability angle gives you that great advantage. Sri Popuri, Ph.D: So on on the different note, actually, I mean, as we were discussing earlier about the AI, right? This artificial slash superintelligence, that really comes back to the where these kind of sophisticated systems can help to discover the better materials, like be it quantum computing, be it DFT calculations, whatever, so on. So what's your view on that? How how this super ingredients is going to impact the best performing materials, particularly from the battery perspective, in your opinion? NIKHIL KORATKAR: Yeah, I think that's a great question. so I'm an experimentalist myself, right? I've I spent my whole career doing experiments. And the way we make discoveries is very often intuition-based. then, for example, trial and error. So a combination of intuition and trial and error. And sometimes I get my best

ideas while doing a shower, right? Because when you're showering, you turn off your phone, you turn off your computer, Your mind has the space to think, right? Normally all the time responding to email, because on your phone all the time. So that's why people I think get their best idea in the shower. But but but this intuition-based model is probably not the most efficient model, right, for for making discovery. so I think that there is a really an opportunity with AI and machine learning. To create a formal design process by which you can rapidly screen through hundreds, hundreds of thousands of possibilities, right? And then identify those maybe five or ten that are the most promising, which can then be subjected to a trial and error kind of process. Right now it's intuition that drives the trial and error, whereas if there's a formal design process, a design structure that drives that. So that's where I think AI and machine learning can be very valuable. the problem is that AI machine learning needs data. You need good data, right? And where do you get the data from? So typically you get the data either from either from computational models like density functional theory. But density functional theory is a very inefficient way to create data because very slow takes a long time to run DFT calculations. So DFT is not a smart way to create a lot of data. the other way to create a lot of data is through experimental studies. Now, there have been thousands of battery publications, right, over decades. however, that data is not curated properly. So, in fact, we are trying to get some funding where we are trying to propose taking all the battery literature and in sort of ingesting all that data. And after ingesting the data, organizing it properly, And making sure all the units are matched up, everything is matched up properly, everything is consistent. And then after you do this organization of the data, then you have to extract this the AI models have to be able to extract this data. So they need access to this data. You need data security, you need so many things, data governance, how do people add data to this database? So you need there's a lot of issues that are still need to be worked out to create the kind of databases that AI can then tap into. to do the kind of reliable discovery that we need, right? so I think there is a great opportunity here. but a lot of work has to be done in terms of curating the data properly. there's also quantum. Now, quantum is again very exciting buzzword, but one has to be careful where you apply quantum and where you don't apply quantum. So in the battery context, what I can think of is that. any problem where you're looking at the resting state, like for example, a battery electrode in equilibrium where all the ions have come in and settled into that structure, right? when you look at a battery electrode in equilibrium in its resting state, DFT does a great job. You know, it is very accurate in terms of kind

of predicting the operating voltage, in terms of predicting the capacity. there's no justification to try to use quantum in these kind of things. In fact, quantum will be more computationally expensive. In some ways. But when you have problems that involve transients, so any kind of ion migration process that's happening, that's much more complex because now you have bonds breaking at one end, you have new bonds forming as the ion is migrating. And in these kind of situations, the electrons are get localized. anytime you have a delocalized electron cloud, a DFT does very well. so I think in certain types of problems where you have transients, using quantum makes a lot of sense. it could be much more accurate than DFT. so you have to identify those problems. But the other issue with quantum is that. you need in order to denoise the system, in order to have low noise, you need a lot of qubits. so for example, if you're modeling an ion, let's say diffusing through a complex oxide material, you might need to model hundreds of electrons, right? So, for that, if you have a computer that has 200 qubits, that's just not enough, right? So, you're going to need hardware to improve to the point where you know you have hundreds of qubits, thousands of qubits, you need to get to that point where the quantum will st will will work and give you reliable results. So you gotta identify the right problems, but you also have to have improvement in hardware, quantum computer hardware, to get to the point where you can really make a difference. Sri Popuri, Ph.D: So I'd like to really finish the discussion with the one of your strong expertise as a as an academic, right? Mentorship. Mentorship of PhD students, mentorship of postdocs, as a student, I'm asking you if I'm just graduating from university and if I want to join your team as a PhD student, so what are the criteria you follow to select? NIKHIL KORATKAR: yeah, so I think the academic institution from which the student comes is important. you know it has to be like a high-ranked school, it has the student has to have a good GPA. if the student doesn't have a good GPA, there's some other compensating factor. Like for example, the student has worked in industry, So if GPA might be low, but if the student has published some nice papers, then that compensates. So you look at everything. The most important thing for me is I need people who are self-motivated. Because I'm not the kind of person who's constantly micromanaging people, looking over their shoulder, checking is the person working hard enough. That's not what I enjoy doing. I give people problems and I give them space and freedom, So unless they are self-motivated, that's gonna be a problem. So I'm really looking for people who are ambitious, self-motivated, driven. very often you can realize that through even a Zoom interview, you can talk to the person and kind of figure out does the person have those qualities? Sri Popuri, Ph.D: So one of the points you

mentioned just now, they need to come from the potentially from the elite universities or elite institutions. why that should be criteria to choose for example. NIKHIL KORATKAR: Yeah, so it's just because of risk mitigation. So the issue with in the US is that most of the money from my grant goes into the student cost. So if I get a you know like a hundred thousand dollar grant per year, 90,000 of that will go into supporting the student, the stipend in the tuition of the student. so student is a very risky. Like if you have a bad student that can really sink your research, sink the grant, make the sponsor very unhappy. If you hire a person from some unknown university or or a lowly ranked university, it could be a gem. The guy could be a gem. He could be really outstanding. But you're taking a risk. While if you hire a guy from a person from a well-ranked school, high-ranked school, you mitigate that risk to some extent. Everyone is off a certain minimum standard, at least, right? So now among those, you also need to find the right people. But it's just a risk mitigation approach. Sri Popuri, Ph.D: So based on your experience, what makes a PhD thesis or PhD student successful? NIKHIL KORATKAR: Yeah, that's a very good question. so I think firstly it's up to the advisor. it's up to the advisor to give the student an important problem. just defining the problem is something that the student doesn't do usually in the US. I don't know how it works in in, for example, Europe or India but or China. But in the United States, it's usually the advisor who defines the general area that you want to work in. So the advisor has to come up with important problems that academia cares about, that industry cares about. That's the responsibility of the advisor. The advisor also has to push the student to think in a non-incremental way. Because it's easy to do incremental research, right? It's easy to take someone's paper that already works and to tweak it a little bit and to show that little bit of tweak improves it by 5%, 2%. that's easy to do, right? Because the chances of it working are very high. Because you're working with a baseline that already works. But the high risk, high payoff research is to try something that nobody's tried before, or something that people have tried and have always failed at. so when you do high risk research, chances of failing are also very high. So 90% chance you're going to fail. which can be difficult for the student to handle, right? Because the student needs a thesis to defend. And if the advisor gives this very high-risk, high payoff work and nothing works, then the student is in limbo. So what I do is I typically push the student to do high-risk, high payoff work, but I at the same time define a lower risk project, which can be like a fallback option that the student can pursue in parallel, which is more likely to work. I think the advisor's role is very, very important here. But from a student's point of view, the student

has to be open, has an open mind to try new things. You don't want a student with a closed mindset, right? You you've got to be willing to try new things, you've got to be bold, and you've got to work hard, obviously, you've to be motivated. I said self-motivation is a very important. And very often the student has to overcome hurdles. So, After maybe a couple of years of research, the student is more hands-on than me, right? if the student runs into a problem, something isn't working, the student, if he just comes back to me and says, you know, this is not working, what do I do? chances of me coming up with a solution may be less efficient than the student talking to other students in the group, talking to other students in the university, brainstorming, trying to figure out a way to overcome, circumvent that problem. So problem solving, independence, these are also very important qualities. ability to write. Again, when you write a paper, some people have a natural gift of expressing themselves in a very lucid manner. Some people unnecessarily they complicate stuff. so the art of writing some people have some people don't have. so I think that's that's also important. but I I think it's very important also for the advisor to support the student, to back up the student, right? So, for example, that's something I learned in my journey, right? Because I have graduated 35 PhDs in 26 years. I've been doing this now for 26 years. my 35th PhD student just graduated last week. And so there's more than one per per year, Sri Popuri, Ph.D: Congratulations, please. NIKHIL KORATKAR: right? And you know, one thing I've learned Is that yeah the papers are important and the publications are important and the patents and you start a company and all of that. But what is also very satisfying is the when you see these students, they come to your lab and they are in a very fragile state when they come, right? They have no money in the bank, they're coming from a different country, they are in In a very vulnerable state. And to carry them through the PhD and then to see them join all these top companies is very satisfying. And to see them flourish, you know, rise to high high posts in the company, have families, get married, have children, to see all that process is extremely satisfying. And one thing that I have really become better at over the years is in supporting these students. So I encourage them to do internships. in companies. usually that's six months. They're gone for six months, the project suffers, the research suffers. But I still I never say no to a student who comes to me and says, Hey, I want to go on an internship, because that really helps them to get a job. Because when they go to their company, they establish relationships, they build their CV. it is much easier for them to find a job after they go. Or even the last year of their PhD, they're already hardwired into jobs. So I think it's very important also for the advisor to think about the welfare of their students, right? Not only, not Sri Popuri, Ph.D: Yeah. NIKHIL KORATKAR: just your welfare, right? Because you already made it, right? You already got this all these publications and funding. You're not in the vulnerable state. It's the student who's in the vulnerable state. And so I've become way better at this now than I was when I was, let's say, pre-tenure. I was only thinking of myself. But now I think more about I put the student's interest above my own interest. In most of the cases, I I've been doing that. And the student then has loyalty for you. Because when the student sees that you got their back, then the student has loyalty for you. I have so many students who graduate and then their leftover PhD work that they've not published. They write papers on weekends, they'll work, they'll complete those papers, and we'll publish papers a year, two years after the student has been working in a very high stress job like Intel or IBM. Why? Because the student has a loyalty to me. He knows that I sort of helped him graduate on time. you know, I could have kept him in at RPI for another two years to churn out all this leftover work into into papers. But then that would have affected that student's possibility to get a job. So that loyalty is so strong that So many of the students they publish after they've gone out, after they've left the group. I think this is something I now when I observe my own father. So my father was an army officer, he served in the Indian Army, and the Indian Army there's a creed that the ease, comfort, and safety of your men are always come over and above your own ease, comfort, and safety. Each and every time. So this is like a creed that he always followed. And I can never be anywhere close to him. But I've seen the way he treated the soldiers who served under him. Where you almost endanger your own life before you risk the life of your soldiers. What I'm doing is nothing even comparable. But but it's the same concept that the way the advisor treats the student is very important. You have to give the student a sense of security. And that security will then allow the student to try crazy ideas and to fail. You have to tell the student it's okay to fail. and you also have to make the student, I think, realize that research is a journey. Right? The journey is very important. the journey is as important as the destination. destination is yes, your project works, you are able to publish in nature or science, you know, that's a fantastic destination. we're all striving for that destination, but If you only enjoy the destination and not the journey, then your PhD journey might be very stressful and hard. But if you actually enjoy all the steps along the way, the successes, the failures, then that's much, I think, more manageable, that journey. in terms of this impact on students' lives that an advisor can make, I had a student from China. so after he graduated, I take my students out for dinner before the

student leaves to you know join a company. So this guy was going to Intel. And he told me while having dinner that for the last, you know, his father, grandfather, great grandfather, even going back like 500 years. Everybody in his village has been a rice farm. Everybody in his family has always been a rice farmer. And this guy was the first person from his village to go to school to do an undergraduate degree in China. Then he came to the US. I hired him and advised him through his PhD and nurtured him through that process. And then he got this job in Intel. And I was thinking to myself that. Even a hundred years ago or two hundred years ago, this would have been impossible, right? That a rice farmer in China could now be doing this high profile job in Intel in Silicon Valley, earning this Sri Popuri, Ph.D: Yeah. NIKHIL KORATKAR: huge, huge salary. So the world has changed, has transformed, it's amazing. It's amazing we live in this time. And then I played some part in this journey, helping this guy go from a being a rice farmer to this high profile engineer in Intel. I had some role to play in making this happen. Whatever however small that role was, that gives me a great sense of satisfaction. And I think that's why the student side of it starts to really grow on you more and more. That, yeah, the publications are important and the citations are important, patents are important, company starting is important. But the human angle, the interaction with these 35 PhD students that have graduated, and The role I've played in their journey, in their growth, that is immensely satisfying and maybe even more impactful than all the publications. Because remember, these students they go to industry and they continue to innovate. So it's not just your innovation, it's all the innovation of these 35 students over the next 20, 30 years that all builds up cumulatively, right? And so your impact as an academic is not just in what you do, but in what all your students are going to do down the road. Sri Popuri, Ph.D: So professor, if you're starting your journey again in 2026 a fresh PhD graduate, how do you plan your career actually? NIKHIL KORATKAR: I think that ultimately nowadays you have to postdoc almost. Like in my time, like I got a faculty position without postdoc. in the last year of my PhD I got a faculty position. nowadays that's almost impossible. You have to do one or two postdocs, some people do three postdocs. that takes a lot of patience, it takes a lot of determination. so you have to be very determined, you have to be very single minded and focused. you have to use postdocs that diversify your knowledge base. So, for example, if you worked on one type of technology in your PhD, you want to do postdocs where you do research in different areas, maybe diversify your knowledge base, maybe do some computations also, so that you have expertise in both computation and experiment. I think one thing I did, which I regret, is that I'd never developed a

computational background. So I always was only a pure experimentalist. Now the things I've told you about AI, machine learning, and quantum is based on my interactions with my friends who are the computational people. but so I would try to build as diverse a knowledge base as possible. Use these postdocs in a very smart way, and then when you start your academic journey. the first five years is is just crazy. You're just trying to make tenure, you're trying to write a lot of proposals, you're trying to write a lot of papers, you're trying to target journals that will accept your work quickly because you have to show in five years you have to show a lot of research output. But then after things settle down, I think it's important to think about what is it that really excites you. Right. Because if you're gonna do this for the next 20, 30, 40 years potentially, you really have to find things that that really make you excited to get out of bed and go to work, right? in my case it was this fundamental discoveries, new properties or materials, but also properties that could be really useful to people, right? So the engineer in me always wanted Societal impact, right? So I was trying to marry this fundamental curiosity with societal impact and to try to marry the two. But every person can be different, right? Some people might want to do more applied type of research, some people might want to do just answer fundamental questions that maybe will not be useful anytime in the near future, but to everyone, to each his or her own, right? but find something that really excites you, that wants to make you go to work every day, get out of bed. my wife jokes to me and says that you know 6 a.m. You just want to jump out of bed and start working. And and and I do. I just want to get out of bed and and start the day because you know I'm excited, excited to meet the students, to talk to them. that's another great aspect about our job, right? As as we age, we stay young because you're interacting with 18, 19, 20 year olds, undergraduates and like graduate students who are 20 in the early 20s. That's a beautiful thing. And and even undergraduates, you know, the joy of teaching, right, it has intensified for me over time. as I've gotten older, I've started putting in more and more time and effort into teaching. slightly dial back the research a little bit and dial up the teaching effort, right? Because these undergrads, they pay a lot of money to come to college. at least in the US. there's a lot of tuition that they're paying. And they really deserve your full effort. I've really become passionate as a teacher also more and more over the over the decades. so as a young academic, I think you also want to give equal importance to the teaching role. it's harder when you you're trying to make tenure, but after that you really Also, realize that your job as an educator is not just to do research and publish, but also to be a very effective teacher and try to understand how do I motivate students right in the classroom. One very common trick is to give practical examples when you teach. So when you're teaching, if it's all theory equations, people often ask, Am I ever going to use this? But if you use that theory to then apply to a practical problem in every lecture or every few lectures you introduce some practical problem where this math can actually be used. I think that really motivates the students. But but for that you have to think, right? You have to spend time and effort into designing the lectures. And very often research active faculty they don't do that. they spend all their time and energy in research. but I think it's very important for us as academics to do justice to that teaching role as well. Sri Popuri, Ph.D: academic institutions are becoming more and more competitive, right, to get a job, particularly for postdocs. So based on your almost nearly three decades of experience, from doing your PhD to becoming full time professor so what advice do you have for the postdocs, who are in the dilemma that whether they should follow their academic passion? or expose their career beyond academia as well. So what's your suggestions there? NIKHIL KORATKAR: So I think one also has to be practical, right? So one has to look at your resume and make impartial evaluation. so again, for a PhD, there are three jobs. There's industry, you can be a researcher in industry, you can join a national lab, or you can become an academic, right? So now in now in academic, also there are two types of roles. You can be a teaching professor, you could be a research professor who also does. Little bit of teaching. So these are all the different types of jobs you can get at the end of a PhD. So if you're doing a postdoc, obviously you're targeting the academic route, right? There, which is either a full-time researcher with a little bit of teaching or mostly a teaching professor. That's what you're kind of targeting. I think you have to be impartial, look at your resume. what are your publications? What is the impact of your publications? What are your citation counts? what would your chances be of landing an academic research type professorship given how competitive it is, right? and if you find that you know you do have a strong case, then you should go after it and take as long as it takes, you know, two, three postdocs, as long as it takes. But if you find that you don't have any eye-catching, eye-popping, you know. Publication, you don't have something that's going to make your resume pop. Because if you're in a search committee, you're going to see a hundred resumes, right? Out of that, you have to decide to interview three or four people. There has to be something in your resume that catches the attention of the committee. If there is no such compelling thing, then I think you should think more about transitioning into a teaching role. if that's of interest to you, There are a lot of lower ranked universities that need

people in a teaching role. Even higher ranked universities now are trying to save money by having a research dedicated faculty and teaching dedicated faculty because the teaching faculty they teach a lot more and their salaries are a lot less than the research active faculty. So just a point from the point of view of running the university, it's much more cost effective to have a lot of teaching types also. So even the higher ranked schools, like my university also has a lot of now teaching active faculty. so if there's nothing in your resume that is popping out, then maybe you think more about a teaching role or think about industry role or think about a national lab, those those kind of so I think you just have to be honest with yourself and honest with your resume. Sri Popuri, Ph.D: So last point from my side, professor, that particularly given the AI, you know, boom that is going on, And the impact it is going to have on the research itself, the way you are pursuing the problems, the way you are delivering the solutions. Even writing academic paper is kind of you know disrupted by these AI tools, What kind of skill sets, PhDs or postdocs should develop? To be in the race for this AI world. NIKHIL KORATKAR: Yeah, so I think at the lower level it's using AI to search to search solutions to research questions. It can speed up that whole process. instead of you reading hundred papers, instead of you finding hundred papers that are relevant, AI can do a lot of that work for you, right? so using AI to speed up your research. So that's something that everybody has to do, including faculty members, right? but then Going one level deeper, I think learning how to use AI models for discovery, I think that's something that that people should try and invest time and effort in. because you know if you have at least some basic knowledge, you can understand what the issues are with use of AI. so just to educate yourself more on How AI can be used as a very powerful tool to drive materials discovery. Actually using some AI models and trying to do some searches, trying to duplicate searches that others have done just to see whether you can reproduce that. so I think getting more educated, more literate when it comes to AI is something that all students and postdocs should do. How is it going to transform research? I think it could just speed up everything. So the speed of discovery, the speed of translation into real world, not just for academia but also for industry, it could potentially speed up that whole process. And I think that could be one revolutionary you know outcome that if it takes ten years right now for a technology to move to market, maybe it could take four years or five years, it could just speed up the whole. So I th I think people need to be educated and try to educate themselves as much as they can. Sri Popuri, Ph.D: So last point from my side, that what is your life motto? What drives you every day? NIKHIL KORATKAR: I think the life motto

has to be to really s sort of enjoy what you're doing. and to enjoy what you're doing is something that will allow you to to perform at a high level and will also allow you to perform, it will give you longevity. So if you want a career of thirty, forty years doing something, you have to enjoy it. So to enjoy it, One thing that's important is to focus on the big stuff. Not get mired in minute things. You know, so for example, I avoid a lot of responsibilities in a department which involves serving on this committee, serving on that committee, because it's just you know routine, mundane kind of work which doesn't make you want to hop out of bed like I mentioned before, right? So I think you need to prioritize. what do really want to do? What should I avoid? What should I not get bogged down in? Right? and if you can really enjoy what you do on a day-to-day basis, that's very important. The other thing is, don't take yourself too seriously. A lot of people get stressed a lot, you know, when their paper gets rejected, or their grant proposal gets rejected, or they suffer a setback, they take it very personally. I think if you take a more larger view, and this is More difficult to do when you're struggling to get tenure, but after doing this for 20, 30 years, I think it's easier to have this perspective. that you're one among billions of humans on the planet, right? what happens to you in the larger scheme of things? the universe doesn't really care that much, right? If your paper is accepted on or if Sri Popuri, Ph.D: Yeah. NIKHIL KORATKAR: your paper is turned down by nature or or if it's accepted by nature, it matters a lot to you, but the universe really doesn't care, right? So I think you've really got to. De-stress yourself, right? And this is very important for an academic, also, in terms of advice for young academics, is you're going to encounter rejection throughout your life. Your papers will get rejected, your grant applications will get rejected. Sometimes students will give you a bad evaluation in your course. So you've got to not carry stress. You've got to let stress just sort of roll off you, because otherwise you won't you won't enjoy yourself. again, if you want to enjoy yourself, enjoy what you do, you have to manage, learn how to manage stress. And one way to do that is to have other interests, right? if you only do research 24-7, you probably are not going to be able to manage stress. So you've got to be able to enjoy sports, you've got to be able to enjoy other activities like running, exercise, gym. you gotta develop other interests that give you a balanced life. a lot of academics don't have a balanced life and I think that that's important. Sri Popuri, Ph.D: With that, I would like to thank you, Professor, again, for joining today on the SP show. I really appreciate it for views and the your patience with a lot of questions NIKHIL KORATKAR: Thank you so much. it was real fun to talk to you. I enjoyed it greatly. Sri Popuri, Ph.D: Thank you, Professor.