 Welcome to this session where we're showcasing some of our top student researchers. Every summer we invite Stanford students to participate in a student energy lecture series in which they speak about their innovative energy research and they receive life feedback during that time from a Stanford or communication student. This year was our 10th year of this series and this summer it featured 14 of our top students who were nominated by their professors. It was also of course the only time that we've conducted this series virtually. The student speakers were judged by a panel that considered the content, relevance and the materials of each talk and I'd like to thank Jenny Mill, Ryan Bartholomew and Steve Egglish who joined me on that judging panel. By the way I'd also like to especially thank Maxine Lim who coordinated the series and the students and I'd also like to thank Jun Li and Rachel Frank who moderated and ran the series and they all did a fantastic job. So from among those many great talks we picked out four and these will be the presentations that you'll hear today. So the way we plan to do this is that I'll give very brief introductions of all the four speakers now and then each of them will speak for about 10 minutes and hold off on your questions and answers because at the end of all four talks then John will coordinate the Q&A session. So let me begin by introducing so the first speaker will be Maha Yusuf. Maha is a 50-year PhD student in chemical engineering and I'll establish a faculty for the future fellow. Her research uses imaging diagnostic tools to mitigate the failure mechanisms of lithium ion batteries and she's advised by Professor Lambertus Hesselink in the Department of Electrical Engineering and co-advised by Mike Tony and Joanna Walker at National Accelerator Lab. She's an MS in chemical engineering from Stamford and a BE from the National University of Sciences and Technology in Islamabad. Next up we'll hear from Angel Young. Angel received her BS in chemical engineering from National Taiwan University in 2015. She's currently a fourth year PhD in chemical engineering in the group of Professor Mateo Connalo. After that Kelly Wu, it's our third speaker, Kelly received her BS in electrical engineering from Caltech in 2018 and she's currently a second year PhD student in electrical engineering working in the lab of Professor Sabanti Chaudhary. And last and certainly in no means least is William Frank. William received his BS in nano-engineering from the University of California in San Diego and is currently a fifth year PhD student in material science and engineering in the lab of each way. And so why don't we get started? So the first talk will be by Maha Yusuf and then we'll move on from there. So Maha, I'll pass it on to you. Hello everyone. My name is Maha Yusuf. I'm very honored to be here today to talk about enabling extreme fast charging of lithium-ion batteries as part of the Stanford Energy Distinguished Student Lecture. I want to start with a bigger picture overview of a word we want to live in, a word that is worth investing in, a word with reduced CO2 emissions, with reduced global warming, a word that is sustainable and has global energy security. And one way to do that is to enable electric transportation. Now that's a word I want to live in. Electric vehicles or EVs reduce global emissions. If you look at the data here, which compares life cycle global warming emissions of a mid-sized gasoline car with that of a mid-sized battery electric vehicle, you can clearly see that there is around a 50% reduction in global warming emissions with most of this reduction coming from the operation of the EV itself. Thus EVs reduce global emissions. However, if you own an EV today, you also know that there are many problems associated with it. For example, if you're driving from Stanford to LA, you need to think about the charging stations to stop at, and you also need to factor in the charging time required. Thus, a potential solution is extreme fast charging or XFC of EV batteries. With this, we are literally looking at around 10 to 15 minutes of charging. Now, this is a goal set by US Advanced Battery Consortium for 2023. Before we delve into the challenges of XFC, let's look into how a lithium-ion battery works. In this schematic, I'm showing a positive electrode known as cathode, a negative electrode known as anode. Now, when you are charging your electric car, you are moving lithium ions from the positive electrode through the electrolyte to the negative electrode, where they get stored between these layers of graphite as electrochemical energy. Now, when you drive your car, you're discharging it, right? So you're moving these lithium ions back from the negative electrode to the positive electrode. Now, what happens when you try to charge your car really, really fast? What you're doing is, you're moving these lithium ions very fast from the positive to the negative electrode. And during that process, lithium or Li-plating develops. In this schematic here, as an example, I'm showing two custom-made battery power cells that were charged between 10 and 15 minutes. And you can see both of them experience the deposition of this whitish material, which is lithium metal. And this process is known as lithium-plating. This process becomes favorable over lithium intercalation in certain regions of this graphite anode, due to the buildup of lithium ion concentration near the anode at high charging rates. Lithium-plating has been identified as one of the key battery degradation mechanisms during XFC, and it needs to be understood well at a fundamental level so that we can develop fast-charge batteries. Therefore, in my research, I investigate where, when, and why does lithium-plating deposit on the battery anode during XFC. In my proposed approach, I'm working with simultaneous neutron and X-ray imaging. It is a non-destructive imaging modality that provides 3D visualization of Li-plating. Because X-rays are sensitive to the electron density and neutrons to the nuclear density, this dual-mode imaging provides us complementary data. I did these experiments at a BD2 imaging beamline at NIST, a schematic of which is shown here. You can see that this red corresponds to neutron beam produced in a nuclear reactor. Blue corresponds to an X-ray beam produced by an X-ray tube. Both of these beams hit the sample here right in the middle at the same time during the imaging experiment. We have a neutron and an X-ray detector that detects the neutrons and X-rays after they have interacted with the sample, providing us a respective neutron and an X-ray image. My spatial resolution was 10 to 15 microns, which is great to look at lithium metal deposits. In my experimental design, I firstly imaged an uncycled graphite anode and secondly a cycled graphite anode to compare changes between them before and after cycling. This is an optical image of a cycled graphite anode. I cut a strip around this red dotted line here and put it in a glass capillary for imaging. For my data analysis, I am using information from my X-ray histogram and my neutron histogram and I use them to create a 2D histogram where I have neutron pixel intensity on the X-axis and X-ray pixel intensity on the Y-axis. Now, because I'm interested in locating lithium and lithium is a lozy material, it would probably lie in this green region here with high neutrons and low X-ray intensity. Here, we are looking at 2D grayscale images of a cycled anode. On the left-hand side, we are looking at a neutron image where we can identify copper strip and the anode pretty clearly. In the X-ray image, we can only identify copper clearly and the anode not so clearly. One thing I want to point out here is that, as you can see in these images, copper delaminated from the anode because of the X-situ destructive nature of the sample itself. I took these 2D images and created a 2D histogram to convert them into my segmented colorized image as shown here. Here, green outer circle corresponds to a glass capillary, yellow corresponds to copper, and pink corresponds to the battery anode. This is an optical image of the graphite strip that I used for imaging, where lithium deposition is pretty evident inside this red highlighted box here. Because I collected this data at many different angles, I can correlate these 2D slices or images to the spatial location on the graphite strip. For example, first and second slice corresponds to the thick yellow line labeled 1 and 2 here. And we can see growth of lithium plating clearly on the graphite anode in the third and fourth slice. Now, these are 2D slices of a 3D volume. So I took them and stitched them together to create 3D visualization that shows me the thickness of lithium plating. In these images, yellow corresponds to 10 microns of copper and purple corresponds to 80 microns of graphite. On the left hand side, we are looking at a 3D visualization of an uncycled or pristine anode, where you can see these 2 strips on top of each other with no sign of lithium plating. However, on the right hand side, where we are looking at the visualization of a cycled anode, you can see deposits of lithium plating on the graphite anode inside these black circles very clearly. This data has been collected on an anode which was harvested from a battery. However, a better way to image lithium plating would be to do so non-destructively, which means without opening the battery, okay. So in my next sample, I took a full battery pout cell assembled at Argonne National Lab, rolled it up as shown here to fit in the field of view for my imaging data. This is a neutron and an X-ray image of the battery pout cell. You can see in my segmented image I was able to identify battery anode and cathode pretty well, which we can see in 3D visualization as well. Now this is important for my future work because I'm interested in locating lithium plating in three dimensions in these battery pout cells non-destructively. To conclude, I use simultaneous neutron and X-ray imaging that provides 3D visualization of lithium plating on graphite anodes. This dual mode imaging is advantageous in its ability to separate different battery components from each other very easily. In this data set, I was able to answer where does lithium plating deposit on the battery anode during X of C in lithium ion batteries. In my future work, I'm currently working on quantifying this amount of plated lithium in three dimensions, but mostly focusing on my non-destructive in-situ imaging work of full battery pout cells at different states of charge to really answer the question when and why does lithium plating deposit on battery anode? We believe that addressing these fundamental questions about lithium plating will lead to rational battery designs that will prevent lithium plating from happening in the first place. With this, I want to acknowledge my three advisors at Stanford, Slack, and University of Colorado Boulder and my funding agencies for funding my PhD. To a sustainable future of electric mobility, I thank you all. Hi. So I guess it's my turn for my presentation now. So hello everybody. My name is Angel and I'm a PhD student in Karniello Group. So today, I'm happy to share my research with all of you which is to identify the active site for propene combustion using nano crystal catalyst. So firstly, I hope to walk you through this title to let you understand this for those of people who are not familiar with chemistry. So basically, we hope to identify this active site where the site is always categorized into the following three categories in terms of geometry, terrors, kink, and step. So we hope to find the site that is able to lower the activation barrier for a specific reaction because with this lower activation barrier, we are able to facilitate this reaction and have a higher rate for this specific reaction. So next, we hope to also target the propene combustion because basically for every gasoline powered car, there exists this catalytic converter which contains this diesel oxygen catalyst to transfer this harmful molecules into less harmful products. And we choose propene to be our model compound because propene is one of the most common harmful molecules in car. And we let it mix with oxygen to produce CO2 in the water which are less harmful to the environment. And next, we hope to use this nano crystal catalyst to find the active site. It is because in the past, people always use a conventional way to synthesize the catalyst which always results in poor defined size, shape, and base. So with this well controlled nano crystal catalyst, we're able to really create a uniform system that is systematically change the surface to help us find the active change with the change in the surface site. So from now on, I'll be working you through how I use this well defined nano crystal to find the active site. So firstly, we choose plating and platinum for propene combustion. It's because these two elements are known to be the best elements for propene combustion. And with this synthesis, we have synthesized a set of materials ranging from 2 nanometers to 10 nanometers of the plating and platinum. Here shows the microscope images of these five different sized nano crystals. And later, we deposit them onto aluminum support because we hope to expose this metal surface to the reaction. And then finally, we use this particle size. We have also conducted the particle size distribution analysis to see that the particle size is maintained before nap to deposition. And you can see the shape of all the materials here are mostly spherical. And the very first thing we hope to ask is that whether we do alloy these two elements. So we have used several techniques to examine that. So from ICP, we see that the plating and plating ratio is very close to one, which is our target ratio. And then from CLK absorption, we see that the dispersion is decreasing with increasing particle size, which is also in line with our expectation. And then from the EDS mapping, we see that the signal plating and platinum are lining atop with each other. So from this several techniques, we see that it's possible that we do alloy these two elements. And we have also have a collaborator in Slack, where we are able to see the X-ray, X-ray absorption spectroscopy result. And from here, we can see that the coordination number of plating is almost the same across different particle sizes. And from the coordination number, we can see that for plating, it's an FCC material. So it's basically surrounded by two of other neighbor atoms. And from the coordination number, we see that half of the neighbor atoms are successfully replaced by platinum. So from the several results, we can tell that we indeed synthesized a random alloy of plating and platinum. And now we can use this set of beautiful materials for the catalysis. So we have put these materials into our reaction mixture, which is composed of propane, oxygen, and water, try to simulate the environment of engine. And before I start, we used this tunnel frequency to measure the activity of our materials, which is basically a number to tell how many of the reactants are successfully transferred into the products. So from the tunnel frequency, we can see that with the larger particle, the tunnel frequency also becomes higher. So this tells us that the propane combustion is a structural sensitive reaction. And now with this, we're able to see now the question becomes, how are we going to use this data to find the active site? So the very first thing is that what is the site that is the larger particle? So we hope to tell that by firstly examine how the particle interacts with the reactants by using rate order. So rate order is basically a quality that is that tells us how sensitive our particle is to the change in the reactants. So we have three reactants in this reaction, propane, oxygen, and water. So firstly, we have changed the propane partial pressure and see the change in the activity to tell the oxygen, the propane rate order. So as you can see, as the particle size increases, the propane rate order becomes more negative. And then as the oxygen partial pressure increases, the oxygen rate order becomes more positive. This tells us that it's possible the larger particle contains a higher fraction of the site that is more sensitive to the change of the reactants. And then finally, from water rate order, we see almost zero rate order for water, which is very uncommon for such reaction because water is always known to point in the nanoparticle. So this water effect is interesting. So we hope to see whether water is playing an interesting role for this particle. So we have found another collaborator in Suncatch, which is able to simulate the particle shape on the different environments. So firstly, they have synthesized or they have simulated this pleidian-plotten nanoparticle on the vacuum. And you can see that the surface is highly faceted. And then we have exposed this particle in water environment, which is done by putting this hydroxyzoate onto the surface. And after that, you can see the particle shape now becomes more spherical, which is very close to our microscope observation for the largest nanoparticle. And then this tells us that with the presence of water, this step site is now exposed on the larger nanoparticle. And using a more quantitative way, it means that the surface energy of the step site is now stabilized in the presence of water. And then with the DFP calculation, we can also see that several different representative reactions are also having lower activation barrier on the step site. So this tells us that it's possible the step site is the one that has a lower activation barrier. And the larger particle has a higher fraction of the step site, which is stabilized by the presence of water. And now we really want to answer how the step site, watch the step site looks like for the larger nanoparticle. So we have constructed a model which is able to help us count the percentage of different ensemble of the sites, the particle. So we see that basically this is the change of these ensembles with the change in particle size. And then to compare that change without another frequency, which is the experimental data, we see that this is this edge ensemble 7-7 is in line with the slope of the terminal frequency. So this tells us that the 7-7 ensemble site is the active site for this reaction. And in the future, we should find a way to engineer this site in order to maximize the rate for this reaction. So the takeaway of this is that firstly, we have used this well-defined materials to identify the structure sensitivity. And then next, we see that the water is able to drive this reconstruction of the surface and help expose the more active step site. And then the active site 7-7 is active for this reaction. So with that, I would like to thank our collaborators and Akane Yellow Group members. And I'm happy to take any questions. Thank you. Hi, everyone. My name is Kelly. And thank you for joining me today. It's a pleasure to be here to be able to share my research with you on photo conductivity in diamond done in the wide bandgap lab advised by professors for Monte Chaudry. So in this presentation, I will first talk about how wide bandgap materials are suitable for use in power electronics and specifically diamond as such a semiconductor material. Afterwards, I will introduce photo-conductive switches and talk about the characterization setup for our doped diamond devices. And finally, I will present some preliminary results we have measured recently from our fabricated samples. So power electronics is the process of using semiconductor switching devices to control and convert electrical power flow from one form to another to meet some specific need. These kinds of systems have applications in energy conversion in power grids, electric drives, in industrial motors, power supplies for data centers, and electrical vehicles, among many other applications. And what leads to our interests for wide bandgap materials in this space are their intrinsic properties that allow for reduced energy losses, high voltage operation, high temperature operation, and high frequency operation. The potential energy saved annually by the U.S. by switching from silicon, which we know to be the most traditional semiconductor to wide bandgap materials in power systems is summarized in this table from a study done by the Oak Ridge National Laboratory done in 2017. And perhaps to put into perspective the importance of engineering efficient devices, we know that around 10% of of energy is lost in inefficient power conversion. So while we can think silicon for the electronics we know today, it has reached its operational limits and we must now look towards new materials such as diamond for the next generation of high power electronics. We are choosing to look into diamonds since it appears to be the ideal wide bandgap semiconductor, at least on paper. You may have heard of gallium nitride and silicon carbide, which are more well known wide bandgap materials that are commercially available. And diamond on the other hand is still in its earlier stages of development. In comparison to silicon, which has a bandgap of 1.12 electron volts, diamond has a much larger bandgap of 5.47 electron volts. And the wide bandgap of these kinds of materials gives them a high breakdown voltage. This allows devices to handle much higher voltages with a fraction of the size as pictured here. And more uniquely, diamond also has high thermal conductivity, which helps prevent heating effects and potentially eliminate the need for cooling systems. So all these material properties together allow for lighter, simpler, and more efficient systems using these new wide bandgap materials. So since diamonds are newer material, I just wanted to give a brief introduction on how they're grown and their dopants. So diamond single crystalline substrates are grown by high pressure, high temperature methods, and by chemical vapor deposition. And then to control the conductivity of a semiconductor, we can dope the material by adding impurities, which can source free electrons or holes that can change the conductivity of the material. So in diamond, boron, nitrogen, and phosphorus have been used as dopants, and they each have their respective activation energies, which define the energy required to release their carrier for conduction. And as doping technology in diamond has improved through research in recent years, we are now able to fabricate devices such as diodes and transistors based in diamond technology. For example, previous work done in our group has shown a PIN diode blocking over one kilovolts across 8.5 micron thickness before breaking down, which corresponds to where this current spikes here. And in other places around the world, such as Japan and Europe, diodes have also reached hold up hold off voltages of up to two kilovolts. So with many communities actively invested in diamond, its high voltage capabilities have been demonstrated. So now in my research, we are studying diamonds, photo conductive properties. The basic operation of a photo conductive switch is pretty simple. The conductance between two terminals on a semiconductor is modulated by the absorption of the optical radiation. This photon energy then creates free carriers in this material to generate current under the influence of a bias voltage applied to the contacts. So more simply put, when the laser is off the switch ideally is considered open in a circuit. And when we shoot a laser at the device, it would ideally act as a closed switch, allowing current flow for approximately the duration of the laser pulse. And our motivation for developing photo conductive switches is for high power applications. These switches have been shown to be high speed, compact, with very low jitter, and they most immediately have applications in pulse power. However, as we are moving towards electrification of our energy, we will rely more and more on safe mechanisms for making and breaking circuits that carry large amounts of power, which photo conductive switches can play a role in. When we create switches that are triggered by light, it isolates the control from the output, eliminating the potential for an invert and switch triggering, and also simplifies the control circuit. So now I would like to go into a bit more detail on the principles of photo excitation. As I mentioned earlier, in photo conductive switching, the incident light is absorbed in the semiconductor, exciting carriers into their conduction or valence band. In intrinsic photo conductive switches, in which the material is not doped, it depends on the absorption of a photon with energy greater than or equal to the band gap energy. However, in extrinsic photo conductive switching, which is our focus, impurities in the samples allow carriers to be excited by sub band gap photon energies based on their activation energies. So to excite carriers across the band gap, a UV light source is needed, while to excite carriers of dopants, less energy is needed, which would allow for greater flexibility in choosing our excitation source, which is helpful for engineering applications. And so throughout these past months, I fabricated samples, which consists of a couple of photo conductive switches on a three by three millimeter dope diamond substrate. The basic optical setup, we have for photo excitation uses a 1064 nanometer ND Yag laser. The second harmonic is generated to produce 532 nanometer light, and then the beam is directed to the sample being probed under the scope. So the fabricated sample is connected to a test circuit that is pictured over here, and this results in a pulsed photo response that is measured across the load, resistor corresponding to the laser interaction with the photo conductive switch that is in series with the load. And so some preliminary testing was done using past Dope Diamond samples in the group before starting my own fabrication. And so the data is shown on the left, you can immediately see that there is a pulse in the photo response that corresponds to the laser pulse. The rise time is quite sharp while the fall time is slightly slower, and this might be due to heating effects from the laser causing the resistivity to remain lower as it is cooling down. We have since done further testing on fabricated photo conductive devices which produced a much sharper response and order of magnitude higher in voltage, and the current on over off ratios was calculated to be on the order of 10 to the 7 and greater. So more comprehensive studies are to come that will allow us to identify how different dopant species and concentrations affect the photo responsivity among other factors. And this will then allow us to engineer a more optimized photo conductive switch. And based on our preliminary results, I think it's very promising and it's encouraging us to keep exploring diamond as a substrate for light triggered switching. So this concludes my presentation for today, but I would like to acknowledge the Stanford Precourt Institute for Energy for funding this research. And I would also like to thank my advisor, Sir Banti Choudhury, my research partner on this project, Muhammad Ali Malakutian, and the members of the Wideband Gap Lab. Thank you. Hello, everyone. My name is William. I'm a fifth year PhD here at Stanford and the lab of Professor Itzway in the material science department. And I'm going to tell you guys a bit about some of my recent work using electron microscopy to characterize the structures within battery materials. So as Maha mentioned, batteries are extremely important for clean energy. But one main impediment to the adoption of electric vehicles in the transportation sector is their driving range. So one solution to increase the adoption is to enable faster charging. But the other direction for that, to address the driving range, is to just increase the energy density within the battery. And one way to do this is by moving beyond lithium ion chemistries. So in lithium ion chemistries, there is no elemental lithium within the battery. You're just shuttling lithium ions between a cathode and a graphite anode. But by instead taking advantage of different chemistries, for example, lithium plating chemistry, which is undesired in lithium ion chemistries, by taking advantage of lithium metal plating and using that as your anode chemistry can actually increase the energy density of your battery. But of course, this hasn't been done because this chemistry is very difficult to actually use in the battery. And the reason why that is is because lithium metal is very inefficient. Your lithium ion battery is actually very good at cycling. But lithium metal batteries are very bad at cycling and will quickly die after only a few cycles. And the reason why that is is due to instability of what's called the solid electrolyte interface or SEI within the battery. And this SEI is basically a surface layer that exists on the battery anode, whether it's lithium metal or graphite, coming from the decomposition of the electrolyte at the very reducing potentials of the battery anode in the lithium-based battery. And what this SEI does is it stops the further decomposition of the electrolyte. So it passivates the electrode electrolyte interface, but it also dictates performance because it controls the ion transfer from the electrolyte onto the electrode. So it's an extremely important structure in the battery. And if we were to build better batteries, for example, lithium metal batteries, we need to understand the SEI as physical chemical properties. So people have done this for many years just using the characterization tools we have available to understand the chemistry and probably the structure of the SEI. And it's commonly used in the field as a technique called X-ray photoelectron spectroscopy, which is a surface analytical technique. And it gives very good spectral resolution, but gives very poor spatial resolution. But using these kinds of advanced characterization tools, people have built models of the SEI, trying to understand its structure and the different chemistry of the components within. This has led to this conventional model of the SEI, where you have this inorganic inner SEI layer consisting of lithium fluorides and oxides and carbonates interfaced with the anode and a more porous organic outer SEI layer interfaced with the electrolyte. This is widely accepted within the field. However, the length scale that these different components exist at are not really understood, whether these components are segregated over the nanometer scale or the micron scale. And these are not really accessible through surface probes like XPS. Ideally, you would use what's called the transmission electron microscope. So electron microscopy is highly useful in material science because it offers resolutions down to the subnanometer, so atomic resolution. But the problem is that these electron microscopes use very high-energy electron beams. You have electron beams coming in on the order of about 300 keV kinetic energy, which is highly damaging to the sample. And that's why TEM, or transmission electron microscopy, has not been widely used in the battery community. So this has recently been addressed through advances in what's called cryogenic TEM, whereby cooling the sample to liquid nitrogen temperature, we can actually stabilize battery materials within the TEM for atomic resolution imaging. And this was first demonstrated by our group here at Stanford. And we published this image in the Journal Science where you can actually see the atomic resolution lithium metal lattice here interfaced with an organic layer of SEI embedded with inorganic materials such as lithium oxide and lithium carbonate. And this looks pretty similar to the SEI models that have been proposed previously, so that's not too surprising. But if you look closely between the experimental image of the SEI developed by or shown through cryo TEM with the conventional model of the SEI, you'll actually notice some differences. And one difference is a certain component in the conventional model not observed through TEM, and this is this component of lithium fluoride. So that's only one component among others. The rest are mostly observed within the TEM. So maybe you're wondering why this is so important. But the reason why is because lithium fluoride is considered the most important SEI component out of this whole mosaic of different SEI building blocks. Lithium fluoride is known to be widely critical to the battery performance and just by adding more fluorine into the electrolytes. So many lithium ion batteries are heavily fluorinated in the electrolyte. You can further stabilize the lithium metal cycling. So the question here being asked across the entire lithium metal battery community is where is lithium fluoride? Because we know it's to be it's so important in the electrolyte lies and not seen in the SEI. So the goal of my research here is to uncover what's happening with lithium fluoride and where it's distributed within the SEI. So to do this we can build some experiments that make good use of fluorinated electrolytes because the SEI chemistry closely matches the electrolyte chemistry. So here we're using a standard battery electrolyte, an ethylene carbonate based electrolyte. This is what's used in the lithium ion battery with 10 volume percent. So a good amount of fluorination through the use through the addition of a fluoroethylene carbonate battery solvent. And this is used in silicon containing batteries along with lithium metal containing batteries. This is a cycling efficiency plot and basically by adding 10 volume percent of fec your cycling stability goes up significantly. And this is widely known to be the effect of fec. If you then take this battery apart and start to do x-ray photoelectron spectroscopy analysis of the lithium metal anode we can see this commonly accepted structure of the SEI where the outer surface is more organic containing more organic carbons and fluorocarbons. And then as you sputter deeper within the sample so approaching the lithium metal surface getting deeper in the SEI there's more inorganic SEI components such as lithium fluoride and lithium oxide. And if you take this structure to be your SEI structure you'll build this kind of a conventional SEI model where you have the inorganic inner SEI and a more organic outer SEI. Now if we then take that same sample and put it inside the TEM this time we're using the TEM the transmission electron microscope in scanning mode coupled with the electron spectroscopy capabilities that can actually map out the chemical distributions at the nanoscale on individual lithium deposits. So lithium deposits in this kind of filamentary structure which is problematic for different reasons but basically we can see that mapping out the structure and chemistry of individual lithium deposits with nanometer precision will find that even across microns of lithium we're not really seeing any fluorine within the SEI we can see oxygen we can see lithium but no fluorine. So basically within this SEI which is this outer oxygenated layer on the lithium surface where there is not this inner SEI rich in lithium fluoride that XPS would suggest. So what's going on here and this is the major discrepancy shown in the TEM and XPS and the whole field really. So in principle this led me to think about the different length scales being probed in XPS a surface analysis technique and that in microscopy and basically it's in principle possible that this lithium fluoride component could be depositing outside of the SEI. So to probe this outside of the compact SEI layer on lithium metal. So to probe this we build a model system where we can have a very well defined electrochemically active surface. So here we're using a carbon membrane just evaporated onto a TEM grid and we can use that as a working electrode in a battery and build a intentionally build an SEI on this very flat substrate and basically what we find is if we take this substrate build that SEI within this highly fluorinated electrolyte we can see now that there's a whole new regime of SEI that's not being probed on the lithium metal surface through microscopy but it's being probed through XPS. Basically what we see is that all these particles are depositing within the SEI at the micron scale no longer the nanometer scale and these particles are basically where the lithium fluoride is all going and if we look closely at the nanostructure of these different lithium fluoride particles we find that there is a interesting bilayer structure where these particles have this dual inorganic layer rich in lithium fluorides and a more organic outer layer more oxygenated outer layer which matches the structure observed in XPS but not actually observed directly through microscopy and this is basically what appears to be the origin of that discrepancy between microscopy and XPS and through traditional surface analysis techniques it becomes it's impossible to actually probe the true structure of the SEI. Very quickly just by observing the different lithium metal structures a very high surface area deposit it becomes very obvious that these lithium fluoride crystals which are depositing at the micro scale are in fact very dilute on the lithium surface and this is the origin of why it cannot be seen within the SEI layers because these lithium fluoride crystals do not exist within the compact thin film of SEI material but instead are depositing very sparsely and because these lithium fluoride crystals are depositing so sparsely it becomes more in question how much these lithium fluoride deposits can actually affect the battery cyclability but basically through through high resolution probes such as cryo electron microscopy it becomes possible to actually resolve the structure and chemistry of the SEI with nanometer precision and you can only then start to see how traditional surface analysis techniques cannot truly resolve the actual structure of the solid electrolyte interface and lastly I just want to thank my my advisor and collaborators for this work along with the organizers of this symposium for the invitation to share my research thank you thank you all those were four tremendous presentations you should all be proud of your work I think I saw most of these this summer and they're even better this time so we did as almost exactly as planned by Richard we have about five or six minutes for questions so I'm going to try to be brief and maybe just ask one key question to each person combining concepts as we go along so in Mahas talk there were a group of questions and they were basically related to this may be on your research radar agenda now are are you able to to what extent can you ascertain the just where the plating occurs and what the distribution is with an eye towards remedies and ultimately being able to improve both performance and economic competitiveness I know that's a broad question but I'm sure you've thought about at least one of those questions I did see the questions in the Q&A and they're you saw oh great great that makes my job easier sorry I answered like a couple of them I I hope people can see that anyways coming back to your question right now the data I've collected has been on destructive samples so coming back to that question of like exactly where is lithium plating and why it's plating there for that I'm focusing on my in-situ imaging work which will be done in a full battery non-destructively and will enable me to answer why is lithium plating in certain location as compared to the other but I know there are research groups who are working on like other techniques to look at it look at the same problem within even higher resolution great thank you so much for Angel an intriguing question to me maybe because it's not my field but it did sound interesting does the propane reaction change the geometry of the nanocrystals particularly in the quarters and edges yeah so we have to look at the before and after T and images for the nanoparticle and see no change in the corner and kink site so basically the geometry are really similar before and after so yeah I believe there's no significant change on that great thank you and for Kelly a lot of interest in what the negatives are for these diamond wide bandgap devices leading towards improvement of the economics why aren't they more widely adopted now and where do you see the most promising applications yes so that is definitely a fair question and I think just like any research topic it takes time to develop the technology to a point where um perhaps the costs of getting new infrastructure to support these new materials becomes worth it in terms of better performance efficiency so it is not widely used basically I guess we're here doing research to hopefully show that the benefits are worth like the commercial interests of perhaps investing in these new technologies and so I think the applications would be similar to what gallium nitride and silicon carbide are currently taking over some like charging applications applications in electric vehicles general high power switching I think by using these new materials we would have a lot smaller devices that are more efficient so to me it would probably be for diamond which is the most immature of these materials probably still like a decade looking forward but pretty much looking at the I would say the costs and the benefits that people would have to see whether this is this pans out in the future which I believe it will be worth it yeah so I guess you're saying that as a basic researcher you're trying to create options that could lead to marketable technologies and of course whether or not it's marketable depends on what what the markets are doing so actually a similar question was asked for William how do you see this space obviously a lot of our audience is kind of market business rated I have noticed hanging out with Richard and Brian Bartholomew is that a lot of you who are in the lab now 10 years from now will be CEOs and CTOs are very successful companies so William do you have inclinations in that direction or what what advice would you give colleagues who might be so inclined yeah um thank you for this question we'll see about uh CEO or CTO but for for the um for the battery field there's a lot of very interesting technology startups out there um along with uh larger companies such as uh Tesla so Tesla is probably going to be the largest producer of lithium ion batteries soon but for kind of next generation lithium metal type lithium metal batteries for example or sodium ion sodium metal batteries there are a lot of really interesting startups out there doing a lot of work trying to commercialize these technologies um the hardest part about in my opinion the hardest part about making these next generation batteries is not getting a great chemistry or getting a great material the hardest part that's actually like that's sort of step zero in a sense of actually commercializing your technology um and the hardest part really is manufacturing these these materials at uh say like the gigawatt hour or terawatt hour scale um and this is a problem that Tesla among their suppliers Panasonic LG Chem um they've all recognized that this is an extremely difficult task um taking basically something from uh the lab to the factory um but still I am optimistic for these these uh next gen chemistries lithium metal especially along with silicon anodes which are kind of the next generation of lithium ion battery anodes uh but recently if you follow lithium based startups this can be called quantum scape which is actually started here at stanford um trying to commercialize solid state lithium metal batteries and they recently announced that they'll be having a public offering and I think um to will um in Tesla's recent um public um announcement they announced that they're replacing graphite with silicon and they'll be using that for their lithium ion batteries the silicon anodes it was just announced like I think few uh a month ago yeah uh just playing along this storyline uh I actually knew Tim home from quantum escape when he was a grad student and he was in the lab at that point and now as you saw in the an energy seminar a little over a year ago uh he's uh writing high so I look forward to seeing uh all of you uh doing so I I do think as I look at your advisors um often they do do the breakthroughs usually working with you in the lab but they seem to be some of the best people to ask about the manufacturing and scale up problems because they're able to take the science that they've developed and say well there's a trade-off here so if you could save a lot of money we'll give you back a little bit of performance and so on that seems to be the ticket so I do look uh for big things from all of you in the future and I'd like to one last time thank maha angel Kelly and William and especially Richard for putting this session together thank you so much and thank you to the participants for going through the very intense competition during the summer program and being victorious and sharing the results of your research with us here today thanks again one last time