 And then just share it with you. Okay. Okay, so hello everyone. Thank you all for joining us today in another session of Fusion Breakthroughs. I'm Matteo Berberino for the International Atomic Energy Agency. And I'm here today with my colleague Joanne Liu from the IAEA Office of Public Information and Communication. Hi, I'm Joanne and I am a writer and editor at the IAEA, also co-host of the podcast Nuclear Explains. So if you have a chance, we recently did a podcast on fusion energy if you'd like to check it out. Thank you, Joanne. So in this special episode today, experts from the US DOE Lawrence Livermold National Lab, the National Ignition Facility will present results from the historic Fusion Breakthrough, which was announced on December 13, 2022. Scientific energy gain for the first time in a fusion experiment. Extraordinary. So scientists that the NIF ignited a fusion reaction and produce about 3.15 megajoule of energy from 2.05 megajoule energy output of the 192 lasers. An effort that achieved net energy for the first time in a fusion experiment. So many thanks to our guests today and for their availability. We have Dr. Homer Hurricane, chief scientist for the inertial confinement fusion program design physics division. Dr. Jean Michel de Nicola, chief engineer for the NIF laser system. Dr. Annie Critcher, principal designer and team lead for the integrated modeling for integrated modeling. Dr. Abbas Niquro, target fabrication program manager. Dr. Dave Schlossberg, science lead, NIF nuclear diagnostics. Dr. Bruno von Wundergem, NIF operations manager. And Dr. Alex Zilsra, the principal experimentalist. So the format of today will be a little bit different than what we have at what we have usually. There will be one talk. So we'll begin with one talk from Dr. Hurricane on how ignition and target gain larger than one was achieved in inertial fusion. And then we'll go into a Q&A panel discussion, which will be moderated by Joe Andrew. Please type your questions and comment to the chat box during the technical talk and we'll go through your questions at the end. I'd like to ask to the other panelists to make sure they are on mute and to turn off their videos while we go through the first talk. So Omar, thank you for being here and I hand it over to you. Super. Thank you very much for the introduction. I'll get right into it. So our recent ICF experiments on the NIF are an existence proof of laboratory ignition and target gain. And that's pretty exciting. No physics mystery obstacles apparently stand in the way of ignition. And that's what we mean by an existence proof. What we mean by ignition is the explosive thermodynamic instability of the fusion plasma. And by gain, we mean energy out greater than energy in. What's really exciting is the theoretical prediction of the physics parameter regime. When I'm referring to the loss in triple product where ignition was expected to occur is actually consistent with our results. And that's why there's no apparent mystery in getting to ignition. Additional energy at fixed laser power was very beneficial. The implosion physics was more sensitive to engineering control of the laser targets than originally thought. And so far, very high gain, high compression targets have not worked as expected, which actually was a big frustration for us and forced us into a different strategy. And all the breakthroughs which we're going to talk about in this talk over the past decade have actually used low gain designs. But in the end, the last bullet here, which is the remarkable takeaway is that we can now talk about burning plasmas, ignition and scientific break even in the past tense. And that's exciting. Okay, so in order to get high fusion yields, what we need to do in ICF is assemble our fusion fuel into a configuration that can stop alpha particles in the fusion plasma. We're aiming for a configuration that looks like this image over on right, where we have a cold assembly shell, a spherical assembly of cold fuel in a shell. That's the blue DT. And in the center of that, we have a hot spot of hot DT. And what we are trying to do is implement the DT fusion reaction, which combines deuterium and tritium to produce a neutron and alpha particle. The neutron carries 80% of the energy, the alpha particle carries about 20% of the energy. And we're trying to capitalize on the fact that the refusion reaction rate shows in blue increases the function of temperature. So we can get this configuration and stop say 70 to 80% of the alpha particles from the DT fusion in this hot spot. By stopping those alpha particles, we get their energy, we get their heat. So when we get their energy and heat, that causes the temperature to go up. Because of the nature of the DT fusion reaction rate increasing, if the temperature goes up, we get more reactions. If we get more reactions, we get more alpha particles and we can stop more of those alpha particles in the hot spot. We get yet more heat out and we get the temperature to go up again. And by doing this process essentially repeatedly, we get what's called self heating. And if it is intense enough, we get ignition. The conditions that we need for this to occur are a hot spot aerial density, which is a product of density times radius of 0.3 grams per centimeter squared. It makes central density of about 100 grams per centimeter cubed or more and a central pressure of over 400 gigabytes, which is over twice the pressure at the center of the TUM at the center of the sun. So if these conditions are met, we get a thermal feedback loop, which is what is described here and ignition is generated. So the way we create that assembly of fusion fuel is to use what we call indirect drive for inertial confinement fusion, which is illustrated in this cartoon here. Where we use x-rays to accelerate and to ablate and accelerate a capsule of fusion fuel to extreme velocity. So here is an example target configuration. Our fuel capsule is the center object. The exterior object here is a Hallrom. It's a high Z metal can. Laser beams enter through the top and bottom of that metal can through apertures that we call the laser entrance hole and deposit energy on the inside of the Hallrom. Those laser beams are absorbed by the inner surface of that Hallrom wall and they're converted to a bath of x-rays and that heats the environment inside the Hallrom much like an oven, except in this case it's an x-ray oven. That intense bath of x-rays causes the surface of the capsule to explode because it ionizes and it explodes with a pressure of about 150 megabars, 150 million atmospheres of pressure. And that crushes the capsule inwards upon itself and accelerates at the high velocity of order 300 to 400 kilometers per second. We call that stage the implosion. That kinetic energy when the implosion actually runs out of anywhere to go because the volume is shrinking as it accelerates inwards. Eventually it runs out of anywhere to go and that kinetic energy from the acceleration is converted into internal energy and we call that stagnation. And if the conditions are right again we can get ignition and to make this all work it requires incredibly precise control of the design of the laser and have the target physics parameter. So a lot of finesse goes into making this work. So let's cover some terms real quick because the terms burning plasma ignition and gain all actually mean something physically different yet the terms are often misunderstood and conflated. So for the purposes of this talk and many of our publications, these are the definitions we use. So burning plasma for ICF burning plasma occurs when the self heating that I described earlier exceeds the PDV work coming from this implosion process that heats and compresses the DT. This is very much analogous to the magnetic fusion definition where for magnetic fusion a burning plasma is one where the self heating exceeds all external heating of the DT. And the difference between ICF and MFE is that while MFE has multiple ways of heating the plasma and ICF we just have one, the PDV work of the implosion. Ignition, by ignition we refer to the loss and criteria and this is a condition where the self heating power exceeds all the DT plasma power losses. That's the condition that sets up the plasma to have this thermodynamic instability. The losses that the plasma are competing against or the heating is competing against are radiative losses, electron heat conduction and negative PDV work where the implosion actually becomes an explosion and that sucks energy out. But if the loss and criteria is achieved and a quote ignition is achieved, you get thermodynamic instability, which by that we mean we get an explosive increase in temperature and fusion yield generation. Target gain is a separate idea. It stands for when the fusion yield exceeds the laser energy into the target. Now, unfortunately, the 1997 National Academy of Sciences committee that reviewed the proposals for NIF construction actually used target gain as the definition of ignition in their report and that was later adopted by the Department of Energy. So that didn't help with this conflation. So when you look at official documents, just be aware, they're actually referring to target gain and calling that ignition. But for us, we have this distinction that I outlined here. Okay, so here's the National Ignition Facility. It's a remarkable piece of technology. It's about the size of three football fields in length and width. So it's a gigantic facility. What you're looking at here in the schematic, you're looking at capacitor banks on each side where most of the energy enters the system. And we store, you know, 300 to 400 megajoules of electrical energy in those capacitor banks to drive any given experiment that on each side of the facility are laser bays of 96 laser beams on each side. A set of preamplifiers at the top end here and the main laser beam lines are at the lower end. This configuration down here on the lower right is what's called the switch yard, which has mirrors in it that deflect the laser energy into the target chamber, which is shown on the right. Okay, so the NIF actually delivers a frequency tripled three omega light into the target chamber. So that 300 to 400 megajoules of electrical energy that goes into the capacitor banks gets turned into three megajoules of red light. And just before entering the target chamber, that red light go passes through what's called the tripler, which converts the one omega frequency red light into three omega frequency blue light. That to the blue light actually comes out as two megajoules, not three megajoules, it's not a perfectly efficient tripler. So we inject two megajoules of blue light into the hallroom, which is shown here on the right with a person's fingers to give you an idea of the scale. It's about one centimeter in length, roughly speaking. That two megajoules of blue light that enters the hallroom is converted into x-rays as I mentioned a moment ago. But again, that process is not 100% efficient. So only about 200 to 250 kilojoules of that energy is converted into x-rays and is absorbed by the capsule, which is shown here on the lower right again with a person's finger for scale. The capsule is about two millimeters in diameter. Because that capsule ablates, most of that capsule is turned into plasma, so very little of that 250 kilojoules of x-ray energy actually gets into the DT fusion fuel in the center of the capsule. So there's an extreme loss of energy going from several hundred megajoules to 20 kilojoules before the fusion fuel even sees that. And well, why do we do this? Well, the whole process of inertial fusion is to sacrifice energy, knowingly sacrifice energy for energy density. So that's the way inertial confinement fusion works. But the result of that is most of the energy is lost before the fuel sees it. Because there's such a large loss of energy from the beginning of the laser system to what the fusion fuel sees, that actually leads to several energy gain metrics that are relevant for ICF. And they're illustrated in an equation form in the diagram on the right. Again, we have our hallrom with our capsule inside and inside the capsule is the fusion fuel. So you can define measures of yield out compared to energy in for different layers of this system. So the target gain is defined as the fusion yield divided by the energy deposited into the hallrom by the laser. And it basically represents the energy in and out of this largest volume. The capsule gain is the fusion energy out compared to the energy absorbed into the capsule. So it's basically this volume here. And then finally inside you can define a fusion fuel gain, which is the yield over the energy that actually made it into the fusion fuel. So in the plot on the left, you see a plot of fuel gain as a function of a complicated parameter that involves the plasma pressure inferred from the experiment, the confinement time, which is the tau, the temperature, the time average temperature from the experiment to a funny power, and an inference of the energy that was put into the fusion fuel versus the fusion hotspot. So if you plot things along this parameter, basically all our data falls on one curve, which are the gray points going from the lower left to the lower right. The key takeaway here is that over the last 10 years, we've had a 5000 times increase in fusion fuel gain from where we started back in 2012 at the end of the National Ignition Campaign, back at the beginning of the National Ignition Campaign until now or just last month. And that's quite a remarkable set of achievements. But again, because they're different layers of the target, again, they're different fuel, they're different gain metrics. The fuel gain metric was actually passed in 2014. The capsule gain metric was passed in mid-2021. The target gain was the most recent result in 2022. So after 10 years as of December 5th, 2022, we have a target gain of about one and a half, which is also what we refer to as scientific break-even, a capsule gain of 12 and a fuel gain of 160. Now, when you talk about gains, it's important to remember that even having a target gain greater than one is not net energy, because that doesn't account for the energy used by the facility. It only includes the energy injected into the target, so keep that in mind. All right, so how did we do this? Well, it was actually a lot of work to get things working correctly. There were many processes that frustrated our progress. The key processes that frustrated us are illustrated here. We had instability control problems, symmetry control problems. We had an issue with getting sufficient energy coupling to our system, to our target. We had target quality problems, and we continue to have problems with achieving ultra-high compression, which forced us into a new strategy. So the way we control some of those issues is through what's called the laser pulse shape, and this has been understood for a while. We inject a laser, a temporal, designed temporal laser pulse into the Hall-Rom, which is shown in blue. This is laser power versus time shown on the left. That is converted into X-rays in the Hall-Rom, which is the red curve shown on the left. And by designing the temporal profile in time of the laser pulse, we can control key elements of our target physics performance. And I'm just highlighting a few here also to get you familiar with some of our jargon. So this early part of the laser pulse, we call the foot. And what the foot does is it actually controls the stability and the majority of the fuel entropy, which we sometimes call the 80-abatt, a measure of that is the 80-abatt alpha IF, which stands for alpha in-flight or 80-abatt in-flight. We have the peak power part of the laser pulse. That actually tends to control the velocity of our implosion. And since kinetic energy, as you'll see, is important to us, that peak power is important. And then finally, this last part of the laser pulse, we call that the coast period or the coast time. And it turns out that seems to be key to controlling the efficiency of the conversion of kinetic energy into DT internal energy in our implosions. And it's linked to something called the radius of peak velocity, which is where the implosion acquires its peak velocity. The reason why this pulse shape has influence over the performance of the implosion is because the implosion is driven by this pressure on the outside called ablation pressure. And the ablation pressure is related to the radiation temperature of the hallrom, which is the red curve. So it responds to that. And it also depends on what the material of the capsule is made out of through the Z, the atomic number, the atomic weight, and the albedo, the way the ablator responds to that path of X-rays. So that's how we have some chance of controlling some of the properties of the implosion. Okay, so where did we start? Well, back in 2010, 2012, we were using a plastic ablator, what we called low-foot implosion. So that early part of the laser pulse was, quote, low. And that was designed that way to have a low adiabat and very high compression. They were intended to produce high yield, but those implosions actually underperform for many reasons. So we'll be going through a set of plots like this, where I'm showing you neutron yield in the total number of fusion neutrons on the left plot versus DDI on temperature on the X-axis. And on the right plot, I'm just showing basically the same fusion yield except converted into energy units rather than neutron number, just so one of those might give you a better feel for what's going on. And this fusion yield on the right plot is plotted versus the inferred hotspot pressure, which is in gigabars, which is billions of atmospheres. Okay, so these implosions produced much less than one megajoule of energy. They didn't seem to have as much of a pattern in their behavior as was desired. So there were definitely some problems associated with them that had to be sorted out. One of the key problems which was hypothesized at that time and ended up being proven later was that hydrodynamic instability was really problematic for those implosions. And it's going to be illustrated by this movie here, which I'll show in a moment. But basically, this is the process where small perturbations in the surface of the capsule are growing exponentially in time under these enormous accelerations that an ICF implosion undergoes. So there's an example, an analytic formula which expresses this instability process. This is the growth rate for the instability. It goes with the square root of the wave number, the acceleration. And that's what we call Rayleigh-Taylor instability. It's mitigated somewhat by density gradients and ablation velocity. There are numerous forms of this equation, but when you have a positive growth rate, again, small perturbations in the surface become exponentially large in time. And that will be illustrated by this movie here. So obviously we want to have like a shell of high density material around a hot, central hot spot. If you have a lot of instability, the implosion tends to want to turn itself inside out. And you defeat the hot spot in the middle, and the shell of material on the outside is basically torn to pieces. Now, of course, we need to get high velocities. So that means high accelerations, but that's destabilizing. So the only way to really fight against that is to try to increase the density gradients and to increase the ablation velocity with an increased temperature in the hallrom to mitigate the growth rate of the instability. And those ideas led to the high foot implosion. So the high foot implosion is one that we implemented in the 2013 to 2015 times scale. And like the name says, we just raised the power in that part of the laser pulse called the foot to increase the stabilizing influence of a blade of Rayleigh Taylor of ablation and of the density gradients in that blade of Rayleigh Taylor formula. That actually seemed to work. We improved the performance by about an order of magnitude. And we were able to increase the pressure by a factor of three in those implosions. And it allowed us to explore other behavior and learn what else might be frustrating the performance of our ICF implosions. So while these high foot implosions actually increased the fusion yield by a factor of 10, they also had repeatable behavior by controlling instability better. But we saw that symmetry control was still an issue. So here's a set of experiments. And there are two sets of experiments shown here. The first three are lower power, lower energy experiments where we start with a baseline shot and we do a couple of repeats. And then we have a set of experiments where we are using higher laser power and higher energy to get higher implosion velocities. Again, we have a baseline shot and then we do a couple of repeats. And here's an example of the repeatability of these old high foot experiments. In a qualitative sense, you can see the symmetry. This is, by the way, this is X-ray emission. So you're seeing the hotspot emission from peak compression on this first row. The second row, you're seeing neutron emission. The 14 MEV neutrons are the red and the downscattered lower energy neutrons are the blue. So that gives you some sense of the hotspot and the cold fuel around the hotspot. Fusion yield is shown on the lower row and the DT ion temperature is also shown on the lower row to give you an idea of the repeatability. So we get a fair degree of repeatability in the shape of the implosions, a good degree of repeatability in the yield, and the temperature except for one shot where the laser misfire and it underperformed. So by having a fairly repeatable implosion because we mitigated instability to some degree, we could then do studies on some other processes. One of the things we wanted to study is the late-time part of that laser pulse, which again we call the coast time period. So what we did in a series of experiments is we tested higher power, shorter duration laser pulses versus lower power, longer duration laser pulses conserving total laser energy. And again, that's compared to a baseline shot in blue. You can actually have the same implosion velocity when you change the peak power in this way. So you can take an implosion velocity out of the equation and just see what the late-time behavior of the laser pulse does to the implosion. As you can see, the red curve, which is the one where we kept the laser power lower, but extended the duration of laser power leads to an extended duration of radiation temperature at late-time in the hall room. That ends up being key because of the ablation pressure formula I showed earlier. So in the end when we did a whole series of experiments, this is a plot of pressure versus coast time, we found that by reducing the coast time we got a very rapid increase in stagnation pressure in our implosion, which helps us with our loss in criteria metric. And in fact, we later understood the physics behind this by understanding that this very late period of increased laser energy and radiation temperature forces the implosion to acquire its peak velocity at a smaller radius. We call that the radius of peak velocity. And that ends up being key to increasing the stagnation pressure and other quantities that are important for fusion to occur. All right, so with that knowledge, we moved forward to a new set of experiments that were meant to address the asymmetry problems we saw in those earlier plastic implosion designs. We switched to ablator material from plastic to high-density carbon. The reason why that ends up being key is high-density carbon has a density of about three times that of plastic. If you have a density that's three times higher, you can make the thickness of the ablator three times thinner, and that allows you to use shorter laser pulses. And we were already suspecting that the length of the laser pulse was a serious issue for our symmetry control. Those implosions performed quite well. We were able to push up the temperature and yield by about a factor of two in yield as compared to the high-foot implosions. We were able to push the ablation pressure up almost by a factor of the hot spot pressure up by about a factor of two compared to the high-foot implosions. But we still weren't igniting. We were only about a factor of two higher in yield than the high-foot, and it still orders the magnitude below what needed to ignite. So it turns out, even though we had made a lot of improvements on the symmetry control, it was still frustrating. Symmetry control actually is still challenging even today, and this is an illustration of some asymmetries that occur in our implosions. The data from fluence compensated downscattered images on the top, synthetic images from simulations on the bottom, and simulations of what the density looked like on the bottom. So this is just kind of a zoology of different asymmetries that can occur. Now, why was symmetry so important to getting these fusion implosions to perform correctly? The reason actually comes down to energy. In ICF, it's actually essential to maximize the conversion of implosion kinetic energy into hot spot energy. We saw this in our database, so on the left plot, you see total fusion yield plotted versus hot spot internal energy. And you can see this is a log scale here, but we get an increase in fusion yield as we increase the energy deposited into the hot spot, and the scaling goes as energy to the 3.3 power. There's scatter in the data because of things we can't control as well. Again, the symmetry and the target quality. But essentially, we knew that hot spot internal energy was important to getting higher fusion yields. But when you look at simulations and a lot of work that's gone back over many decades, symmetry always seemed to degrade the fusion yield if you had any. So if you had increased asymmetry, the yields would go down. If you had decreased asymmetry, the yield would go up, and simulations had shown that for a long time. You can put those two things together, and you see that the net energy that goes to the hot spot is a combination of the kinetic energy of the implosion, just from freshman physics, the one-half MV squared that the implosion acquires, but it's degraded by a factor related to a measure of asymmetry, which we tend to term normalized RKE, and RKE will stand for residual kinetic energy. So this principle observation led to something we call the hybrid strategy. We realized, well, we're going to be stuck with controlling asymmetry imperfectly and other parts of our implosion imperfectly. Let's just try to maximize the energy that we can get to the hot spot. We were already pushing our implosions to very high implosion velocity, so we didn't think we could squeeze much out of that, much more out of velocity because we might start having stability control problems again. So that caused us to focus on the mass of the shell. Let's just make the shell of the implosion bigger. You can make it bigger by making the radius bigger, you can make it bigger by making it thicker, and so on. And so that led to a strategy we called the high yield, big radius implosion design, and we moved on from there. So before we move on, let's talk about the physics of asymmetry because that's kind of interesting. You can actually understand why symmetry control is important using a basic kind of freshman physics model. And the way you can understand why symmetry is important is to look at the simplification of an implosion. Where instead of having a complicated spherical implosion, you just look at two pistons surrounding a hot plasma. You write down Newton's laws for each of these pistons, so f equals ma, they have some initial velocity. They come together and hit each other, squeezing up the gas inside. If the pistons are symmetric, there's no net center of mass motion because they come in with the same momentum. They squeeze up the gas, kinetic energy is converted into internal energy and they stop. But if they're asymmetric, there's a net center of mass motion. And because of conservation momentum, the kinetic energy associated with that center of mass motion never goes away. And so if you go through the freshman physics to evaluate this system, you find that the pressure in the implosion is again related to the mass of the pistons, the velocity squared, so the kinetic energy. The minimum volume that that gas is squeezed up to, but there's this wasted kinetic energy that is related to the center of mass velocity squared over the implosion velocity squared, and that's the residual kinetic energy. Now why this model is interesting, it looks very oversimplified, but it matches our simulations and data very well. So it allows one to kind of understand what are the key parameters and why, understand why asymmetry is so important. That's a 1D sort of picture of asymmetry. You can extend that to three dimensions. The physics actually ends up being the same. Even though there's no net center of mass motion, you can imagine a configuration of multiple pistons. In this case, the weighted harmonic mean of the shell aerial density ends up being the key parameter, and that also matches our simulations very well. So you can see as you get implosions that look more and more asymmetric, that kinetic energy or that residual kinetic energy just continues to go up, and that's basically wasted energy, and energy is precious. Okay, we actually had some idea of some of the levers that are important for controlling asymmetry in an indirect drive ICF experiment. Beam pointing, which is illustrated here on the left, where you have a hall rom with our capsule inside and you have laser beams coming in and hitting the wall in different positions, allowed us to have some control over the x-ray flux that the capsule sees. And let's blow that up. And that was the primary way that ICF designers and workers have controlled symmetry and implosions for many decades. What we learned during this period, though, during 2015 to 2018 is that that beam pointing control of asymmetry is not sufficient in itself. In particular, the outer beams from the NIF hitting the outer section of the hall rom wall would create a plume of plasma. That plume of plasma would move into the hall rom and intersect the inner beams that we're trying to deposit here closer to the implosion. And thus, the energy would not end up going where we want it. So there were two key realizations during this 2015 to 2018 period that allowed us to understand what has been frustrating our symmetry control and then gave us a better ability to control it. One is this asymmetry model due to Callahan and Joe Ralph, where it was recognized what the key parameters controlling low-mode asymmetry were. I'm going to mode two in particular, which is the most primitive asymmetry of either being a pancake or a sausage implosion. We found out that that symmetry control was influenced by the energy that goes into the picket incident on the inside of the wall here, the spot size on the inside of the wall, the gas density inside the hall rom, the duration of the laser pulse, the radius of the hall rom, the radius of the capsule, and the radius of the hall rom again. And if you plot all our data against that parameter, all our data basically collapsed on one curve. So this gives you a way to design roughly an implosion where you have some hope of controlling asymmetry. The other key tool developed during this period was the recognition or the proof that cross beam energy transfer that had been used previously in high-gas-filled hall rom also worked in low-gas-filled hall rom. And this basically is a process where you're leveraging a laser plasma instability to shift energy from one set of beams to the others in this section where the beams cross. And that's very effective at taking energy and putting it into the beam that's losing energy due to this plasma and forcing more energy in that direction. And the initial set of experiments were done by Annie Kreitscher on that, and it was followed up by some additional experiments by Louisa Pickworth and her and the team that I actually called the hybrid C team. Okay, so now with that understanding of better symmetry control in hand, we thought we could now really tackle this hybrid strategy. And so we did scale up the capsule radius and it didn't work. So it didn't work nearly as well as we thought. So those are the red and purple points. We attempted to use this strategy of getting more energy into the hot spot by increasing the mass or the size of the capsule. And we thought we can control the symmetry, but it struggled initially. And so the yields really didn't go up that much. The pressures were actually even lower than the earlier set of experiments. And so we were struggling. One of the things that surprised us was that the target's quality, the capsules in particular got worse when we scaled the capsules up. And that was kind of the opposite of our expectation because you would think defects basically stayed the same size when you make the system bigger in a relative sense. The defects are less impactful, but that didn't happen in this case. So there are a number of problems that we identified in the capsule. So again, here's our capsule, which is an HDC high density carbon ablator. What we found that there were a bunch of voids and particles in the through thickness of the capsule. There were problems with pits on the surface of the capsule. There were sometimes big holes in the through thickness of the capsule. And all those are seeds for Rayleigh Taylor instability because again, any little perturbation is amplified by acceleration. And here's a little movie of what happens. You can actually see a hydro feature that was seeded by these features entering our hotspot. And that's a problem because there's usually more than one. And I'll explain in a moment why that's damaging. So the reason why having those defects in the capsules is damaging is because they cause what we call mixing, which is the injection of high Z material into our low Z plasma. And that costs us energy. So if you look at a cartoon of our implosion here on the right, we have a high density shell that we're trying to surround the hotspot with. And if it's just DT, you have a competition between alpha heating, getting the plasma hotter, thermal conduction, cooling it off and brems losses, which are X-ray losses, cooling it off. Now, the issue with the brems is that the X-ray losses go up as a rapid power of atomic number. In particular, it goes up as the square of the average atomic number. So where you might have had a DT plasma, which has a Z bar of one in the nominal case, as soon as you inject higher Z material such as carbon or higher Z materials like tungsten into the hotspot, it causes it to radiate like mad. And that steals energy from the system. And as I tried to emphasize earlier in the talk, energy is precious because we get hardly anything into the fusion fuel by the nature of an ICF system. So we really can't afford to lose any energy. So if you now look at a model of what happens to the fusion yield amplification as a function of kinetic energy. If you had a case where you didn't have any mix and you got to a certain level of kinetic energy and you get this explosive increase in yield, this kink upwards is basically ignition. But now you take that system and you inject mix into it because you don't have perfect control of the targets or you don't have perfect control of the physics. That moves this ignition curve to the right in energy. And that's an energy cost that you may or may not be able to pay. In particular, the scaling here shows that the kinetic energy shifting to the right goes as the Z bar to the point six power, which is kind of interesting. So that tells you the cost of mix and it's a cost we usually can't afford. But in this particular case of the experiment of last month, after years of effort by the laser engineering team, we actually got more energy from the NIF. And it's something that was planned back in like 2015 or 2016. It goes back quite a ways. But we were able to only implement it and try it out this year. So they were able to move the peak laser energy of the national ignition facility to from 1.9 megajoules to 2.05 megajoules. That's enabling the most recent success. So again, so here's our neutron yield versus temperature neutron yield versus hotspot pressure. And the more recent, the most recent experiment from last month with all the press is this one that's kind of shown in the bullseye. But there was some other experiments shown here as well. So we're finally able to get the yields well above megajoules. Again, the most recent one is 3 megajoules. And it took this 10 years of effort by addressing problems and steps, building our understanding, using kind of basic principle ideas, a lot of computer simulation. And it's coupled with a lot of work on the design optimization and trying to finesse things in a regime where mother nature is working against you. And that's pretty exciting. Okay, so the concept that has resulted in kind of the most recent, the series of breakthrough results over the last couple of years is what we call the hybrid E. There was a hybrid B, C, D, and E. So it's the hybrid E, which has been the one that's kind of gotten kind of the pinnacle of all of this understanding put into it. So we've obtained burning plasmas and ignition in the laboratory using this design. In a rust sense, it looks the same as those cartoons of ICF designs from many years ago. It's still a haul run with the capsule in it, but it's really the details of the dimensions of all of this and the laser pulse that really have made it work. So the key elements have been 20% larger capsule radius than the previous diamond to blader designs. A reduced LEH size. Again, that LEH is the laser entrance hole for better x-ray confinement with the symmetry control physics understanding that I was talking about a few slides back using cross beam energy transfer and trying to use the pointing as best as we could. And that understanding of the gold bubble frustrating our symmetry control. Lowering the peak laser power, but extending the duration of peak power using that coast time physics was key. And all of this resulted in increasing the hotspot energy and pressure and pushing us well past what's required by the loss and criterion for ignition. So the most recent result, again, the one from last month in particular implemented an 8% thicker blader again increasing the mass of the shell with the corresponding 8% more laser energy that was mentioned a moment ago and improving the symmetry or controlling the symmetry. And that took the previous record of 1.37 megajoules to 3.15 megajoules and here's a three dimensional neutron image from the previous record shot from it's August 8th 2021 shown here. And the same sort of neutron image from a month ago is shown here and you can kind of see you don't have to be an expert that's better symmetry. And, you know, it had more energy in the hotspot because of the thicker shell. Alright, so the corresponding laser pulse that goes with this thicker shell is shown here on the left. So I'm overlaying the two laser pulses again from the previous record shot 210808 shown in red and the most recent shot 221204 in blue. And again, it's the basically the same duration, same peak power, but just an extending the duration of peak power a bit. That's really the only change, although there's some details here that experts can see but it's that's the significant change here. So if you look in loss and parameter space of pressure times confinement time versus temperature. That most recent shot is shown over here with a p tau of about 50 atmosphere seconds and 12. Kev temperature and the previous record is here so we basically push the temperature up through this strategy. And you can see this whole this whole talk has been a story of making success not by revolutionary improvements but by incremental evolutionary improvements where we we take a step. We falter, we learn from that and we take another step forward and we were able to basically climb our way up towards the ignition boundary and then pass through it. So almost done. One outstanding remain one outstanding problem remains or one significant outstanding problem remains that is going to need to be addressed in order to get the significantly more fusion yield. And that is this issue that's been plaguing us for a decade of compression. So, if we look at our database of experiments, we've always had a problem of our computer models predicting more compression than we actually observe in reality. And so this plot here illustrates this problem. This is a this is a complicated axis, but this is basically measured compression here on the y axis and the expected compressibility based on the entropy or the 80 bat on the x axis. Theory line shown in black and you can see most of our data falls below that theory line so we usually don't get the amount of compression that we expect and that that has been frustrating. The reason why it's so important is the burn efficiency of an ICF plasma is related to the fuel row R and the fuel row R is directly related to your ability to compress the plasma. So if you are limited in your ability to compress, you will always be limited in terms of the burn efficiency. So this is a problem that we need to address. Nevertheless, we've gotten ignition and again greater than unity bypassing this problem for now. So I'd say this is like the end of the beginning. There's more work to do. This is actually just one of the many things we want to explore going forward. But it's remarkable now that we can talk about burning plasmas ignition and scientific break even in the past tense. You know, it's kind of funny because the the the joke about fusion always being 20 years away. And I'll just read through this again. There were no mystery of physics optical standing in the way of ignition or gain our theoretical predictions of the physics parameters regime where ignition was expected to occur is consistent with our results. Additional laser energy was was really key. And it's very beneficial and the implosion physics ended up being more sensitive to the engineering control laser and targets than originally thought. So that that's why we had so much effort to make things work. And so far that the you know very high compression target designs have not worked as expected. So that's a primary focus of research in the future. All the breakthroughs we have made have basically used low gain designs. So with that, I'll stop and I hope my time is good. Yeah. Okay. So. Mateo, are you. Omar, thank you. I mean, it was a great excellent explanation of how energy game was achieved. And it is also exciting to be here with the experts who made this breakthrough happen. So we have about 11, 10 minutes left for q amp a. Here comes some of the questions. I'll actually just start with a very basic one first before I go to the question box. Should I leave the slides up by the way? Yeah, I think that'd be a good idea. So maybe in case you want to refer to it, but have. Have you tried has your team tried or are there plans to try to reproduce the ignition shot? Annie or Alex, do you want to take that one? Yeah, I can, I can speak to that. So, of course, we will try to reproduce the level of performance. There are always challenges and things that can go wrong on any particular shot. So we can't quite say when that will be. We're still limited in how many of the 2.05 megajoule laser shots that we can do. For now, about once a quarter. So the next one of those will be coming up in February. Okay, so coming up next month then as we do have, or as speakers and panelists come on to answer, you can also turn on your camera so you can see you. One question we have here about the indirect drive. Is it using a gold Haloram and is it suitable for potential reactor power plant design? So what happens with the cold Haloram doing during a shot? I think this could also be a design question. So. Furthermore, do you need to go back to the direct drive working on the cemetery again? Yeah, sorry, I was muted. It is a goal. And if you could turn on your camera. Great. Thank you. Can you see me here? Yeah. Okay, great. Yeah, I can see it. It is a gold Haloram that we're using here in these experiments. There are concepts for reactor designs that both use direct drive and also some that use indirect drive. So that both of those are being looked at and the question as to what happens to the gold during the shot. It's totally obliterated. Thanks, Annie. And going to lasers. How much is the uncertainty in the timing between all lasers? Can you get the lasers sufficiently synchronized or is there room to improve? I'll take that one. There is definitely room to improve. And in terms of accuracy being delivered to the target from a timing standpoint. We are in the realm of tens of picoseconds. So it's extremely precise timing. But the team is investigating ways to make it even more precise. Thank you. And for Omar, we have a question have non thermal distribution of the fusing nuclei had a significant role. Yeah, that's okay. I kind of I think I know who the question is. They don't they don't have a significant role. They have a definitely there's a signature of non thermal particles and actually Dave Schlossberg can actually follow up on this question to I believe in terms of the fusion yield say for the recent shot. We are quoting a 3.15 megajoule fusion yield that includes subtracting out the contribution from the part of the plasma that we think is non thermal, which I believe was about 50 kilojoules of the 3.15 megajoules was was non thermal. So there is definitely a signature of non thermal, but we think it's a minor contribution. Dave, do you want to follow up on that and correct me if I misstated something? Omar, you got it. You got it right. As we've timed higher and higher in performance, we've some of the diagnostics have started to see hints of a small population of non thermal ions in the distribution. But as more more mentioned at the at the moment, I think it's the small effect, but it's an exciting effect to investigate and see. So we're going to need to do that. Okay. And I'd also like to follow up on that on another question about laser systems. This might go to you, John Michael. How do you increase or how do you get more energy from a laser? You mentioned there was an increase from 1.9 to 2.05 megajoules. It's an excellent question. It's we have been constantly increasing since the commissioning of NIF. The delivered energy and power. So it's not something we have done overnight. You know, Omar mentioned the fact that we have been working on that over many years, which is absolutely true. So to get to your point, the way that you get to increase the laser energy. Is to play on multiple parameters. First of all, we're minimizing losses in the laser as well as in the infrared section of the laser. It goes with developing new coatings, new technology to make that possible. Then second, we also investing a lot in material resistance, material fabrication processes to make those optics. Fardened and more sustainable to take shots at higher fluencies. Without incurring damage. And then last, but not least, what we do is we're working on all aspects of the beam or shaping or beam in time as Omar indicated. To achieve the epics, but we're shaping them in frequency as well and in space. So we are flattening the beam as much as possible. So for a given laser aperture, you can truly optimize the delivered energy without having hot spots that would compromise the integrity of the optics. And we're about to increase by yet an additional 8%. Going from 1, you know, 1.9 to 205 megajoule was 8%. As Omar mentioned, but we're about to increase yet by another 8% over the summers, which will provide more margin for ignition experiment. And could this increase, could this have happened already 10 years ago? Or why now? Well, it's a combination, as I said, of, you know, technology improvements and interest, you know, and momentum that we have behind us. But we have been working on that for years. I mean, we have a couple of paper 2019 nuclear fusion explain how we did it. Excellent. And this is a general overall question we've also had in our chat, but what needs to be done to scale up laser based fusion and perhaps for the future fusion energy being a commercial product? Well, you know, our facility is a research facility. We're not a fusion reactor. I think there's a number of things that have to be done on the laser and the targets. The targets in particular wondering if maybe a boss, you might be able to expound on the target needs to make IFE practical. Yeah, just very simple terms. We need to shoot at about a million times a day. And at this point, we shoot about one time a day. And that it's actually that one time is a culmination of about seven days of trying to get the DT fuel into an ice form. So as Alex and Omar and company, we say, we wish we could do it more often, but that's what the facility can do at this point. So there has to be major engineering advances to be able to make these targets at that rate, which basically comes 10 times a second. And be able to shoot the laser that many times and deal with, you know, all the hot stuff in the chamber and so on. So there's a lot of studies on that, by the way, and like Jean-Michel said, it's lots of papers on that from the past and that work is beginning to be ramped up again. But the targets is one challenge, the laser shooting 10 times a second. And then in the physics side, I think we need a gain of about 100 or so to make this thing work. So we got another two orders of magnitude to go. And also make the laser more efficient because as Omar pointed out such that there is no confusion, we drew about 300 to 400 megajoule out of the electrical grid. And if was not designed to be energy efficient, it was designed to demonstrate scientific breakeven, but we know now how to make laser way more efficient at the tune of 10 to 20% while plug efficiency. So we would need this type of lasers firing at 10 hertz or so to make it viable. Thank you. We do have a few minutes left. I really want to thank you, the panelists and Owen for the presentation and hand it over back to Matteo. Since we have one minute left, I'll take this time. I have two questions. I wanted to ask, I listened to the live streaming of the announcement and could you maybe it was any, I think, who answered this question, could you tell us about how, like, what was the process of, I know you use AI machine learning for some of the design for fine tuning, could you just tell us a little bit more about what went into this? Yeah, so I can sort of walk you through. In this case, we didn't actually use AI to design the experiments. But after we had already determined the design parameters, we used AI to provide a probability distribution for achieving gain. So the process that goes into this, like many other things has been a culmination of years of building up understanding and developing models and cross checking those models against experimental data. Both analytical models as well as complex rad hydro radiation hydrogynamic simulation so over many, many years we've been calibrating our understanding to our experimental data. And then using those calibrated models to make design improvements. So, specifically going into this regime, I used a combination of analytical scaling, some analytical models, radiation, hydrodynamic simulations. And that's how I determined, for example, how much thicker to make the target with the, with the given laser energy upgrade as well as figuring out how to fix the asymmetry. And so that's quite a complicated process and we do rely on our scaleings and also on our simulation tools to do that because you have basically laser beams propagating through plasma that's expanding at a different rate into the hall room. And so trying to get a symmetric radiation drive at the percent level is quite challenging. So we do need to use a combination of many tools. So I set the design parameters and then I worked with our team to they took that information and came up with an independent assessment for they confirmed. What I had predicted is that we would achieve gain more than 1 and they also confirmed that and put a probability distribution to it. Thanks and very interesting. And I just want to ask question where 1 minute later, maybe Omar, this can be quick. I think it's for you. It sounded like the physics of the cost time was a real game changer. It was significant. The funny thing about the coast time was it's it's not something that's common knowledge in ICF. So it was it took some understanding to realize how important it was. So it was something that people thought was minor that actually ended up being a major factor. But one of the points I tried to make in the talk, it's ICF is no one thing. It's all of these elements have to come together to get ignition in an ICF target to work. So, and you have to be lucky that they all work correctly at the same time. So, so, so, yes, coast time is important. It was neglected largely before we understood it better. But it's 1 of many things that has to be rolled in together to make this work. Okay. Thank you all very much. This was fantastic to have you all here. Thank you for this 1 hour. Congratulations again for this extraordinary results. Thank you, Joanne. So we'll try to make this recording available after of course doing the checks with with the NIF and the Lawrence Livermore National Lab. And if we do make it available, we will announce it to all the registered participants. And we let you know for the about the next episodes in this fusion breakthrough series. So thank you for staying with us and thank you for the for attending to all the panelists. Thank you for your invitation to speak and thanks to my colleagues for taking the time this morning to be online. Okay, bye bye. Bye bye. Bye bye.