 It's my great pleasure to introduce my colleague, Professor Francesco Volpe. Francesco used to be a professor of plasma physics at Columbia University in the US, when he realized that he would have a lot more fun studying a private company and making fusion power reality. He's now running one of the few, I believe, three fusion companies that are private fusion companies in all of Europe. And he's specializing on stellarators with liquid walls and high-temperature simple conductors. And he will talk to us today about the technology behind this company and hopefully also a little bit about this company. I'm excited about this because I believe that if we can solve the problem of having cheap energy with little impact on the environment that will make many of our other problems go away. And I believe that fusion power eventually is the solution or at least part of the solution for this. So let's see what Francesco can tell us about it. Thank you very much, Michael. Thank you very much, Raul, for the introduction. It's a big honor to be here. We were joking over lunch on the fact that with my high school professor, he wanted me to become a mathematician and I never dared to tell him I wanted to become a physicist and eventually I became an experimental physicist and I do fusion where I work at the boundary between physics and engineering and most recently I became a sort of businessman. So my descent to lower and lower grades was complete. But jokes apart, there are a lot of interesting theoretical, mathematical and physical problems that I would like to share with you today. Let me go back to the first slide. I'd like to share with you today that hopefully we'll be inspiring, hopefully we'll prompt new collaborations or we'll inspire some of the students and postdocs in this audience and in the remote audience to join this endeavor. As Raul said, the company is called Renaissance Fusion, it's located in Grenoble, France. We are a very international group at the moment we are 18, you see 14 of us here and four more just joined recently. In fact, we are in a moment of significant expansion where we are growing to 60 people by the end of the year. And again, I cannot stress enough this point about the fact I'm looking for talents, we are looking for talents. So in this presentation, I will give a brief reminder on fusion and plasma confinement, a motivation on why fusion and why specifically our approach to fusion, our three pillars as I call them on high temperature superconductors, you can metals and accelerators. Let me show you, let me use this mouse. Yes, here we go. Recent results and next steps. So first a quick reminder of fusion. The reactions used in the sun are shown, sorry about this, just for analyzing with the mouse, here we go. The reactions used by the sun are shown here at the bottom left relying on the weak nuclear force and you see plotted, apologies, and you see plotted on the right the fusion reactivity. If you multiply this quantity by the densities of the two reactants, you obtain the number of reactions per unit time and unit surface. You see this quantity plotted against the temperature of the ions for various reactions shown here. The reaction happening in the sun is very low efficiency, very few reactions take place per unit time and unit volume. Yet the sun is producing a huge amount of energy per unit time. On earth, we need to do something more compact. So we rely on other reactions, one of which the most often used in today's laboratories is the fusion of deuterium and deuterium with a branching ratio of 50% in these two families of products. You see the reactivity of this reaction in orange here, much more effective, several orders of magnitude more effective than proton-proton. But even better than that and what we want to use in reactors, in actual power plants and what has been used in some experiments in the past and will be used in few years by ETER, the large dogamac under construction in Provence, an international collaboration you might have heard of. That reaction is deuterium tritium, giving an alpha particle and a neutron. You see here the reactivity shown in blue much higher than the others. The tritium for this reaction is sourced by the same neutrons that are produced in the reaction reacting with a blanket of lithium and the producing indeed tritium which is re-injected in the plasma infused. This is the good part, the neutrons carry the heat, most of the heat, 80% of the heat and we want to capture them to extract this heat indeed. But at the same time, they create issues in terms of activating the solid part of the reactors which is why other public and private endeavors try to exploit other reactions shown here in green, in purple, in red, less efficient but they have the attractive feature of producing fewer or no neutrons at all. In reality, at the National Institution we have a solution by which we can deploy the deuterium tritium reaction without the disadvantages of the fast neutrons as I will show later. Now, if you noticed the numbers in these slides you saw MEV with a capital M all the time. This is the energy produced in every single reaction as opposed to MEV with a small M with a lower case M or EV produced in chemical reactions. So this is the reason that makes fusion so energy dense and this is why we need the such tiny amounts of deuterium tritium fuel to produce the equivalent of many, many barrels of oils or many, many car loads of coal. It is also by the way very efficient in respects to power per unit surface unlike for example solar. So we want many reactants ACC on the left. So of course we want the density of reactants. You want these reactions to take place and before the strong nuclear force prevails you have to defeat Coulomb repulsion. So you want this nuclei to collide at high energy which is why you want high temperature. And last but not the least you want to make an efficient use of this heat both externally injected as well as internally produced by fusion reactions because at the end of the day what matters is the temperature but what you're spending for is the heat and just like in your apartment you care about the temperature but you want an efficient heating system you want to spend little heat and therefore you have a small electricity bill or gas bill in a fusion reactor it's exactly the same. You want to well confine your energy so you want good thermal insulation or good energy confinement which is why I'm sketching the system as a closed or nearly closed system with very modest leaks of energy. This can be summarized by the requirement that we want a high product of these three quantities density, temperature and the metric of confinement called energy confinement time the scale, the time scale over which energy leaks away. For the terium tritium you need to exceed this threshold for net energy to be produced. The threshold for other reactions is even more challenging. A typical power plant will consist of a plasma most likely the terium tritium producing these neutrons interacting with a blanket of lithium that produces the tritium which is re-injected in the plasma. This blanket of lithium becomes hot under neutron bombardment and this heat can be extracted by a heat exchanger. This is basically a furnace so it's the core of any thermal power plant be it coal or fission which is in fact very interesting it offers opportunities of retrofitting coal power plants fission power plants something we are looking very actively in. And in the value chain of fusion even though the ultimate goal is to produce electricity, net electricity for which when all the losses of the power plant from the previous slide are factored in you need actually to produce much more heat than you inject again to compensate for losses. For net electricity you need a fission gain of 20 20 times heat out more heat out than heat in but at capital Q fission gain of five so non-electricity but net heat you can already do very impactful things in terms of for example producing the nitrogen or decarbonizing some industrial processes which are very heat intensive so production of paper, steel glass, cement, etc. In one step back with very modest amounts of fission gain just a small amount of fission nutrients fast nutrients you can produce scarce medical isotopes or treat fission waste and one further step back this is magnetic confinement fusion the magnets that we are developing for fusion are large, very accurate and strong they generate strong magnetic fields these three requirements are encountered also in other fields from energy storage to medical imaging so there is an opportunity to make something good for society in the case of energy storage and wind energy and mobility for the environment before even making any net fission electricity. So as you said before we need to exceed this threshold in triple product if you remember the plots on reactivity there is an optimal temperature at which to operate of the order of 10 kilo electron volts so in reality our knobs are the density and the energy confinement time there are three main approaches to fusion in one initial confinement you push this knob very hard you work at very high densities higher than solid you compress basically a capsule of fuel by lasers or other means until exceeding the triple product threshold on the opposite end of the spectrum is magnetic confinement in which we act on the energy confinement time we maximize that by confining the plasma with strong magnetic fields and or by making the device large enough this increasing size relative to the normal radius maximizes the time spent for particles and energy to travel from the core to the edge of the plasma and eventually be lost what is interesting now as you would see later is that now we can make the normal radius much smaller thanks to strong magnetic fields and it has just the same effect in a wind tunnel spirit as making the device larger but it's more yes the energy confinement time is the time scale over which the energy is lost from the system from the moment you turn off the heating systems so if we suddenly turn the air conditioning in this room it will take some time for the temperature to stabilize to a new value a certain time scale the same in the plasma it is dependent on the density because among other things it depends on collisionality and on turbulence and both processes depend on density so indeed yes it is a function of density and other things the magnetic field the helical twist of the magnetic field yes this is a simplified picture but very good point thank you so the magnetic confinement between magnetic confinement and inertia confinement you have the hybrid of magneto inertial in which as the name suggests you magnetize a plasma and compress it as you would do in inertia confinement trying to get the best of the two worlds or the worst we focus on magnetic confinement because it's the process that has approached more closely q equal one capital q equal one net heat and there are two champions in this field of magnetic confinement called the tokamak and the stellarator they are both toroidal devices where magnets pictured in blue coils pictured in blue confine a plasma shown in purple a toroidal plasma the field shown in black is helically twisted at the first proposals of toroidal confinement of a plasma by magnetic fields Fermi immediately realized that due to the one over R R being the major radius non-uniformity of the toroidal magnetic field the larmor radii will also be non-uniform and this will result in a vertical drift of charged particles the positive ones in one direction the negative ones in the other your particles are lost due to this radial non-uniformity of the toroidal magnetic field the solution to that is helically twisting the magnetic field this property is called rotation of transform and the difference between the tokamak and the stellarator is how you impose this twist this rotation of transform the tokamak there is a solenoid acting as the primary of a transformer and your plasma is the secondary of the transformer so a current is induced in the plasma shown here in the toroidal direction that generates a field in the poloidal direction and then the twist on the other side instead we have the stellarator where the same is obtained by deforming the plasma by helically twisting the coils and thus the plasma and ultimately the magnetic field in the tokamak the disadvantage is that even though in spite of the apparent simplicity of the coils this current is inevitably pulsed you have a linear ramp of current in the primary to get a constant current in the secondary and you cannot make an infinite linear ramp so inevitably this device is pulsed you can make long pulses of half an hour or so hopefully longer but it's pulsed secondly it is unstable we are talking on a very large current resulting in very large current driven magnetodynamic instabilities the most dangerous of which is the disruption a sudden loss of confinement in few milliseconds the energy that you have accumulated in seconds is suddenly lost in few milliseconds with consequences for the device possible regulatory issues instead here the way a stellarator works is everything is DC you turn on your coils you evacuate your vacuum chamber you inject a gas you ionize it you inject enough heat to warm it up to fusion temperatures and voila you have fusion energy everything is steady state so everything is simple to operate in fact it's also very energy efficient because you don't need so-called current drive methods to non-inductively sustain the current as instead you do in the tokamak the disadvantage of the stellarator and the one we address at renaissance fusion is these coils on the right are more complex than the coils on the left and in modern stellarators they are even more complex more difficult to build so we came up with an idea that simplifies that but before doing so a quick comment on the energy confinement that the colleague was asking about a plot a quick plot just to show a couple of things first experiments agree with theoretical predictions so theoretical expectations and measurements of these energy confinement time are in good agreement to the data points lie on the diagonal and the data points from tokamaks in gray are clustered together with the color points for stellarators tokamaks and stellarators of the same size and magnetic field achieve comparable confinement without the disadvantages like these disruptions I was referring to now the motivation for what we're doing renaissance fusion and in general infusion is that even though I'm a big supporter of renewables renewables are not the only solution to the problems we're facing in our society renewables I used to live in Manhattan and that this is the amount of power consumed per unit surface continuously the city that never sleeps this is the amount per unit surface produced by a farm of wind turbines intermittently so there is obviously a discrepancy here and about energy storage with your cell phone is powered by a lithium ion battery your laptop your car but seasonal storage of energy for a city of several millions in habitants unless we find a completely different solution that problem is not solved yet this is a much energy needs to be stored to for 24 hours for one million people city comparable to a nuclear bomb amount of energy and by comparison this is the output of a gigafactory of ion batteries per year so again another discrepancy fusion the energy of the stars is the solution to this and to several other problems apologies something the fonts from one computer to another get got messed up but yes there are several advantages all shown here from geopolitical advantages to safety compared to fission for example abundancy free efficiency and efficiency one advantage that is not underlined well enough in my opinion is the ability to follow the demand of energy from society as opposed to for example just provide a base load so we hear this term continuously base load base load so that's when you turn on a power plant and you keep it on at a fixed level of output power with minimal changes or none at all fusion can certainly do that but even better it is a load following power source meaning that if there is a fruition in supply pardon me in demand it can respond on a time scale of the energy confinement time few seconds it can respond with the fluctuation in supply now we leave exciting definitely exciting time someone would argue expensive times because after a long progress from the 60s to the 90s in that triple product that I was mentioning before and correspondingly in the fraction of output heat to input heat shown here on the right after so much progress now we are approaching this elbow here in this dependence on one upon the other this is either the large tokamak international tokamak I mentioned before and exciting but the expensive part is that to get there you need large devices costing tens of billions the role of the startups like Renaissance Fusion is to recalibrate this horizontal axis try to build for billions what was previously built for tens of billions which by the way was out of the scope of public clubs the mission of it is to prove the scientific and technical feasibility of fusion it was not and it is not to address its economical feasibility or attractiveness this is where startups can play and should play a role combined with the acceleration to reach markets for fusion which are not just electricity but as I should be for heat or medical for example so there is a lot of excitement about around fusion at the moment and fusion startups in particular due to a combination of three factors an obvious need good enough technical readiness of fusion good enough for starting industrial research and development and breakthroughs in related fields as magnets computers and lasers and one message here is if you want to contribute to carbon neutrality by the 2050s we need to start commercializing by the 2030s there has been an explosion of startups you see the timeline here we are the last one in this timeline the most recent but we are catching up quickly thanks to good ideas and good people and you see on the right papers from specialized journals as well as mainstream journals excited about fusion there are about 35 actually nearly 40 they keep appearing one week after another I need to update this slide nearly 40 companies all over the world distributed geographically as shown by these columns and technologically as shown by these rows these rows are ordered by readiness with the tokamak being the most ready and the accelerator admittedly the second most ready for historical reasons related once again to the complexity of the coils but catching up if those are resolved they can group they can be grouped into two large families concepts that are working and companies are working to make them cheap or concepts that are cheap and people are trying to make them work we prefer the first approach working means developing prototypes in our case since we started very recently we are as I will show in more detail later we are focusing on two enabling technologies superconducting superconductors and liquid metals but these companies are actually building devices are not only making designs or they are actual machine builders maybe I will take one more second to expand the answer to say that one thing is building the experimental devices another is operating the experimental devices that's an area in which I believe in fact I strongly believe these companies should work very closely with public institutions and let's say we are working we are taking steps in those directions so building machines and then operate them jointly with public institutes so of the companies pursuing the deuterium tritium reaction which is once again the most efficient three are pursuing deuterium tokamaks or stellarators ourselves in Europe convoy fusion systems a spin-off of the MIT on the tokamak and the tokamak energy working on the spherical tokamak in the UK but we are indeed the only one working on the deuterium tritium stellarator with all the advantages listed before as we go down in that funnel I listed the advantages but there are also challenges that we need to face as we go down in that funnel so fusion neutrons from the deuterium fusion can induce do induce reactivity in the solid parts of the reactor we address that by coating the inside of the fusion reactor with a layer of liquid metal a solution containing lithium that absorbs most of the neutrons before they reach the solid parts of the reactor and make them radioactive tokamaks and stellarators tend to be large but we address that by strengthening the magnetic field at the end of the day again is the ratio of the larmor radius to size that matters and stellarators tend to be complex as you see in the cartoon on the left but I will present a technique that greatly simplifies that and at the same time the HTS manufacturing process so in a way we think we're making the stellarators safer smaller and simpler so let's start with simpler specifically simpler HTS manufacturing the traditional way of making high temperature superconductors is by means of tapes similar to audio cassette tapes for those of you who remember audio cassette tapes like myself traveling back and forth in various machines where they are coated with various layers the one in black is the superconducting one the others have electrical, mechanical, protection properties after this tape has been produced several tapes are stuck together and helically twisted into a cable which is wrapped on a non-planar table to produce a planar coil for the stellarator and finally the coils are carefully positioned around the fusion device the vacuum vessel so both this manufacturing and positioning process require a few resist below a millimeter I was going to say yeah thanks for that's that's good by the way that we are aligned we are where tuned okay I was going to say that these devices are few meters in major radius and yet you need these coils to be manufactured and positioned with the sub-millimeter precision so instead what we do is we skip altogether the concept of tape and cable in a way it's like skipping altogether the concept of transistor and making an integrated circuit so what we do here is directly deposit the HTS material on a cylinder in just two machines as opposed to seven as a matter of fact we deposit a stack of many layers of HTS intertwined with the other layers that I mentioned before this multiple stack is our equivalent more elegant I would argue of winding multiple times the tape or the cable to make the coils we just deposit multiple layers one on top of the other the result of this deposition process is a cylinder that superconducts in any direction which is not what we want in reality what we want is specific current patterns that generate the specific magnetic field geometry needed for stellarator confinement so what we do is with a laser we ablate this in specific locations effectively we make grooves in the cylinder these grooves bound the current when you energize these coils the current will follow a specific pattern as imposed by the boundary conditions by these insulating grooves one point I would like to stress there is superconductor nearly everywhere except then in the grooves not the other way around that these are not current filaments these are current stripes bounded by narrow grooves once these cylinders are ready and by the way please notice that the 3D sculptures from the previous slide are now replaced by two 1D movement a rotation of a cylinder and the translation of a laser and finally the cylinders are shipped to the power plant manufacturing site and assembled all together for motorists which is at the same time the prior start of SL and the poi set people are interested in video on details of this process but in the interest of time I will move on and comment on why this is also a simplification of the stellarator so historically the coil winding surface so the surface of which these coils have been designed has become more and more complicated as you see in this illustration the point is that we want to accurately generate the magnetic field that confines the plasma in yellow and use a conformal coil winding surface conformal in the sense of parallel to the plasma boundary and you stay close enough to the plasma this is typically considered to be good you don't want to be away from the plasma because I order multiples decay more rapidly and for the same amount of plasma shaping you need more coil shaping the effort of staying close to the plasma I think we have overdone it we have made our life unnecessarily complicated in the sense of we have chosen we came into the paradox of adopting a complex surface in blue and then on that surface simplify the current pattern so here comes the idea why not do the opposite let's adopt simple surfaces we know already that the patterns on those surfaces will be more complicated we expect that already but who cares it's just it just means that the programming of the laser would be a bit more complicated the trajectories will be more zigzaggy so to say to heal the same results and this is a proof that this is indeed true so the plot on the left and the plot on the right correspond to the two configurations on the left and on the right of the previous slide shown on the vertical axis is a metric of quite complexity indeed if you adopt a piecewise cylindrical envelope you obtain higher values in this metric a more complex pattern compared to adopting a conformal surface but the point here the important message here is that this level of magnetic field accuracy one part in 10 to the three is tends to build a satisfactory or more than satisfactory stellarator it's also another aspect which is flexibility with respect to the distance between the coiled surface and the plasma boundary giving enough space for that neutron shielding and treatment breeding or liquid metal and last but not the least a code has been recently published actually yeah it is actually already published on a journal but the blueprint is available on archive in collaboration with our mathematician friends in Paris and Strasbourg by which now we will optimize the coiled winding surface these cylinders were geometrically constructed but if you now start moving left or right or elongate or otherwise deform the cylinders we expect to achieve even higher accuracy of the order of one part in 10 to the four and in fact we think that the best results will be obtained if we have a prism of non-circular cross-section an extruded non-circular cylinder as shown in this picture on the right stellarators are made smaller by the please of course four or five slides ahead I will give you the answer thank you so stellarators are smaller thanks to the strong magnetic fields so this is how much the size reduces shrinks if you increase the magnetic field from five to ten tesla the major radius reduces as such I should clarify however that this is also made possible by the lower aspect ratio so making a plasma which is fatter basically the aspect ratio is defined as the ratio between the minor radius of the torus and the major radius of the torus that's also important a topic which I hope will be of interest to this audience is that we can so for a given target field in the volume and for a chosen surface coiled winding surface there is only one exact solution of current on that surface exactly giving that magnetic field but there are many approximate solutions of current on that surface and out of these many approximate solutions we can choose the one that strikes a good compromise between field accuracy quite complexity and and this is new force minimization this is something that was not studied in other stellarators before because they were operating at lower fields but if now thanks to a temperature superconductors we go to ten tesla let's say we double the field in principle we quadruple the forces if we don't change anything else in reality actually compared to present experiments we are quadrupling the field and therefore in principle going up in a factor of 16 in forces again if we don't change anything else but based on this fact that there are many approximate solutions you can properly define an optimization metric and a force minimization metric force a force metric the plus force metric either surface you know the surface average of the root minus square or you can introduce a non-linear penalty basically forces below some value are allowed between a soft and a hard limit are penalized and above a hard limit the penalty is infinite so the forces are completely forbidden if you do that you can achieve a good compromise shown in the third column between simplicity of the points you see that this current pattern is simpler than this one for example this is the current shown in a unwrapped surface of the torus so poloidal coordinate here poloidal coordinate here good field accuracy so small field field error shown here and lower forces these forces here are a factor of two lower than here so also helping in our roadmap to smaller accelerators regarding the question about the hts being damaged by by radiation by neutrons by heat and not only the hts by the way the vessel itself we protect both the hts and the vessel by means of liquid metal walls these are walls of lithium lithium hydride as shown in the next slide the mhd the simple physics that allows us to have it stick to the wall is that in this case the liquid metal is injected from the top and is forced to flow along the surface by means of jacros b forces pushing regularly outwards remember we are in the presence of a strong b so just applying a current in the liquid metal by means of electrodes in proper direction if forces pointing upwards at the end of the day it's parallel currents attract each other you have these parallel currents in the hts coils on the outside you have parallel currents in the liquid metal which is a deformable medium on the inside the currents on the inside tend to move towards the current on the outside and so the liquid metal tends to stick and either by and it does so better and better if you increase the force and or the the metronomic force and or the velocity and thus the central in this animation in this video one decrease this video you see a liquid metal flow speaking to a substrate that that caused me laughing like a maniac ready yes so when you turn off the jacros b the liquid metal flow detaches from the substrate um this is a one centimeter thick flow of gallium in gluten the convenient material for experiments every temperature because it is really temperature it is not what we want to use in the plasma in the plasma instead in contact with the plasma to be used before i described our lithium solution a quick word of stability uh it is inevitable to be concerned about this lithium um this liquid metal flow out you know at ease that can develop in this flow uh droplets that can develop in this flow regions of depletion so to contrast both the depletion and the bulging of the liquid metal to stabilize effectively predator can be animals and other instabilities we have a simple solution which is an array of electrodes immersed in the liquid metal by which simply by controlling the potential of these different electrodes we can design the current pattern uh in the metal such that in regions where the liquid metal is bulging we apply a slightly stronger jacros b in regions where it is depleting a slightly weaker jacros b this is an idea by the way inspired by what we do in plasmas when the plasma is bulging out you apply a correction with a magnetic coil that could non-axisimetic correction coils used at columbia for this purpose and so that's exactly the same thing with the different actuator electrostatic as opposed to magnetic now um our material is a solution of lithium lithium mine right so lithium is needed for the breeding of the tritium as you saw before um and it's a very good absorber of neutrons at low energy but a very lame actuator uh at high medium and low energies so if you follow the history of a neutron in a layer of lithium what happens is that a long distance is traveled for the neutron to gradually slow down and finally in few centimeters the lithium is absorbed um we are asking or they other people are asking to one material to do two different jobs but this material is only with the one and not at the other so there is a reason why they use water in fission reactors hydrogen so protons have comparable mass to neutrons they are much better at the metals by collisions compared to lithium in fact they have a perception 80 times higher than lithium so they are 80 times better than lithium at slowing down these neutrons hence the idea of instead of just lithium use a solution of lithium lithium migrate okay much more effective as slowing down the neutrons as I said um and eventually the lithium absorbs them and you know produces the tritium um now although hydrogen is a good actuator at medium and low energies you need an additional material to moderate the high energies this is uh which serves a dual purpose it also multiplies the neutrons so there is an end to end reaction and this is important because some of the neutrons will escape from the reactor but if we really want to make a close look uh where for each reaction in the plasma we produce a tritone which is injected in the plasma uh we need to compensate for these neutron losses so you need the neutron multiplier the lead does just that and this uh layer of materials protects back to your question sir uh protects the superconductor uh from damage by neutrons uh from nuclear heating etc this is important of course because the superconductor relies on a precise crystalline structure for its superconducting properties if these points of cooperate of sorry of copper oxide are permitted with bitumen barium uh are damaged the superconducting properties are lost or at least degraded um it is not super clear from the sketch but this layer of lead which seems to magically float on top of the lithium-bithium hydride is in fact a granular fluid is a tumbled of pebbles containing lead which at those temperatures is molten but the outside of the pebbles is solid and it is a ceramic which is electrically insulating and so what happens is when we apply our current to the deepened metal these electrically insulated pebbles are not subject to the jacuzzi force pushing outward so instead they can't accumulate inward where they are most needed both for the attenuation at high energy and for the multiplication this choice of materials in a relatively modest thickness of 46 centimeters allow us to uh attenuate uh to absorb 99 percent of the neutrons so only one percent of the neutrons passes through and a much smaller fraction of the neutron energy passes through this energy spectrum neutrons as a function of distance traveled across this uh medium there is an additional advantage this solution of lithium-bithium hydride also allows us to extract the tritium so shown on the right is the phase diagram of a solution of lithium-bithium hydride concentration on axis temperature on various phases and uh lithium hydride deuteride and tritide are chemically very similar to each other so um this is in fact a concentration of hydrides plural um check the liquid metal at 700 Celsius in the device and their neutron bombardment it warms up to 900 Celsius you extract it from the stellarator before re-inject it in the stellarator you want to remove the heat which you do by heat exchanger basically you put it down in the process of putting it down you can also you do also extract the tritium why because um you make a phase transition to this phase here where there is a precipitate which uh in hydrides including the tritide and if you do it carefully as a matter of fact similar to a distillation you can extract just the tritide because there are slight differences in precipitation hydrides so as an intermezzo um probably nearly the end of my work as a matter of fact because I'm not doing too well with time this is a symbiosis between the three parts in which this is not just a random collection of three ideas ideas they really work very well together and they help each other and they are symbiotic to each other and without entering in all the details on the on this triangle I will just say that the strong field generated by the hts helps the liquid metal sticking to the wall and on the other hand the liquid metal protects the delicate hts similar synergies exist quite still a little and I should also mention that regardless of our approach succeeding or not liquid metals and hts are good for fusion in general these are some pictures from other companies working on fusion various magnetic linear approaches most of them nearly all of them in fact need hts fancies in these areas are important for the whole field just few quick words on recent progress for example in stellarito theory in obtaining equilibrium between the kinetic pressure that tends to expand the plasma and the magnetic pressure that tends to compress it for non axisymmetric configurations of aspect ratio of unusually low aspect ratio between three and four so fat plasmas basically so these are small image radius but they are fat in many radius this is exactly what you want for a compact reactor you can see by the way the shrinking from standard designs so this by a factor of five or so as you see equilibria obtained by our stellarito equilibrium theories that also exhibit high rotational transform that properties that I was described for low effective ripple this is a metric classical transport so uh of transporting to a little configuration so it's really a very promising configuration and there is a similar one shown here for three periods as opposed to four which you saw in the big fight the trend yes are numerical results yes so I was a bit fast but these are numerical results and the color contours are contours of magnetic field strength on the last close flux surface and so compared to so we took the HSR 3 stands for helical axis stellarito reactor 3 stands for the number of periods and this is a design one of these gigantic designs that I showed in one of the original slides gigantic if you do it the low field if you do it the tie field so you see the field much higher here than here other mouse this pointer if you are here much higher than this one that allows to shrink ice significantly for completeness these other contours on the right this is the field strength again on the last close surface this if you take the torus and you wrap it and you plot the field strength as a function of the poloidal and troidal coordinate you get these nice stripes and so this is a quasi isodynamic stellarator which basically means your particles even where they are trapped between maximum magnetic field are not drifting away from the plasma they stay close to it and at the same time is quasi helically symmetric so the field patterns as a function of the toroidal and poloidal coordinates that both properties that the when it looks like that when it looks like stripes it's a good sign alpha particle losses so the alphas that we produce in the detection is that we want them to stay in the plasma so the idea is 80 percent of the energy goes in the neutrons we capture the neutrons they are obviously not subject to the magnetic field alphas are confined by the magnetic fields that's good because it means 20 percent of the fusion energy stays in the plasma and keeps it hot for more reactions to take place basically we don't have chain reactions in fusion at least not mediated by neutrons but we have a sort of chain reaction not quite but one reaction can promote the next one because part of the heat from that reaction keeps the plasma hot and fosters more reaction so it's a heat other than neutron mediated so bottom line we want the alpha particles to stay in the plasma and so an important calculation is what is the fraction of alpha particles lost as a function of time and the device that I showed before the one in red exhibits comparable trajectories with other mainstream device there is a new quasi-symmetric configuration this is generated by a colleague in Maryland and so the next thing we will do is take that and scale it up in field size for the recipe but the point is this performs just as well as traditional devices another interesting development lies the number of parameters to describe these complex surfaces and so normally you need something of the order of 400 free coefficients but a colleague is working with colleague a colleague at Renaissance fusion is working with other colleagues a Princeton to develop a machine learning artificial intelligence method that based on this input it encodes and distills only between 40 and 100 coefficients to reproduce the same equilibrium important because when you optimize when you look for plasma shapes that better confine your plasmas you don't have to move away in a 400 dimensional space but only in a 40 dimensional space and yes I know at first site it might be counterintuitive but let me go back one slide or two so the poloidal and toroidal coordinates shown here are in fact booster coordinates as we call them they are not exactly the toroidal and toroidal coordinate in the in the lab frame they are combinations of them basically it's a curvilinear coordinate system that you can define on this story if you properly define your variables the system is in fact axisymmetric or quasi-axisymmetric or more generally quasi-symmetric set of coordinates so in these three if you don't think in again the lab frame coordinates but you think in these other coordinates major radius quasi-poloidal quasi-toroidal also called booster coordinates or magnetic coordinates in this set of coordinates the stellarator is actually bidimensional one of the three dimension is ignorable just like in the tokamak the toroidal coordinate is so they don't look bidimensional to our eyes but in a proper set of coordinates they are bidimensional this bidimensionality lowers the dimensionality of the system improves transport and gives good confinement so they don't need to be two-dimensionally in the lab frame the advantage is better confinement i i don't have it in this presentation unfortunately but i can explain in words and maybe send you more references later so i alluded briefly to the effective repo that's a metric of transport neoclassical well sorry for the general uh conditional transport into toroidal geometries neoclassical as we say so on the net of turbulence which is a whole other class of you know nuances but on the net of turbulence if you just think about uh conditional transport into toroidal configurations other stellarators are actually better at confining the plasma than tokamaks again counterintuitive per site but if you try to calculate this metric which is called effective repo it is actually lower in in a stellarator than in a tokamak like with memories can i make a sketch quickly okay so if you plot the field strength the function of the coordinate along the field line so this field line is helically twisted right so it moves from the inside from the inside to the to the outside of the plasma and in doing so it moves between region of high and low magnetic field okay so a particle moving along the field line experiences maxima and minima of magnetic field okay and therefore maxima and minima of magnetic energy potential energy um if it has enough energy pass it will keep moving along the field line if it does not it bounces back and forth between turning points it's trapped exactly that's exactly it's the particle is trapped uh that metric the epsilon effective the power of three half it's a metric of uh distrapping it's a metric of the ripple in that field line and although it's counterintuitive at first site you can make it smaller in a stellarator in a stellarator the physics is much richer is much more complex because in addition the one over our dependence so that gives you a kind of you know slow variation along the field line so if you have i don't know if you're if you're field line um helically twist three times you have three times three minima and three maxima okay but but exactly because in a stellarator you have due to the complex shape you have additional maxima and minima you have at least three or four families so there are many uh so there is a there is a slow medium and short length scale by the way and if you look at the details of particles globally trapped or locally trapped actually these if property optimized can be better than the tokamak confined in the plasma now um i i believe there are many but there is i i'm sure there is a fraction of um high energy particle physicists in the audience or remote audience so i would like to stress that hts can be very good for accelerators and actually we have a collaboration uh with the synchrone in uh we are starting a collaboration with the synchrone in grenob so one quick step back is that if you print field geometries for different applications medical energy storage etc that technique that i showed you before just requires different corrugations so for different b you need different corrugations okay you can even do tapes if you want you just cut all the way through the sitting there um and some of these applications are useful for accelerators one idea for example is to deposit the hts on the beam line directly on the beam line pattern it and by different patterning you can do quadruple examples optimal spending magnets you name it um in particular we are excited about taking plates and patterning them with serpentines as shown on the left which incident coincidentally is also an energy storage concept so when a colleague from the synchrone saw that he said oh that's an undulator uh so you know for wigglers for um for uh sorry it's an undulator for for synchrotrons and it turns out but i will not enter into too many details because i think i'm off i'm nearly done with my time uh it turns out that only with this technology our technology you can generate very strong fields on short distances typically these two things are mutually exclusive but in uh in a good undulator you want to to let's say aggressively bend the trajectory trajectories of your electrons but for reasons of you know cohesion some of your x-ray signal you want to sum up over several periods therefore for the same total distance you want many short periods long story short with our technology this is possible and here you see that by operating at 20 Kelvin you can generate a little 1.2 tesla without getting too close to quenching so superconductors can be as you know a big deal if you're superconductor suddenly loses superconductor and by operating at lower temperature it can superconduct safely superconductor up to 15 1.5 tesla and yet be oscillating on just one centimeter uh yeah i'm not sure i have the time to enter in details maybe one quick comment about the fact that in addition in addition to the main uh corrugations that gives you this is a piece of the serpentine by the way so you have to imagine this mirror mirror than periodicized um so if you now do grooves inside the serpentine you avoid points of accumulation of current current in superconductors takes shortcuts just like in regular conductors of course it's a different type of shortcuts they don't take the pattern of minimum resistance because resistance the pattern the path of minimum energy which sometimes is kind of intuitive but they tend to accumulate that they tend to you know cut corners i mean slalom as quickly as possible and accumulated dislocations and you can reduce that by properly uh guiding the current with proper grooves by the main path and also another trick which our colleagues are very happy with we added an ion your top and bottom of the superconductor so now the the ion thanks to attract the superconductor the two superconductor the two serpentines tend to attract each other but strategically positioned ferromagnetic material stop and bottom attract the superconductors in the opposite direction if you properly design it these superconductors stay in place you get balance of forces um one final slide or two and again i'm sorry this is a bit over time if i can i just two slides okay um so um our roadmap so for the first three years our main focus experimental focus is um building and demonstrated uh cylinder one meter in length only one meter in diameter externally coated with the hts with that technique that you saw before and internally coated by a flowing layer of liquid metal 20 centimeters initially um because this will be the building block to build first a small then a medium and eventually a large stellarator as we learn to make bigger and bigger cylinders as you saw this is also this technology whether planar or cylindrical uh has also spin-offs in accelerators energy storage and other applications in parallel we are as you saw you know from those equilibria that we were discussing before we are designing our plasma shape and correspondingly we need to solve the inverse problem of even this plasma equilibrium uh what is the pattern of grooves that heals bearing in mind all those superconducting physics of minimum energy path which as i said so that's an important design and modeling work that is being carried out in parallel with these two experimental activities so that when we are ready to build a stellarator we are ready also with a design for the stellar or in the you know an optimal plasma configuration that we want uh so part of the reason why i'm here is to look for talents who might want to join us and this is a list from the department head to group leader to physicists level and these are some emails a web page when you can consult openings quick warning we need to some of the positions are not even listed yet so dynamic or the best way is i'm sure or approach me after this talk it would be a pleasure but as well as about collaborations by the way and that's it i stop here thank you very much for your attention i'm super happy to take more questions formally what sorry uh at that level we're not making a plasma so i mean we're making a plasma of the tritium surrounded by a liquid uh it's at the liquid state um yes you're thinking yes okay you're thinking of the electrochemistry okay um yes but okay this is um so in electrochemistry you apply an electrical current in this liquid metal we are applying indeed the current in the azimuthal direction for the sake of imposing radial forces which is good the only situation in which this free moving electrons might have a negative impact which i can foresee now but i i need to think more about the problem but the only situation i can think of is a radial current but it will have to be in the form of convective cells uh unless you know the now it's mostly in the form of convective cells unless you have a source in the sink but whether it is so this radial currents really have no reason they're not applying a current in the radial direction neither it is forming and they're you know as a result of the interaction with the plasma and if it does the simple solution to that is to bias the solid substrate to minimize this current we have these electrodes in the plasma anyway again to control the radial force i wonder if any happens so control the radial force which means controlling the azimuthal j if anything happens with the radial j in principle we could use those electrodes also to you know minimize those radial currents uh this metal is definitely in the presence of a strong field maybe instead of electrochemistry we should think about magnetochemistry but electrochemistry anyway thought-provoking question thank you very much it's um we are so the short answer is yes and not as much as we need in my opinion one of our openings is actually for a material scientist slash chemical engineer slash um corrosion expert uh you know it's a lot of uh yeah as I said material science science chemical engineering and so forth chemistry that needs to be addressed we have a for example a collaboration with the savannah river national lab uh in the u.s. they will measure for us these um phase diagrams because for lithium-lithium iodide they are known and they are approximately known for lithium deuteride and lithium tritide but but the detailed measurements when you have the three of them were never done before and actually there is a very interesting chemical physical calculation based on measurements with the lithium-lithium deuteride you can extrapolate what will happen with lithium deuteride lithium tritide so it's a very it is a very rich field uh theoretical and experimental that has been only the surface has been scratched there is a lot of work to do um collaborations that some of them exist but we definitely need more of them so from a business point of view the the fusion product will be ready here so at the end of the electrify phase the idea is to retrofit the idea is to build a demonstrator together with public partners in a dismissed fission power plant it could be a cold power plant actually um this is to save cost and time that would be our demonstrator our working prototype before we can commercialize fusion um to be honest I wasn't expecting this question from an audience of fellow physicists but I have a slide in another presentation but I will explain in words um so we don't see ourselves as operators of power plants in fact not even builders of power plants but rather upstream in the supply chain procurers of reactors emphasize reactors not power plants so the core the the stellarator plus the heating systems because there are other companies that build the pines generators switch yard etc so in the power plant we focus on the reactor or even at a smaller level on subsystems such as the htis coils or the liquid metal solutions so our ultimate goal our product just about commercialization our main product will be the cylinders okay with htis on the outside liquid metal on the inside even though the cylinders per se will be ready in three years the fusion relevant cylinders demonstrated that you know you can actually make fusion with them they will be ready in a longer period of time um on the way to the fusion product um in terms of business the idea is are magnets and they can be used for a ton of applications for medical and energy storage are particularly promising magnets for accelerators synchrotrons etc are also another important one so our idea is okay in three years as soon as we prove these magnets we can start uh i don't feel like the word telling it's it's more about having impact on society we can use them to to benefit medical science energy storage etc and in parallel we develop something that will have an even greater impact on energy in a longer period of time so it's high temperature superconductor sometimes it's just a high tc high critical temperature uh yeah sorry i said it quickly at the beginning so yeah high temperature superconductors so these are this is repco rare air barium copper oxide materials basically cooperate um exhibiting conductivity temperatures of the order of liquid nitrogen and higher so hence the name high temperature superconductors uh in reality the property that attracts us is not so much the ability to work at high temperature but rather to generate strong field the three-dimensional problem is not just the temperature but also the magnetic field and the current so in this three-dimensional space low temperature superconductors like now guanty now in titanium they occupy a small corner uh hds exactly they occupy a bigger space and the one axis that we are mostly interested in is actually the strong field rather than the high temperature exactly yes yes it's a big yeah it's precisely that it's a big inductor um but so you you when you design it you design in a way to maximize the inductance that's one aspect but another aspect is a much current you can inject and hence uh how much energy you can store and the other axis in which these cells are very you know i have a huge operational space current or current that seem to be precise so if you can piece a lot of current in these inductors or equivalently a lot of magnetic field in those volumes surrounded by the inductors you can store a lot of energy oh that's remarkably well it can be remarkably quick or slow it depends on the L over r time of your circuit the L of the inductor and the r of you know your charging or discharging system so it's very flexible but but it can be if you need energy in short periods of time you can have it as similarly if you want to store it quickly you can do that in fact this is one of the most attractive pictures the technology exists already is called smess superconducting magnetic energy storage we have some innovative ideas on how to uh on how to maximize the inductance uh on seeing their own plates etc it has a lot of interesting ramifications in fractal geometry if you're interested but but the technology is a family of the the basic idea let's say the basic principle has been around for some time but one of the limitations uh a low temperature superconductors can only generate so much field p even if magically they could be operated the tire field the forces on the inductor would be so high that your device breaks unless you have a huge supporting structure but uh in the process of process okay here we go in the process of minimizing the forces on the stellarators i mean this minimization for the stellarator is actually also good for energy storage the torus so the stellarator is a torus optimized for plasma confinement uh the energy storage system could be a solenoid or it could be a toroid preferably a toroid and you optimize it with a different geometry for maximum energy storage volume integral of the squared over to u0 and you can do this optimization under the constraint of tolerable forces just or you can create a figure of merit where as you see here the forces are part of the metric or the energy storage system just as we did for the stellarator so you can you out of many approximate solutions you choose the one with minimal forces so this this issue of you know the the system breaking apart due to huge forces is is thank you