 Hello everybody, so this is part two of our talk on tailoring metal hydride for practical application. In this talk we will see the effect of reducing the crystallite size on the hydrogen storage capacity and also we will see the use of other techniques to obtain nanostructure. So first let me talk about iron titanium alloy. This is an old and well-known alloy. This was discovered as metal hydride in the 1970s. It could absorb hydrogen at room temperature, but the first hydrogenation, what we call activation, is very difficult. So even if the alloy could store absorb hydrogen at room temperature, usually the first hydrogenation, you have to go to high temperature like 450 degrees C in vacuum. After that you anneal at a certain pressure like 7 bar, you cool to room temperature and then you go at high pressure and you do that a couple of times. So it's a lengthy and complicated process. So what will be ideal will be to be able to do the activation, the first hydrogenation directly at room temperature under low pressure. So all you could do that, well, you could substitute some element. Usually people are substituting on iron. So they replace iron by other transition elements, for example manganese, and then you could have a first hydrogenation that is quite quick. But when you do that, you could change the thermodynamics and the capacity. So that's not the ideal way. Or you could add some other element or add an alloy, another alloy and have like a nanocomposite, or you could get a nanostructure. So this is what we will see in this talk. So the work that I'm showing you is the ball milling of iron titanium with some addition of zirconium. And that work was done in collaboration with the Sabrina Sartory of the University of Oslo. So again, here we add 4% of zirconium to make the alloy easy to activate. We know that, but we would like to know what is the effect of ball milling. So how does this compare with ball milling? So we cast the alloy by arc melting, and then we mill it on a specs machine, a high energy milling machine, and we mill for 5, 15, 13, 60 minutes. And also we try cryo milling. So milling at low temperature, but this time it was in air. All this milling here was done in argon, but the milling at low temperature was done in air. And we will see that this made some problem. So this is a morphology of our powder after ball milling. So this is ass cast. So we cast a sample and then we just crush it with mortar and a pestle. And you have this big chunk of iron titanium. And then we ball mill it for 5 minutes, 15 minutes, 30 minutes, 60 minutes. And this is cryo mill. So you see that after just 5 minutes of ball milling, you have a huge reduction of particle size. But after if you continue to mill for 15 minutes, you see that you have agglomeration of smaller particles. And this agglomeration will continue for 30 minutes and even 60 minutes. And now you see that all these particles are agglomeration of the small particle that you see here. And so you don't see any small particle anymore because you have agglomeration of all of that. And for cryo mill, you see that the particles are very small. But we think it's because it was mill in air and you have a lot of formation of oxide and it prevents agglomeration. And it makes the degradation of the particle. So this is the X-ray diffraction pattern. So this is the ass cast. You have iron titanium and also a little bit of iron to titanium. And when you mill, you see the peak are getting broadened because it's getting nanocrystalline structure. But we don't have formation of any new phase. So you see the cruciate size ass cast is 23 and ball mill 5 minutes were already reduced by a factor of 2. And it continues to be reduced. And here for 15 minutes of cryo mill, we have about 15 nanometers. This is bigger than this because this milling machine is not as energetic as the specs machine, the machine that we use for milling at room temperature. So that's why this crystallite size is bigger than this one because they are taking on different machine. So this is the first other generation at room temperature and under 45 bar of hydrogen. So you see this is the ass cast here and you see it activate and takes about two hours and a half to fully absorb hydrogen. If we cryo mill for 15 minutes, you see that the sample is totally dead. It's because we we mill under our own in air and then we have formation of oxide. And as you as we saw in parts one, in order to to hydrogenate the hydrogen has to go through the oxide layer. And here the oxide was probably too thick and it was impossible. So in the future, we will do this experiment again, but under argon. And then we're confident that the sample will then absorb hydrogen. But for milling under in at room temperature, you see that this is five minutes and 15 minutes, 30 minutes and 60 minutes. And you see that as the milling time is getting bigger, the the kinetic is getting faster. So you have a faster kinetic longer milling time give you faster kinetic, but you see that the capacity is getting down. So kinetic is getting better, but capacity is getting worse. So how could you explain that? So we took with plot. So this is the hydrogen capacity as was measured as a function of milling time. And you see that the capacity is going down. And here this is the crystallite size again as a function of milling time. And you see that the crystallite size is also decreasing and it looks similar. So probably these two are related. So this is what we'll see in the next slide. So if you consider that the crystallite again from the first part, we saw that the crystallite is the current domain, the crystallographic current domain. And so in order to break that that currency, all you need is one unit cell that is not current with the unit cell that is beside it. So that's more or less the small, the thinner grain boundary you could have is to have one unit cell that doesn't match the unit cell beside it. So for iron titanium, the unit cell is a triangstrom. So we assume that the grain boundary is 0.3 nanometers. And then if we assume that the crystallite are spherical and of course a crystallite that we measure by X3 is an average. But we assume that all the particles are at this average. And we assume that all of them have the same grain boundary thickness. Okay, and then we could calculate the volume of that grain boundary compared to the total volume of the particle. So this is a very simple equation. So on first order, this is three times. That's the thickness of the grain boundary and that's the radius of the crystallite. Okay, so if you do that, you'll see that if you express the volume of the grain boundary over the total volume, these are with respect to the crystallite size. So these are the number and that's the capacity, the hydrogen capacity loss that we measured. And you see that these number are very close to these numbers. So it looks like the loss of capacity is scaling exactly as the volume, the total volume of the grain boundary. So then we could explain the loss of capacity. So when you create a new grain boundary, this act as an hydrogen path as we saw in the first part. So this means that it's fast kinetics because now you have more grain boundary. So you have more path for hydrogen to go into your crystallite. But the grain boundary give a fast diffusion, but it could not store hydrogen. So hydrogen is going through the grain boundary but doesn't stay there. So you have a loss of capacity. And of course, the more grain boundary you have, the less capacity you will have. So and you see that for iron titanium, this scale perfectly well. But I have to tell that this loss of capacity depends on the system. So for some alloy, this loss could be minimal because when you hydrogenate the hydrogenate, you may recreate a currency. So your crystallite size may increase a little bit because the hydrogen will force the unit cell that is on the grain boundary to adopt the same to get coherent with the unit cell beside it. So this loss of capacity with the grain boundary is not true for all systems, okay? But for some system, we see that it scales very well. So the other aspect that I would like to see in this talk is the other method to obtain a nanostructure. So in the field of metallurgyride, we have a couple of methods that are widely used by the researcher. Maybe the one that is mainly used as equal channel angular pressing. So I will not go into detail, but here you have a piston, you have a die and there is an angle in the die. So you take your sample. So this is your sample and you push it through the die. So it has to go through this angle. And then by going through this angle, it will sustain a lot of strain. And so you could get a high-less strain material and also you could get nanocrystalline structure that way. The advantage is that depending on, because you could do that, you could repeat that many times. And depending on how you push back the sample, if you rotate it 90° or 180°, you could induce deformation in different planes. So this is interesting because according to the crystal structure of your material, you could say, well, I would like to induce deformation mainly in that plane and not in that other plane. So from a fundamental point of view, this is quite interesting. The disadvantage is that the bulk sample has to be filed. So after you process that, you have, so typically this dimension is maybe one centimeter. So you have a block of one centimeter by one centimeter by maybe a couple of centimeters long. And you have to file this block. So you more or less modify your sample by doing that. So you induce some other deformation by just filing, by reducing in powder. And also this could be difficult to scale up. The other technique that is more and more used now is high pressure torsion. So here your sample is a thin disk that you put between two anvil and you rotate it. You rotate the anvil and you press on it very hard. So the advantage is you have a very high strain. So maybe this is the technique that you have the highest strain. But the disadvantage is you have a very small sample. So maybe it's one centimeter of diameter by maybe one millimeter of thickness. So your sample is very, very small. And finally, we have cold rolling. So the advantage of that, it's a well known technique. And I think it's very easy to scale because the industry is doing a lot of cold rolling. The disadvantage is usually you have to use a plate. But in our case, we solved this problem because we took the cold rolling machine and we rotated it 90 degrees. So then we could process powder because usually our material, our metal hydride is in form of powder and not plate. So the example that I will show you today is lentinon nickel five. So cold rolling of lentinon nickel five. So we took commercial lentinon nickel five. We roll it in air. And after rolling, we did the first hydrogenation, the activation at 50 degrees and 15 bar. And the desorption was also done at 50 degrees and five kilopascals. And the theoretical capacity of lentinon nickel five is about 1.5 percent. And again, I have to stress that the cold rolling was done in air. And we compare that with ball milling because ball milling is used by everybody and everybody knows what this ball milling is doing to metal hydride. So we use a high energy mill and the milling was done in argon, contrary to here, that was done in air. And we milled for 15 and 60 minutes. So this is a morphology. So we're after, so the as receive one is like big block. So this is 50 micron. So this is almost 100 micron. So you have big particle. And when you call roll, basically what you do, you take all these particles and you make a small plate with it. OK, so you see that the and then we took this plate after we're rolling once we take the plate, we fold it in two and we'll roll again. So after each rolling, we have a reduction of 50 percent of our of our plate. OK, so so this is after 512 and 25 roll and you see the morphology of the plate do not change so much. But of course, we will have reduction of crystallite size and strain because we are every time after every roll, we reduce the the thickness of that plate by 50 percent. And this is ball mill 15 minutes. So you see, as we saw previously, you have the agglomeration of smaller particle here. And after 60 minutes, you more or less you take these particles and you break them again. So you see the particle are smaller and you have even smaller particle here. So this is the first other generation, the activation at 50 degree and 15 bar. So this is the as received. So you see the as received. It's very difficult. It's it takes forever to to activate. And if you ball mill 15 minutes, you see that it's very fast. You it's that the sample is taking hydrogen right away. OK, but you have a loss of capacity here. OK, but if you ball mill more, if you ballman 60 minutes, you see that you have a reduction of capacity and the sample is almost dead again. It's even worse than than the as cast. But if you do cold rolling, you see five times cold rolling. It's quite fast, not as fast as as ball milling. But you see, we have almost full capacity. And if we roll 12 time or 25 time, you see that 12 time. Actually, it's getting a little bit slower. The incubation time is longer. But we are we have almost no loss of capacity. But if we keep cold rolling, we have a loss of capacity because again, we do every all of that in the air. So as we roll, we will start to produce oxide and to make oxide and then it will reduce the capacity. So this is the desorption. So you see that this is the desorption of the as receive. And this is the ball mill 15 minutes. So it's quite fast. And that's a ball mill of 60 minutes. It's slow. But what is interesting is that all of them, the ball mill 15 minutes and all the cold rolling, they are all of them have more or less the same kinetic. They just the only difference is the capacity, but they all have the same kinetic. So the point is after one cycle, they almost have all the same capacity. The same kinetics. And this is what you see on the second absorption. So on the second absorption, you see that all of them are absorbing at the same rate. So the difficulty is the first hydrogenation. Then you have a difference of kinetic between all of them. But after the first hydrogenation and this is a common example for many metal hydride, after the activation, then the kinetic is the same. The only thing that is different is the capacity. So this is the diffraction pattern. So as receive, cold roll 5 times 12 times 25 and ball mill. And you see that 15 minutes of ball mill, the peak are already quite broad compared to 5 cold roll. The peak are broader, but not so much. So ball milling is much more efficient to reduce into nanocrystalline structure. And you see it here. So as synthesized, so this is as received, it's 50. This is in nanometer. So 15 nanometer and ball milling 15 minutes, you reduce that by a factor, almost a factor of 10. So it's 6.9 and 60 minutes, you go down to 5 nanometer. While for the cold roll, you see it's 21 nanometer and then 12 and then 9. So you have a reduction of crystallite size, but not as important as for the ball milling. And this is after two cycle of hydrogenation. So you see that even for the as receive, the crystallite size will reduce in this case because of the decrepitation of your material. But you see here, you go from 21 to 16. So you have a reduction, but here you have an increase. And here it's more or less constant, but it doesn't absorb hydrogen so much. But so you see that it looks like you have like a natural size of crystallite. After cycling, it will tend to have like a certain size of crystallite. So the conclusion for the lentinominal 5 is that cold rolling sample present is a faster activation despite having similar crystallite size than ball milling. So the point that I want to give you a year is that nanocrystalline is important for kinetics and for activation, but it's not the only parameter. Okay, some other characteristics also have an effect on the hydrogenation kinetics. Okay, so cold rolling five time at the highest reversible capacity and the shortest incubation time of all the cold roll sample for ball milling. If you ball mill for too long, it has a detrimental effect. So it's not good to ball mill for too long and to try to reduce the crystallite size at the very, very small. Sometimes it's better to have a crystallite size that is a little bit bigger, but you will still have faster kinetics. Up to now, the exact reason for activation and enhancement with cold rolling, it's still unclear to us, but it's known that cold rolling increase a number of eye-angled grain boundary. But in our case, it still has to be proved. We still have to do some experiment about that. So in conclusion of that course, I hope I show you that metal hydride are attractive material for hydrogen storage, but we need more development. We need to reduce the cost to get better capacity and also to improve the cycling. Nanocrystallinity is a good way to improve the kinetics, but it has an impact on capacity. And we saw that many techniques could be used to get a nanocrystalline structure. So with this, I will end my course. I would like to thank you for your attention and I hope you will have a better understanding of metal hydride now. Thank you very much. Bye-bye.