| Battery2030+ Excellence Seminar - Laurence Croguennec | |
|---|---|
| ID | OSW7d5ef82cc31c49dd99a0df7439fe21dd |
| UUID | 7d5ef82c-c31c-49dd-99a0-df7439fe21dd |
| Label | Battery2030+ Excellence Seminar - Laurence Croguennec |
| Machine compatible name | Battery2030ExcellenceSeminar |
 |
|
| Statements (outgoing) | |
| Statements (incoming) | |
|
|
|
| Keywords | webinar excellence seminar |
Description
An instance of the Battery2030+ Excellence Seminar series featuring Laurence Croguennec on the topic, “Solid state chemistry, a source of innovations in the metal-ion battery field”
| Item | |
|---|---|
| Type(s)/Category(s) | Event with People |
| Event |
|---|
| Event | |
|---|---|
| Event series | |
| Start date | 2024-11-28 |
| End date | 2024-11-28 |
| Duration | |
| Location | |
| URL | https://battery2030.eu/news/happenings-events/battery-2030-excellence-seminar-28th-of-november/ |
| Organizer | Uppsala University |
| Minutes taker | |
| Project(s) | Battery 2030+ |
| Associated OU(s) | |
This seminar, featuring guest speaker Laurence Croguennec, will highlight the latest research in battery materials, mechanisms, and solid-state battery applications. Laurence Croguennec is CNRS Research Director at the Institut de Chimie de la Matière Condensée in Bordeaux (ICMCB-CNRS, France). She graduated (PhD) in 1996 from Nantes University at the Institut des Matériaux Jean Rouxel (France) and spent one year as a Post-Doc at the Bonn University (Germany).
Recording
Transcript
Thank you, Erik, for your own nice introduction. And I would like also to thank the Battery 2030 team, and especially Christina and you, Erik and Camilla for arranging this talk today. So as you mentioned, I am the CNRS Director of Research at the University of Bordeaux and especially at ICMCB, and I will give a talk on--I will try to highlight how solid-state chemistry can bring a lot in the innovation of metal-ion battery next generations.
Overview of ICMCB
First of all, I will give a few words on ICMCB, our laboratory.
This institute is localised at the University of Bordeaux. And as you can see on this slide, we have three sites the main site being this one. And we have expertise in solid-state chemistry, material science, chemistry and processes, with the support now also more and more of artificial intelligence. And the director of this lab is Cyril Aymonier since 2022, and I am the Deputy Director of this lab.
So in fact, for those who would not know, Bordeaux is localised here in France, so this lab is there. And we are about 250 people in average belonging to these three institutions: CNRS, University of Bordeaux and Bordeaux INP. As you can see, we have 110 permanent researchers, engineers and technicians and all these here are hired PhDs, non-permanent post-docs, engineers or technicians and today a few masters will join the lab from January 2025 as each year. So our expertise is really based on the knowledge of condensed matter and of course, I would say, taking into account the resources, and all our expertise, solid-state chemistry, chemistry and processes and material science is really dedicated to propose new materials or to optimise materials in the field of energy, environment and health, electronics and photonics with the support of life-cycle assessment and also artificial intelligence.
So we have seven research groups, as you can see here. I will not go into the detail of these seven research groups. I just wanted to mention that our group is the group “Energy, materials and batteries” and since 2022, Prof. Dany Carlier is leading this group with, as you can see here, about 45 people in this group at the present time.
So what is interesting here for you, I think, is to know the keywords concerning our research. We are involved in batteries and supercapacitors chemistry, especially proposing positive electrode materials and solid electrolytes. And our expertise is really crystallochemistry, but also I would say from the synthesis of the materials to the behaviour in the battery or in the supercaps. So we are really focused on the characterization of the mechanisms.
We really want to address the role of defects and then also to do the link with the chemical balance and the electronic structures in these materials, combining and doing that, combining our expertise in diffractions, spectroscopy, microscopy and theoretical calculations. So the first part of my talk will be dedicated to highlighting how solid-state chemistry can still bring new results on high-voltage spinel, LNMO, as positive electrode material for the next generation of batteries. Like this I wanted to highlight how our expertise can really bring some new knowledge in that field. And I wanted also today to focus on especially raw materials and the need to develop materials that take into account, I would say, this constraint.
Challenges in Metal-Ion Battery Chemistry
So first of all, as you know, metal-ion battery chemistry has in fact a series of challenges, because we have the materials that the crystallochemists develop that will be involved in an electrode with carbon additive binders on the current collectors, but then this one will be integrated in a battery, and this battery will deliver a capacity over a number of cycles. So it can be, of course, the autonomy for a car, for instance, but depending on the use of this battery, of course, we will have more or less fast ageing, depending on the temperature, depending on the C-rate of charge and so on. So we have a lot of reactions that can occur in a battery as illustrated here. I will not go into detail, but what we have to retain is a battery is a metastable system and, I would say, the material by itself, so here the positive electrodes, but also with this interaction with the electrolyte, the electrolyte itself being also metastable.
So we have really to understand all the reactions involving the material, its interfaces and I would say even the cross-talking with the other electrodes in order to optimise the chemistry of a battery.
Material Research and Technology
So of course, in our group, we developed material research in order to try to respond to these challenges, and for instance, to propose new materials for high energy densities, for fast-charging safeties and taking also into account the raw materials availability. So as mentioned, I will focus today mainly on that point, mentioning some materials for lithium-ion batteries, but also some new technologies, such as sodium-ion batteries.
Generations of Positive Electrode Materials
So of course the batteries are made of two electrodes and an electrolyte, and you have different scales you can look at, here I would say the electrode, here the material, and in our case we are more focused, I would say today, on this scale here, the crystal structures and the atomic and electronic structures of the materials, of course in interactions of the electrodes with the electrolytes.
So here it's a kind of recap for lithium-ion batteries of generation of positive electrode materials we have worked on. You can see here that we do not have so much structures still developed in batteries. We are moving here from layered oxides to olivine, so it has LFP here and LFMP then to spinel and then again to layered oxides with lithium in excess in these materials. We can see that you have different chemistries here along this arrow here.
And at first the motivation was really to go towards, I would say, longer life, more safety for all these materials here, moving from lithium cobalt oxide to NCA, nickel cobalt aluminium, and NMC, nickel manganese cobalt. And in that case, we were really playing with the cationic redox. Now in order to answer to more and more demanding applications in energy, we are also moving to other chemistries, such as lithium-rich layered oxides, rock-salt type structures and high-voltage spinel. And in that case, here for the lithium-rich and layered oxides, we are even jumping in structures, compositions where we have cationic redox and in addition, anionic redox.
At the present time, we are all highly interested in these materials due to the fact that all these materials answer to another constraint when we develop materials, its resources. And here you can see that all of them are rich in manganese, lithium iron manganese phosphate, a high-voltage spinel that is rich in manganese and nickel and all these lithium-rich are also rich in manganese. So today, in fact, I will highlight our expertise and new results we obtained on these LNMO materials, these manganese-rich materials.
Sodium-Ion Technology
And then I will move to sodium-ion technology.
So first, the general context concerning these materials. These materials are often presented with this composition here, lithium nickel 2+ O.5 manganese 4+ 1.5 O4. And in fact we expect for this material only the involvement of the nickel 2+ as the redox couple in these materials. You can see here nickel 2+, nickel 3, nickel 4.
And you can see first that if we compare here the electrochemical curve of these materials, here in black, versus a classical, a conventional NMC here in red, we can see that with this LNMO materials considering the energy density, we can really compete. We have, considering the material, 650 Watt-hour per kilo versus 700 for the NMC chemistry. But I would say the great interest for this material is really the fact that it's a manganese-rich material and it's also a lithium-poor material. You can see that here, on the contrary to NMC, we have a lithium over transition metal ratio that is 0.5 versus 1 for NMC, and we have a nickel over transition metal that is smaller than 25% and manganese-rich versus often more than 60% in an NMC.
So in fact, if we consider the metals, you can see here that we have a decrease in lithium content of 50% for this chemistry. Of course, no cobalt, a decrease of 60% in nickel and of course an increase to compensate in manganese. So, of course, this material is highly attractive. So then if we come, I would say, to the structure of this material, this material is based on a cubic structure, a spinel-type structure, and we have in fact two structures for this material.
We have a disordered structure, it means that here on the octahedral site of this spinel structure we will have manganese and nickel that will be statistically distributed, whereas lithium will sit on a tetrahedral site. So this polymorph is obtained at a higher temperature than 900° when we do the synthesis at this temperature. But in fact here I use this tilt here because in fact, when we are doing this temperature at high temperature, we have always a tendency to lose oxygen. And so not to have exactly this formula, but to have oxygen loss, reorganisation of the material with generation of impurity and then for charge compensation, reduction of manganese.
We'll come back to that later. And we have a second structure for this LNMO that is also cubic. And when we reanneal this material at lower temperatures, 700, we have a rearrangement of the transition metals, I would say, within the samples. And in that case, the pure phase can be obtained with an ordered structure with nickel in octahedral sites, only nickel, nickel 2+ in that site 4b, and then manganese in the other site, 12d, octahedral also.
So in fact, I would say the phase equilibrium when we are doing the synthesis of the spinel oxides can be really described by this equation here. We are really always in balance between the stoichiometric phase with only nickel 2+ and manganese 4+ and with deficient phase in manganese that does contain manganese 3+ for charge compensation and impurities rich in nickel in that case, with, of course, an evolution of oxygen. And then the other question here is, when we speak about ordered and disordered structure here, the question is also, and it was addressed by our colleague Montse Cabanas and coauthors in 2016, the fact that, in fact, we are not really able to describe exactly how is organised this order and disorder at the particle level and even at the powder level. You see here if we are in the 1-1 who play in this material, disordered is described here with this tactical distribution on the 16d sites of the nickel and manganese.
When we move here to the ordered, we have again this 4B and this 12d. But in fact in the material we don't know if it is a coexistence like this of ordered domain, more or less big, I would say, and next to them the disordered domain, or if it's this, much more subtle than that, with antiphase boundary, as you can see here, or only antisite defects, and I would say in that case, in the ordered phases, and in that case, I would say, kind of punctual defects. So in order to have more information on that kind of disordered-ordered model, we need special information, and I will give you some clues about that in my talk. So then, another information that is very important here is also with the ordered LNMO, you have perfectly ordered, you have this green signature here.
And so as you can see only the involvement of the nickel in the reactions, whereas when we move to disordered phases here we have here the involvement of the redox compose manganese 3/manganese 4 in addition that had higher voltage than again the nickel redox. But there are also still debates. In the literature we can find different, I would say statements about what is better, to have an ordered, to have a disordered phase, And so the ordered, of course, is only a manganese 4+ phase, when it is a perfectly ordered material and in that case it is associated with higher structural stability. When we have manganese 3+, we have always problems with manganese 3+ dissolution upon cycling.
And the involvement of this manganese is the redox processes. On the other side, disordered is often presented as having higher capacity retentions, higher C-rate stability due to this mixed balance we have here between manganese 3+ and manganese 4+. So still a lot of debate, and often the phase nickel 0.4, manganese 1.6 is presented as the best one I'd say. So as mentioned, still a complex system, considering especially that all these are highly correlated.
So the oxygen loss we have during the thermal treatment will control the nickel of a manganese stoichiometry in the active phase. Considering that of course the stoichiometry in the compound, in the nickel of a manganese does not change. But depending on if we have impurity or not, next to the active phase, we can have a very different stoichiometry in the active phase and very different impurity next to it. And depending also on the stoichiometry, we have, as just mentioned, the presence of manganese 3+.
I mentioned also the importance of this material. This material is a high-voltage material, so it means that we are here doing the redox reaction at 4.8 volts. So in a voltage domain where the electrolyte is rather unstable, mostly the electrolytes. Today we’ll not mention, but in fact, in the literature and in our labs we have a lot of research or so in order to improve and to modify this electrolyte with additives in order to stabilise the interfaces of this material with the electrolyte and the stability of the material.
Because you can see here that as soon as the electrolyte, your salt here is degraded, we form some HF and of course here we can have an acid digestion of our material at the surface and specially a solubility of manganese 2+, generated by the dismutation between 2 manganese 3+ giving 1 manganese 2+ and manganese 4+ and then migration of this one at the negative electrodes and being the source of loss of performance for this material. So of course, we want really to mitigate the amount of manganese 3+ in this material. I just wanted also to mention that of course coating can be a part of the solutions also in order to stabilise these interfaces. So today in fact I will mention the results especially obtained in the thesis of Ilia Tertov that was funded by the European programme InnovaXN and ILL and Ilia will defend next Monday.
And also a short view about the work done by Gozde Oney. We developed a 4D-STEM analysis in order to get spacial resolutions of this material. It was done in the frame of an ANR programme. So the question is, is disordered LNMO always the one that performs better? Cannot we have some ordered LNMOs that perform rather well and competitive with this disordered, and so close to the perfect stoichiometry? Or are distributed ordered and disordered domains and which perspective can we give to this work? So in order to try to answer to this questions, we collaborated with Umicore in the frame of this work and with LRCS and in fact we did large batches of materials 18 samples that were performed by Ilia in Umicore, and starting with a hydroxide metal precursor from Umicore, you can see here.
And in fact, in these synthesis conditions here, at a very high temperature for the first thermal treatment, in order to really promote the growth of the particles and so to promote low specific surface area and thus, I would say, less interaction with the electrolyte and more stability versus degradation. And then we did a lower thermal treatment in order to try to promote the formation of the ordered phase. And that we did. And here just to highlight, 75/25 corresponds in fact to the composition of manganese and nickel.
So here we are in the stoichiometric conditions, air in the atmosphere. And so we have really varying the manganese over nickel ratio, deficiency in manganese, stoichiometry excess in manganese. We have varied the composition in lithium, deficiency in lithium, stoichiometry excess of lithium and we have used different air versus oxygen. So I will not give you all the details of the results obtained.
I will just focus on a few results in order to highlight the main results. So first of all, I would say the originality of this work is also that we combine synchrotron X-ray and neutron diffractions in order to really take advantage of both techniques. And here I took four samples in order to highlight this. So 75/25 here under air or oxygen in red, so are the stoichiometric materials and 77/23 in air or under oxygen are the manganese excess materials.
So when we use synchrotronic X-ray, of course, we can distinguish lithium versus transition metal ions, but we cannot distinguish nickel and manganese. And you can see also that using synchrotron we are really able to detect very tiny peaks. So here in blue it highlights peaks associated to rock-salt type impurity, nickel-rich. But you can see that it highlights also already by X-ray peak 3 2 O associated to an ordering between nickel and manganese.
So it already highlights that we have an ordering between nickel and manganese in at least these materials here that were obtained under oxygen. So now if we jump to neutrons, the benefit of neutrons is also really that we can distinguish nickel and manganese in these materials. So we are able to distinguish what I have called at the beginning a disordered structure with the statistical occupancy of the octahedral site and the ordered with one site for one transition metal. You can see here in grey, I would say the ordered superstructure line and you can see that for all of them we can see the superstructure line, whatever the synthesis conditions, I recall you that the second thermal treatment was done at 725°, so in the domain of stability of the ordered phases.
But it is also interesting to highlight that depending on the conditions, you can see that the superstructure line associated to this ordering between nickel and manganese are very different broadenings, moving from, I would say, a sharp line to a much broader line and even here, much more diffuse, and that, you will see, it will be associated to the size of the ordered domain. So here I just wanted to highlight the fact that here it's really a dome of a synchrotron X-ray diffraction pattern that we were really able to identify the impurity at least from a structural point of view, and to identify the presence of rock-salt impurity as one can observe in that kind of materials, but also of different layered oxides impurity. You can see here different peaks associated here in the backgrounds. So here this graph here gives an evolution according to the composition in manganese and nickel here, the stoichiometry here, deficiency in manganese, excess in manganese and here at the bottom you have the stoichiometry in lithium versus transition metals.
You can see that for this, the more green it is, the purer is the LNMO phase, as you can see here. Here we are close to the purity and we are here under air with oxygen here we have the full purity of this material here. So you can see that depending on the conditions, we can really have highly impure materials such as this one, with different natures of impurity as you can see here, spinel, rock-salt. And for this one here for instance, we have layered and rock-salt.
So for this one here, the blue one, the manganese excess, stoichiometric and lithium, we have the presence of a tiny amount of rock-salt impurity. So you can see that we have three types of impurities in that kind of materials. So then it is also interesting to do the link, of course, with the electrochemical performance and we have tested this material higher in C-rate conditions or in long-cycling protocols I had to have versus lithium in that case. So here you have I would say, low C-rates, D/10, fast C-rate, so 20-minute charge, and here you have the Coulombic efficiency.
And again, the Y axis is still the same, the manganese-nickel composition. Here the X axis is still the lithium over metal ratio. And here you have still the green, of course, is the better. And you can see that the better performance is obtained for manganese excess compositions, so 77/23.
If we look more carefully at the electrochemical curve, here you have the air materials, oxygen materials and here the stoichiometric and here the excess of manganese. You can see that we have a slight modification of the electrochemical signature. You can see that for this one, stoichiometric composition under oxygen, we are really able to have almost pure stoichiometric materials and ordered signature, with this derivative here, despite a small amount here of manganese 3+ as you can see here by this signature here and by this jump here. And here, for this other composition, you can see that we have a much more complicated evolution here, with two peaks here in the derivative curve that can be observed for this one obtained for manganese-rich materials.
And this was associated to the presence of manganese 3+ in the materials, but on the two sites, 4b and 12d, and so with the redox activity of 1 manganese 3+ in one site versus the other activity at higher voltage in the other site. Interestingly also, you can see that at C-rates, we have some materials that behave much better than the others. Especially here you can retain the colours, orange and green, light green and strong green, that behave much better upon long-run cycling, but also together when we consider C-rate behaviours. So it's of course very interesting to go deeper in the understanding of the structure of these materials.
And it's also interesting to highlight here that you have a much worse behaviour for the fully stoichiometric materials obtained under oxygen that shows no impurity in that case. So if we go further in the structure of these materials, I just put these materials, this X-ray diffraction pattern, just to convince you that the results were, I would say, of good quality, but we will focus really here on these results here. So if we consider the manganese-rich material, whatever the atmosphere we are using to prepare this material, in fact it does not change so much versus the structure. You can see that the same parameters are very similar, and I would say the distribution among the two different sites also.
We have a manganese site that is full of manganese and we have a mixing in the nickel site between manganese and nickel, and of course, manganese 3+ and nickel 2+. But you can see that depending on the atmosphere in a similar annealing condition, we can have a difference in the size of the ordered domain, that I had associated to the superstructure line broadening here. Now if we move to these stoichiometric materials here, in orange and red, you can see that for the red one, stoichiometric under oxygen, we are almost at the perfect distribution between the two sites, and we have also a big size for the ordered domain. And again under air, a much smaller domain size and similar distribution and mixing versus the manganese-rich materials.
So in fact, this combination of X-ray synchrotron diffraction, and neutron diffraction allows us to really describe these materials. In fact, all of them have to be described in the ordered space group P4332, and not in the disordered space group, Fd-3m, with a more or less extended size of the ordered domains. And then we have shown also that we have a manganese partial occupancy on the nickel 4b site. You can see it can vary from 2% to 5% for that kind of materials.
And then we have also different sizes of ordered domains. So that was for given synthesis compositions, playing, I would say, with the compositions of the materials. But now what happens if we take these manganese-rich materials obtained under air and if we do the thermal treatment at different temperatures, moving from 625° to 725°? How do these parameters evolve? Do we have always the same tendencies? And in fact, I will not show here the performance, the electrochemical performance, but in fact we have really the same tendencies. Look here in blue.
In blue we have the manganese-rich materials versus the red here, the stoichiometric composition materials. In fact, you can see that in these materials, we have bigger ordered domains obtained here with the manganese-rich versus for the same temperatures smaller domain when we are at the stoichiometric compositions. And you can see also that we have a higher mixing when we are-- we have a higher mixing here between the nickel and the manganese site. These parameters here being a kind of scale for the amount of mixing between the two sites.
When it is no mixing, this parameter is 1, and with the increase of the mixing we are moving towards 0.5. So you can see that we have more defects for these manganese-rich materials despite, I would say, the bigger domain. So of course, it's interesting to determine how these domains are organised, and for that we did some microscopy, because as you have seen, really these materials are those which behave the better in the battery. So of course it's interesting to understand how the domains are organised and to propose a hypothesis why it happens like this as better electrochemistry.
So as mentioned just before, this crystallographic plane allow really to see, I would say, the arrangement between the nickel and the manganese and thus to have a highlight about the extent of the ordering or defects in these materials. So this work was performing collaboration with EMAT in Antwerp and Umicore, and Dmitry and Mylène. And they did this cutting here, FIB cuttings, along these directions here in our particles in order to have thin layers, 50 to 100 nanometers thick, and to be able really to have information from the surface of the particles, from here to the bulk of the materials. And they did some different analyses of the images that they got by microscopy, combining different fast Fourier transform, masking and so on, and they were really able to show that we had in our materials some, as you can see here, area where as shown here by the electron diffraction pattern that we are here in a disordered area.
So it means indexed by the Fd-3m space group. And here we are more in ordered, where we have this additional spots here showing that we are in this P4332 space group. So what does it mean if we go further? In fact, if we look to these images here, we have from the surface to the bulk of our materials a mosaic organisation in the thickness of the particles in the bulk, and you can see that when we, I would say, screen the particles, we have whatever from the surface to the bulk the same size, average equivalent diameter for this domain here, all these grey white are associated to the ordered domain. And what is very interesting much more is on the contrary to what we have-- what we have seen or we could have expected a core-shell effect due to this oxygen loss during the synthesis, we have in fact a mosaic effect with ordered domain here and disordered here at the boundaries.
And I spoke about this effect of temperatures, when the annealing is performed at 625° and moving to 700°, you can see that we have exactly the same mosaic effect, whatever the temperatures. And you can see that we increase the size of this ordered domain from roughly 10 nanometres to 30 nanometres here. And if you remember, we had for this material also performed neutron diffraction experiments that allow us to evaluate the size of this domain. And for this four materials, we have a very good agreement between TEM and NPD.
So I wanted also to highlight that you could say these materials were obtained 1000° first and then low temperatures, in given conditions. It's perhaps specific, yes. It's also this technique that is used by microscopy. So here I wanted to highlight another result that we obtained also very recently in collaboration with LRCS, Arnaud Demortière.
And these platelet particles here, LNMO, were obtained in different conditions in molten salts a thin platelet, in order to be able to use the 4D-STEM analysis and to do cartography on the platelets in order also to be able to localise ordered and disordered domains. And you can see for this particles here we have a link to this materials here very close to the stoichiometry and as shown here also in blue by this lithium MAS NMR spectra, we have a single signature for lithium in this material in good agreement with the almost perfect stoichiometric material. And here, for this spinel that shows already defects ideal for disordered manganese-rich, we have a much more heterogeneity in this particle with here disordered domain and here ordered domain. So just to conclude this part 2, I wanted to say that in this material we are not in disordered materials and they behave very well in the battery electrochemistry upon long-run cycling at high C-rates, but we are in materials where we have this mosaic effect of ordered domains and as you can see, percolated by disordered small domains.
And we think that both contain in fact manganese 3+ and that it's source of mixed balance, and it's in fact the key for optimised electrochemical properties. Of course, we still have manganese 3+ in this material, but you can see a very small amount of manganese 3+ versus the composition reported in literature. We have also a small specific surface area versus the electrolytes. So of course, the next step for this study is really to test this material in full cell at 55° in order to check for the stability.
And I wanted also to highlight after this first part that here I have also illustrated the necessity for implementing spatially resolved techniques such as 4D-STEM in order to come to these answers here. So then I will move to the sodium-ion batteries technology, and I wanted to say that, in fact, the motivation for the development of this sodium-ion battery technology, of course, is the localization of the resource, I would say not equally in the world, and especially more in America, in China and in Australia. And you can see it is in fact produced in a very limited number of countries, with 95% of the production controlled by a few number of countries. And you can see also that we have, as we know, a lot of pressure sometimes on these lithium carbonate and nickels.
So in fact, as you know, in our lab and others, we are all working a lot now on sodium-ion battery technologies that can be really an alternative sister technology for lithium ions, despite intrinsically a smaller energy density. But it is also attractive due to the fact that in fact, here we have this raw material abundance. You can see that here. So the more it is orange, the more abundant it is.
If we compare sodium to lithium, it's much more abundant. If I remember, three times, yes, three orders of magnitude. And the other good point is also that sodium does not form any alloy with copper, so we can really use also aluminium at the negative electrodes. So it means that we replace copper, that is less abundant, by aluminium, that is more abundant.
And of course, we have also sodium in Earth crust, but also, of course, in seawater. So in fact, this technology is based on much more abundant materials. Of course, it's a complex system. I will not go into detail in this technology.
We have at the present time three main chemistries developed on the positive electrodes. We have also a lot of work that is done on this hard carbon in order to study the best structure for this graphitized nano domain for hard carbon. And also a lot of work, I would say, especially to do also on the electrolyte due to the key role of this electrolyte for the behaviour of the sodium-ion batteries. And so again, a lot of research was done on these materials in order to develop new materials at the positive electrodes.
You can see it's also interesting to highlight that in fact here you have the energy density, the specific energy and here you have different chemistries. At the top, you have the lithium-ion batteries and here, I would say, the sodium-ion batteries. You can see that versus half-carbon here or versus different negative electrode. You can see that for some of these chemistries, we can really compete with, for instance, the LFP or the LMO versus graphite lithium-ion batteries as an alternative.
So the growing activity around sodium-ion batteries, I would say, even at industrial levels. And today I will give some information about these materials and about new chemistries that we prospect beyond I would say these materials. I will not go into detail also. This is just to show you how we bring this material from the lab to the introduction in a battery that is now developed by Tiamat.
So this material is very interesting. It's a 3D material. You have here this P octahedral unit and you have here along this P octahedral unit fluorine, and here in the square of this octahedral, oxygen. And we can play in fact with the oxygen fluorine stoichiometry in this material.
You can see that ordered chemfistries are developed layered by and Prussian blue. So here I wanted also to highlight that here also the solid-state chemists play a big role in the development of these materials because you can see that in the literature we had all that kind of electrochemical signatures and I would say even more this kind of electrochemical signature for the material I will call NVPF in the following. You can see that in fact, here we have very different content in oxygen and fluorine in these materials, and that this signature is really the one of any 3-D2PO4 2 times F3. So only fluorine here.
And what happens when we have this electrochemical signature here? In fact, we are moving from this P octahedral unit, vanadium 3+ rich fluorine along the direction of the P octahedral unit and we increase here the oxygen. We have oxygen defects in this material, and we do not control the fluorine stoichiometry. And in fact we are oxidising vanadium 3 to vanadium 4 and in fact creating these vanadium type bands here. And here I wanted also to highlight that in order to really detect these defects and to be sure of the stoichiometry in oxygen and fluorine here, the phosphorus solid-state NMR is a really good probe.
You can see that when you increase the oxygen in this material, you have more NMR signals in good agreement with the fact that when you do not have oxygen, your phosphorus here is only surrounded by vanadium 3+ octahedra. And the more you have oxygen, you increase the number of vanadium 4+ octahedra around it and you change the transfer, the electronic transfer from the vanadium to the phosphorus. And so it means this shift from high shift in NMR to diamagnetic shift in NMR. And I wanted also to highlight that this technique also here allows to really detect very small defects in this materials here, you can see here.
So already at that time in collaboration in the frame of the Alistore network, we were able here in full cells versus carbon with collaboration with the group of Maria Palacin and Patrik Johansson to demonstrate that this material was very good at high rate. You can see here in the lab prototype, we had almost no loss of capacity moving from slow rate to high rate here. And I would say that we were able already at that time to improve the performance, changing the chemistry of the electrolytes. So then we merged efforts with CEA-Liten in order to do the upscale of this material and especially to be able to promote the formation of this material with a control carbon coating at the surface.
The homogeneous coating is the key of the performance of this material. Of course, it brings the conductivity within the material, within the electrode, but it brings also the stability of the CEI and especially also versus vanadium dissolution. And we were able to upscale this material. And also we have shown that playing with the synthesis, we can really play with the microstructure of this material, moving from nanosphere here to platelets and sand roses here.
And depending on the microstructure, we can also play on the electrochemical performance. And we have shown very recently that by mechanical grinding, we can really increase the tap density of this material also with, I would say, maintained performance in that case. So the first prototype was performed by CEA in order to demonstrate the interest of the technology. And you can see that it is interesting here when we are in a prototype for power prototype here, you can see in blue that we can really discharge the battery at very high rates and charge here also in 6 minutes the battery.
And we had demonstrated also at that time a very long cycling for this material. So stability upon cycling. So just a few slides on this technology that I obtained with the courtesy of Tiamat. So for this technology, you can really do a fast charging, high-power density and superior safety of this material.
We can really store the battery at 0 volt without damaging the performance, and no fire, no thermal runaway, long lifetime and based on abundant materials. So here it was a work performed, I would say, from the material proposed first by the solid-state chemists brought to maturation and now integrated in a technology that is developed by the industry. So I will finish by highlighting two recent results obtained in our group, one on manganese and substituted NVPF and one on new NaSICON materials.
Recent Advances
So in order to show you that solid-state chemists are also, I would say, a force of proposition for new materials again for this sodium-ion technology.
So in fact, Gaël Minart funded by the ANR programme developed during his PhD the new materials, so same structure as NVPF and in fact the same materials. But as you know, we really want to promote to maintain the structure but with more abundant materials and he was able through this topochemical reactions we had used a few years ago to synthesise this material, NVPF oxyfluorine. In fact, by topochemical reactions here, these are using this precursor vanadium-rich, in unique liquid media as source of fluorine and also as media where we can do a thermal, a solar thermal reaction at low temperatures, in order to really use this topochemical reactions with an exchange between-- with integration of sodium and fluorine structures and vanadium reductions and formation of these NVPF materials. So we were only able to substitute manganese in these structures with exactly the same approach it means to prepare this precursor here.
And we were able to substitute up to 25% of manganese for vanadium in these structures. So in fact, we were also able to develop a new series of materials with a variation of composition in manganese and vanadium, but also in fluorine and oxygen and also in sodium, as you can see here with the lattice parameters versus the composition in sodium in this material. You can see here a series of materials. And interestingly, these materials here behave very well.
You can see we have very small polarisation, very good reversibility of these materials. We have activity of vanadium and manganese in this material. So it means no loss of capacity versus the vanadium only material, very good behaviour at high C-rates and upon long-run cycling. Of course, it's preliminary tests in half-cells, so these could be interesting to continue for better maturation.
So then I wanted also to highlight these results concerning these new NaSICON material that we recently published. So this work was performed by Sunkyu Park in Tiamat. And I would say starting from a very old material. You know this NaSICON material Na3V2PO4 3 times in that material, you have vanadium 3+ material.
And this reaction was this material behaved very well. We have a flat plateau. We have the exchange of two sodium per formula unit and one electron per vanadium, so vanadium 3, vanadium 4. And you can see that we have very good performance at high rates.
So already demonstrated. But in fact a colleague, Piero Canepa, did calculations on this material and he told us, "Here I think we have an intermediate phase and it's in fact Na3 Na1, it's not a B-phase reaction, but two successive B-phase reactions. So of course, we were interested in order to prospect these reactions and first we prospected literature and we saw that we had some already intermediate phases in some results reported by other colleagues, but not mentioned and not analysed. And in fact, we did this experiment at synchrotron Alba, and as you can see here, when we move from Na3 to Na1, we have intermediate phased Na2.
That is observed when we are in non-equilibrium conditions. And you can see also that we do not have a symmetric behaviour between charge and discharge. And this Na2 phase could be isolated in synchrotrons, but could not be isolated when we recover the material from the battery. It's a metastable structure.
And despite that, of course, this structure could be determined. Of course, we were interested to stabilise and to search more around this material Na2V2PO4 3 times. It's interesting because here we have a mixed balance for vanadium 3, vanadium 3, vanadium 4 and we have only 2 sodium per formula units, so I would say efficient sodium. And in fact, we did this chemistry here, a mixture between Na3 and Na1, and we did it in temperatures in synchrotrons and you can see that we moved from 2 phases here to a single phase, but when we cool down, on the contrary to what we had in the battery, we have and we maintain a single phase that is stable at room temperatures.
And we did this for different compositions and mixtures between Na3 and Na1, and we were able to stabilise like this a series of materials. Why are these materials interesting? In fact, they are interesting because on the left, you can see here the conventional electrochemical signature of NVP, Na3,V2,PO4 3 times, a flat plateau at 3.45 Volt. And you can see here the signature of the new NaSICON material, Na2V2PO4 3 times, a much higher working voltage at 3.7 Volt. It means +10% in energy density.
We have a solid solution versus a plateau. We have much less volume expansion upon cycling, 6 versus 8%, and we are able to play between Na2 and Na0 that was obtained for the first time. And as I mentioned, we can even have full new families of materials of NaXV2P04 3 times, X varying between 1.5 and 2.5, with very different electrochemical signatures depending on the content initially in the structure. Of course, the Na2 is the most interesting.
So then I wanted to conclude this part by highlighting that it's only the characterization in depth of the materials, I would say from the local to the long-run structures and knowledge of defect that allows you to understand the materials. We have also a lot of non-equilibrium phase and mechanisms that are observed. And so for that we need the development of in situ cells and we are really still able to identify original mechanisms, new families of materials of interest for sodium-ion batteries. And I will conclude my talk by giving two perspectives to say that very recently, we have studied all these mixed anion chemistries, here with phosphate, tavorite structures as you can see here for lithium, NVPF structures for sodium and KTP-type structures for potassium-ion batteries.
So these chemistries, not only the cation chemistry, but also the anion chemistry, are really a source of new materials for batteries, because depending on the composition in anion, you can really play on the voltage and on the electrochemical curve, I would say the signature, and you can play on the potential of your batteries, here lithium, sodium or potassium.
Conclusion
And I will finish this really by mentioning also that it's worth, of course, to consider, I would say, much more abundant materials such as iron polyanionic materials, and especially here I focus on sulphate and sulphate phosphate materials. We found new phases recently and, of course, the challenge here is really to stabilise these materials and to optimise these materials in order to have very good performances. But they are very interesting, because here we have sodium, iron and sulphur or phosphate systems, so only abundant elements.
And then I'll thank you for your attention and thanks to all my colleagues from ICMCB, Dany, Jacob, Emmanuel and François, from LRCS, Christian, Jean-Noël and Arnaud, and from the large-scale facilities, Antonella, François, Lorenzo and Emmanuel. And of course, I would say all my best acknowledgement to Ilia, Gozde, Sunkyu, Gaël, Romain and Anastasia. You saw either longer or briefly their results, and you all for your attention and all the funding. Thank you all.
Yes, thank you very much.
AI-Generated Content Disclaimer
jsondata
| type |
| |||||
|---|---|---|---|---|---|---|
| uuid | "7d5ef82c-c31c-49dd-99a0-df7439fe21dd" | |||||
| label |
| |||||
| description |
| |||||
| start_date | "2024-11-28" | |||||
| end_date | "2024-11-28" | |||||
| url | "https://battery2030.eu/news/happenings-events/battery-2030-excellence-seminar-28th-of-november/" | |||||
| name | "Battery2030ExcellenceSeminar" | |||||
| keywords |
| |||||
| organizer |
| |||||
| associated_ou |
| |||||
| project |
| |||||
| image | "File:OSW808f1f524f2d4d7c97e21a70fb56e88e.png" |