1Mr Administrator,
Dear Colleagues and Friends,

2Although I have already delivered an Inaugural Lecture in this same room in 2010, as the Annual Chair of Sustainable Development – Environment, Energy and Society, I can assure you that the exercise is still just as intimidating. Allow me, however, Mr Administrator and my dear colleagues, to thank you for your trust and for honouring me by welcoming me once again in your midst in this prestigious institution, as Chair of Chemistry of Materials and Energy. The creation of this Chair marks an important time in the history of the Collège de France. This is the first time that it hosts three Chairs of chemistry. I am delighted at this initiative and would like to thank Marc Fontecave and Clément Sanchez for initiating it. Three Chairs are clearly needed if we are to enhance the image of this science. While it is occasionally in the headlines, it is still all too often disparaged despite the considerable advances it has afforded in health, agriculture, textiles, food and many other fields. The International Year of Chemistry (IYC) in 2011 certainly did restore the prestige of this wonderful science, yet an enormous challenge remains, which I will gladly take up in this renowned institution driven by the mission of spreading “research in the making”.

3Chemistry is comprised of various components, including, amongst others, organic chemistry, which studies the construction of molecules, biochemistry, dedicated to the molecules of life, and solid-state chemistry, more specifically concerned with the synthesis, structure and properties of solid materials. My teaching and research at the Collège de France will focus on the latter, which I will now discuss.

4France has played a crucial role in the rise of solid-state chemistry since the 1950s, mainly thanks to Paul Hagenmuller and some of his students. As some were my teachers, I can only be a proponent of this discipline, which has already been taught at the Collège de France by eminent professors such as Jean Rouxel, who left us too soon, followed by Jacques Livage. Both made significant scientific advances and inspired me.

5In this Inaugural Lecture, I will first describe how solid-state chemistry, initially a series of recipes similar to culinary art, has evolved over time, especially following the nineteenth-century scientific discoveries, to become a fully-fledged science of materials and its transformations. Drawing on a few examples, mainly from my own research, I will then show how, through experiments, chemists can reveal the secrets of matter and better understand it, so as to process it systematically and create new efficient materials to store energy.

6The development of humanity has always been linked to that of materials and chemistry. Take two examples from the field of solid-state chemistry. The first concerns the synthesis of high purity silicon (Si) rods in the sixties, which has led to the current boom in microelectronics and informatics. The second relates to the development of optical fibres in the seventies. These light conductors, which convey information to your TV, your phone, and your computer, have revolutionized the way we communicate and interact.

7What challenge facing today’s society does chemistry and particularly solid-state chemistry need to tackle? All global governance institutions¹ agree that the primary challenge for our planet over the next fifty years will be energy, as we will need to double production whilst limiting CO₂ emissions. Great hope is therefore put in the replacement of fossil fuels, which are nearly depleted, with renewable energies from the sun and other natural sources. However the success of this energy transition will depend on our capacity reliably to manage, convert and store these renewable energies, on a large scale and at a low cost. This management of energy capture and use involves a wide range of collection and conversion technology, all of which is limited by the lack of suitable materials. In fact, materials are crucial. The US Department of Energy clearly observed this back in the sixties – “technology is always limited by the materials available” –, and it still holds true today. In this context, the “Chemistry of Materials and Energy” coupling is therefore vital for both fundamental and applied research, to meet the challenges surrounding energy: its collection, conversion and storage.

The chemistry of our ancestors

8There is however nothing original about this “chemistry of materials and energy” coupling. According to archaeologists it dates back to Antiquity, as excavations have uncovered an object from the third century bce referred to as the Baghdad battery (Fig. 1).² This is a hollow piece of pottery that may have contained an acid solution used as an electrolyte, and within which an iron rod and a copper cylinder were plunged. Although this type of structure has been shown to produce low current, there is still controversy surrounding the use our ancestors may have made of the object.

Figure 1. Sketch of the three pieces that make up the Baghdad Battery.

9Irrespective of its use, this discovery is proof that our ancestors mastered the art of pottery and knew how to produce iron and copper through processes that date solid state chemistry back to the origins of civilization, as we shall see in what follows. Potters’ work is probably the best expression of this link. When ancient potters turned a mix of clay and water into a paste, which, once shaped, was fired to make pottery, they were already exploiting the ceramic process that is used so extensively nowadays in solid-state chemistry for preparing materials. Moreover, by combining fire and pigments, they knew how to enhance the aesthetic nature of their potteries. At the time smelters shared this command of fire to prepare metals. They did so by reducing metals at high temperatures without knowing their formula, but identifying them by their colour. They produced copper using malachite, and obtained bronze by mixing malachite and cassiterite (Fig. 2). Over time, through their command of kilns, craftsmen were able to reach very high temperatures and thereby expand the range of metals obtained. Thus over the centuries, the Stone Age led to the Copper Age, then to the Bronze Age and the Iron Age and even to the Silicon Age, attesting to the close relationship between the development of humanity and that of materials.

Figure 2. The origins of solid-state chemistry: preparing metals using minerals in Antiquity.

10Whether driven by beliefs, religion, the search for identity, the quest for artificial beauty of the body or other motivations, the Ancients used minerals, sometimes of different colours, for cosmetic purposes. Cinnabar (HgS) was used as lipstick despite its toxicity. Likewise, black from stibnite (Sb₂S₃) and blue from azurite were used as eye shadow. The scarcity of certain highly prized pigments, like blue, prompted the Egyptians to produce them. The first artificial pigment, “Egyptian blue”, was synthesized 5,000 years bce. To do so, the Egyptians used a wise selection of precursors composed of red native copper shavings (Cu), sand (SiO₂), sea salt (NaCl), and water which, once shaped into balls, were heated for a few days at high temperatures in an atmosphere controlled through a careful choice of fuels, so as to obtain the desired blue. By some irony of fate, this jealously guarded recipe was lost over time. It was only in the early nineteenth century, with the rise of modern chemistry, that the chemical formula of the compound obtained (CaCuSi₄O₁₀) was identified, allowing for the restitution of the chemical reactions involved in the synthesis. Other red or black pigments were subsequently obtained through heating, respectively in air and in a reducing atmosphere, from a yellow mineral: we now know that these are Fe₂O₃ and Fe₃O₄, prepared from FeOOH. Apart from these processes involving firing treatments, ancient writings show that our ancestors also mastered solution-based synthesis (in other words a sort of soft chemistry) of pigments, particularly with the synthesis of lead minerals (Pb) such as phosgenite Pb₂CO₃Cl₂ and laurionite PbClOH, which can also be used for medical applications (collyrium).³

11Learning by trial and error, our ancestors thus developed reliable recipes for creating pigments. Even though they did not understand the underlying science, their creative power and their ingenuity produced recipes admired by today’s researchers.

Solid-state chemistry: from recipe to science

12The shift from our ancestors’ recipes to modern chemistry, governed by scientific laws, was the fruit of long-term work. It was not facilitated by alchemists, who associated magical chemistry, mysticism, and spirituality, and whose overriding intention was to transform all metals into gold. To prepare their materials, alchemists combined the four elements – fire, air, earth, and water –, associating each one with a Platonic solid. They perpetuated empiricism in chemistry, exposing themselves to fierce criticism, as is shown by an edict of the Roman Emperor Diocletian declaring that all chemistry and magic books should be burned. It took over 2,000 years to move from these four compound elements to the chemical elements of which they are comprised, and then to the 33 elements proposed in 1789 by Antoine Lavoisier, who is universally recognized as the founding father of modern chemistry. Lavoisier was never rewarded for it during his lifetime, as he was guillotined while uttering the following words, a last plea for more time to finish his experiments underway: “The Republic needs neither scholars nor chemists”. His work, however, triggered an effervescence of works in all aspects of chemistry.

13The race to identify new chemical elements had begun, and I cannot help here but mention former Collège de France professors who actively contributed to it: Nicolas Vauquelin, who discovered beryllium and chrome; Louis Jacques Thénard, who isolated silicon and discovered boron; and Antoine J. Balard, to whom we owe the discovery of bromine. Nevertheless, the fact remains that during this same period the discipline of solid-state chemistry was still lagging behind, despite the work of a few pioneers who laid the foundations of crystallography. Jean-Baptiste Romé de l’Isle introduced the laws of symmetry and of constancy of interfacial angles deduced from the observation of crystals’ habit; in 1813, in his Treaty of Mineralogy, René Just Haüy proposed the concept of “crystallographic lattice”. However this discipline lacked analytical techniques, which limited knowledge of the structure of matter. The US chemist Harry Clary Jones summarized this frustration in his 1913 book, A New Era in Chemistry: “We do not know what is the formula of solid sodium chloride or rock salt; or of solid water or ice; and we have no reliable means at present of finding out these simplest matters about solids. Our ignorance of solids is very nearly complete”.

14This was the last time such an assertion was made, thanks to two successive discoveries. The first was Wilhelm Röntgen’s discovery of X-rays, which made skeletons visible. He demonstrated this by X-raying his wife’s hand rather than his own, wary as he was of the radiation’s secondary effects. The second was the work of Max Von Laue who, in 1912, illuminated a nickel crystal with an X-ray to obtain the first spotted diffraction pattern. For the first time, this revealed the presence of order and symmetry in the group of atoms that makes up crystal. The X-ray diffraction technique thus saw the light of day. It was soon exploited by the Bragg father and son, who not only established its fundamental rules, but also solved the structure of hundreds of compounds, showing how atoms are organized within compounds familiar to us, such as rock salt (NaC1) and many others. For example, and to provide an order of magnitude, a 1mm³ grain of rock salt contains as many atoms as there are stars in the universe. At last, science had made progress in its understanding of atoms’ positions, but knowledge of the forces binding them was yet to be discovered.

15During the same period, atomic theory, based on the quantum approach to infinitely small particles, was developing and finally solved the mystery of the atom’s structure. The electron’s place was identified within orbitals whose shape and electronic density were determined by the resolution of the Schrödinger equation. This allowed Linus Pauling to discover the nature of the chemical bond connecting two atoms and to introduce the notions of “orbital mixing” – called hybridization in chemists’ lingo – and “electronegativity”.

16One more step remained, however, for solid-state chemists: knowing what happens to the electron with the shift from the molecule, which is comprised of two atoms, to the population of ordered atoms that makes up crystal. The answer was found by extending the quantum approach adopted at the molecular level to crystal, through judicious approximations. The final result was the formation of a continuum of energy levels called bands, capable of hosting the electron and establishing an electronic structure for each compound. This structure constitutes the compound’s fingerprint, and is instrumental in determining its physical properties (electric, magnetic, etc.).

17Building on this earlier work (Fig. 3), solid-state chemists were then able to associate composition, structure, and properties, and make predictions. New compounds could be designed on the basis of structural affiliation by following a deductive reasoning. The relations between structures and properties allowed for predictive models to be built, and to confer specific electric, magnetic or electrochemical properties to a material. Solid-state chemistry thus enjoyed a euphoric revival in the fifties, despite the remaining unknowns. This trend was fuelled not only by the strong interactions that had developed with solid-state physicists who were fascinated by materials – as my theoretical physicist colleague Antoine Georges brilliantly expressed in this very room, during his inaugural lecture in 2009 –, but also by an industrial context that was highly conducive to the design of new materials with remarkable physical properties, in a world undergoing a major technological revolution. In the mid-twentieth century, solid-state chemistry thus became a science founded on the properties-structures relationship, allowing for the development of new materials with specific functionalities.

Figure 3. The four key advances that revived solid-state chemistry.

The chemist’s tool-box

18Now that we have seen the foundations of solid-state chemistry, let us look at how chemists conceive of their materials. The periodic table (Fig. 4) is to the chemist what the easel and pastels are to the painter. They prepare new compositions based on its content, though certain elements cannot be used because of their non-reactivity, limited abundance, cost and toxicity. Just as a slide rule could be used to perform operations, so this periodic table allows chemists to make an informed choice of elements to combine, provided they understand the information it hides. This is qualitative information concerning the variation of the atoms’ radius, amongst other things, which increases from right to left and top to bottom, not unlike electronegativity, which is linked to the relative position of the energy bands associated with each element and which varies in the opposite direction. Electronegativity is of crucial importance, since it governs the ionic-covalent nature of the metal-ligand bond, on which the physical properties (magnetic, electrochemical, etc.) of a material heavily depend. With this information, chemists can control the chemical bond and thereby carefully choose the appropriate elements to design new compounds.

Figure 4. The periodic table: the chemist’s tool-box.

19For solid-state chemists, the conceptual design of a new compound therefore amounts to a game of bowls (the atoms) and bands (the chemical bonds), following certain rules. In the case of an ionic model, these include the “tailor’s rule”, which takes into account the respective size of the cations and anions. It indicates whether the cations, depending on their size, will prefer tetrahedral or octahedral sites, with large cations preferring octahedral sites. Now, if the anion’s size is reduced, chemical pressure is created which confines the cation in a smaller volume, and may promote a higher state of oxidation. Hence, 3D metals generally present higher oxidation states in oxides than in sulphides, as sulphur is larger than oxygen. The respective size of the anions and cations, as well as their interactions, govern the stacking of the polyhedra, which may share vertexes, edges or sides. The final arrangement produces structures with a symmetry that makes them similar to works of art.

20The “tailor’s rule”, for example, explains natural phenomena linked to the evolution of the structure of the silicates within the Earth’s mantle, a very important problem for our geophysicist colleagues. Under the effect of pressure, in other words, as it moves towards the centre of the Earth, the pyroxene structure containing SiO₄ tetrahedra turns into a perovskite structure containing SiO₆ octahedra connected by vertexes, and then into a post-perovskite structure, where SiO₆ octrahedra are linked by edges and vertexes. In this specific case, note that while Si⁴⁺ and O²⁻ both contain 10 electrons, since oxygen is much larger than silicon, it is especially affected by pressure, thereby leading to an increase of the cation/anion ratio and a shift in its coordination number from 4 to 6.

21With these notions of atomic sizes, solid-state chemists can play their favourite game, cationic or anionic substitution, to change the material’s properties. Let us take a very obvious example: zirconia (ZrO₂) as a substitute for yttrium. In this case, for reasons pertaining to electroneutrality, the substitution of 2Zr+4 by 2 Y+3 in ZrO₂ leads to oxygen deficiencies in the anionic network. These deficiencies create an empty volume that facilitates the movement of oxide ions O²⁻ from one site to another and creates compounds with high anion conductivity, used as membranes to transport oxygen anions in fuel cells.

22These substitutions are sometimes so minute that they are described as doping, and have equally spectacular consequences in terms of optical properties. Consider two structures, alumina (a white powder) and beryl (transparent material), each containing Al³⁺ in octahedral sites. Replacing a few ppm of Al³⁺ with Cr³⁺ causes a drastic change in these compounds’ colour, producing ruby in the first case, which is red, and emerald, which is green, in the other. This doping breaks the local symmetry of the crystal, which translates into a different ionicity for the Cr-O bond, and of course modifies the optical spectrum. With this example, note that the defect often becomes dominant in steering the optical, electrical or magnetic properties of a compound, thus affording chemists a degree of additional creative freedom.

23Aside from the atoms, chemists can also play with band structures, which they know how to handle with respect to the electronegativity of the elements they use. Take for example an insulating compound for which the d and p bands are far apart. If the structure is maintained (in the case of a rigid model), the replacement of the sulphur with less electronegative elements, such as selenium (Se) and tellurium (Te), results in the p band penetrating the d band. Moving right or down in the periodic table, the reverse occurs: the d band penetrates the p band. The two processes cause the material to shift from a semi-conducting behaviour to a metallic or even superconducting behaviour.

24Let us now look at how solid-state chemists prepare the materials they imagined conceptually. Like potters and smelters in Antiquity, they use the ceramic method, which requires high temperatures and involves repeated grinding and heating sequences – hence the name of the method, shake and bake, in English-speaking countries. In solid form, reactivity is governed by mass transport; atoms must move to combine. This movement is governed by a scattering law from which, by the way, it can be deduced that it will take an ion 320 years to travel a distance of 1mm at 25°C, whereas it will take only eleven days at 900°C; hence the value of high temperatures. Similarly, pressure encourages contact between grains, affording greater reactivity and therefore a smaller reaction time. Ultimately, the laws of thermodynamics prevail and allow chemists to produce the most stable material.

25Though it requires a lot of energy, for centuries this high-temperature chemistry was solid-state chemists’ favourite method to prepare most compounds studied until then, whether they were substituted or not. It was ultimately a societal issue, once again, namely the 1973 oil crisis, which spurred its diversification by leading solid-state chemists to address the need to find a solution to the energy cost of materials. In France, this demand led to the development of low-temperature chemistry, now more commonly known as “soft chemistry”. Jacques Livage and Jean Rouxel, former Collège de France professors, were its pioneers, with the one using molecular precursors⁴ and the other solid precursors.⁵

26What is soft chemistry actually? It relies on the principle of topotactic reactions, in other words, reactions that preserve the structural skeleton of the initial precursor and therefore do not break chemical bonds, which explains the low energy input needed to trigger these reactions. The most widespread – those underpinning the functioning of Li⁺ ion accumulators – are intercalation reactions. Similar to millefeuilles, they consist in inserting lithium ions between the layers of the host structure (Fig. 5). Through grafting reactions, these host structures can also fix organic monomers, in order to obtain thicker millefeuilles with applications in the domain of catalysis or as flame retardant or even intumescent materials.

Figure 5. Insertion reaction governing the functioning of lithium-ion accumulators: analogy with the millefeuille.

27It was during this period of profound change in the early eighties that I began my scientific career. High-temperature solid-state chemistry was taking a major turn towards soft chemistry processes, as the second oil crisis intensified the need to reduce the temperature of materials preparation. Reading journal articles, the words of one of our eminent solid-state chemists, Robert Collongues, stayed with me: “Is the chemist’s role not to transform matter to tailor it to our needs and make it a useful solid?” This question was to guide my research, which I devoted to magnetic, superconducting, ferroelectric, optoelectronic and redox materials.

28Although all these materials are fascinating, in what follows, I will limit myself to those of interest for energy, whether it be to save or store it. These first include superconducting materials with no resistance below a certain temperature, called critical temperature (T_c), which gives them the advantage of transporting electricity without loss due to the Joule effect in this temperature domain. They can thus prevent the 11% energy loss due to the Joule effect associated with the transport of electricity by our high voltage lines. These materials also include electrode materials designed to store energy in chemical form and convert it into electricity through a system that we call a battery. The choice of these types of materials is not random; it is informed by the synergy between them, as the same compound can serve as both an electrode material and a superconductor, hence the name bi-functional materials. Moreover, their properties are governed by redox reactions enlisting electron transfers.

29At this stage, readers with a taste for concrete facts may lose patience and rightly ask what exactly these materials are, and how they are developed and characterized. This is the question I will now try and answer, drawing on three examples from my research activities, namely:

• superconductors seen through the “soft” approach, so as to show the versatility of this chemistry for the synthesis of new materials;

• lamellar oxides designed for energy storage, in order to illustrate the benefits of the synergy between fundamental and applied research;

• the development/design of new electrode materials in the framework of sustainable development, to describe the adaptive aspect of solid-state chemistry.

Superconductors

30The first classes of bi-functional materials that caught my attention include the Chevrel⁶ phases of general formula M_xMoX₈, named after the French scientist who discovered them in the seventies. These are compounds built from Mo₆S₈ clusters, which delimit tunnels within which copper atoms sit, and which lend themselves to a varied soft chemistry involving either the cations (M), or the molecular entities (Mo₆X₆), or both. Copper (Cu) can for example be extracted from this compound through chemical oxidation in the presence of an oxidizing agent (I₂), to produce an open frame called a host material for: i) the electrochemical insertion of various cations and the development of Li, Na or Mg ion accumulators, and ii) the absorption through reaction in a gaseous state of highly volatile metals to obtain lead and tin phases and even thallium (Tl) phases, which had never been prepared and proved to be the superconductor with the highest critical temperature T_c. This attests to the bi-functional nature of these compounds, which can be used as both electrode materials and superconductors.

31Cu⁺ can also be driven out of the Chevrel phases and turned into metallic Cu through electrochemical reduction by lithium, which allowed us to reveal a new reaction mechanism and envisage the fabrication of electrodes relying on displacement reactions. But these reactions still needed to be totally reversible, which was not the case. It was only eight years later, following my collaboration with Jean Galy and Patrick Rozier, that we found the ideal material. This is the lamellar phase Cu_2.33V₄O₁₁ containing Cu⁺ and Cu²⁺ in the interlamellar space (Fig. 6). The electrochemical reduction of this crystallized compound by lithium led to a massive displacement of the copper, as attested to by the appearance of particles looking like an octopus, with the dendrites of the metallic copper (Cu) for tentacles. Most spectacular, however, was the total reabsorption of these dendrites during the oxidation stage to return a perfectly crystallized compound, reflecting the total reversibility of the displacement reaction. Thus was born a new reactional concept for the elaboration of electrode materials.⁷

Figure 6

32The possibility of using the Mo₆X₈ cluster as an elementary building block is another asset of the Chevrel phases. The condensation of two, three, four or even n blocks allows for example for the preparation, at moderate temperatures (700-900°C), of compounds of general formula M_xMo_6+3nX_8+3n with a degree of one dimensional character which increases as n increases. The last of the series is the mono-dimensional compound M₂Mo₆Se₆, formed of chains of Mo₆Se₆. While this Lego chemistry was widely practised in inorganic synthesis, its feasibility at low temperatures remained to be proven. We reached this next stage when we understood that the phases consisting of negatively charged linear chains strongly shielded by cations could be conducive to exfoliation. In the presence of highly polar solvents with a high dielectric constant, it was possible to solubilize them to obtain gels or, in the case of diluted solutions, unique fully solvated chains of Mo₆S₆ with a 6 Å diameter, which behaved like real mineral macromolecules or inorganic polymers, a breakthrough at the time.⁸ Most exciting though was the possibility of recondensing these highly reactive fibres around different ions, by adding diverse alkaline salts, to create new materials. Ultimately, through this set of reactions, we had simply mimicked the reactional concepts of Jean-Marie Lehn’s supramolecular chemistry. Among these new phases, the compound Li₂Mo₆Se₆ proved to be an excellent electrode material due to its amphoteric nature. In 1985 it allowed for the first demonstration of the concept of symmetric lithium-ion technology (in other words using the same material for the two electrodes).⁹

33Although I have limited myself to cationic species until now, I would not like to mislead you into thinking that this intercalation chemistry is only a chemistry of cations. It can also involve anions, as is shown by the example of the superconducting cuprate family, the different phases of which share the common characteristic of having an oxygen non-stoichiometry. For reasons pertaining to electroneutrality, the latter controls the respective quantity of Cu³⁺/Cu²⁺ and, thereby, these materials’ superconductive properties. Steering away from classical high-temperature methods, colleagues in Bordeaux¹⁰ were thus able to demonstrate the possibility, through electrochemical oxidation in an aqueous medium at room temperature, of inserting oxygen in these materials, and thereby modifying the Cu³⁺/Cu²⁺ ratio. For the insertion of 0.15 oxygen per formular unit, this allowed for a critical temperature of 42 K to be reached for the La₂CuO_4.15 phase. This is an elegant approach showing how, through electrochemical insertion, chemists can precisely control the number of electrons injected and thus minutely and continuously follow the effect that the injection of anions or cations can have on the electric, magnetic, and other properties of various materials. In my opinion, this is an opportunity that physicists have not sufficiently explored.

34I will stop here with my first example; I hope that I have enabled you to understand that soft chemistry is an exceptional synthesis tool, as it offers the possibility of inserting cationic and anionic species in host networks, even of displacing ions, or of condensing clusters to stabilize new phases. Most encouraging of all, after thirty years in existence this chemistry still offers vast and original prospects for the synthesis of eco-compatible materials.

Lamellar compounds

35I will now discuss a fundamental subject relating to a concrete technological challenge, which I think will resonate with you as it pertains to materials found in your phone and laptop batteries. This is the LiCoO₂ lamellar phase, used in the first commercial lithium-ion accumulators, with a layered structure as pointed out earlier.¹¹ Although LiCoO₂ presents a theoretical capacity of 290 mAh/g, in practice only half, that is, 150 mAh/g, can be used, due to the collapse of the structure caused by interlamellar repulsion when too many Li⁺ ions are removed. To alleviate this structural stability issue, chemists successfully reverted to cationic substitution – their favourite game. Through the partial substitution of cobalt (Co) with manganese (Mn) and nickel (Ni) within the metallic layer, layered oxides with capacities of 180 mAH/g called NMC (acronym of Ni-Mn-Co) were obtained and are currently integrated in commercial batteries. The additional introduction of lithium as a partial substitute of cobalt within the metallic layer recently allowed for lithium-rich NMC compounds to be obtained with performances reaching up to 270 mAH/g. Although these developments strongly contributed to the development of Li-ion technology, they nevertheless raised many fundamental questions, first of all regarding the possibility of removing all the Li from the initial LiCoO₂ phase.

36Several approaches were available to us. Chemical oxidation using strong oxidizing agents (particularly Br₂) was unsuccessful, as it proved to be incomplete. Removing Li⁺ ions completely, however, was possible through electrochemical oxidation, which led to the CoO₂ compound with a lamellar structure, but with layers that had shifted, with respect to the initial material LiCoO₂ to minimize their repulsions. At this stage, it was important to know which redox pair we had activated within LiCoO₂ through the total removal of Li. Had we driven the Co³ to its higher degree of oxidation, as is customary, or had we oxidized the anionic network, a rather unusual situation in chemistry? As the Co³⁺ was not totally oxidized into Co⁴⁺, as determined by magnetic measurements, the latter option was preferred. We attributed this to the migration of the material’s band when the Li is removed, leading to an extreme situation where there are holes on the oxygen with the formation of a peroxo group (O₂)²⁻, the same entity found in hydrogen peroxide (H₂O₂), whereas water (H₂O) has superoxide (O²⁻). This was therefore quite a provocative scenario, in which copper (Cu) became more electronegative than oxygen (Fig. 7).

Figure 7. The band structure of layered oxides.

37This idea, certainly too daring for the community of researchers on batteries, was controversial, which was surprising as the chemistry of holes had already been brilliantly developed by Jean Rouxel for sulphides.¹² It had also been mentioned in the case of high T_c superconductors which, like CoO₂, have a high hybridization of the p and d bands. Moreover, it was on the basis of this similarity that, in an article written in 1999 in memory of Jean Rouxel, I suggested that the layered Co phases could be superconductive.¹³ Four years later, this was confirmed by Professor Tanaka’s group, which showed that the compound Na_xCoO₂.H₂O effectively became superconductive at 5 K.¹⁴

38Let us now look at how this proposition on the existence of peroxo groups in the compound CoO₂ – an idea that was controversial at first – was to be realized in the world of batteries fifteen years later, with the rise of the new generation of Li-rich lamellar oxides.¹⁵ These compounds, with a layered structure where Ni, Co and Li coexist in the MO₂ layers, have an enhanced capacity of 280 mAh/g, which was to allow us to envisage a 20% increase in our mobile phones’ autonomy. However our poor understanding of their enhanced capacity as well as the gradual drop in their average potential during consecutive charges/discharges were major obstacles to their commercialization. From a fundamental point of view, the challenge was therefore twofold. For these materials to be commercially viable, we needed to understand not only the origin of this enhanced capacity, which exceeded theoretical capacity, but also the drop in potential.

39As is often the case in science, the first stage of our work amounted to simplifying the material, in other words, to reducing its number of redox centres whilst preserving its millefeuille structure. Once again, we resorted to the benefits of chemical substitution by replacing the elements in the third period with those in the fourth period of the periodical table to prepare Li₂Ru_1−xSn_xO₃ lamellar phases. These, like the previous ones, have an enhanced capacity of 270 mAh/g with the major difference, however, that their average potential in relation to lithium remains constant during cycling. With only one redox (Ru) centre left, since tin (Sn) is electrochemically inactive, these phases are model materials for analytical studies seeking to understand their reactional mechanisms.

40Among the range of characterization techniques applied, two proved highly fruitful. X-ray photoelectron spectroscopy (XPS) first allowed us to identify, for highly oxidized compounds, alongside the presence of the O²⁻ anions of the crystalline network, the presence of another type of oxygen, less negatively charged, which we attributed to the peroxo group O₂ⁿ⁻. This attribution was confirmed by electron paramagnetic resonance (EPR) measurements – a technique that is used to probe the entire material and that is able to detect electrons unpaired to metals or to radical species. The identification by EPR not only of the presence of the peroxo group but also of its concentration allowed us to discover the reactional mechanism of Li insertion-extrusion in these Li-rich lamellar compounds, which amounts to a game of band structure overlapping. There is, first, the oxidation of Ru⁴⁺ into Ru⁵⁺, and then, due to the penetration of the d band of ruthenium in the p band of oxygen, there is the spill over of the electrons of the p band into the d band, creating holes in the p band associated with the formation of peroxo groups.¹⁶ At last, all this provided irrefutable proof that the enhanced capacity afforded by these compounds stemmed from the contribution of the anionic redox, thus confirming the idea proposed fifteen years earlier.

41This work opened up a new era in research on electrode materials, which is already proving fruitful. Professor Shinichi Komaba’s group has just reported a new lamellar compound which presents an exceptional capacity of 300 mAh/g and which also builds on the electrochemical activity of the anionic network. This compound also includes a redox element (Nb) in the fourth period, which was previously overlooked, as it was too heavy. We can thus see that this activity of the anionic network broadens the range of efficient battery materials. And so it is that often, in the world of research, new concepts benefit from the emulation created by internationalization, and open up new horizons when they reach maturity.

42Bearing in mind Robert Collongues’s words, however, we still needed to understand the reasons behind these minerals’ drop in cycling potential if we wanted to make them commercially viable. To get further insights into these issues, we turned to high-resolution electronic microscopy, a technique to visualize atoms of a size and distance of the order of the angstrom, i.e. 1/10 of a millionth of a millimetre. Up to this point, I have been talking about the layers and the place of the atoms as they were deduced by X-ray diffraction. Now, thanks to high-resolution microscopy, they can be visualized directly, with both heavy atoms appearing as two white spots due to their significant electronic charge, whereas the lithium is not visible given its low atomic number. The question then was a matter of knowing what happened to this arrangement when lithium was inserted into and then extruded from the material during electrochemical cycling. Unexpectedly, during the charge we observed a mass migration of the cations, which nevertheless returned roughly to their initial position upon the following discharge, thus confirming the reversibility of the system but providing no clues as to the drop in potential. What happens however about after 100 cycles (which, for a user, amounts to charging their battery 100 times)? For tin compounds (Li₂Ru_0.75Sn_0.25O₃) that do not present a drop in potential, the picture is very similar to the starting product. This contrasts with the Li-rich metal compounds in the third period (Li_1.2Co_xNi_yMn_zO₃), which present a drop in potential, since we can observe the presence of white spots indicating the migration of atoms in sites outside the layers. What are these sites? Looking at the crystallographic structure, we can see that they are tetrahedral sites. What happens, therefore, is that when the atoms migrate from the layer to the interlayer space during cycling, some of them stay trapped in the tetrahedral sites, which explains the drop in potential during cycling. In retrospect, this result can once again be explained by the “tailor’s rule”: due to its large size, tin does not get trapped in these sites, whereas manganese, which is small, does. Now that we have found the origin of the problem, we are in the process of developing customized materials to bypass it.

43In addition to revealing a major advance in the field of electrode materials, the aim of this second example was also to make you aware of: i) the perseverance needed for new concepts to be accepted, concepts which are especially challenging to the community in view of their subsequent importance; and ii) the benefits of resolving fundamental scientific problems emerging from concrete technological challenges. I think it would be wiser for our institutions to promote a complete synergy between science and technology, so as to answer rapidly the problems faced by society, rather than setting fundamental research in opposition to applied research, as is unfortunately often the case.

A progressive chemistry serving sustainable development

44This last example will show how the current strategy, founded on a progressive solid-state chemistry, allows for new materials to be developed and designed for energy storage in the context of sustainable development. I will first point out that a compound, in itself, is useless. It is only the material, which should be seen as the assemblage of a chemical composition, of a means of development and of a function that can be useful. In the case of sustainable development, this implies that: i) its composition must contain only abundant and non-toxic elements; ii) its production must involve only low-energy-consumption processes; and iii) its performance for the targeted application must be appealing (for example, with respect to potential in the case of electrode materials).

45How do chemists proceed to identify these compounds? There are two options. They can either use combinatorial experimental chemistry, a tedious approach with random results, or practise deductive chemistry. The second path is the one I chose and am describing to you now, which consists in: i) proceeding by analogy; ii) taking advantage of the strong structure-property coupling; and iii) drawing on the understanding of reaction mechanisms to make an informed synthesis. To illustrate this approach, I will take the electrode material currently most prized, LiFePO₄, the aim being to increase its potential, which is only 3.45 V. Based on the established electrochemical property-structure relations, we know that the potential is especially high given that the Fe-O bond is ionic. Based on the periodic table, this therefore means the phosphate entity needs to be replaced with the sulphate entity, with the concomitant addition of fluorine, which is more electronegative than oxygen. While the compound LiFeSO₄F met our criteria, it still needed to be synthesized, using low-energy means. Even though the aqueous medium is ideal, sulphates are soluble therein, meaning that we needed to find an alternative. Although they are used very little in inorganic synthesis, ionic liquids – which are none other than the salts that make up organic cations and anions, but liquid at room temperature – seemed to offer an interesting avenue.

46Through the reaction of a ferrous sulphate monohydrate with lithium fluoride in an ionic liquid medium, after heating at 280°C for 20 hours, a new phase of tavorite structure was obtained, which proved to be LiFeSO₄F. While the isolation of this new compound is certainly a progress, the most exciting was determining its formation mechanism, in other words, the key to the reaction. The solution is simple if we take into account the fact that the structure of the hydrated precursor and of the final product are almost identical, thus attesting to a topotactic reaction in which the oxygen of the H₂O molecule is replaced with an F⁻ ion with, furthermore, the addition of an Li⁺ ion to maintain charge neutrality. Once we had grasped this mechanism, we were able to generalize it, and pretty soon over 20 new Li and even Na and K fluorosulphates were obtained. Understanding reactional mechanisms is therefore crucial in chemical synthesis.

47In no way am I trying to suggest that in solid-state chemistry we control everything. In fact I will point out that if a reaction identical to the one described above is carried out without ionic liquid, the result is a compound with the same formula, LiFeSO₄F, but with a very different structure: a polymorph with a triplite structure. Solid-state chemistry is therefore a science at the crossroads between the expected and the unexpected. While the expected is of course intellectually pleasing, the unexpected is just as interesting, as it always opens up new perspectives. This is confirmed here, as this new polymorph presents rather exceptional electrochemical properties, particularly with a Fe³⁺/Fe²⁺ redox potential of 3.9 V, the highest ever obtained for ferrous inorganic compounds. LiFeSO₄F can thus favourably compete, in terms of performance, with the electrode material LiFePO₄. An extension of this work¹⁷, based on a controlled ionothermal synthesis and solid-state reactivity, allowed for a whole new class of electrode materials to be developed including, in addition to fluorosulphates: oxysulphates, hydroxosulphates and lithium-bearing sulphates, made of transition metals, which were unknown three or four years ago (Fig. 8). From an environmental point of view, these are attractive electrode materials as they are constituted of abundant elements, iron and sulphate, and are synthesized at temperatures below 250°C.

Figure 8. Comparison of lamellar oxides and polyanionic compounds in terms of electrochemical performance.

48At this stage we were faced with another challenge: how to know whether such compounds could be prepared at room temperature. Solid-state chemists generally turn to the chemistry of life to try and achieve this. Nature is full of examples of bio-mineralized materials, in other words materials precipitated at ambient temperature by microorganisms with highly efficient enzymatic/catalytic systems. The prime example relates to well-known unicellular algae, diatoms, which are able to concentrate the silicon contained in sea-water in order to create highly textured silica shells. To broaden the spectrum of biomineralizable materials, we turned to the use of other simpler microorganisms, such as bacteria, and even yeasts, which are unicellular fungi. Let us look at two examples.

49The first concerns the synthesis of LiFePO₄, a material that we tried to prepare by biomineralizing an iron sulphate and lithium phosphate mix in the presence of a Bacillus pasteurii-type bacterium. Its role here was mainly owed to the fact that its enzyme, urease, can hydrolyse urea to produce the medium basicity needed to precipitate LiFePO₄. The change in colour of the reactional medium, after 8 hours of slow stirring at 60°C, suggested the presence of a reaction. This was confirmed by high-resolution microscopy, showing that the bacterium is surrounded by a biofilm inside of which small fine needles can be observed, with its diffraction pattern indicating the presence of LiFePO₄. It took a long time for this approach, though elegant, to make it beyond laboratory curiosity, due to issues with reproducing and upscaling it.

50The bacterial approach can also be used to prepare textured materials, as we shall see with this second, simple but striking example, of the synthesis of the hematite α-Fe₂O₃.¹⁸ This time we use a ferro-oxidizing bacterium of the Acidovorax kind which, in the presence of a Fe²⁺ solution at 30°C, triggers the formation of nanoparticles of FeOOH that preferentially precipitate within the highly confined inter-membrane wall (40 nm). The precipitated nanometric particles (Fig. 9) are then arranged to form a shell matching the shape of the bacterium. Briefly heating it in air at 700°C eliminates the organic/living part of the compound, and the FeOOH is transformed into α-Fe₂O₃ hematite, whilst preserving not only the form of hollow shells but also hematite nanoparticles’ organization into stringers. Thanks to this original alveolar structure induced by the bacterium, these textured hematite samples display interesting electrochemical properties in terms of potential behaviour when they are used as electrode materials. In other words, accumulators designed with such materials could be charged and discharged 50 times faster than those which use non-textured materials.

Figure 9. The biologically assisted synthesis of textured electrodes.

51Although all these results are still preliminary, they demonstrate the value of biologically assisted syntheses for the elaboration of electrode materials. This eco-compatible synthesis approach, currently dependent on our capacity to control bacteria’s physiological conditions and to accelerate their reaction kinetics, is appealing for the development of specific lithium-ion accumulators with a small carbon footprint.

52I have reached the end of my third example, from which it is important to remember two crucial advances: first, the benefits of the ionothermal approach, in inorganic synthesis, with the discovery of a new family of low-cost and efficient sulphate-based electrode materials; and second, the possibility of preparing certain electrode materials with bio-inspired methods (biomineralization). These two aspects highlight the fact that solid-state chemistry is a highly adaptive science, which can meet societal demands in the framework of sustainable development.

53Although I discussed low and high temperature syntheses separately, I would like to point out, to avoid any confusion, that the combination of the two can also allow for the elaboration of materials with value added. In this context I can cite the recent work of a Korean group which successfully prepared lamellar oxide particles with a concentration gradient, by combining soft chemistry and high temperatures.19 Within the same particle, these compounds thus cumulate different functions: the shell (which is manganese rich) ensures its electrochemical stability whereas the heat affords it a good electrochemical performance. This is another booming aspect of solid-state chemistry.

The future of solid-state chemistry

54As you will have noticed, superconductors aside, all the materials I have discussed here today are designed to improve electrochemical systems and, more specifically, energy-storing batteries. Researchers’ growing interest in this approach, simultaneously developed by many laboratories worldwide, has allowed for the creation of increasingly efficient materials over the years. The best evidence thereof is probably the emergence of the electric vehicle. What was long thought of as an elusive idea will play a major role in the car industry in coming years.

55The development of the electric vehicle will however have to rely on the use of electricity derived from renewable energy for the “green” vehicle to be fully realized. This will raise new challenges for solid-state chemists, particularly regarding the development of better alloys/compounds, even magnets for wind turbines, or more efficient semi-conductors and reflectors for photovoltaic cells.

56I hope that, with this journey through time, I have made you aware that solid-state chemistry, which is a science of materials and their transformations, is highly dependent on our society and its technological options. I am convinced that technological challenges bring about new scientific problems, and that they can be resolved through fundamental research. Whatever the emerging technology, the fact remains that it is always dependent on solid-state chemistry’s sometimes capricious capacity to provide more efficient materials, better suited to the additional constraints imposed on them. The materials and systems we are currently designing must be more sophisticated, miniaturized, recyclable, environmentally friendly, energy saving, highly reliable, and cheap. Therefore, only with an interdisciplinary chemistry will we be able to advance in this quest for ideal materials for systems of varied complexities. The scope of solid-state chemistry is currently extending to new domains such as biology, with the development of materials produced using bio-inspired synthesis processes, and the chemistry of organic-inorganic hybrid materials.

57These are all interesting materials and concepts, which I unfortunately do not have the time to describe here. However, I am fortunate to have this time ahead of me, thanks to the years of research and teaching I have been granted at the Collège de France. I will therefore use this opportunity to delve further into the issue of energy. In this context, my teaching will discuss the advances and challenges, both fundamental and applied, linked to the production, conversion and storage of energy, so as to identify the different “technological deadlocks” that remain in existing materials for each industry. I will seek to provide knowledge of the synthesis methods and tools needed to design tomorrow’s materials and systems. Of course, the chemical bond, the common denominator of the society of atoms constituting crystal, will be the guiding theme.

58I have thus reached the end of this lecture. I will conclude by returning to the question of time. As I pointed out in my introduction, we need to double our energy production. Our hopes are riding on materials and we must be optimistic about our capacities to design better ones. Yet, unlike past generations, we have neither thousands of years nor centuries, but only thirty to forty years to double our energy production, with the additional constraint of sustainable development. What are the odds? What can we hope for? Fortunately, we have a periodic table full of elements. This is certainly a great advantage, as we can design and sculpt new materials as we please, with properties exacerbated by eco-compatible approaches. But it can also rapidly turn into a nightmare: given the large number of possible combinations, it is difficult to find the winning composition. That is why we put so much hope in informed combinatorial theoretical chemistry, so as to imitate, with materials, what happens in the genome with DNA – which has only four nitrogen bases and five elements –, and to establish a genome of materials. US President Barack Obama has made this aim one of his five scientific priorities for the next decade.

59Theory aside, another way to understand better is to see better. X-ray diffraction has allowed us to understand the arrangement of atoms, and microscopy has allowed us to see them. Throughout this Inaugural Lecture, I have shown you that most energy storage and conversion technology mainly relies on chemical/electrochemical reactions that involve a transfer of electrons. Why not dream and hope that we will one day be able to see these famous electrons? This may seem like an adventurous gamble. But the SEM imaging that we are currently performing on electrode materials, in collaboration with Hervé Vezin from Lille, seems to point to its possibility, though we are still at the stage of micron resolution. This is an ambitious dream, the realization of which would trigger a scientific revolution comparable to the observation of the atom through microscopy. It would radically change our material design and elaboration strategies.

60Despite the feats that can be expected from the predictive theoretical approach, the solid-state chemist will always have to design clever synthesis pathways to develop them. If we follow Linus Pauling’s advice – “The best way to have a good idea is to have a lot of ideas” –, an abundance of ideas is the first ingredient needed to succeed on the journey. To make this idea materialize, the chemist will however have to associate it with experimentation, as another quote of the US Nobel Prize laureate emphasizes: “Facts are the air of scientists. Without them you can never fly”. This is an important message that I wish to repeat again and again for our future generations, who all too often blindly limit themselves to theoretical calculations, to the point of forgetting about experiments. Theory and experiments certainly make a wonderful pair that satisfies our intellect and allows for new compounds to be prepared. But it is worth nothing if it is not used to make useful materials. That is why systemic approaches based on cooperation and multidisciplinarity are necessary. No proverb illustrates this point better than this citation of Kenneth G. Wilson, a Physics Nobel Prize laureate (1982): “The most complex problems in science can only be resolved through the full collaboration of the entire international scientific community”. This is a struggle I am relentlessly pursuing within our community, by creating entities such as the RS2E, driven by the motto: “integrating, federating, gathering, and bringing closer better to innovate and develop”. All with the clear objective of better managing the energy resources of our planet and preserving it for future generations.

61Speaking of cooperation, I would like to thank all my colleagues with whom I have had the privilege of working, in the research organizations for which I have worked in the United States (Bell Labs and Bellcore) and in France (LRCS). I would also like to thank all the national research institutions, particularly the Ministry of Research and Higher Education and the CNRS, for their support in setting up the energy storage network RS2E. I extend my warmest thanks to all the members of this network, as well as those of the European network ALISTORE-ERI that I created in 2003; they have made these adventures fascinating on both a scientific and a human level. Many thanks as well to all the brilliant and talented researchers with whom I have collaborated throughout the world. Without them, many of the advances described in this inaugural lecture could not have been exploited. Finally, my thoughts go to my wife Régine and my son Adrien for their support: I dedicate this lecture to them.