Manganese-based cathode materials for Li-ion batteries Von der Fakultät Chemie der Universität Stuttgart zur Erlangung der Würde eines Doktors der Naturwissenschaften (Dr. rer. nat.) genehmigte Abhandlung Vorgelegt von Yuri Surace aus Cinquefrondi (RC), Italien Tag der mündlichen Prüfung: 29.10.2015 Institut für Materialwissenschaft der Universität Stuttgart 2015 Hauptberichter: Mitberichter: Prüfungsausschussvorsitzender: Frau Professor Dr. Anke Weidenkaff Herr Professor Dr. Joachim Bill Herr Professor Dr. Thomas Schleid ~ 1 ~ Alla mia famiglia Luigino, Maria Teresa e Luca ~ 2 ~ Declaration The work described in this thesis was carried out by the author in the Laboratory of Solid State Chemistry and Catalysis at Swiss Federal Laboratories for Materials Science and Tech- nology (EMPA) in Dübendorf, under the supervision of Prof. Dr. Anke Weidenkaff from Sep- tember 2012 to October 2015. The content is the original work of the author except where in- dicated otherwise and has not been previously submitted for any other degree or qualification at any academic institution. Dübendorf, 07/08/2015 Yuri Surace Erklärung Die vorliegende Doktorarbeit wurde vom Autor selbst in der Abteilung Festkörperchemie und Katalyse an der Eidgenössische Materialprüfungs- und Forschungsanstalt (EMPA) in Düben- dorf, unter der Leitung von Prof. Dr. Anke Weidenkaff im Zeitraum von September 2012 bis März 2013 angefertigt. Der Inhalt ist die eigene Arbeit des Autors, Ausnahmen sind gekenn- zeichnet, und wurde noch nicht zur Erlangung einer Qualifizierung oder eines Titels an einer akademischen Institution eingereicht. Dübendorf, 07/08/2015 Yuri Surace ~ 3 ~ Contents Acknowledgements ................................................................................................................................. 5 Abstract ................................................................................................................................................... 6 Zusammenfassung ................................................................................................................................... 8 Sommario .............................................................................................................................................. 10 Chapter 1 ............................................................................................................................................... 12 1. Introduction ................................................................................................................................... 12 1.1 Energy and batteries ................................................................................................................ 12 1.2 Electrochemical cells and redox reactions .............................................................................. 13 1.3 Definitions and concepts ......................................................................................................... 17 1.4 Li-ion batteries ........................................................................................................................ 20 1.5 Components of Li-ion batteries ............................................................................................... 23 1.5.1 Cathode materials ............................................................................................................. 23 1.5.2 Anode materials ................................................................................................................ 27 1.5.3 Electrolyte ........................................................................................................................ 28 1.5.4 Current collector ............................................................................................................... 30 1.5.5 Separator ........................................................................................................................... 30 1.6 Mn-based cathode materials .................................................................................................... 30 1.6.1 Advantages of Mn-based cathode materials ..................................................................... 30 1.6.2 Crystal Field Theory (CFT) in Mn-based cathode materials ............................................ 31 Chapter 2 ............................................................................................................................................... 35 2. Materials synthesis and characterization methods ........................................................................ 35 2.1 Materials .................................................................................................................................. 35 2.1.1 Synthesis of µ-Ca2MnO4 and n-Ca2MnO4 ........................................................................ 35 2.1.2 Activation of µ-Ca2MnO4 and n-Ca2MnO4 ...................................................................... 36 2.1.3 Synthesis of SSR-Li3MnO4 (solid state route).................................................................. 36 2.1.4 Synthesis of FDR-Li3MnO4 (freeze drying route) ............................................................ 37 2.1.5 Preparation of Li3MnO4 cycled electrodes........................................................................ 37 2.1.6 Incorporation of vanadium in Li3MnO4 ............................................................................ 37 2.1.7 Synthesis of FDR-LiMnBO3 ............................................................................................ 38 2.1.8 Synthesis of FDR-LiMnBO3/rGO .................................................................................... 38 2.2 Physico-chemical characterization .......................................................................................... 39 2.2.1 X-ray diffraction (XRD) ................................................................................................... 39 2.2.2 Thermogravimetric analysis (TGA) ................................................................................. 40 2.2.3 Scanning electron microscopy (SEM) .............................................................................. 40 2.2.4 Transmission electron microscopy (TEM) ....................................................................... 41 2.2.5 X-ray photoelectron spectroscopy (XPS) ......................................................................... 41 ~ 4 ~ 2.2.6 Raman spectroscopy ......................................................................................................... 42 2.2.7 Surface area determination ............................................................................................... 42 2.2.8 Particle size determination ............................................................................................... 42 2.3 Electrochemical characterization............................................................................................. 43 2.3.1 The equipment used for electrochemical measurements .................................................. 43 2.3.2 Electrode preparation ....................................................................................................... 44 Chapter 3 ............................................................................................................................................... 45 3. Manganese in octahedral coordination: activation of Ca2MnO4 for Li intercalation .................... 45 3.1 Introduction ............................................................................................................................. 45 3.2 Acid treatment of µ-Ca2MnO4 and characterization ................................................................ 46 3.3 Acid treatment of n-Ca2MnO4: influence of the particle size and comparison with µ-Ca2MnO4 ....................................................................................................................................................... 58 Chapter 4 ............................................................................................................................................... 65 4. Manganese in tetrahedral coordination: Li3MnO4 as cathode material. ........................................ 65 4.1 Introduction ............................................................................................................................. 65 4.2 Characterization of Li3MnO4 synthesized by FD .................................................................... 66 4.3 Capacity fading in Li3MnO4 .................................................................................................... 79 4.4 Vanadium incorporation in Li3MnO4 ...................................................................................... 89 Chapter 5 ............................................................................................................................................... 98 5. Manganese in square pyramidal coordination: h-LiMnBO3 .......................................................... 98 5.1 Introduction ............................................................................................................................. 98 5.2 FD synthesis of LiMnBO3 and LiMnBO3/rGO composite .................................................... 100 Concluding remarks ............................................................................................................................ 107 References ........................................................................................................................................... 109 Additional declaration ......................................................................................................................... 115 Curriculum Vitae ................................................................................................................................. 116 ~ 5 ~ Acknowledgements At the end of this three-year long path, there are many people I would like to thank. First of all, Prof. Dr. Anke Weidenkaff for accepting me as PhD student in her group, for her always creative ideas, and useful and constructive feedbacks. Dr. Simone Pokrant for her guidance, supervision and trust in me during these years. She was a mentor for me regarding scientific and management issues. Dr. Mario Simoes for his extremely helpful scientific support regarding electrochemistry and batteries. Both Mario and Simone contributed to my personal and professional growth, making me be- come a better person and a better scientist. Dr. James Eilertsen for his introduction concerning synthesis and XRD during the first months of my PhD and also for being a great advisor with the English language. Dr. Lassi Karvonen for the very interesting discussions about solid state chemistry and his support with TGA measurements. Dr. Songhak Yoon for the help with XRD and Dr. Santhosh Matam for the introduction in Raman Spectroscopy. Mr. Matthias Trottmann and Mr. Oliver Brunko for solving all technical issues in an excellent manner. Ms. Stephanie Looser for her always present administrative contribution. I would like to thank all old members of the Solid State Chemistry and Catalysis Lab and all new members of the Laboratory Materials for Energy Conversion. Dr. Ulrich Muller and Dr. Roland Hauert for the introduction and the assistance in XPS measurements. Dr. Cedric Pitteloud, Dr. Yoann Mettan and Dr. Jose Antonio Gonzalez Martinez from Bele- nos Clean Holding for the financial contribution and the wise advices regarding battery mate- rials during the FAMSADI meetings. Prof. Dr. Joachim Bill for taking on the task of co-examiner and Prof. Dr. Thomas Schleid for being the Chairman of the examination committee. My girlfriend Monica for the precious chemistry-related advices, and her every day love. In the end, my deepest thanks go to my family, who supported me with enthusiasm, happiness and love to whom I dedicate this thesis. ~ 6 ~ Abstract Li-ion batteries are one of the most commercialized solutions to store electrochemical energy, but until now their broad use is limited to small electronic devices. Higher specific energy and longer cycle life are needed to open the way to a broader range of applications (i.e. electric vehicles or stationary batteries). The specific energy of Li-ion batteries is a function of the anode and cathode capacity for lithium intercalation and the cell voltage. However, capacity and voltage of current state-of-the-art cathode materials are the main specific energy-limiting factors of Li-ion batteries. For this reason, much of the attention during the past few years fo- cused on cathode materials with either high voltage or high capacity or in the best of all cases both, coupled with high stability. Manganese is one of the most common transition metals used in battery materials due to its multiple (and at least partially accessible) oxidation states, its low toxicity and its high availa- bility. Mn-based cathode materials benefit from the Mn3+/Mn2+ or Mn4+/Mn3+ redox couples which allow obtaining a potential range between 3.0 V and 4.2 V vs Li+/Li depending on the crystal structure and the chemical composition. The aim of this work was to study unexplored and scarcely explored Mn-based cathode mate- rials and to improve their electrochemical performances through structural, morphological and chemical modifications. In the initial part of the thesis, a study of calcium manganite Ruddlesden-Popper phases Ca2MnO4 was carried out. Although the pristine material was not electrochemically active, Ca2MnO4 was activated for Li intercalation by Ca extraction using a novel and simple treat- ment with sulphuric acid. The influence of the amount of Ca extracted, and of the particle size were studied and correlated with the electrochemical properties. It was proposed that the acid treated materials had a bi-functional crystalline-amorphous structure, composed by a Ca2MnO4 crystalline bulk phase for the stability and an amorphous MnO2·xH2O surface for the electrochemical response. For each 25at% of calcium extracted, capacities of 40 Ah/kg and 55Ah/kg were obtained for micron-sized particles and for nano-sized particles, respec- tively. A stability improvement of a factor of 10 was reached in comparison to bare amor- phous hydrated manganese oxide. The work focused then on Li3MnO4, a lithium rich phase containing manganese (V). Develop- ing a novel freeze drying (FD) synthesis-route, the micro- and nanostructure of the material ~ 7 ~ were modified with relevant consequences on the electrochemical properties. Smaller parti- cles size in conjunction with smaller grains size allowed obtaining a first discharge capacity of 290 Ah/kg with an improvement of up to 31%, in comparison to Li3MnO4 synthesized by the solid state route. Moreover, measurements carried out at different cycling rates showed improvements in rate capability. In addition, this new route allowed reducing the reaction temperature and time. However, considerable modifications in the Li3MnO4 structure oc- curred during the first cycle and the capacity improvement vanished after a few cycles due to structural instability of this material under cycling. To gain deeper insight into the reason of the capacity fading of this material, a post mortem analysis was carried out which allowed to create a model for the degradation mechanism. Briefly, the lithium extraction or insertion in the structure caused the amorphization of the material with conversion to the more stable amorphous manganese oxide. In the last part of this thesis, preliminary studies on lithium manganese borate LiMnBO3 were carried out. It was shown in a proof of concept study that the FD synthesis was applicable for this material as well. Nanocrystalline material was obtained with electrochemical performance comparable to the state of the art by gaining in synthesis simplicity. ~ 8 ~ Zusammenfassung Li-Ionen Batterien sind die am stärksten kommerzialisierte Lösung zur Energiespeicherung. Aber bis heute ist ihr flächendeckender Einsatz auf kleine elektronische Geräte beschränkt. Höhere spezifische Energien und länger Lebenszeiten sind nötig, um den Weg für eine breite- re Anwendungspalette zu öffnen, wie zum Beispiel elektrische Fahrzeuge oder stationäre Energiespeicher. Die spezifische Energie von Li-Ionen Batterien ist eine Funktion der Ano- den- und Kathodenkapazität für Li-Interkalation und der Zellspannung. Insbesondere die Ka- pazität und die Spannung von dem heutigen Stand der Technik entsprechenden Kathodenma- terialien begrenzen die spezifische Energie von Li-Ionen Batterien. Aus diesem Grund wurde in den letzten Jahren verstärkt nach Kathodenmaterialien geforscht, die entweder hohe Span- nungen oder hohe Kapazitäten oder am besten beides besitzen, gekoppelt mit hoher Stabilität. Mangan ist eines der häufigsten Übergangmetalle, das in Batterien benutzt wird wegen seiner multiplen und zumindest teilweise zugänglichen Oxidationszustände, seiner niedrigen Giftig- keit und seines hohen Vorkommens. Mn-haltige Kathoden profitieren von Mn3+/Mn2+ oder Mn4+/Mn3+ Redoxpaaren, die es erlauben, Zellspannungen zwischen 3.0 V und 4.2 V gegen Li+/Li zu erhalten, abhängig von der Kristallstruktur und der chemischen Verbindung. Das Ziel dieser Arbeit war es, nicht oder nur wenig erforschte Mn-haltige Kathodenmateria- lien zu untersuchen und ihre elektrochemische Leistung durch strukturelle, morphologische und chemische Modifikationen zu verbessern. Im Anfangsteil dieser Doktorarbeit wurde eine Untersuchung an der Ruddlesden-Popper Pha- se Ca2MnO4 durchgeführt. Obwohl das reine Material nicht elektrochemisch aktiv ist, konnte Ca2MnO4 durch die Extraktion von Ca für die Li-Interkalation aktiviert werden mit Hilfe ei- ner neuartigen und einfachen Behandlung mit verdünnter Schwefelsäure. Die elektrochemi- schen Eigenschaften wurden in Abhängigkeit von der extrahierten Kalziummenge und der Teilchengrösse untersucht. Darauf aufbauend wurde ein Modell entwickelt, nämlich dass die säurebehandelten Materialien eine bifunktionale kristallin-amorphe Struktur besitzen, die aus einem kristallinen Ca2MnO4 haltigen Festkörperkern bestehen zur Stabilisierung und aus amorphen MnO2·xH2O Oberflächen, die für die elektrochemische Funktionalität zuständig sind. Für jeweils 25% extrahiertem Ca wurden einerseits Kapazitäten von 40 Ah/kg für mik- rometergrosse und andererseits 55 Ah/kg für nanometergrosse Teilchen gemessen. Eine Stabi- litätsverbesserung um einen Faktor 10 wurde erreicht im Vergleich zu amorphen hydrierten Manganoxid. ~ 9 ~ Die Arbeit konzentrierte sich dann auf Li3MnO4, eine Li-reiche Phase, die Mn (V) enthält. In- dem eine neuartige Syntheseroute über Gefriertrocknen entwickelt wurde, konnte sowohl die Mikro- als auch die Nanostruktur des Materials modifiziert werden mit relevanten Konse- quenzen für die elektrochemischen Eigenschaften. Kleinere Partikelgrössen kombiniert mit kleineren Kristallitgrössen erlaubten es, Kapazitäten von 290 Ah/kg für die erste Entladung zu erhalten, was einer Verbesserung von bis zu 31% im Vergleich zu Li3MnO4 synthetisiert über traditionelle festkörperchemische Methoden entspricht. Darüber hinaus zeigten Messungen bei verschiedenen Ladegeschwindigkeiten eine Verbesserung in der Geschwindigkeitsfähig- keit. Zusätzlich erlaubte die neue Syntheseroute die Reaktionstemperatur und die Reaktions- zeit zu verringern. Allerdings fanden gewichtige Modifikationen der Li3MnO4 Struktur wäh- rend des ersten Zyklus statt und die Verbesserung in der Kapazität verschwand nach einigen Ladungs-/Entladungszyklen wegen der strukturellen Instabilität dieses Materials unter Belas- tung. Um den Grund für den Kapazitätsschwund dieses Materials zu verstehen, wurde eine Post-Mortem Studie durchgeführt, die es erlaubte ein Model für den Degradationsmechanis- mus zu entwickeln. Die Li-Extraktion oder Insertion führte zu einer Amorphisierung des Ma- terials mit einer Umwandlung zum stabileren amorphen Manganoxid. Im letzten Teil dieser Doktorarbeit wurde eine vorläufige Studie an LiMnBO3 durchgeführt. Es wurde gezeigt, dass die Syntheseroute via Gefriertrocknen auch für dieses Material an- wendbar ist. Es konnte dadurch nanokristallines Material erhalten werden, das in seinen elekt- rochemischen Eigenschaften vergleichbar zum Stand der Forschung ist bei einem gleichzeiti- gen Gewinn an Einfachheit in der Synthese. ~ 10 ~ Sommario Le batterie agli ioni di litio costituiscono una delle principali soluzioni presenti in commercio per immagazzinare energia elettrochimica e finora il loro uso è stato limitato a piccoli appa- recchi elettronici. Per il loro utilizzo in una più vasta gamma di applicazioni (es. veicoli elet- trici o batterie stazionarie) sono necessari una maggiore energia specifica e un ciclo di vita più lungo. L’energia specifica dipende dalla capacità del catodo e dell’anodo d’intercalare ioni li- tio e dal voltaggio della cella. Tuttavia, allo stato dell’arte, essa è limitata dalla capacità e dal voltaggio dei materiali catodici. Per questo motivo, negli ultimi anni si è posta molta attenzio- ne sui materiali catodici che abbiano sia un alto voltaggio che un’alta capacità, o nel migliore dei casi entrambi, in congiunzione con una maggiore stabilità. Il manganese è uno metalli di transizione più comunemente usato nei materiali per batterie a causa dei suoi multipli (e parzialmente accessibili) stati di ossidazione, la sua bassa tossicità e l’alta accessibilità. Materiali catodici basati sul manganese beneficiano delle coppie redox Mn3+/Mn2+ or Mn4+/Mn3+ originando un voltaggio fra 3.0 V e 4.2 V vs Li+/Li, il quale dipen- de a sua volta dalla struttura cristallina e dalla composizione chimica del materiale. Lo scopo di questo lavoro è stato quello di studiare dei materiali catodici sconosciuti o poco conosciuti basati sul manganese e migliorare le loro prestazioni elettrochimiche attraverso delle modifiche strutturali, morfologiche e chimiche. Nella parte iniziale di questa tesi è stato effettuato uno studio sul Ruddlesden-Popper calcio manganite Ca2MnO4. Nonostante il materiale di partenza non fosse elettrochimicamente atti- vo, Ca2MnO4 è stato attivato per intercalare ioni litio attraverso l’estrazione di calcio usando un nuovo e semplice trattamento con acido solforico. E’ stata studiata quindi l’influenza della quantità di calcio estratta e della dimensione delle particelle e la loro correlazione con le pro- prietà elettrochimiche. E’ stato proposto che i materiali trattati con l’acido abbiano una strut- tura bi-funzionale cristallina-amorfa, in cui il bulk è composto da Ca2MnO4 cristallino per migliorare la stabilità e la superficie è composta da MnO2·xH2O amorfo il quale dà origine al- la risposta elettrochimica. Per ogni 25 at% di calcio estratto, sono state ottenute capacità di 40 Ah/kg e 55 Ah/kg per particelle di dimensioni micrometriche e nanometriche, rispettivamente. Un miglioramento della stabilità di un fattore 10 è stato raggiunto in confronto con il materia- le composto da solo ossido di manganese idrato amorfo. Lo studio si è concentrato quindi su Li3MnO4, una fase ricca in litio e contenente manganese nello stato di ossidazione (V). Attraverso lo sviluppo di una nuova strategia di sintesi basata ~ 11 ~ sul freeze-drying (FD), sono state modificate la micro- e nanostruttura del materiale con im- portanti conseguenze sulle proprietà elettrochimiche. Particelle di dimensioni minori in con- giunzione con una minore dimensione dei cristalliti ha permesso di ottenere una capacità du- rante la prima scarica di 290 Ah/kg con un miglioramento del 31 % in confronto con Li3MnO4 sintetizzato attraverso una reazione in stato solido. Una maggiore capacità è stata anche regi- strata testando il materiale a differenti velocità di ciclaggio. Inoltre, questa nuova strategia di sintesi ha permesso di ridurre la temperatura e il tempo di reazione. Tuttavia, la struttura di Li3MnO4 subisce notevoli modifiche durante il primo ciclo e il miglioramento sparisce dopo alcuni cicli a causa dell’instabilità strutturale del materiale durante il ciclaggio. Per capire la causa della diminuzione della capacità di questo materiale è stata effettuata un’analisi post- mortem che ha permesso di creare un modello per il meccanismo di degradazione. Brevemen- te, l’inserzione e la de-inserzione di ioni litio nella struttura causano un’amorfizzazione del materiale ed esso converte nel più stabile ossido di manganese amorfo. Nella parte finale della tesi, sono stati effettuati degli studi preliminari su litio manganese bo- rato LiMnBO3. E’ stato dimostrato che la sintesi via FD può essere utilizzata anche per questo materiale. E’ stato ottenuto un materiale nanocristallino con delle performance elettrochimi- che confrontabili con lo stato dell’arte e, al tempo stesso, guadagnando anche in semplicità di sintesi. ~ 12 ~ Chapter 1 1. Introduction 1.1 Energy and batteries Nearly all developed countries depend on fossil fuels (coal, oil, natural gas) as a primary en- ergy source. Heavy consumption of these limited resources, mainly for electrical-energy pro- duction and transportation, produces greenhouse gas emissions that lead to global climate change. Therefore, it is imperative that alternative, renewable and sustainable energy technol- ogies must be further developed in order to reduce the dependence on fossil fuels and de- crease greenhouse gas emissions [1]. In the last few years, the steady increase in the global energy demand triggered a considerable progress in several renewable energy technologies as solar energy, wind power, biofuels and hydropower. This resulted in a growth of the total renewable power capacity by a factor of 7 (the most significant has been in photovoltaics by a factor of 70). By 2013, 19% of the world’s final energy consumption was supplied by renewables sources [2]. However, renewable energies production is not demand-oriented and variable over time. Therefore stationary energy storage solutions are needed as essential components to guarantee the reliability of future energy systems [3]. Batteries are the most discussed component of sta- tionary applications. Indeed, electrical energy needs to be stored in batteries during times when production exceeds the consumption, and batteries have to supply electrical energy when consumption exceeds the production. In this way the electricity production could be maintained at more constant levels and the cost could be lower [4]. Another field of interest where batteries play a primary role is mobility. A contribution to a greener world can be obtained only if vehicles driven by internal combustion engines (ICEs) are phased-out. An alternative are electric vehicles (EV) powered by batteries. The progress which has been made during the last decade lowered the battery cost. In parallel the infra- structure dedicated to electric vehicles (EVs) was increased. This is expected to result in an ~ 13 ~ 83% increase of EVs sale in the next 10 years [5] with a positive impact on the global envi- ronment. In addition, batteries are also extremely important in everyday life since they are one of the most commercialized power sources for portable applications such as mobile phones, laptops, etc. For all the above mentioned applications batteries are the key factor. Higher gravimetric ener- gy, higher volumetric energy and longer cycle life are compulsory to go beyond the state of the art. This would open the way for the conception of new energy technologies and the de- velopment of the existing one with their wide-spreading all over the world. For example, more efficient energy storage solutions would allow using the full potential of renewable en- ergy sources and they could enable countries to run on 100% renewable energy in the next fu- ture [6]. Consequently, this thesis was focused on the study of materials for battery applica- tions. 1.2 Electrochemical cells and redox reactions An electrochemical cell is a device capable of producing electrical energy from spontaneous chemical reactions or inducing non-spontaneous chemical reactions through the consumption of electrical energy. In the first case the cell is called galvanic cell and in the second case electrolytic cell [7]. An electrochemical cell consists of two electrodes immersed in an electrolyte solution. The electrodes are electronic conductors, while the electrolyte solution is an ionic conductor. At the interface between electrodes and electrolyte occurs an oxidation-reduction reaction (redox reaction) which produces electrical current. The electrode where the oxidation occurs is called anode. The electrode where the reduction occurs is called cathode. A redox reaction is a chemical reaction which involves the transfer of electrons between chemical species. The species involved in a redox reaction have to be able to change their ox- idation states. In a common reaction two species are usually involved: the reductant (or reduc- ing agent) and the oxidant (or oxidizing agent). The reductant transfers electrons to the oxi- dant. Thus, during the reaction, the reductant or reducing agent loses electrons and is oxidized (increasing its oxidation number), and the oxidant or oxidizing agent gains electrons and is reduced (decreasing its oxidation number). Redox reaction of a specific element are expressed as half reaction and, by convention, listed as reductions. A reduced and an oxidized species of ~ 14 ~ a specific element form a redox couple (Table 1). Each of these half-reactions can be associ- ated with a standard electrode potential, E0. However, since half‐reaction potentials cannot be measured in an absolute sense, every half reaction has to be coupled with a standard refer- ence electrode. The standard hydrogen electrode (SHE) is used as reference electrode and its potential is set to 0.0 V. The SHE electrode is constituted of a platinated platinum electrode flushed with hydrogen in a 1 mol/l HCl water solution (T = 25 °C, p = 1 bar, all active species at unity activity) [8]. Table 1: Standard electrode potential for some redox couples. Half-reaction Redox couple Standard Electrode Potential (V) vs SHE Cu2+ + 2e− ↔ Cu Cu2+/ Cu 0.52 2H+ + 2e− ↔ H2 H +/H2 0.0 Zn2+ + 2e− ↔ Zn Zn2+/Zn -0.76 Mn2+ + 2e− ↔ Mn Mn2+/Mn -1.18 Al3+ + 3e− ↔ Al Al3+/Al -1.66 Na+ + e− ↔ Na Na+/Na -2.71 Li+ + e− ↔ Li Li+/Li -3.04 For non‐standard conditions the Nernst equation is used to determine the potential of the half- reaction at equilibrium [8] Eq.1.1: E = E0 + RT nF ln∏ ai υi i Eq. 1.1 Where: 𝐸: electrode potential [V] 𝐸0: Standard electrode potential [V] R: ideal gas constant [J K-1 mol-1] T: absolute temperature [K] 𝑛: number of electron exchanged 𝐹: Faraday constant [C mol-1] ai: activity of species i 𝜐𝑖: stoichiometric coefficient species i. ~ 15 ~ The cell voltage of an electrochemical cell is calculated from the electrode potentials (reduc- tion potentials) of the half‐reactions. The overall theoretical cell voltage ΔE of an electro- chemical cell is obtained by the difference between the half‐cell potential of the reduction (cathode) and the half‐cell potential of the oxidation (anode): ΔE = Ered − Eox = Ecat − Ean Eq. 1.2 The first example of galvanic cell was the “Volta cell”. The electrodes were made of a Zn metal piece and a Cu metal piece immersed in an electrolyte solution of sulphuric acid which can be represented as follow: Zn|H2SO4|Cu By convention each interface is represented by a vertical stroke and if the electrochemical chain includes several successive electrolyte media, then the abbreviated notation | | is often used to denote the separation zone between two electrolytes. Another simple example of galvanic cell is the “Daniel Cell”. It includes two compartments containing respectively a ZnSO4 aqueous solution in contact with zinc metal and a CuSO4 aqueous solution in contact with copper metal. These two compartments are electrically con- nected by a third aqueous solution, e.g., a concentrated KNO3 solution, which is called a salt bridge. Zn| ZnSO4|| CuSO4|Cu In the case of the Volta cell the reactions are: Cathode reaction 2H+ + 2e− ↔ H2 E 0,cat = 0 V Anode reaction Zn ↔ Zn2+ + 2e− E0,an = −0.76 V Total reaction 2H+ + Zn ↔ H2 + Zn 2+ ΔE0 = 0.76V In the case of Daniel cell the reactions are: Cathode reaction Cu2+ + 2e− ↔ Cu E0,cat = 0.52 V Anode reaction Zn ↔ Zn2+ + 2e− E0,an = −0.76 V Total reaction Cu2+ + Zn ↔ Cu + Zn2+ ΔE0 = 1.28V Nowadays, galvanic cells, more commonly called batteries, are divided in primary batteries and secondary batteries. Primary batteries are based on irreversible electrochemical reactions; they produce energy for a limited period of time and after discharge they have to be disposed ~ 16 ~ (e.g. Alkaline). Secondary batteries are based on reversible electrochemical reactions and they can be recharged, converting electrical energy in chemical energy during the charge process (e.g. Lead-acid, Ni-Cd, Li-ion) [7]. Since the discovery of the “Volta cell”, batteries based on different chemistries have been de- veloped over the years (Fig.1). The most common commercial batteries with their voltage, specific energy and redox reaction are listed below, and a comparison of the volumetric and specific energy is shown also in Fig.1.1: Alkaline (1.4 V, 120 Wh/kg) Cathode reaction: 2MnO2(s) + H2O(l) + 2e − → Mn2O3(s) + 2OH −(aq) Anode reaction: Zn(s) + 2OH−(aq) → ZnO(s) + H2O(l) + 2e − Total reaction: Zn(s) + 2MnO2(s) ↔ ZnO(s) + Mn2O3(s) Lead-Acid (2.0 V, 40 Wh/kg) Cathode reaction: PbO2(s) + HSO4 −(aq) + 3H+(aq) + 2e− → PbSO4(s) + 2H2O (l) Anode reaction: Pb(s) + HSO4 −(aq) → PbSO4(s) + H +(aq) + 2e− Total reaction: Pb(s) + PbO2(s) + 2H2SO4(aq) → PbSO4(s) + 2H2O (l) Ni-MH (1.2 V – 100 Wh/kg) Cathode reaction: NiO(OH)(s) + H2O (l) + e − → Ni(OH)2(s) + OH −(aq) Anode reaction: MHx(s) + OH −(aq) → M(s) + H2O (l) + e − Total reaction: NiO(OH)(s) + MHx(s) → Ni(OH)2(s) + M(s) Ni-Cd (1.2 V – 60 Wh/kg) Cathode reaction: 2NiO(OH)(s) + 2H2O (l) + 2e − → 2Ni(OH)2(s) + OH −(aq) Anode reaction: Cd(s) + 2OH−(aq) → Cd(OH)2 + 2e − Total reaction: 2NiO(OH)(s) + Cd(s) + 2H2O (l) → 2Ni(OH)2(s) + Cd(OH)2 ~ 17 ~ Figure 1.1: Battery chemistry over the years (left) [9] and comparison between different batteries technologies (right) [10] 1.3 Definitions and concepts The performances of an electrochemical cell are expressed using theoretical values or practi- cal values. Theoretical values are calculated from the thermodynamics of the electrochemical cell reaction and are thus independent of a particular cell design. Practical values are related to the total mass of the full battery, including the mass of electrodes, electrolyte, separator, current collectors, terminals and cell housing. They depend strongly on the cell design and on the conditions of discharge [11]. In Li-ion batteries, electrodes are usually composites made of electroactive material (EAM), carbon and binder, therefore the redox reaction involves on- ly the EAM which act as reactant. In this work, the specific charge always refers to the weight of the EAM only. Cell voltage The cell voltage can be calculated from the Gibbs free energy of the corresponding chemical reaction: U0 = ∆E0 = −ΔG0 nF Eq. 1.3 where: ∆E0: Standard cell potential ΔG0: Standard Gibbs Free Energy n: number of electron exchanged F: Faraday constant ~ 18 ~ Current density The current density j is calculated by dividing the total current I flowing through an electrode by the electrode area A. The normal case is that the geometrical area is used: j = I A [ A m2 ] Eq. 1.4 Capacity The capacity Q is the total amount of charge obtainable from a cell: Q = ∫ I ∙ dt t2 t1 [A ∙ h] Eq. 1.5 Theoretical Specific Charge The theoretical specific charge, qth, is the amount of charge per kg of reactant. It is usually based on the molecular weight of the active materials and the number of electrons transferred in the electrochemical process [12]. It can be calculated via Faraday's law. qth = nF∙1000 W∙3600 [ A∙h kg ] Eq.1.6 Where: n: number of electron exchanged F: Faraday constant 96485.34 [C mol-1] W: molecular weight of the substance [mol g-1] Very often in the battery community this quantity is called also capacity. Practical specific charge The (practical) specific charge is the total charge obtainable under specified discharge condi- tions from a practical cell in one discharge cycle divided by the total mass of the cell (mc). q = | Q mc | [ A∙h kg ] Eq.1.7 ~ 19 ~ Theoretical and practical charge density The theoretical charge density (QV,th) is the amount of charge divided by the volume of the reactant. The practical charge density (QV) represents the total charge obtainable from a prac- tical cell divided by the volume of the cell. The unit measure is very often Ah/l. Theoretical specific energy The theoretical specific energy, wth, can be calculated from the Gibbs energy change ∆G 0 of the electrochemical cell reaction divided by the sum of the stoichiometric masses of the reac- tants (mR). wth = | ∆G0 ∑ mR,ii | = nF∆E0 ∑ mR,ii [ W∙h kg ] Eq.1.8 Practical specific energy The (practical) specific energy is the total electrical energy (Wc) obtainable from a practical cell in one discharge cycle divided by the mass of the respective cell (mc). w = Wc mc where Wc = ∫ U(t)I(t)dt t 0 [ W∙h kg ] Eq. 1.9 Theoretical and practical energy density The theoretical energy density, 𝑊𝑉,𝑡ℎ, can be calculated from the Gibbs energy change ∆𝐺 0 of the electrochemical cell reaction divided by the sum of the volumes of the reactants. The practical energy density, 𝑊𝑉, is the total electrical energy obtainable from a practical cell under specified discharge conditions divided by the volume of the cell. The unit measure is Wh/l. Specific power The specific power is the capability to deliver power per mass of a primary or secondary bat- tery. The specific power of a cell depends on the discharge current and decreases during dis- charge. The unit measure is W/kg. ~ 20 ~ Power density The power density is the power divided by the volume of the cell. The unit measure is W/l. Coulombic efficiency For secondary cells, the Coulombic efficiency (CE) represents the ratio of charge released during the discharge Qdis to the charge necessary for charging the battery Qch. ΦQ = Qdis Qch Eq. 1.10 Both Qdis and Qch are obtained by integrating the respective currents over the charging and discharging time, respectively. They depend on the conditions for charging and discharge. Values of CE lower than 100% are directly related to an irreversible capacity as results of side reactions inside the cell. In scientific literature on batteries, one frequently encounters the term C-rate, which describes the current required to charge or discharge a cell in 1h. Using twice the amount of currents (i.e., at 2C-rate) the cell is completely charged or discharged in half an hour. The amount of current at certain C-rate is closely related to the specific charge of the material. For example, for a material with 200Ah/kg theoretical specific charge, 1C would mean to charge or dis- charge the material at 200A/kg [12]. 1.4 Li-ion batteries The need for batteries with higher specific and volumetric energy boosted the development of new battery technologies like lithium ion batteries. The use of lithium in a battery seemed very attractive because lithium is the lightest metal on earth (m.w. = 6.941 g mol-1, density = 0.534 g/cm3, qth=3861 Ah/kg) and it has also the lowest absolute electrochemical potential (E0 =-3.04 V vs. SHE). The first paper on metallic lithium as electrode was reported by Lewis in 1913 [13] but it is only in the early 70s that primary Li- ion batteries were introduced into the market. These batteries were based on lithium metal an- ode and manganese dioxide (MnO2), sulfur dioxide SO2, or polycarbon monofluoride ((CFx)n) as cathodes in an organic electrolyte as propylene carbonate [14]. However, issues regarding the stability of metallic lithium in organic electrolyte became soon evident. Due to its high reducing power, lithium decomposes the electrolyte creating a film ~ 21 ~ made of electrolyte decomposition products on the surface of the electrode [15] . If, on one side, the formation of this film, called solid electrolyte interface (SEI), can give rise to some advantages, since it is permeable to Li ions and prevents lithium to further corrode; on the other side lithium deposition and dissolution can give rise to dendrite formation [16] (Fig.1.2a). Figure 1.2: (a) Rechargeable Li-metal battery with lithium metal anode and (b) rechargeable Li-ion battery with intercalation cathode. [17] The growing of dendrites becomes a problem, when they perforate the separator placed be- tween anode and cathode, shortcutting the cell. Then all stored energy of the cell is instanta- neously transformed into heat. Explosions and fires are possible, due to the low melting point (ca. 180°C) and the high reactivity of lithium [14]. For these safety reasons, in the early 80s lithium intercalation materials with an electrochemi- cal potential close to that of Li were proposed as anode for lithium-ion batteries (Fig.1.2b). The development of such low-voltage intercalation materials was successful. A well-known example is lithiated graphite LixC6 [18, 19]. In the same time period, the “intercalation” con- cept was also applied to cathode materials. Layered di- and trichalcogenides first, and layered transition metal oxides [20] later, were largely studied as positive electrodes. This new system, the so called “ rocking chair system” [21], because lithium ions travel back and forth between cathode and anode intercalation materials, set the foundation to today’s Li- ion commercial batteries. The term “lithium-ion cell” refers to the working mechanism based on the highly reversible electrochemical reaction, usually called a “lithium insertion” or “lithium intercalation” pro- cess, which may be described as the insertion/extraction of mobile lithium ions into a host structure (Fig.1.3). ~ 22 ~ The basic setup of a Li-ion battery is composed of three main components: a positive elec- trode, a negative electrode and an organic electrolyte between them assuring Li+ conductivity (not electronic conductor). The positive and negative electrodes are referred to as the cathode and anode during discharge, and vice versa during charge. During the discharge process lithi- um ions are extracted from the anode which is oxidized and inserted in the cathode which is reduced. The process is reversed during the charge [22]. Figure 1.3: Schematic representation of a Li-ion battery during charge/discharge. The right side shows the graph- ite sheets and the left side the layered structure of LiCoO2 [10]. The charging/discharging process for a Li-ion battery with lithium cobalt oxide as the positive electrode and graphite as the negative electrode material is illustrated in Scheme 1.1. Scheme 1.1: Reactions occurring at electrodes during charge and discharge in Li-ion battery with graphite as an- ode and LiCoO2 as cathode. ~ 23 ~ 1.5 Components of Li-ion batteries 1.5.1 Cathode materials The positive electrode of a lithium ion battery has to fulfill some basic requirements proposed by Whittingham [20]: 1) The material contains a readily reducible/oxidizable ion; for example a transition met- al ion. 2) The material reacts with lithium in a reversible manner and the lithium host structure does not charge during intercalation. 3) The material reacts with lithium with a high free energy of reaction (high capacity, preferably at least one lithium atom per transition metal atom and high voltage, prefer- ably around 4 V vs Li+/Li). This leads to high energy density. 4) The material reacts with lithium very rapidly on both insertion and removal, this leads to high power density. 5) The material is a good electronic conductor. This allows for reaction at all contact points between the cathode active material and the electrolyte rather than at ternary contact points between the cathode active material, the electrolyte, and the electronic conductor i.e. carbon black. 6) The material is stable, i.e. does not change structure or otherwise degrade, to overdis- charge and over charge. 7) The material is inexpensive and environmentally benign. Three main cathode families have been extensively studied over the years: layered oxides LiMO2 (M=Co, Ni, Mn, etc), spinels LiM2O4 (M=Mn, Ni, etc), and olivines LiMPO4 (M=Fe, Co, Ni, Mn, etc) [23]. LiMO2 oxides have α-NaFeO2 layered structure. In a ccp oxygen array Li+ and M3+ are dis- tributed in the octahedral interstitial sites in such a way that MO2 layers are formed consisting of edge-sharing [MO6] octahedral. In between these layers lithium resides in octahedral [LiO6] coordination, leading to alternating (111) planes of the cubic rock-salt structure [24]. Due to strong M-O bonds the MO2 layers are relatively inert against electrochemical reduc- tion/oxidation. On the other hand, the weak interlayer bonding interaction (comprised of an interplay of electrostatic repulsion and attraction among negatively charged MO2 layers and ~ 24 ~ positively charged Li+ cations together with a weak van der Waals interaction between the MO2 layers) allows a reversible insertion/extraction of lithium in between the MO2 layers [25]. Figure 1.4: Crystal structure of LiCoO2 (left) (adapted from [26]) and voltage profile of LiCoO2 (right)(adapted from [25]). LiCoO2 (Fig.1.4) was first published by Goodenough [27]. This was at the origin of the com- mercial success of the high energy Li-ion batteries, mainly due to its cycling stability over thousand cycles. Lithium cobalt oxide exhibits a redox potential of 3.9 V- 4.1 V and a capaci- ty of 140 Ah/kg [28]. This is due to the extraction/insertion of only 0.5 lithium equivalent in a reversible manner. The extraction of Li>0.5 equivalents above 4.2 V lead to higher capacity (170Ah/kg) but also to structural distortion and capacity fading [25] . However, the cost, tox- icity, and safety of cobalt based batteries led to the need for its replacement. Efforts to reduce the amount of cobalt by substitution with other transition metals, e.g. Mn and Ni, have result- ed in a solid solution between LiCoO2, LiNiO2 and LiMnO2 to form LiCo1/3Ni1/3Mn1/3O2 (NMC) [29] with improved performance compared to LiCoO2. Its higher specific charge more than 200 Ah/kg, its broad redox potential between 3.6 V-4.4 V and its excellent rate capability allowed its use in commercial batteries (Fig.1.5). Figure 1.5: Voltage profile of NMC (adapted from [29]). ~ 25 ~ Another important novel material is the so-called integrated layered-layered cathode 0.5Li2MnO3-0.5LiMn1/3Ni1/3Co1/3O2 that shows the composite structure of the two phases with a mixing at the atomic level. The Li2MnO3 is initially an inactive component so that the composite electrodes comprising these materials, which are cycled at a potential below 4.5 V vs Li+/Li, only demonstrate the electrochemical activity of the LiMO2 component. Upon acti- vation of these materials, on the first cycle at potentials higher than 4.7 V vs. Li+/Li, a pro- nounced irreversible structural change occurs that includes delithiation and partial loss of ox- ygen[30]. This irreversible process activates the Li2MnO3 components and forms a new active material which exhibits a broad redox potential window of 3.0 V - 4.5 V, thermal stability at elevates temperatures and capacities higher than 250 Ah/kg [31] (Fig.1.6). Figure 1.6: voltage profile of 0.5Li2MnO3-0.5LiMn1/3Ni1/3Co1/3O2 (adapted from [32]) LiM2O4 oxides have MgAl2O4 spinel-type structure. In a ccp array of oxygen atoms, Li + oc- cupies 1/8 of the tetrahedral sites and M3+/4+ occupied ½ of the octahedral sites. A strong edge-shared octahedral [M2]O4 array permits a reversible extraction of the Li + ions from the tetrahedral sites without provoking a collapse of the 3-dimentional [M2]O4 spinel framework [25]. Figure 1.7: Crystal structure of LiMn2O4 (adapted from [33]) and voltage profile of LiMn2O4 (adapted from [34]). ~ 26 ~ The LiMn2O4 spinel [35] has a redox potential around 4.0 V and a capacity around 120 Ah/kg due to extraction/insertion of lithium ions from/into tetrahedral sites transforming LiMn2O4 in λ-MnO2. When one additional lithium equivalent is inserted into the structure a flat plateau around 3.0 V occurs (Fig.1.7). This is the result of the transition from cubic Li[Mn2]O4 to te- tragonal Li2[Mn2]O4 caused by the Jahn-teller effect of the Mn 3+ ion [25]. Although the inser- tion of two lithium equivalents increases the capacity up to 250Ah/kg, the 3V region limits the cyclability of LiMn2O4. In recent years Mn has been partially substituted by Ni to improve this type of spinel. LiMn1.5Ni0.5O4 has higher redox potential around 4.7-4.8V and a higher capacity of 140 Ah/kg [36] (Fig.1.8). Figure 1.8: voltage profile of LiMn1.5Ni0.5O4 (adapted from [37]). LiMPO4 oxides have an Mg2SiO4 olivine-type structure. In an hcp array of oxygen atoms Li + and Fe2+ occupy ½ of octahedral sites and P5+ occupies 1/8 of tetrahedral sites. Corner-shared MO6 octahedra are linked together in the bc-plane, while LiO6 octahedra form edge-sharing chains along the b-axis building up the channel from where the lithium ions can be removed. The tetrahedral PO4 groups bridge neighboring layers of MO6 octahedra by sharing a common edge with one MO6 octahedra and two edges with LiO6 octahedra [38] (Fig.1.9). Figure 1.9: Crystal structure of LiFePO4 (adapted from [38]) and voltage profile of differents olivine cathode materials (adapted from [39]). ~ 27 ~ Within this family, the compounds where M is Fe, Mn, and Co exhibit the following redox potential vs Li+/Li: 3.5 V, 4.1 V and 4.8 V respectively (Fig 9). The advantage of these mate- rials in comparison with other cathode materials is that lithium insertion/extraction occurs with a very flat plateau due to two-phase process during lithiation/delithiation. However, the electronic conductivity has to be improved by carbon coating [40]. The low surface reactivity of the olivine gives the possibility to create nano-LiMPO4 particles to improve Li intercala- tion/deintercalation. LiFePO4 [41] reaches a capacity of 160 Ah/kg, it is cheap and has a good cyclability but due to its low redox potential, researchers in this field have turned their attention to the Mn-containing analog with higher voltage and a capacity of 140 Ah/kg [42]. 1.5.2 Anode materials It is generally accepted by the battery community that graphite electrode is the most suitable material due to its unique characteristics in terms of safety, high capacity (372 Ah/kg), cycla- bility and low voltage range (0.25-0.05V vs Li+/Li) for the lithium insertion/extraction reac- tion [40]. Figure 1.10: 1st and 2nd cycle of graphite (adapted from [24]). Graphite can be reduced to LiC6 (Scheme 1.1) upon lithium intercalation. The reversibility of this reaction was initially compromised by the choice of the electrolyte, in fact, the use of propylene carbonate (PC) –based electrolytes leads to the intercalation of solvent molecules between graphite layers with consequent exfoliation and poor cyclability. Since 90s the use of ethylene carbonate (EC) and dimethyl carbonate (DMC)-based electrolyte allowed a reversi- ble intercalation of lithium ions in the graphite[24]. ~ 28 ~ Figure 1.11: Schematic presentation of the formation of the SEI layer by decomposition of EC-based electrolytes [43]. During the first charge the capacity exceeds the theoretical one because of the SEI formation. EC reduction forms a passivation layer on the external graphite surface in the early stage of reduction, this film prevents excessive solvent co-intercalation. In the second cycle 85-90% of the theoretical capacity is recovered (Fig 1.10 and 1.11). New materials based on Sn or Si alloy have been investigated recently. Tin and silicon behave similarly upon alloying with Li, with similar stoichiometries and >300% changes in volume [40] . Between them, the Si is more promising because it is more abundant and can reach higher capacity (Li4.4Si: 4200 mAhg -1 vs Li4.4Sn: 900 mAhg -1). It was found that the use of nanowires of Si or Sn improves in the accommodation of the mechanical strain that occurs during the volume changes [44]. Metal oxide with low voltage vs Li+/Li are also being studied as anode materials as for exam- ple Li4Ti5O12 and Li3VO4. The former, with a redox potential of 1.5 V vs Li +/Li and a capaci- ty of 160 Ah/kg [45] was recently overtaken by the latter which has a redox potential of 1.0 V vs Li+/Li and a capacity of 300 Ah/kg [46]. 1.5.3 Electrolyte Conventional electrolytes for lithium-ion batteries consists of an inorganic salt dissolved in organic solvents with a large electrochemical stability window (1.3 V- 5.0 V vs Li+/Li) (Fig 1.12). The electrochemical window is defined by the energy separation Eg between HOMO and LUMO of the solvent; To prevent oxidation and/or reduction of the electrolyte, Eg has to be larger than the difference in electrochemical potential between anode and cathode [43]. A suitable electrolyte should have high ionic conductivity (> 10-4 S/cm) and low electronic con- ~ 29 ~ ductivity (<10-10 S/cm), high chemical stability, low cost and assure safety. Solvents with low melting points, high boiling points and low vapor pressures are highly desirable. The best choice was found to be a mixture of alkyl carbonates like ethylene carbonate (EC) and either dimethyl carbonate (DMC) or ethyl methyl carbonate (EMC) with a Li salt like LiPF6. Lithi- um hexafluorophosphate LiPF6 is, at present, the electrolyte salt for most commercial lithium- ion batteries because it is highly soluble in alkyl carbonates solvents forming a high Li+ con- ducting solution. However, it is expensive to produce in the high purity needed and it is also prone to hydrolysis, forming the highly toxic hydrofluoric acid (HF) [39]. Figure 1.12: Organic solvent commonly used as electrolytes in Li-ion batteries (left). Schematic presentation of the electrochemical window of various solvent families with Li salt (right) (adapted from [39]). Room temperature ionic liquids (RTIL), have been proposed and investigated as safe solvents of Li-ion battery electrolytes. They are composed of an organic cation (i.e. imidazolium cati- ons (RRIm+), pyridinium cations (RRPy+), tetraalkylammonium cations (RRRRN+)) com- bined with a variety of large anions having a delocalized charge (PF6-, BF4- , N(F2SO2)2 -). The main advantages over organic electrolytes are 1) higher oxidation potential ~5.3 V vs Li+/Li, 2) safety features (non-flammability and non-volatility) but the main drawbacks are high vis- cosity and low ionic conductivity at low temperatures. The final electrolyte is usually com- posed by an IL and a lithium salt which in most of the case include the anion composing the electrolyte [47]. Recently, solid electrolytes are also being studied due to their very wide electrochemical win- dow (0.0 V – 5/6 V vs Li+/Li) and safety. The most important families are: 1) perovskite – type structure like lithium lanthanum titanate (LLTO) Li3xLa(2/3)−x□(1/3)−2xTiO3 (099%) and Mn(NO3)2·4H2O (Sigma-Aldrich 98%) with a CA:Ca:Mn ratio of 8:2:1. The precursors were dissolved in a round bottom flask containing 50 mL of high purity water. The solution was subjected to a reflux for 4h, then poured into a borosilicate glass bowl and heated in an oven according to the following temperature pro- gram: the furnace was heated to 100 °C with a heating rate of 20 °C/min, and held for 12h, and then the temperature was increased to 300 °C with a heating rate of 20 °C/min and held for 4 h. During the first step, the solution dried sufficiently to from a gel. During the second step, the decomposition of the precursor nitrates created a solid foam. The solid foam was ground in an agate mortar, and loaded into an alumina crucible and calcined to remove resid- ual organics: the sample was heated at a rate of 30 °C/min to 800 °C and held for 12 h. The resulting powder was ground, pressed into a pellet, and annealed at 1000 °C for 18h and again at 1100 °C for 12 h to obtain pristine Ca2MnO4. Nano Ca2MnO4 (n-Ca2MnO4) was prepared by a soft chemistry method [72] using analytical grade citric acid (CA) (Sigma-Aldrich 99%), ethylene glycol (EG) (VWR 98%), Ca(NO3)2·4H2O (Sigma-Aldrich >99%) and Mn(NO3)2·4H2O (Sigma-Aldrich 98%) with a CA/EG ratio of 1 and a CA/metal ion ratio of 1. The precursors were dissolved in a round bottom flask containing 50 mL of high purity water. The solution was heated to allow the evaporation of water and the formation of the gel. The gel was heated at 250 °C for 6 h with a heating rate of 1 °C/min. The solid foam formed was ground in an agate mortar, loaded into an alumina crucible and calcined at 800 °C for 12 h. ~ 36 ~ 2.1.2 Activation of µ-Ca2MnO4 and n-Ca2MnO4 The modification of Ca2MnO4 was achieved by suspending the pristine powder in an aqueous solution of H2SO4. In all reactions the acid concentration was kept constant by setting the pH value to pH=2 and the same quantity (0.5g) of pristine material was used for each sample. Theoretical calcium extractions of 25at%, 50at%, 75at% and 90at% were obtained by chang- ing only the volume of the solution. The samples were named throughout the text as µ- Ca2MnO4-25%Ca-extr, µ-Ca2MnO4-50%Ca-extr, µ-Ca2MnO4-75%Ca-extr and µ-Ca2MnO4- 90%Ca-extr. The pH was monitored and the reaction was stopped after 24 h. During this time a steady pH was reached. The solution was then filtered and the residue (a black solid) was washed with distilled water and dried in an oven for 2 h at 75 °C in air. For comparison amorphous hydrated manganese dioxide was prepared by mixing 1.58 g of KMnO4 dissolved in 60 mL of high purity water with 3.68 g of manganese (II) acetate dis- solved in 100 mL of high purity water as already reported [75]. A blend between pristine µ-Ca2MnO4 and MnO2·xH2O (supposing to contain roughly 1 eq. of water[76]) with an atomic ratio Ca/Mn of 1 was prepared mixing the right quantities in a mor- tar. The modification of n-Ca2MnO4 was performed as described above. Theoretical calcium ex- tractions of 25 at%, 50 at% and 75 at% were obtained. The samples were named throughout the text as n-Ca2MnO4-25%Ca-extr, n-Ca2MnO4-50%Ca-extr, and n-Ca2MnO4-75%Ca-extr. 2.1.3 Synthesis of SSR-Li3MnO4 (solid state route) SSR-Li3MnO4 was prepared by a solid state route, typically used to synthesize this material [67]. A ground mixture of LiMnO4·3H2O and LiOH·H2O (Alfa Aesar, 98%min) in a 1:2 ratio was introduced into a furnace under oxygen flow. The sample was heated at 1°C/min from RT to 70 °C. At this temperature the sample was re-ground and heated by steps of 10°C until 125°C. After 1h at 125°C the heating was continued at 1°C/min to 170°C and then heated for 3h. The LiMnO4·3H2O precursor was prepared by ion exchange from KMnO4 as described elsewhere [77] . ~ 37 ~ 2.1.4 Synthesis of FDR-Li3MnO4 (freeze drying route) FDR- Li3MnO4 was prepared by freeze drying route in a two-step reaction. In the first step the synthesis of LiMnO4·3H2O was performed. At this stage LiOH·H2O was added to the solution containing LiMnO4·3H2O (ratio of 2:1 for hydroxide: permanganate) (Eq. 2.1). The solution was initially cooled in liquid nitrogen and then maintained under vacuum (0.6 Pa) to remove the water by sublimation. As a result, a freeze dried purple powder was obtained. The powder was then introduced in the furnace and was subjected to the same heating steps as SSR- Li3MnO4. LiMnO4(aq) + 2LiOH(aq) 𝐅𝐃 → mixture ∆ → Li3MnO4 Eq. 2.1 2.1.5 Preparation of Li3MnO4 cycled electrodes SSR-Li3MnO4 sample and the equipment described in subchapter 2.3 were used to prepare the cycled electrodes. The electrodes were freshly prepared as described in section 2.3.2 and then underwent to the following cycling: the batteries were cycled in galvanostatic mode following the cell voltage until the intended electrochemical reaction was completed, then the cycle was stopped. The cells were quickly transferred into the glove box, where they were opened al- lowing collecting the electrodes at different charge. These electrodes were subjected to further analysis as described below. The analyzed electrodes were named through the text as: Li3MnO4-fresh for the fresh electrode (not cycled), Li3MnO4-1, Li3MnO4-2, Li3MnO4-3, and Li3MnO4-4 for the cycled electrodes where the electrochemical reaction was stopped in the points 1, 2, 3, and 4 respectively. Point 1 and 4 are charged states at 4.2 V after initial charge, and after discharge to 1.5 V followed by charging to 4.2 V, respectively. Point 2 and 3 are discharged stated at 1.5 V after charge to 4.2 V followed by discharge to 1.5 V, and after ini- tial discharge, respectively. 2.1.6 Incorporation of vanadium in Li3MnO4 First, the synthesis of lithium permanganate (LiMnO4) was performed by ion exchange reac- tion from potassium permanganate [77]. Then, lithium hydroxide (LiOH·H2O) was added, to the solution containing LiMnO4 creating a solution with pH around 12 (Eq.2.2). NH4VO3 was then added in the desired amount. The highly basic solution allowed the deprotonation of the 𝑂2 ~ 38 ~ ammonium ion (NH4 +) to form gaseous ammonia (NH3). The solution was then freeze dried to obtain a purple powder. The powder was heated and ground by steps of 10°C from 70°C to 125°C then 1h at 125°C and 3h at 170°C under O2 flow. (1 − x)LiMnO4(aq) + (2 + x)LiOH(aq) + xNH4VO3(aq) 𝐅𝐃 → mixture ∆ →Li3Mn1−xVxO4 Eq.2.2 2.1.7 Synthesis of FDR-LiMnBO3 FDR-LiMnBO3 was prepared by a freeze drying route (Eq. 2.3). Lithium citrate tetrahydrate Li3(Cit)•4H2O (Fluka analytical >99.5%), manganese acetate tetrahydrate Mn(CH3COO)2•4H2O (Merck > 99%) and boric acid (Sigma-Aldrich, > 99,5%) in a ratio 1/3:1:1 were dissolved in 20ml of high purity water. The solution was initially cooled in liq- uid nitrogen and then maintained under vacuum (0.6 Pa) to remove the water by sublimation. As a result, a freeze dried white powder was obtained. The powder was pressed in a pellet and introduced in the furnace for 12 h at 300°C under Ar/H2 95/5 mol% atmosphere. The pellet was then ground and the powder reheated at 400°C for 12 h in the same atmorphere. Li3(Cit) • 4H2O(𝑎𝑞) +Mn(CH3COO)2 • 4H2O(𝑎𝑞) + H3BO3(aq) 𝐅𝐃 → mixture ∆ → LiMnBO3 Eq.2.3 2.1.8 Synthesis of FDR-LiMnBO3/rGO FDR-LiMnBO3/rGO was prepared by a freeze drying route (Eq. 2.4). Lithium citrate tetrahy- drate Li3(Cit)•4H2O (Fluka analytical >99.5%), manganese acetate tetrahydrate Mn(CH3COO)2•4H2O (Merck > 99%) and boric acid (Sigma-Aldrich, > 99,5%) in a ratio 1/3:1:1 were dissolved in 20ml of high purity water. 10 ml of a GO solution (20g/L) was then added to obtain a ratio LiMnBO3/GO of 80:20 wt%. To reduce GO agglomeration, a few mg of LiOH were added to reach a lightly basic pH around 8 and obtain an agglomerate-free fluid solution. The solution was cooled in liquid nitrogen and maintained under vacuum (0.6 Pa) to remove the water by sublimation. As a result, a freeze dried brown powder was obtained. The powder was pressed in a pellet and introduced in the furnace for 12 h at 300°C under Ar/H2 95/5 mol% atmosphere. Li3(Cit) • 4H2O(𝑎𝑞) +Mn(CH3COO)2 • 4H2O(𝑎𝑞) + H3BO3(aq) + GO 𝐅𝐃 → mixture ∆ → LiMnBO3/rGO Eq.2.4 𝑂2 Ar/H 2 Ar/H 2 ~ 39 ~ 2.2 Physico-chemical characterization 2.2.1 X-ray diffraction (XRD) Powder XRD measurements were performed for all samples on a PANanalytical X`Pert PRO system equipped with a Johansson monochromator (Cu-Kα1 radiation, 1.5406 Å) and an X`Celerator linear detector operating in Bragg-Brentano geometry (θ/2θ). The diffraction pat- terns in Chapter 3, Chapter 4 subchapter 4.3 and 4.4 were recorded between 10° and 80° (2θ) with an angular step interval of 0.008° and a time per step of 45s. The diffraction patterns in Chapter 4 subchapter 4.2 were recorded between 10° and 80° (2θ) with an angular step inter- val of 0.004° and a time per step of 150s. The diffraction patterns in Chapter 5 were recorded between 10° and 80° (2θ) with an angular step interval of 0.033° and a time per step of 200s. In-situ high-temperature powder XRD patterns were obtained using a PANalytical X´Pert PRO θ-θ system with Cu-Kα radiation (λ = 1.5418 nm). The powder samples were placed in the heating chamber (XRK 900, Anton Paar, Austria) and a gas flow was applied through the mass flow controller (5850 TR, Brooks instrument, Ger- many). Roughly 0.03 g of sample powder was heated following material-specific temperature program. In chapter 3, the measurement was carried out in N2 gas (99.999%, Messer, Switzer- land) with a flow of 50ml/min from room temperature (RT) to 900°C with a heating rate of 20°C/min. In chapter 4 subchapter 4.2, the measurement was carried out in O2 gas (99.999%, Messer, Switzerland) with a flow of 50 ml/min from room temperature (RT) to 170°C with a heating rate of 1° C/min. In chapter 4 subchapter 4.4, the measurement was carried out in Synthetic Air gas (99.999%, Messer, Switzerland) with a flow of 50 ml/min from room tem- perature (RT) to 700°C with a heating rate of 20 °C/min. Data treatment and crystal phase identification were carried out with X’pert HighScorePlus software. The diffraction data in Chapter 4 subchapter 4.1 were analyzed by a profile match- ing mode (Le Bail fit) as implemented in the program Fullprof. Peak shapes were described by a Thompson-Cox-Hastings pseudo-Voigt function. The crystallite size was extracted from the microstructural analysis within the Fullprof software. The diffraction data in chapter 5 were analyzed to extract the crystallite size by Williamson- Hall method. ~ 40 ~ 2.2.2 Thermogravimetric analysis (TGA) Thermogravimetric analysis experiments were performed with a NETZSCH STA 409 CD thermo-balance coupled with a quadrupole mass-spectrometer (Netzsch QMS 403 C Aeolos) sampling the thermobalance exhaust. The oxygen content of the pristine µ-Ca2MnO4 phase in Chapter 3 subchapter 3.2 was deter- mined from the relative sample weight loss (r). r was measured while heating 0.1 g of sample with 7.5 ºC/min rate from 40 ºC to 1200 ºC under 50 ml/min flow of 5% H2 in He. The oxy- gen content (4-δ) was then calculated from r using Eq. 2.6 based on the reaction stoichiometry as stated in Eq. 2.5: Ca2MnO4-δ (s) + H2 (g)  3Ca2/3Mn1/3O (s) + (1-δ)H2O (g) Eq. 2.5 4-δ = 4-[(1- 4r)MO - 2rMCa - rMMn]/[(1-r)MO] Eq. 2.6 where Mi is the molar mass of the constituent element i. The Ca2/3Mn1/3O (s) solid-solution of the residual was verified using XRD. To specify the H2O-related volatile species present in the H2SO4-treated samples, an addition- al study was made using the sample ‘µ-Ca2MnO4-25%Ca-extr’. The r was measured under both 5% H2 in He and N2 atmospheres. H2O emissions from the sample with temperature were monitored using the mass-spectrometer. The reactively reducing 5% H2 in He atmos- phere is expected to facilitate the extraction of OH- species, while H2O should have atmos- phere-independent extraction behaviour. The same heating program as described above for the pristine phase was employed with both atmospheres (maximum measurement temperature was 800 ºC). TGA data in Chapter 4 subchapter 4.2 were acquired on around 0.10 g of each powder. The sample was placed in an alumina crucible and heated at a rate of 20°C/min to 450°C in N2 with a flow rate of 50 mL/min. 2.2.3 Scanning electron microscopy (SEM) The morphology of the samples was studied with a Scanning Electron Microscope (SEM) FEI Nova NanoSEM 230 using ETD (Everhart-Thornley), TLD (Through-Lens), BSED (Back Scattered electrons) detectors with electron-beam energy of 5-10 keV. The chemical composi- tion was determined by Energy Dispersive X-ray Spectroscopy (EDX) using electron-beam ~ 41 ~ energy of 20 keV in spot mode. The samples were sputtered with carbon or platinum in order to avoid charging effects in the images. 2.2.4 Transmission electron microscopy (TEM) Transmission Electron Microscopy was carried out using a Philips CM30 (300 kV) equipped with a LaB6 filament in bright field image and in selected area diffraction mode. A JEOL JEM 2000FS TEM/STEM (200kV) equipped with a high-angle annular dark field (HAADF) detec- tor and an in-column Omega-type energy filter, was used to assess the morphology and local crystallinity of the samples by low-magnification imaging, and high-resolution electron mi- crographs (HREM). HREM images were further processed in the Digital Micrograph routine. Sample preparation was performed by depositing the powders suspended in ethanol or hexane on a holey carbon grid. 2.2.5 X-ray photoelectron spectroscopy (XPS) X-ray photoelectron spectroscopy (XPS) spectra were acquired on a Physical Electronics (PHI) Quantum 2000 photoelectron spectrometer using monochromatic Al Kα radiation (1'486.6 eV) and a hemispherical capacitor electron-energy analyzer equipped with a channel plate and a position-sensitive detector. The electron take-off angle was 45°. The detail spectra were acquired with a pass energy of 58.7 eV and a step width of 0.250 eV for all measure- ments. The beam diameter was typically 150 µm. A charge neutralizer system was used for all analyses. Data treatment was performed with CasaXPS software. As for the XPS spectra in Chapter 4, subchapter 4.2, the C 1s signal position at 285.0 eV binding energy (BE) was used as an internal standard for calibration of the XPS peak posi- tions. As for the XPS spectra in Chapter 4 subchapter 4.3, the C 1s of the adventitious carbon was not detectable because of the high amount of conductive carbon present in the electrodes. For this reason before and after measurements the peak position was calibrated with the signal of Au 4f7/2 at 83.95 eV as external standard. The samples were loaded on a XPS plate in a glovebox. The plate was closed in a sealed box to avoid contact with air during transportation from the glovebox to the XPS instrument. The plate was extracted from the sealed box and loaded in the pre-chamber under vacuum as fast as possible. ~ 42 ~ 2.2.6 Raman spectroscopy Raman spectra in Chapter 3 were measured on a Renishaw 2000 spectrometer equipped with holographic notch filters for elastic scattering and a CCD array detector. The WiRE software was used to collect the data. Sample powder was loaded onto a glass substrate, and excited with a red laser (632.816 nm). The laser was focused onto the sample using the optical micro- scope. The instrument was calibrated with a Si single crystal (Raman band at 520 cm-1). The spectra were recorded at room temperature with an exposure time of 60 s and an accumulation number of 5. The range analyzed was between 1000 and 100 cm-1. Raman spectra in Chapter 4 subchapter 4.3 were measured using a Bruker Senterra instru- ment. Sample powder was loaded onto a glass substrate, and excited with a green laser (532 nm). The laser was focused onto the sample using a 50 times magnifying objective of the mi- croscope, the laser beam power was 0.2 mW. The spectra were recorded in air, at room tem- perature with an exposure time of 15 s, an accumulation number of 5. The range analyzed was between 2000 and 100 cm-1. 2.2.7 Surface area determination Nitrogen adsorption and desorption experiments were per-formed at −196 °C on a Mi- cromeritics ASAP 2020 Surface Area and Porosity Analyzer. The samples were evacuated for 1 h at 80°C under vacuum. The specific surface area (SSA) was determined using the Brunau- er–Emmett–Teller (BET) model with N2 gas (99.999%, Messer, Switzerland). 2.2.8 Particle size determination Particle-size distribution was qualitatively analyzed by dynamic light scattering spectroscopy (DLS) with a Beckman Coulter LS230 Laser Diffraction Particle Analyzer (0.04 µm – 2000 µm) equipped with polarisation intensity differential scattering (PIDS) technology. Prior to measurement, the samples were suspended in water for Ca2MnO4 and toluene for Li3MnO4 and ultrasonicated for 5 min (Bandelin Sonoplus HD2200). ~ 43 ~ 2.3 Electrochemical characterization 2.3.1 The equipment used for electrochemical measurements Galvanostatic measurements were monitored by ASTROL, a computer software by Astrol Electronic AG. A potentiostat (BAT-SMALL battery cycler) was linked to a personal com- puter running Windows 7 by a serial-to-analog converter and by a serial cable. The assembled batteries were in-house built, two electrode cells. The technical drawing of the test cells is displayed in Fig. 1. The description of all parts and the used materials are given in Table 1. The inner vertical stack was composed of a titanium current collector containing the EAM acting as working electrode, a polypropylene (PP) Cellgard 2400 (25 µm Polypropylene mon- olayer membrane) separator to prevent an internal short cut by lithium dendrites, a silica foam spacer filled with Merck® LP 50 electrolyte (1M LiPF6 in EC/EMC 1:1) or LP30 electrolyte (1M LiPF6 in EC/DMC 1:1) and covered with the lithium anode (Alfa Aesar 99.9% metal ba- sis) acting as reference and counter electrode. The battery test cells were assembled inside an argon filled dry glove box under inert condi- tions(< 0.1% H2O and < 0.1% O2). The potential values given through the text are referenced against Li+/Li. Measurements were usually acquired at 10 A/kg or 50 A/kg. Figure 2.1: Technical drawing of the in‐house test cell design. ~ 44 ~ 2.3.2 Electrode preparation Electrodes for electrochemical measurements in Chapter 3 were prepared by first mixing 60wt% of the electroactive material (EAM) with 30wt% of carbon black (Super P® Li from TIMCAL) and 10wt% of polyvinylidene difluoride (PVDF, Fluka powder Mw ~530k) in tet- rahydrofuran (THF, Alfa Aesar 99%). This suspension was then dispersed ultrasonically for a few minutes until a dark homogenous suspension was obtained. The desired volume was transferred to a mortar and gently stirred with a pestle at room temperature. The THF evapo- ration leads to a very dark viscous solution that was then drop casted on the titanium current collector. The electrode was dried in air before being heat treated at 100 °C for 10 minutes. Electrodes containing 5 to 10 mg EAM were prepared by this method. Electrodes for electrochemical measurements in Chapter 4 were prepared with 60/30/10 ratio between EAM, carbon and PVDF in toluene (Alfa Aesar 99%, anhydrous). After ultrasoni- cation the desired volume was drop casted directly on the titanium current collector. The elec- trodes were dried in air before being heat treated at 80 °C for 1h in a vacuum oven. Electrodes containing 3 to 4 mg EAM were prepared by this method. Electrodes for electrochemical measurements in Chapter 5 were prepared as follow: for FDR- LiMnBO3 60/30/10 ratio was used between EAM, carbon and PVDF in a solution THF:Toluene 1:1. In the case of FDR-LiMnBO3/rGO the powder was directly suspended in THF. These suspensions were then dispersed ultrasonically and drop casted directly on the ti- tanium current collector. The electrodes were dried in air before being heat treated at 80 °C for 1h in a vacuum oven. ~ 45 ~ Chapter 3 3. Manganese in octahedral coordination: activation of Ca2MnO4 for Li intercalation 3.1 Introduction Mn-based Ruddlesden-Popper phases are promising materials for new battery cathodes due to their layered structure and often benign composition. The Ruddlesden-Popper (RP) phases are highly accommodating for transition-metal-site substitution [78-80], and form distinctly lay- ered structures that are susceptible to Li insertion [81, 82]. The calcium manganate RP phases with the formula Can+1MnnO3n+1 (n=1,2 etc) or CaO-(CaMnO3)n (n=1,2,3...∞) are based on an array of corner sharing MnO6 octahedra. When n=∞, the typical orthorhombic perovskite structure is formed. When n=1, Ca2MnO4 forms the two dimensional (2D) K2NiF4-type struc- ture with CaMnO3 perovskite blocks alternating with CaO rock salt layers [72] (Fig. 3.1). The synthesis of Ca2MnO4 is usually performed by solid state reaction or sol-gel reaction. Figure 3.1:Structural representation of the Ruddlesden-popper Ca2MnO4. ~ 46 ~ Ca2MnO4 is very interesting since Ca and Mn are non-toxic and readily available elements and also Mn is an easily reducible/oxidable ion. On the other hand its extremely high resistivi- ty [83] (>20 MΩ) prevents the use as battery material. An acid treatment was designed to activate the material by extracting calcium, i.e. creating cation vacancies, in order to further facilitate lithium intercalation. Most reactions designed to extract/exchange interlayer ions rely on HCl [84]. However, the strong diprotic H2SO4 acid (with the double negatively charged SO4 2−counter ion) should facilitate the Ca2+ extraction [85]. To our knowledge, Ca2MnO4 based oxides are not used as cathode material due to their poor performance. Therefore the aim of this research was to explore possibilities to enhance the electrochemical activity of pristine low cost Ca2MnO4 with an acid-treatment. 3.2 Acid treatment of µ-Ca2MnO4 and characterization Pristine µ-Ca2MnO4 shows a basic behavior in high purity water (dissolution of 0.05 g of the compound in 0.05 L water results in a pH of 9.7) suggesting an easy removal of O2- ions. When the acid treatment was carried out, Ca2+ ions were also extracted from the structure. The sulphate anions act as limiting reagent. The total calcium oxide extraction is explained by the following equation (Eq. 3.1): OxHxCaSOMnOCaSOxHMnOCa xx 24)4(24242    Eq. 3.1 The extracted Ca2+ ions reacted with sulphate anions, therefore CaSO4 is recovered in the dried filtrate, as identified by XRD. The O2- ions reacted with hydronium ions 𝑂2− + 2𝐻3𝑂 + → 3𝐻2𝑂 leading to a neutral pH at the end of the extraction reaction for all cases. Consequently complete H2SO4 consumption during the reaction was concluded. The theoreti- cal calcium extraction was calculated from the mass of the pristine µ-Ca2MnO4-δ used, the ini- tial pH and the volume of solution based on Equation 3. The XRD pattern of the pristine Ruddlesden-Popper µ-Ca2MnO4 and the Ca 2+-extracted com- pounds are shown in Figure 3.2. ~ 47 ~ 10 20 30 40 50 60 70 80 (3 1 0 ) (2 0 8 ) (4 0 0 ) (2 2 1 2 ) (3 1 6 ) (2 0 1 2 ) (2 2 4 ) (2 2 0 ) (0 0 1 2 ) (1 1 1 0 ) (2 0 0 ) (1 1 6 ) (1 1 2 ) (e) 2(degree) In te n s it y n o rm . (a .u ) (a) (b) (c) (d) (0 0 4 ) Figure 3.2: XRD pattern of µ-Ca2MnO4 pristine (a), µ-Ca2MnO4-25%Ca-extr (b), µ-Ca2MnO4-50%Ca-extr (c), µ-Ca2MnO4-75%Ca-extr (d), µ-Ca2MnO4-90%Ca-extr (e). All reflections in Figure 3.2(a) were indexed with the Ca2MnO4, K2NiF4-type structure be- longing to tetragonal crystal system (space group I41/acd). The relative weight loss resulting from reductive TG measurement was measured as r = 7.91(5) % indicating an oxygen stoi- chiometry for the pristine phase as (4-δ) = 3.98(1). A tendency towards oxygen deficiency is typical for the oxide compounds of mid- and late-transition 3d metals having several relative- ly stable oxidation states [86-89]. The XRD patterns of the µ-Ca2MnO4-25%Ca-extr and µ- Ca2MnO4-50%Ca-extr compounds (Figure 3.2(b) and Figure 3.2(c)) revealed no perceptible 2θ-shift in comparison to the pristine compound. However, the peaks intensity decreases for the µ-Ca2MnO4-75%Ca-extr and µ-Ca2MnO4-90%Ca-extr (Figure 3.2(d) and 3.2(e)) com- pounds. Additional XRD analysis was carried out by in situ high-temperature XRD (HT- XRD) experiments (Figure 3.3). ~ 48 ~ 10 20 30 40 50 60 70 80 (3 1 0 ) after cooling 25°C 2(degree) In te n s it y n o rm . (a .u .) 900°C (a) (2 0 8 ) (4 0 0 ) (2 2 1 2 ) (3 1 6 ) (2 0 1 2 ) (2 2 4 ) (2 2 0 ) (0 0 1 2 ) (1 1 1 0 ) (2 0 0 ) (1 1 6 ) (1 1 2 ) (0 0 4 ) 10 20 30 40 50 60 70 80 (3 1 0 )             after cooling 900°C 25°C 2(degree) In te n s it y n o rm . (a .u .) (b)  (2 0 8 ) (4 0 0 ) (2 2 1 2 ) (3 1 6 ) (2 0 1 2 ) (2 2 4 ) (2 2 0 ) (0 0 1 2 ) (1 1 1 0 ) (2 0 0 ) (1 1 6 ) (1 1 2 ) (0 0 4 ) Figure 3.3: XRD pattern of µ-Ca2MnO4 (a) and µ-Ca2MnO4-25%Ca-extr (b) before heating, at 900°C and after cooling. The peaks corresponding to the impurity phases are marked with asterisks. No phase transitions or evolution of impurity phases were observed. Aside from an expected shift in diffraction-peak position due to thermal expansion, the patterns were identical before, during, and after heating (Figure 3.3(a)). The diffraction pattern of the µ-Ca2MnO4-25%Ca-extr compound unmistakeably showed the formation of additionally Ca2Mn2O5, CaMn2O4 and Mn2O3 phases starting at around T=800 °C, which remain present upon cooling (Figure 3.3(b)). The diffraction peaks of µ-Ca2MnO4 were not influenced by the formation of these phases during the heating and cooling process indicating that the impurity phases crystallized from the amorphous component present in the sample were not detectable by RT-XRD. These phases have a lower Ca/Mn stoichiometric ra- tio than the pristine compound indicating calcium losses upon the acid treatment. ~ 49 ~ Thermal analysis was performed for the µ-Ca2MnO4-25%Ca-extr compound (Figure 3.4) to specify the volatile H2O-related species possibly intercalated into the sample crystal structure during the Ca2+ extraction process. 100 200 300 400 500 600 700 800 86 88 90 92 94 96 98 100 3 rd step 2 nd step W e ig h t( % ) Temperature (°C) N 2 5% H 2 in He 1 st step Figure 3.4: Relative weight loss with respect to T of the µ-Ca2MnO4-25%Ca-extr compound under N2 and 5% H2 in He atmospheres. Under N2 a two-step weight loss associated with an extraction of H2O was observed between T = 100 °C and 600 °C, while under 5% H2 in He a third step is added, related to the reduc- tion of the sample into Ca1-xMnxO. The first step terminates at 230 ºC regardless the em- ployed atmosphere, implying the accommodated species to be a coordinated H2O. The second weight-loss step, on the other hand, appears to be facilitated under the reductive atmosphere, which is understood on the basis of the easier convertibility of the OH- species into extracta- ble H2O in the presence of H2 (g). Similar observations of the decomposition behaviour were made by others for the MnO2 hydrate prepared reducing NaMnO4 in an aqueous solution of fumaric acid and H2SO4 [90]. In the case of MnO2 the first weight reduction from the as- prepared compound concluding at slightly higher temperature was understood as the extrac- tion of crystal water, while the second step was considered to be related with oxygen extrac- tion due to conversion of the MnO2 into Mn2O3. SEM images of the samples (Figure 3.5) showed that the microstructure of pristine µ- Ca2MnO4 and acid-treated samples consisted of agglomerates of irregularly shaped particles smaller than 1 micron. The morphology of higher calcium extraction samples (>50at%) was altered. Additional roughness on the surface of the samples was observed. In general, the ag- glomerates and particle dimensions were similar before and after acid treatment. ~ 50 ~ Figure 3.5: SEM images of µ-Ca2MnO4 pristine (a), µ-Ca2MnO4-25%Ca-extr (b), µ-Ca2MnO4-50%Ca-extr (c), µ-Ca2MnO4-75%Ca-extr (d), µ-Ca2MnO4-90%Ca-extr (e), high magnification of µ-Ca2MnO4-75%Ca-extr. Local chemical compositions were obtained by SEM/EDX. EDX spectra were recorded in different zones of the samples. Qualitative analysis confirmed the presence of a constant cal- cium manganese ratio on each spot showing the respective Kα1 and Kβ1 lines. Although no chemical inhomogeneity was observed within each sample, a relative quantification of the atomic ratio between calcium and manganese in the pristine and the Ca2+-extracted com- ~ 51 ~ pounds showed a significant difference. The quantification is in agreement with the theoreti- cal Ca2+ extraction calculated from ionic equilibrium. The results are summarized in Table 3.1. Table 3.1: EDX quantification Compound Ca %at Mn %at Theoretical ratio Ca:Mn Experimental ratio Ca:Mn (EDX) µ-Ca2MnO4 pristine 67.1±0.3 32.9±0.3 2 2.04 µ-Ca2MnO4-25%Ca-extr 59.2±0.5 40.8±0.5 1.5 1.44 µ-Ca2MnO4-50%Ca-extr 51.7±0.6 48.3±0.6 1 1.07 µ-Ca2MnO4-75%Ca-extr 34.7±0.5 65.7±0.5 0.5 0.53 µ-Ca2MnO4-90%Ca-extr 14.5±0.3 85.5±0.3 0.2 0.17 DLS measurements were performed on all samples to determine the particle size distribution. The major particle size of between 2 µm and 8 µm was found in all samples. Combined SEM and DLS indicate the samples consist of particles having dimensions of around 1 micron along with larger agglomerates. It was found no difference in particle size as a result of acid treatment. The TEM images of pristine and µ-Ca2MnO4-50%Ca-extr compounds are shown in Figure 3.6. The pristine compound consisted of particles with well-defined facets. However, the µ- Ca2MnO4-50%Ca-extr compound had a rough surface suggesting that an amorphous coverage was formed. This result was in agreement with SEM characterisation. Figure 3.6: TEM images of µ-Ca2MnO4 pristine (a) and µ-Ca2MnO4-50%Ca-extr (b). ~ 52 ~ Raman spectra were recorded on all samples to clarify the composition of the amorphous cov- erage (Figure 3.7). 1000 800 600 400 200 1000 800 600 400 200 Raman shift (cm -1 ) In te n s it y ( a .u .) (f) (a) (b) (c) (d) (e) Figure 3.7: Raman spectra of µ-Ca2MnO4 pristine (a), µ-Ca2MnO4-25%Ca-extr (b), µ-Ca2MnO4-50%Ca-extr(c), µ-Ca2MnO4-75%Ca-extr (d), µ-Ca2MnO4-90%Ca-extr (e) and amorphous MnO2·xH2O (f). The spectrum of pristine µ-Ca2MnO4 (Figure 3.7(a)) showed a main peak centred at 560 cm -1 related to the ν(Mn–O) stretching vibration of Mn4+ in Oh coordination and small peaks be- tween 180 cm-1 and 400 cm-1 related to skeletal vibrations [91]. The two main peaks centered at at 630 cm-1 and 560 cm-1 in the hydrated amorphous MnO2 spectrum were also due to the ν(Mn–O) stretching vibration of Mn4+ in Oh coordination [91]. These two spectra can be seen as the end members of the calcium-extracted compound series, while the others represent in- termediate states between the two extreme compositions. Looking at Figure 3.7 from the top to the bottom, namely increasing the quantity of extracted calcium, the peak present at 630 cm-1 in the MnO2 spectra was correlated to the MnO2 content on the compounds with extract- ed Ca2+. The intensity of the band at 630 cm-1 seemed to be directly related to the amount of extracted Ca2+ from µ-Ca2MnO4. To summarize, the calcium deficiency was a function of the amount of acid used in the reac- tion as confirmed by SEM/EDX measurements, while XRD confirmed the presence of the ~ 53 ~ pristine Ca2MnO4 phase. The deficiency was indirectly confirmed in HT-XRD by the for- mation of new phases with Ca/Mn ratios lower that in the pristine material after heating up at around 800 °C. They were attributed to two factors: 1) the reduction of MnO2 to Mn2O3 and 2) the enhanced diffusion of calcium from the Ca-rich phase to the manganese oxide at high temperature. An amorphous hydrated manganese dioxide MnO2·xH2O coverage was identi- fied by Raman spectroscopy. The lower intensity of the µ-Ca2MnO4-75%Ca-extr and µ- Ca2MnO4-90%Ca-extr reflections in the XRD pattern was attributed to a decrease of the crys- talline primary phase in the centre of the particles as a consequence of formation of the outer amorphous layer. The amorphization of the surface was observed also by TEM and by SEM for high calcium extraction. The overall characterization carried out on acid-treated samples revealed the presence of an amorphous material distributed uniformly on the surface of the crystalline particles. This treatment attacked the surface of the particles, but the global morphology was preserved. The removal of calcium and consequently of oxygen provoked a collapse of the structure with a loss of the long range order in the surface layers, while the crystallinity of the bulk remained the same as in the pristine materials. Electrochemical measurements were carried out to study the lithium intercalation in the acid- treated compounds. The first galvanostatic cycle for all compounds is shown in Figure 3.8. 0 50 100 150 200 250 1.5 2.0 2.5 3.0 3.5 4.0 0 50 100 150 200 250 (a) Capacity (Ah/kg) V o lt a g e ( V v s L i+ /L i) 1 st discharge 0 50 100 150 200 250 1.5 2.0 2.5 3.0 3.5 4.0 0 50 100 150 200 250 (b) Pristine µ-Ca 2 MnO 4 µ-25% Ca-extr µ-50% Ca-extr µ-75% Ca-extr µ-90% Ca-extr MnO 2 xH 2 O Blend Ca 2 MnO 4 +MnO 2 1 st charge Capacity (Ah/kg) V o lt a g e ( V v s L i+ /L i) Figure 3.8: Discharge (a) and charge (b) profile for the 1st cycle between 1.5 V and 4.2 V at 10A/kg of µ- Ca2MnO4 pristine (black), µ-Ca2MnO4-25%Ca-extr (red), µ-Ca2MnO4-50%Ca-extr (blue), µ-Ca2MnO4-75%Ca- extr (green), µ-Ca2MnO4-90%Ca-extr (magenta), amorphous MnO2·xH2O (brown) and µ-Ca2MnO4/MnO2·xH2O blend(orange). ~ 54 ~ The pristine compound showed negligible electrochemical activity in the selected potential range. Although a discharge capacity of around 4 Ah/kg was measured, it is unlikely due to an intercalation reaction. All the acid-treated compounds showed identical discharge profile, but a different capacity depending on the amount of calcium extracted. An approximate discharge capacity of 40 Ah/kg could be achieved for each 25 at% calcium extraction. The lithium in- tercalation is reversible. During charging, roughly 30 Ah/kg can be deintercalated for each 25 at% calcium extracted material. The shape of the charge curve is similar for all acid-treated compounds. Upon charging, the measured capacity value is in good agreement with the de- pendency on the amount of calcium extracted, as observed during the discharge. A large dif- ference between the first and second discharge was measured and capacity losses between 20% and 40% were observed for all compounds (Table 3.2). This behaviour was more promi- nent when increasing the calcium extraction and was also present in amorphous manganese oxide. The 30th cycle was selected for comparison with the 1st cycle since all samples have reached the maximum of coulombic efficiency (>95%) and showed stable capacity values in the fol- lowing cycles (Figure 3.9). The discharge and charge curves show the amorphous-like profile as observed during the 1st cycle. The lithium intercalation is still reversible. The capacities are lower than at the 1st cycle although in accordance with the amount of calcium extracted. 0 10 20 30 40 50 60 70 80 1.5 2.0 2.5 3.0 3.5 4.0 0 10 20 30 40 50 60 70 80 Pristine Ca 2 MnO 4 25% Ca-extr 50% Ca-extr 75% Ca-extr 90% Ca-extr MnO 2 xH 2 O Blend Ca 2 MnO 4 +MnO 2 Capacity (Ah/kg) V o lt a g e ( V v s L i+ /L i) (a) (b) 0 10 20 30 40 50 60 70 80 1.5 2.0 2.5 3.0 3.5 4.0 0 10 20 30 40 50 60 70 80 Capacity (Ah/kg) V o lt a g e ( V v s L i+ /L i) 30 th discharge 30 th charge Figure 3.9: Discharge (a) and charge (b) profile for the 30th cycle between 1.5 V and 4.2 V at 10A/kg of µ- Ca2MnO4 pristine (black), µ-Ca2MnO4-25%Ca-extr (red), µ-Ca2MnO4-50%Ca-extr (blue), µ-Ca2MnO4-75%Ca- extr (green), µ-Ca2MnO4-90%Ca-extr (magenta), amorphous MnO2·xH2O (brown) and µ-Ca2MnO4/MnO2·xH2O blend(orange). ~ 55 ~ A plot of discharge capacity vs. number of cycles for all compounds is shown in Figure 3.10(a). 0 10 20 30 40 50 60 70 0 25 50 75 100 125 150 175 200 225 250 C a p a c it y ( A h /k g ) Cycle number (a) 0 10 20 30 40 50 60 70 0 20 40 60 80 100 120 C a p a c it y ( A h /K g ) Cycle number 0 10 20 30 40 50 60 70 0 25 50 75 100 125 150 175 200 225 250 C a p a c it y ( M n O 2 c o n te n t) ( A h /k g ) Cycle number (b) Figure 3.10: (a) Discharge capacity vs cycle at 10A/kg of µ-Ca2MnO4 pristine (■, black), µ-Ca2MnO4-25%Ca- extr (▲, red), µ-Ca2MnO4-50%Ca-extr (▼, blue) , µ-Ca2MnO4-75%Ca-extr (◄, green), µ-Ca2MnO4-90%Ca- extr (►, magenta) and amorphous MnO2·xH2O (●, brown). Inset: cathodic specific charge vs cycle of µ- Ca2MnO4-50%Ca-extr (▼, blue) and µ-Ca2MnO4/MnO2·xH2O blend (♦, orange). (b) Discharge capacity vs cy- cle normalized to MnO2 content for all compounds. A very fast degradation occurred for amorphous manganese oxide MnO2·xH2O. The capacity retention on this compound after 35 cycles was almost zero. As for the calcium-extracted compounds, they reached the stability after 30 cycles delivering a capacity of ca. 20 Ah/kg for µ-Ca2MnO4-25%Ca-extr, 40 Ah/kg for µ-Ca2MnO4-50%Ca-extr and 60 Ah/kg for both µ- Ca2MnO4-75%Ca-extr and µ-Ca2MnO4-90%Ca-extr. The capacity for lithium-ion intercala- ~ 56 ~ tion on the Ca-deficient compounds was lower when compared to the initial value of amor- phous MnO2·xH2O, while the capacity retention under cycling was significantly improved. Aiming to better understand the stabilizing effect of the crystalline core of the acid treated compounds under cycling, a blend was prepared by mixing pristine µ-Ca2MnO4 and MnO2·xH2O for comparison. The inset in Figure 3.10(a) shows the plot of discharge capacity vs. number of cycles considering only µ-Ca2MnO4-50%Ca-extr and the µ- Ca2MnO4/MnO2·xH2O blend. As expected the discharge capacity of the blend electrode dur- ing the first discharge was almost half of the pure MnO2·xH2O since only half of manganese was electrochemically active. The degradation of the blend followed the same trend of MnO2·xH2O and the capacity retention was much lower than µ-Ca2MnO4-50%Ca-extr. The analysis of the electrochemical results for all compounds is summarized in Table 3.2. Table 3.2: Resume of the electrochemical data for all compounds In Figure 3.10(b) the same data were plotted assuming that only the percentage of calcium ex- tracted material gave rise to electrochemical activity. The compounds having 25at%, 50at%, 75at% and 90at% Ca extraction, normalized to 100at% MnO2-content, showed a stable dis- charge capacity around 90-100 Ah/kg over 70 cycles. Taking into account the characterization and the electrochemical measurements discussed above, the following model was proposed to explain the electrochemical behaviour: the acid treatment removed calcium oxide from the samples leaving a crystalline bulk of µ-Ca2MnO4 covered with an amorphous MnO2·xH2O layer. XRD measurements on compounds obtained Compound Specific Charge [Ah/kg] % Specific charge lost between 1st and 2nd cycle % Capacity retention after 70 cycles 1 st cycle discharge 70th cycle discharge µ-Ca2MnO4 pristine 4 4 n.d n.d µ-Ca2MnO4-25%Ca-extr 45 20 18 44 µ-Ca2MnO4-50%Ca-extr 85 41 25 48 µ-Ca2MnO4-75%Ca-extr 117 59 43 50 µ-Ca2MnO4-90%Ca-extr 185 61 29 33 MnO2·xH2O 220 8 21 4 Blend µ-Ca2MnO4 + MnO2·xH2O 102 15 10 15 ~ 57 ~ after electrochemical cycling revealed that the crystal structure of the crystalline part of the sample did not change (Fig. 3.11). Since no differences were observed in the XRD pattern be- fore and after battery cycling, we conclude that lithium ion intercalation occurs only into the amorphous structure formed on the surface. 10 20 30 40 50 60 70 80 -50% Ca-extr In te n s it y ( a .u .) 2(degree) Figure 3.11: XRD pattern of µ-Ca2MnO4-50%Ca-extr after 70 cycles. Higher calcium extraction causes a higher quantity of manganese oxide on the surface of the particles and therefore more available sites for lithium ion intercalation. The electrochemical activity data of the first cycle confirmed the structural finding that the active compound con- sisted of amorphous MnO2·xH2O (Figure 3.10b). However, the crystalline bulk appeared to have a stabilizing influence on the capacity retention under cycling. A considerable improve- ment of the stability was noticed on the Ca-extracted compounds, compared to bare amor- phous MnO2·xH2O and the µ-Ca2MnO4/MnO2·xH2O blend. The functionalization of the surfaces of the particles obtained by controlled calcium extraction was synthesis-dependent and cannot be obtained by mixing the two different materials having different properties. With this bifunctional crystalline-amorphous structure, composed by a µ- Ca2MnO4 bulk phase for the stability and an amorphous MnO2·xH2O surface for the electro- chemical response, it was possible to reach a stability improvement by a factor of 10. After 70 cycles this structure was still stable and capacity retention around 35%-50% is reached for all compounds. This new structure leads to an increased stability during charge and discharge cy- cling of amorphous MnO2·xH2O as a host for lithium intercalation in Li-ion battery cathodes. ~ 58 ~ 3.3 Acid treatment of n-Ca2MnO4: influence of the particle size and comparison with µ-Ca2MnO4 In the previous subchapter Ca2MnO4 having micron-sized particles has been already synthe- sized, modified for lithium intercalation, and characterized electrochemically. It is well known, however, that lowering the particle size increases the electrode/electrolyte contact ar- ea improving total capacity and rate capability [92], especially in materials with low conduc- tivity. For this reason, Ca2MnO4 was synthesized as nanoparticles and subsequently activated removing calcium ions from the surface of the samples by acid treatment similar to the treat- ment described in subchapter 3.2. The activation of n-Ca2MnO4 towards lithium intercalation was performed in two steps. In the first step, nanoparticles of Ca2MnO4 were obtained by Pechini method. Synthesis parameters as CM (CA to metal ratio) and CE (CA to EG ratio) [93] and calcination temperature were set to obtain the smallest particle size possible. In the second step, n-Ca2MnO4 was treated with H2SO4. The XRD patterns of the pristine n-Ca2MnO4 and the Ca 2+-extracted compounds are shown in Fig. 3.12. All reflections in Fig. 1(a) were indexed with the Ca2MnO4, K2NiF4-type structure. The XRD patterns of all Ca2+-extracted compounds revealed no 2θ-shift in comparison to the pristine compound. However, a decrease in crystallinity and consequently of the peak intensi- ties can be observed for n-Ca2MnO4 -50%Ca-extr and n-Ca2MnO4 -75%Ca-extr. 10 20 30 40 50 60 70 80 (d) (c) (b) 2(degree) In te n s it y n o rm . (a .u ) (3 1 0 ) (2 0 8 ) (4 0 0 ) (2 2 1 2 ) (3 1 6 ) (2 0 1 2 ) (2 2 4 )( 2 2 0 ) (0 0 1 2 ) (1 1 1 0 ) (2 0 0 ) (1 1 6 ) (1 1 2 ) (0 0 4 ) (a) Figure 3.12: XRD patterns of Ca2MnO4 pristine (a), Ca2MnO4 -25%Ca-extr (b), Ca2MnO4-50%Ca-extr (c), and Ca2MnO4-75%Ca-extr (d). ~ 59 ~ Figure 3.13 shows the SEM micrograph of the pristine n-Ca2MnO4. It is composed of nano- particles having irregular shape and dimension between 50 and 100 nm. EDX spectra were recorded in different zones of the samples. Quantitative analysis showed a Ca to Mn ratio in very good agreement with the theoretical one (Table 1) highlighting the occurred calcium ex- traction. SEM micrographs of the Ca2+-extracted compound showed a higher degree of ag- glomeration but no changes in overall particles dimension. Figure 3.13: SEM micrographs of (a) n-Ca2MnO4 pristine and (b) n-Ca2MnO4-50%Ca-extr. Table 3.3: EDX quantification. Compound Ca %at Mn %at Theoretical ratio Ca:Mn Experimental ratio Ca:Mn (EDX) n-Ca2MnO4 pristine 67.5±0.5 32.5±0.5 2 2.08 n-Ca2MnO4-25%Ca-extr 60.4±0.5 39.6±0.5 1.5 1.52 n-Ca2MnO4-50%Ca-extr 50.0±0.6 50.0±0.6 1 1 n-Ca2MnO4-75%Ca-extr 33.0±0.6 67.0±0.6 0.5 0.49 TEM images of pristine and n-Ca2MnO4-50%Ca-extr compound are shown in Fig.3.14. n- Ca2MnO4 was composed of crystalline nanoparticles. Lattice planes could be clearly identified up to the surface of the particles. On the other side, n-Ca2MnO4-50%Ca-extr particles showed an inner crystalline bulk and an amorphous outer layer separated by a well-defined interface. ~ 60 ~ Figure 3.14: TEM images of (a) n-Ca2MnO4Ca-extr pristine and (b) n-Ca2MnO4-50%Ca-extr . Line scan EDX analysis was carried out by STEM/EDX to study the change of the Ca:Mn ra- tio in the outer amorphous layer and in the inner crystalline bulk (Fig. 3.15). The Ca:Mn ratio in the outer layer was approximately 0.6 with a composition of Ca 37.5 at% and Mn 62.5 at%. The Ca:Mn ratio changed abruptly in the first 15 nm approach to 1.7 when the inner bulk was reached (Ca 62.5 at% and Mn 37.5 at%). The experimental Mn content of the inner bulk was lightly underestimated due to 3D effect of the measurements but still close to the theoretical value. These results indicate that the inner bulk is constituted of Ca2MnO4 and the calcium extraction involved the surface of the particles. Figure 3.15: (a) BF-STEM image of n-Ca2MnO4-50%Ca-extr and (b) line scan EDX quantification. ~ 61 ~ The calcium extraction, as confirmed by SEM/EDX and STEM/EDX, should theoretically give rise to manganese oxide with formula CaxMnO2 on the surface of the samples. To deter- mine the composition of the amorphous outer layer, Raman measurements were carried out on all samples (Fig.3.16). Amorphous hydrated manganese oxide (MnO2•xH2O) was synthesized for comparison. The spectrum of pristine n-Ca2MnO4 showed a main peak centered at 560 cm−1 related to the ν(Mn–O) stretching vibration of Mn4+ in Oh coordination and small peaks between 180 cm−1 and 400 cm−1 related to skeletal vibrations [94] . 1000 800 600 400 200 1000 800 600 400 200 In te n s it y ( a .u .) Raman shift (cm -1 ) (a) (b) (c) (d) (e) Figure 3.16: Raman spectra of n-Ca2MnO4 pristine (a), n-Ca2MnO4 -25%Ca-extr (b), n-Ca2MnO4-50%Ca-extr (c), n-Ca2MnO4-75%Ca-extr (d) and MnO2•xH2O. Increasing the calcium extraction, namely increasing the amount of amorphous material on the surface of the samples, a shoulder started to appear around 630 cm−1. This shoulder was correlated to the peak present at 630 cm−1 in the MnO2•xH2O reference spectra. The intensity of the band at 630 cm−1 correlated well with the amount of extracted Ca2+ from Ca2MnO4 therefore to the amount of MnO2•xH2O on the surface of the samples. Lithium intercalation was studied by galvanostatic measurements at 10Ah/kg and the results were compared with amorphous hydrated MnO2. The voltage profile for the first discharge of the pristine n-Ca2MnO4 and the Ca 2+-extracted compound is shown in Fig.3.17(a). Ca2MnO4 ~ 62 ~ shows a discharge capacity of 24 Ah/kg which is probably due to capacitive charge on the surface of the particles. All the Ca2+-extracted compounds showed an identical discharge pro- file with a voltage onset around 3.5 V, and an intercalation voltage between 3.5 V and 2.5 V. The capacity depended on the amount of calcium extracted: an approximate discharge capaci- ty of 50 Ah/kg could be achieved for each 25 at.% calcium extraction. Ca2+-extracted com- pounds and amorphous MnO2•xH2O have the same discharge profile. For this reason the ac- tive compound was identified as MnO2•xH2O. 0 25 50 75 100 125 150 175 200 225 1.5 2.0 2.5 3.0 3.5 4.0 n-Ca2MnO4 pristine n-Ca 2 MnO 4 -25% n-Ca 2 MnO 4 -50% n-Ca 2 MnO 4 -75% -Ca 2 MnO 4 pristine -Ca 2 MnO 4 -25% -Ca 2 MnO 4 -50% -Ca 2 MnO 4 -75% a-MnO 2 xH 2 O (a) Capacity (Ah/kg) V o lt a g e ( V v s L i+ /L i) 0 5 10 15 20 25 30 0 50 100 150 200 250 (b) Cycle number C a p a c it y ( A h /k g ) n-Ca 2 MnO 4 pristine n-Ca 2 MnO 4 -25% n-Ca 2 MnO 4 -50% n-Ca 2 MnO 4 -75% a-MnO 2 xH 2 O Figure 3.17: Discharge profile (a) for the 1st cycle between 1.5 V and 4.2 V at 10 A/kg and discharge capacity vs cycles graph (b) of n-Ca2MnO4 pristine (black), n-Ca2MnO4-25% (red), n-Ca2MnO4-50% (blue), n-Ca2MnO4- 75% (green) and amorphous MnO2•xH2O (brown). ~ 63 ~ The stability of the compounds was studied over 30 cycles (Fig.3.17(b)). As can be seen from Fig. 3.17b the acid treatment significantly improved capacity retention (Table 3.4). n- Ca2MnO4-25%Ca-extr, n-Ca2MnO4-50%Ca-extr and n-Ca2MnO4-75%Ca-extr delivered 46, 60 and 70 Ah/kg respectively after 30 cycles. Although amorphous hydrated MnO2 showed higher capacity during the first discharge, it undergoes a very fast degradation during cycling with only 9% capacity retention. XRD was also carried out after cycling and no changes were observed in comparison with the material before cycling. Table 3.4: Resume of the electrochemical data for all compounds. The overall physicochemical characterization carried out on pristine and Ca2+-extracted com- pounds revealed a calcium extraction which was a function of the amount of acid used in the synthesis. The Ca2+ extraction, simultaneously with O2-extraction, gave rise to an amorphiza- tion of the particles surface exposed to the acid, creating an outer amorphous layer of few na- nometers. The inner part of the particles was preserved as crystalline Ca2MnO4. The interface between amorphous and crystalline part of the particles is sharp and the calcium concentration changes in function of the distance from the surface. The outer layer is probable composed of Ca2+-containing amorphous hydrated MnO2 (Fig. 3.18). Compound Specific Charge [Ah/kg] % Capacity retention after 30 cycles 1st cycle discharge 30th cycle discharge n-Ca2MnO4 pristine 26 31 - n-Ca2MnO4-25% 64 39 60 n-Ca2MnO4-50% 131 61 46 n-Ca2MnO4-75% 150 104 70 MnO2·xH2O 220 20 9 ~ 64 ~ Fig.3.18: Graphical representation of the calcium extracted materials. Similar results were obtained for Ca2MnO4 having micron-sized particles. In that case, the in- ner crystalline bulk was preserved for calcium extraction up to 90 at%. However, the presence of big agglomerates did not allow for a uniform amorphization of the particle surface. Taking into account also the electrochemical characterization we explain our findings as fol- lows: lithium intercalation in acid-treated materials occurred only in the amorphous outer lay- er created on the surface of the particles. The more Ca2+ was extracted, the bigger was the amorphous layer and the higher was the capacity of the material. The inner crystalline bulk which was kept intact after acid treatment had a stabilizing influence on capacity retention under cycling. This amorphous-crystalline structure allowed reaching a stability improvement by a factor of 7 in comparison with amorphous MnO2·xH2O. n-Ca2MnO4 in comparison with micro-Ca2MnO4 showed higher capacity (Fig.3.17a dotted lines) for the same amount of calcium extracted. The capacity improvement was correlated to the particles dimension. Due to the smaller particle size, after the acid treatment, the amor- phous electroactive part of the material had a higher surface area exposed to the electrolyte which allowed for a deeper lithiation. At the same time, however, the inner crystalline bulk, which had a stabilizing influence, was much smaller giving rise to lower capacity retention. ~ 65 ~ Chapter 4 4. Manganese in tetrahedral coordination: Li3MnO4 as cathode material. 4.1 Introduction Recently, Saint at al. [67] proposed the possibility to extend the range of the manganese redox couples used in cathodes to Mn4+/Mn5+ or even higher, using compounds containing oxyan- ions 𝑀𝑛𝑂4 𝑛−. As described in the introduction, according to ligand field theory (LFT), the ex- traction of one electron leading to an oxidation state from Mn4+ to Mn5+ can only occur if the manganese ion is in tetrahedral coordination [67]. One example of material belonging to this class is α-Li3MnO4 (low temperature form) with wurzite-type structure and manganese (V) in tetrahedral coordination [95]. The synthesis of this compound is performed only by solid state reaction at low temperature. The compound is very sensitive to water due to disproportionation of Mn5+ to Mn4+ and Mn7+ and decomposes if exposed to air for 2-3 days, therefore it has to be stored in inert atmosphere. Theoretical calculations [67] predicted a topotactic intercalation /deintercalation of lithium in- to Li3MnO4 structure following the pathway described in the following reactions (1) and (2): Eq. 4.1 Eq. 4.2 occurring at 1.9 V and 3.4 V vs Li+/Li, respectively. The total theoretical capacity for the ex- traction of four Li eq from Li5MnO4 to give LiMnO4 is 698 Ah/kg which confirms that Li3MnO4 is a promising cathode material. A novel synthesis route based on a freeze drying (FD) process was developed with the aim to decrease the number of synthesis steps, enable an easier process for doping the structure and improve the electrochemical properties of Li3MnO4. The FD process, used for many decades in polymer science and technology, is be- ~ 66 ~ coming widely used in the synthesis of oxides materials for a broad range of applications; in- cluding Li-ion battery materials [96-98]. The main advantage of this method is the creation of fine powers with 1) high surface area, 2) small crystallite size and 3) small particle size and agglomerates [99]. These characteristics increase the electroactive area of the material ex- posed to the electrolyte improving the full lithiation of the particles; decrease the Li+ pathway promoting a faster lithium intercalation improving electrochemical performances. The structural, morphological and electrochemical properties of the new freeze dried Li3MnO4 are compared to material synthesized by standard solid state reaction. Figure 4.1. Structure of LT- Li3MnO4 (yellow tetrahedral (Li), purple tetrahedral (Mn)). 4.2 Characterization of Li3MnO4 synthesized by FD Structural characterization of the compounds obtained by both solid state and freeze drying routes was carried out by XRD. The XRD patterns of SSR-Li3MnO4 and FDR- Li3MnO4 are shown in Fig. 4.1. 10 20 30 40 50 60 70 80 Li m [Li 2-b Mn b ]O 4 SSR - Li 3 MnO 4 (a) 2(deg) In te n s it y ( a .u .) Observed intensity Calculated Intensity Difference plot Bragg positions ~ 67 ~ 10 20 30 40 50 60 70 80 Observed intensity Calculated Intensity Difference plot Bragg positions 2(deg) In te n s it y ( a .u .) (b) Li m [Li 2-b Mn b ]O 4 FDR - Li 3 MnO 4 Figure 4.1: XRD pattern with Le Bail fitting plot of (a) SSR-Li3MnO4 and (b) FDR-Li3MnO4. The main reflections were indexed with the α-Li3MnO4 structure (low temperature form) (JCPDS-PDF 00-032-0572), belonging to the orthorhombic crystal system (space group: Pmn21). Small amounts of impurities are present in the samples which are attributed to the decomposition product of lithium permanganate and identified as an over-stoichiometric lithi- um manganese spinel structure (main peak at 44.8°) with formula Lim[Li2-bMnb]O4 where m=(2b-b)>1. These types of spinels were named “over-stoichiometric”, because the number of Li ions in tetrahedral sites is higher than 1. However, in over-stoichiometric spinels Li ions reside also in octahedral sites [77]. Le Bail fitting was performed to extract the information on lattice parameters and on crystallite size of the compounds as summarized in Table 4.1. Table 4.1: Le Bail fitting results. A slight lattice expansion was observed along the a axis, and a contraction was observed along the b and c axes, for FDR- Li3MnO4 in comparison with SSR-Li3MnO4. Samples Rwp (%) Rexp (%) 2 a (Å) b (Å) c (Å) Crystallite size (nm) SSR-Li3MnO4 3.40 2.83 1.45 6.3060(2) 5.4218(2) 4.9358(2) 101.3(1) FDR- Li3MnO4 2.71 2.40 1.27 6.3074(2) 5.4212(2) 4.9306(2) 35.5(1) ~ 68 ~ The FD process strongly affected the crystallinity of Li3MnO4. Much broader diffraction peaks were obtained for FDR- Li3MnO4 (Fig. 4.2) as result of a significant change in the crys- tallite size, from 101.3 nm for SSR-Li3MnO4 to 35.5 nm for FDR- Li3MnO4. 22.0 22.5 23.0 23.5 24.0 SSR - Li 3 MnO 4 FDR - Li 3 MnO 4 In te n s it y ( a .u .) 2(deg) Figure 4.2: XRD patterns comparing the (101) reflection of SSR-Li3MnO4 (black) and FDR-Li3MnO4 (red). This effect, as stated before, is a typical characteristic of materials prepared by freeze drying routes and it has been reported in lithium manganese spinel (LiMn2O4) synthesized by a FD method [100]. TEM micrographs (Fig. 4.3) show faceted particles with crystalline domains in the order of 60-100 nm for SSR-Li3MnO4. Much smaller domains having a dimension close to 20-30 nm were detected for FDR-Li3MnO4 and those values are in good agreement with the estimated crystallite size by Scherrer analysis. The Fast Fourier Transform (FFT) in the lower right cor- ner shows single crystallinity for the zone observed in Fig. 4.1(a), while a polycrystalline pat- tern on top of a strongly amorphous background is found in Fig. 4.1(b), for SSR-Li3MnO4 and FDR-Li3MnO4 respectively. Selected area electron diffraction (SAD) was carried out to study the local crystallinity of the compound. The upper right corner inset in Fig.4.3(a) shows very intense diffraction spots up to high order reflections resulting from a small number of highly crystalline grain orientations present in SSR-Li3MnO4 compound. The ED-pattern of the FDR-Li3MnO4 shows less intense reflections originating in several grain orientations on an amorphous background indicating a lower crystallinity (upper right corner inset in Fig.4.3(b)). These TEM results confirm the decrease in crystallinity and crystallite size for FDR-Li3MnO4. ~ 69 ~ Lowering the crystallite size leads to shortening of the lithium ion diffusion path [101] which improves electrochemical performance as already observed for other Mn-based materials [102]. Figure 4.3: TEM micrographs of (a) SSR-Li3MnO4 and (b) FDR- Li3MnO4. The inset in the upper right corners show selected area electron-diffractograms from a circular area with 260 nm in diameter, while the inset on the lower right corner show the FFT of the high resolution images. SEM micrographs (Fig. 4.4) show a pronounced difference in morphology between the two compounds. SSR-Li3MnO4 is composed of micron-sized particles having an irregular shape. Figure 4.4: SEM micrographs of (a) SSR-Li3MnO4 and (b) FDR- Li3MnO4. To determine the particle size distribution, DLS measurements (Fig.4.5) were carried out. A mean particle size of 10 µm was obtained for the SSR material. Results for FDR-Li3MnO4 ~ 70 ~ showed a lower degree of agglomeration and smaller particles, which have been quantified by DLS to have a mean size of 3.5 µm (Fig. 4.4). 0.01 0.1 1 10 100 1000 0 1 2 3 4 5 SSR-Li 3 MnO 4 FDR-Li 3 MnO 4 V o lu m e ( % ) Particle size (m) Figure 4.5: Particle size distribution of SSR-Li3MnO4 (black) and FDR-Li3MnO4 (red) in toluene. The smaller average particle size resulted in a higher surface area for the FDR-Li3MnO4 (11.5 m2/g) in comparison to SSR-Li3MnO4 (7.5 m 2/g), measured by BET. These results show that the microstructure of Li3MnO4 was successfully modified by the FD process. Conse- quently, improved electrochemical properties are expected for compounds with higher surface area. Thermal analysis conducted under inert conditions on SSR-Li3MnO4 and FDR-Li3MnO4 revealed that they start to decompose at around 120°C (Fig.4.6). The decomposition reaction proceeded slowly through the whole temperature range with a final temperature of 392- 395°C. XRD patterns of the residuals after thermal analysis (not shown) confirmed the com- plete decomposition of the compounds. The weight loss for SSR-Li3MnO4 (5.6 wt.%) was in good agreement with the theoretical estimation (5.7 wt.%) relative to the expected decomposi- tion reaction [103]: Eq. 4.3 The change in crystallite and particle size affected the thermal stability of the product depend- ing on the synthesis method. In fact, FDR-Li3MnO4 showed a faster decomposition in com- parison to SSR-Li3MnO4 and a slightly increase in weight loss (5.9 wt.%). ~ 71 ~ 50 100 150 200 250 300 350 400 450 500 93 94 95 96 97 98 99 100 101 SSR - Li 3 MnO 4 FDR - Li 3 MnO 4 W e ig h t( % ) Temperature (°C) 5.6% 5.9% Figure 4.6: Relative weight loss of the SSR-Li3MnO4 (black) and FDR- Li3MnO4 (red) under N2. SSR- and FDR-Li3MnO4 were characterized by XPS to access the Mn oxidation state. The Mn2p spectra for SSR- and FDR-Li3MnO4 are shown in Fig.4.7. Both spectra exhibited two main peaks at 642.8 eV and 654.2 eV attributed to Mn2p3/2 and Mn2p1/2 respectively separat- ed by a spin orbit splitting of 11.4 eV. The binding energies of the main manganese peaks are in good agreement with the Mn(V) oxidation state as already reported [103]. The spin split- ting supports our assignment, since its value resides between the spin orbit values for Mn(IV) and Mn(VI)/Mn(VII) [104, 105]. 660 656 652 648 644 640 636 632 C P S BE (eV) SSR - Li 3 MnO 4 FDR - Li 3 MnO 4 Mn 2p 3/2 642.8 eV Mn 2p 1/2 654.2 eV Figure 4.7: XPS spectra of SSR-Li3MnO4 (black) and FDR- Li3MnO4 (red). ~ 72 ~ In-situ high temperature XRD was carried out to better understand the formation mechanism of the final Li3MnO4 product by the SS and the FD route. 10 15 20 25 30 35 40 45 50 LiOH·H 2 O LiOH Li m [Li 2-b Mn b ]O 4 R.T after cooling after 3h at 170°C after 1h at 125°C 125°C R.T In te n s it y ( a .u .) 2 (degree) (a) 10 15 20 25 30 35 40 45 50 (b) 2 (degree) ** R.T after cooling after 3h at 170°C after 1h at 125°C 125°C R.T In te n s it y ( a .u .) LiOH·H 2 O LiMnO 4* Figure 4.8: In-situ XRD patterns of LiMnO4•3H2O-2LiOH•H2O for (a) SS and (b) FD precursor powders from room temperature (RT) to 170°C. The measurements were performed on the ground precursor mixture used for the solid state route and on the precursor powder obtained after freeze drying employing the same heating procedure for both syntheses. The XRD patterns for the SSR material are shown in Fig. 4.8(a). At room temperature (RT) the pattern exhibit the peaks related to the two precursors (LiMnO4·3H2O and LiOH·H2O). At 125°C the dehydration of the precursors was complete and the pattern showed peaks assigned to anhydrous precursors (LiMnO4 and LiOH). The de- composition of LiMnO4 occurred after 1 h holding time at 125°C. The decomposition product ~ 73 ~ was a overstoichiometric Li-Mn spinel with a broad peak at 44.8°, already detected by RT XRD (Fig.4.1(a)). No changes in the XRD patterns were observed during the heating steps from 125°C to 170°C and after cooling to RT. The final pattern, obtained after cooling down to RT from 170°C did not show the formation of the Li3MnO4 phase. The two main phases present in these patterns were: Li-Mn spinel and LiOH. Fig.4.8(b) shows the XRD patterns recorded for the freeze dried precursor powder at different temperatures. At RT, LiMnO4·3H2O and LiOH·H2O were identified as the two main phases, which was in agreement with those observed for the SS route. After heating to 125°C, the formation of Li3MnO4 was detected. At this temperature the dehydrated precursor LiMnO4 was still present (peak at 23.6°) but Li3MnO4 was already formed as the main phase. After 1h holding time at 125°C, only Li3MnO4 was present and this phase was stable during the heating to 170°C and further cooling down to RT. The in-situ XRD measurements provide important insights about the formation mechanism of Li3MnO4. The reaction is based on the diffusion process of Li + ions in the 𝑀𝑛𝑂4 − permanga- nate framework with a simultaneous reduction of the Mn from (VII) to (V). The tetrahedral coordination of the Mn-ion remains unchanged during the reaction. Instead, the lithium coor- dination changes from octahedral to tetrahedral. In the light of these results the importance of the lithium diffusion process for the formation of this material is revealed. In the case of the SSR, the final compound could not be obtained starting with the ground powder because the reagents are mixed at a macroscopic level and this synthesis procedure requires additional grinding steps to enable the lithium diffusion into the LiMnO4. Using the FD method the reagents are homogeneously mixed at a smaller scale allowing the formation of the final product Li3MnO4 without grinding steps at a lower reaction tempera- ture. The characterization carried out on SSR-Li3MnO4, FDR-Li3MnO4 and related precursors showed: 1) the improvement of the synthesis procedure for the freeze drying route and 2) the influence of the FD process on the micro- and nanostructure of FDR-Li3MnO4. SEM, BET and DLS characterization showed a decrease of the particle size from 10 µm to 3.5 µm and increased surface area. The crystallinity is also affected by the FD process; smaller crystallite size and lower crystallinity were confirmed by XRD and TEM. Galvanostatic measurements were carried out to study the lithiation activity and stability of Li3MnO4, and the effect of the synthesis route on its electrochemical properties. Four data sets were collected for both compounds obtained under defined electrochemical conditions: the cells were charged initially or discharged initially to evaluate the behaviour of the SSR- and ~ 74 ~ FDR compounds upon cycling when lithium is first removed (oxidation) or inserted (reduc- tion) from/into the structure. These experiments serve to establish the reversibility of the lithi- ation/delithiation processes. In addition, measurements at two different current rates (10 A/kg and 50 A/kg) were performed to investigate the electrochemical kinetic behaviour of Li3MnO4. In Fig. 4.9 the 1st galvanostatic cycles of both SSR- and FDR-Li3MnO4 at 10 A/kg discharg- ing initially (solid lines) and at 50 A/kg (dotted lines) are shown. 0 50 100 150 200 250 300 350 400 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 3.8 4.0 4.2 0 50 100 150 200 250 300 350 400 Capacity (Ah/kg) 1 st charge SSR-Li 3 MnO 4 10 A/kg FDR-Li 3 MnO 4 SSR-Li 3 MnO 4 50 A/kg FDR-Li 3 MnO 4 P o te n ti a l v s L i+ /L i (V ) 1 st discharge Figure 4.9: Discharge and charge profile for the 1st cycle between 4.2 V and 1.5 V at 10 A/kg (solid line) and 50 A/kg (dotted line) of SSR-Li3MnO4 (black) and FDR-Li3MnO4 (red). The cells were discharged initially. The material has an open-circuit voltage (OCV) of 3.55 V in both cases. SSR-Li3MnO4 and FDR-Li3MnO4 showed comparable voltage profiles. During the first discharge, lithium inter- calation occurred between the OCV and 2.1 V, and then a plateau appeared between 2.1 V and 2.0 V. This plateau can be assigned to the lithiation of Li3MnO4 as shown by the reaction described in Eq. 2. The final discharge capacities were 290 Ah/kg and 235 Ah/kg for the FDR-Li3MnO4 and SSR-Li3MnO4, respectively. An increase in the discharge capacity of 23% (55 Ah kg-1) was obtained for the FD compound. During the following charge, the plateau at 2.0 V disappeared and the compound showed an amorphous-like polarization curve with no clearly defined intercalation voltage plateau. Lithium intercalation is reversible after the 1st discharge, but the voltage profile significantly changes due to a change in the material struc- ture. The 1st galvanostatic cycle at 50A/kg discharging initially (dotted line) showed the same voltage shape of the measurement at a lower rate (10 A kg-1) during charge and discharge, but ~ 75 ~ an overall decrease in capacity can be seen. This indicates that the lithiation reaction is kinet- ically limited. In Fig.4.10 the 1st galvanostatic cycles of both SSR-Li3MnO4 and FDR-Li3MnO4 are shown at 10 A/kg (solid line) and 50Ah kg-1 (dotted line), when the cells were charged initially. 0 25 50 75 100 125 150 175 200 225 250 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 3.8 4.0 4.2 0 25 50 75 100 125 150 175 200 225 250 SSR-Li 3 MnO 4 10 A/kg FDR-Li 3 MnO 4 SSR-Li 3 MnO 4 50 A/kg FDR-Li 3 MnO 4 Capacity (Ah/kg) P o te n ti a l v s L i+ /L i (V ) 1 st charge 1 st discharge Figure 4.10: Charge and discharge profile for the 1st cycle between 4.2 V and 1.5 V at 10A/kg (solid line) and 50 A/kg (dotted line) of SSR-Li3MnO4 (black) and FDR-Li3MnO4 (red). The cells were charged initially. The voltage profile during the first charge showed a plateau at 3.8 V, 0.4 V higher than the theoretically expected voltage due to the extraction of Li from Li3MnO4 (Eq. 1). The plateau was not reversible during the subsequent discharge. The plateau at 2.0 V was less pronounced comparing to the materials that were discharged first (Fig. 4.9). This result is an indication of an irreversible oxidation reaction occurring when the materials are first cycled at high poten- tial and the lithium is removed from the structure. Lower discharge capacities were obtained when the cells were charged initially, but still a capacity improvement was observed for the FD compound (Table 4.2). ~ 76 ~ Table 4.2: Electrochemical data for all compounds. Compound Discharge capacity [Ah/kg] Charged initially Discharged initially 10 A/kg 50 A/kg 10 A/kg 50 A/kg 1st discharge 1st discharge 1st discharge 1st discharge SSR-Li3MnO4 181 120 235 147 FDR-Li3MnO4 212 158 290 184 Capacity improvement (%) 18 31 23 25 During the 2nd cycle, regardless whether the cells were discharged or charged initially, the voltage profile did not show any of the plateaux that were visible during the first cycle. Lithi- um insertion and extraction was reversible and occurred throughout the whole examined volt- age range. However, the electrochemical features related to an ordered (crystal) structure with well- defined intercalation sites were lost as can be seen in Fig. 4.11 by the absence of any po- tential plateaux. 0 50 100 150 200 250 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 3.8 4.0 4.2 0 50 100 150 200 250 SSR-Li 3 MnO 4 FDR-Li 3 MnO 4 Capacity (Ah/kg)Capacity (Ah/kg) 50A/kgCharging initially Discharging initially P o te n ti a l v s L i+ /L i (V o lt s ) 0 50 100 150 200 250 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 3.8 4.0 4.2 0 50 100 150 200 250 SSR-Li 3 MnO 4 FDR-Li 3 MnO 4 Figure 4.11: Voltage profile for the 2nd cycle between 4.2 V and 1.5 V at 50A/kg of SSR-Li3MnO4 (black) and FDR-Li3MnO4 (red). A plot of discharge capacity vs cycle number is shown in Fig. 4.12(a) for measurements at 10A/kg and in Fig. 4.12(b) for measurements at 50 A/kg. It can be clearly seen, that the mate- rial prepared by FDR has a higher capacity on the first cycle compared to the one from the SSR independently of the electrochemical protocol. In all four examined protocols, a probable ~ 77 ~ reorganization of the material structure occurred within the first cycles, giving rise to a quick decrease in capacity. Then a slower degradation was observed until the 30th cycle. Very high capacities were recorded at 10 A/kg which implied large amounts of lithium being involved in the lithiation processes. We propose that structural stress or a kinetically limited conversion reaction could lead to the poor cyclability of these materials. SSR-Li3MnO4 and FDR- Li3MnO4 delivered around 60 Ah/kg at 10 A/kg after 30 cycles. A better cyclability could be achieved at higher current rates (50 A/kg) resulting in a final higher absolute capacity. The compounds delivered around 80 Ah/kg after 30 cycles when the cells were initially charged. If the cells were initially discharged, however, SSR-Li3MnO4 showed a discharge capacity of 73 Ah/kg and FDR-Li3MnO4 96 Ah/kg. 0 5 10 15 20 25 30 0 25 50 75 100 125 150 175 200 225 250 275 300 10A/kg Charging initially Discharging initially 0 5 10 15 20 25 30 0 25 50 75 100 125 150 175 200 225 250 275 300 (a) D is c h a rg e c a p a c it y ( A h /k g ) Cycle number Cycle number SSR - Li 3 MnO 4 FDR - Li 3 MnO 4 SSR - Li 3 MnO 4 FDR - Li 3 MnO 4 0 5 10 15 20 25 30 0 20 40 60 80 100 120 140 160 180 200 220 240 50A/kg Discharging initially SSR - Li 3 MnO 4 FDR - Li 3 MnO 4 D is c h a rg e c a p a c it y ( A h /k g ) Cycle number Charging initially 0 5 10 15 20 25 30 0 20 40 60 80 100 120 140 160 180 200 220 240 (b) SSR - Li 3 MnO 4 FDR - Li 3 MnO 4 Cycle number Figure 4.12: Discharge capacity vs cycle at (a) 10A/kg and (b) 50 A/kg of SSR-Li3MnO4 (black) and FDR- Li3MnO4 (red). ~ 78 ~ As for the electrochemical characterization, two main improvements can be observed for the FDR material comparing to the SSR: higher absolute initial capacity and faster kinetics. A larger amount of lithium can be inserted in FDR-Li3MnO4; between 18% and 31% higher ca- pacity was obtained for all measurements in comparison to SSR-Li3MnO4. The lithium inter- calation is also faster in the FD material since higher capacity is obtained at higher current rates. The stability over cycling of the two materials is very similar. The initial improvement, ob- served during the first 3-5 cycles for the FD compound, vanished during the following cycles. This can be attributed to a structural instability of Li3MnO4. In fact, XRD after cycling showed an amorphous phase with no evidence of the initial crystal structure (Fig.4.13). 10 20 30 40 50 60 70 80 SSR - Li 3 MnO 4 after 30 cycles FDR - Li 3 MnO 4 after 30 cycles In te n s it y ( a .u .) 2(deg) Figure 4.13: XRD patterns of SSR-Li3MnO4 (black) and FDR-Li3MnO4 (red) after 30 cycles. The material structure can be correlated with the electrochemical properties as follows: lower- ing the particles size and increasing the surface area created more electroactive area for the lithium insertion. In the case of the FD compound, we propose that a higher number of lithiat- ed particles in connection with a deeper lithiation are the most probable causes for the higher capacity. The smaller crystallite size, in addition, helps for better lithium diffusion: the lithi- um path into the material is shorter. Lithiation and delithiation processes can proceed faster than in the SS material. Nevertheless, the structural stability under cycling needs to be im- proved to obtain better capacity retention. ~ 79 ~ 4.3 Capacity fading in Li3MnO4 In this subchapter we focused our attention on a step-wise post-mortem analysis of cycled Li3MnO4. Through a detailed follow-up of the structural and chemical changes occurring dur- ing the lithiation and delithiation process in first full cycle (charge and discharge, discharge and charge) we aimed to understand the degradation mechanism. Li3MnO4 shows a very fast capacity fading during cycling as shown in Fig.4.14a. The materi- al capacity over 30 cycles drops by 66% and 75% when it is charged initially or discharged initially, respectively. The voltage profile of the 1st, 2nd, and 30th cycle is shown when the ma- terial is charged initially (Fig.4.14b) or discharged initially (Fig.4.14c). 0 5 10 15 20 25 30 0 25 50 75 100 125 150 175 200 225 250 0 50 100 150 200 250 300 1.6 2.0 2.4 2.8 3.2 3.6 4.0 0 50 100 150 200 250 300 0 50 100 150 200 250 300 350 400 1.6 2.0 2.4 2.8 3.2 3.6 4.0 0 50 100 150 200 250 300 350 400 discharging initiallycharging initially D is c h a rg e c a p a c it y ( A h /k g ) Cycle number Li 3 MnO 4 charging initially Li 3 MnO 4 discharging initially 1 st cycle 2 nd cycle 30 th cycle (c)(b) P o te n ti a l v s L i+ /L i (V o lt s ) Capacity (Ah/kg) 1 st cycle 2 nd cycle 30 th cycle P o te n ti a l v s L i+ /L i (V o lt s ) Capacity (Ah/kg) (a) Figure 4.14: (a) Discharge capacity vs cycle at 10A/kg of Li3MnO4, (b) voltage profile for the 1st, 2nd,30th cycle between 4.2 V and 1.5 V at 10A/kg for Li3MnO4 charging initially, (c) voltage profile for the 1st, 2nd,30th cycle between 4.2 V and 1.5 V at 10A/kg for Li3MnO4 discharging initially. During the first charge (Fig.4.15a) a small plateau around 3.8 V was observed due to reaction described by Eq. 4.4, as suggested by Saint et al. [67]: 𝐿𝑖3𝑀𝑛𝑂4 → 𝐿𝑖3−𝑥𝑀𝑛𝑂4 + 𝑥𝐿𝑖 + + 𝑥𝑒− Eq.4.4 ~ 80 ~ No plateau was present in the subsequent discharge. Therefore, we conclude that the reaction was irreversible. When the material was initially discharged (Fig.4.15b), a plateau between 2.1 and 2.0 V was present due to reaction described by Eq.4.5 [67]: 𝐿𝑖3𝑀𝑛𝑂4 + 𝑥𝐿𝑖 + + 𝑥𝑒− → 𝐿𝑖3+𝑥𝑀𝑛𝑂4 Eq.4.5 Similar to reaction in Eq.4.4, the following charge showed an amorphous-like voltage shape with no evidence of the previously observed plateau, leading again to the conclusion that also this reaction (Eq.4.5) is irreversible. The voltage profile for the 2nd and 30th cycle clearly showed that the complete loss of structure is not regained in the following cycles and the amorphous voltage shape formed in the first cycle is kept. These are clear indications that the lithiation and delithiation processes associated with the above reactions led to irreversible structural changes and occurred already during the first cycle. To better understand the material modifications during the first cycle, the analysis of Li3MnO4 was carried out by XRD, XPS, SEM/EDX and Raman spectroscopy at different states of charge after specific cycling protocols) Four critical points in the voltage profile were identified which could provide us important in- formation about the degradation mechanism in Li3MnO4 (Fig.4.15):  Point 1: the material was initially charged to 4.2 V.  Point 2: the material was initially charged to 4.2 V then discharged to 1.5 V.  Point 3: the material was initially discharged to 1.5 V.  Point 4: the material was initially discharged to 1.5 V and then charged to 4.2 V. 0 5 10 15 20 25 30 1.6 2.0 2.4 2.8 3.2 3.6 4.0 180 Ah/kg P o te n ti a l v s L i+ /L i (V ) Time (h) 1 2 100 Ah/kg (a) 0 10 20 30 40 50 60 1.6 2.0 2.4 2.8 3.2 3.6 4.0 330 Ah/kg 3225 Ah/kg 4 P o te n ti a l v s L i+ /L i (V ) Time (h) (b) Figure 4.15: Voltage profile of Li3MnO4 for the 1st cycle between 1.5 V and 4.2 V at 50Ah/kg. The cells were a) charged initially and b) discharged initially. In Fig.4.16 the XRD patterns of the fresh electrode and cycled electrodes are shown. Li3MnO4-fresh showed a crystalline orthorhombic structure with the three main peaks be- ~ 81 ~ tween 20° and 25° 2θ degree. When the material was charged to 4.2 V (Li3MnO4-1) a de- crease of crystallinity and peak intensity was observed. The peak intensity further decreased during the following discharge to 1.5 V (Li3MnO4-2) leaving an almost amorphous material after the 1st cycle. In the same way, the first lithiation of Li3MnO4 to 1.5 V (Li3MnO4-3) de- stroyed completely the structure which was not restored in the following charge to 4.2 V (Li3MnO4-4). 10 15 20 25 30 35 40 45 50 10 15 20 25 30 35 40 45 50 4 (2 1 2 ) (2 2 1 ) (1 2 1 ) (1 2 0 ) (1 1 2 ) (2 1 1 ) (0 0 2 ) (0 2 0 )(2 1 0 ) (1 1 1 ) (0 1 1 )(1 0 1 ) (1 1 0 ) Fresh (0 1 0 ) 1 2 3 In te n s it y ( a .u .) 2(deg) Figure 4.16: XRD patterns of the fresh electrode, Li3MnO4-1, Li3MnO4-2, Li3MnO4-3, and Li3MnO4-4. The morphology of the electrodes was studied by scanning electron microscopy (SEM) (Fig.4.17). Since the electrode particles are embedded in carbon based conductive layer, backscattered electron micrographs were acquired to have higher bulk sensitivity and to ob- tain a better view of the particle dimensions and distribution after cycling. All samples were composed of micron-sized particles having an irregular shape with dimension around 10-15 µm. These particles were embedded in a carbon matrix. The amount of carbon used to prepare the samples (30 wt%) gave rise to an homogeneous dispersion of the Li3MnO4 in the carbon matrix (Fig. 4.17a). In all four cases, although we have seen before that cycling leads to amorphization, the particle size morphology did not change significantly. This is an indication that the amorphization did not trigger massive Mn dissolution during the first cycle. ~ 82 ~ Figure 4.17: SEM micrographs of a) and b) the fresh electrode, c) Li3MnO4-1, d) Li3MnO4-2, e) Li3MnO4-3, and f) Li3MnO4-4. X-ray photoelectron spectroscopy was carried out to determine how lithiation and delithiation of Li3MnO4 affect the oxidation state of manganese. In addition, the acquisition of oxygen, carbon and lithium spectra allowed to determine chemical changes involving directly the ma- terial but also the possible formation of new chemical species (Fig.4.18). The O1s spectrum of the fresh electrode showed one main peak at 531.0 eV with a shoulder at 528.9 eV. The former can be assigned to LiOH and the latter to a manganese-oxygen bond. The cycled elec- trodes showed similar features in the XPS spectra depending if they were in the charged state ~ 83 ~ (Li3MnO4-1 and Li3MnO4-4) or in the discharged state (Li3MnO4-2, Li3MnO4-3). Li3MnO4-1 showed a broad peak shifted at higher BE (531.4 eV) due to LiOH with an additional contri- bution from Li2CO3. Li3MnO4-2 and Li3MnO4-3 showed spectra with two contributions at lower BE in comparison to the charged state. The contribution at 530.7 eV can be assigned to Li2O2 and another one at 527.3 eV due to Li2O [106] . The electrode Li3MnO4-4 in the charged state had a main peak due to Li2CO3 at 531.8 eV [107]. The formation of lithium ox- ide and carbonate species are commonly found on the surface of cycled electrode due to de- composition of the electrolyte [108]. The carbon 1s spectra did not provide important information because the amount of carbon used in the electrode preparation was much higher than the amount of C species which could be expected due to electrolyte degradation. The C1s spectra showed only the asymmetric peak typical of carbon black compounds. The main contribution came from the C-C bond centred at 284.3 eV and the asymmetry is due to C-O bonds lying around 286.0 eV [109]. The XPS spectra of the manganese showed the Mn 2p3/2 and 2p1/2 peaks. They were located at 642.3 eV in the fresh electrode and at 642.1 eV for Li3MnO4-1 and Li3MnO4-4 indicating that the active species is Mn4+. However, the intensity of the Mn signal in the discharged state de- creased considerably and for Li3MnO4-3 the Mn signal was almost not detectable. This was already observed for Mn-based cathode materials and it was interpreted as solid electrolyte in- terface (SEI) formation on the surface of the particles [110]. During the following charge, SEI dissolution or creation of cracks in the SEI layer allowed the detection of the Mn signal. The Li1s and Mn3p zone was analysed to confirm our previous findings. The XPS spectra of the fresh electrode showed a Li1s peak at 54.8 eV typical of LiOH species and a Mn3p peak at 49.5 eV due to Mn4+. The electrodes recovered after an electrochemical charge against lith- ium (Li3MnO4-1 and Li3MnO4-4) showed a peak shift towards higher BE of Li1s peak (55.4 eV and 55.6 eV, respectively) in agreement with O1s shift, due to the formation of Li2CO3. However, since the Li1s peak of LiOH and Li2O2 overlap[107], the electrodes in the dis- charged state (Li3MnO4-2 and Li3MnO4-3) showed a Li1s main peak at the same binding en- ergy of the fresh electrode (54.8 eV) which can be assigned to Li2O2 with an additional shoul- der related to the formation of Li2O. The Mn3p peaks were also in agreement with the measurements of the Mn2p energy level. Li3MnO4-1 and Li3MnO4-4 showed a shift towards lower BE (48.9 eV and 49.1 eV, respectively) probably because the 3p orbitals were more af- fected than the 2p but their values were still in the range of the Mn4+ oxidation state. In Li3MnO4-2 and Li3MnO4-3 the manganese was hardly detectable due probably to SEI for- mation as discussed above. ~ 84 ~ 2400 3200 4000 4800 2000 3000 4000 5000 0 2000 4000 6000 1000 2000 3000 4000 545 540 535 530 525 520 515 2400 3600 4800 6000 Li 2 CO 3 LiOH 4 Fresh 1 2 3 O1s Li 2 O 2 Li 2 O M-O B.E. C P S 0 6000 12000 18000 0 3000 6000 9000 0 1000 2000 3000 0 1500 3000 4500 300 295 290 285 280 275 0 3000 6000 9000 C-C C-O C P S C1s 4 Fresh 1 2 3 B.E. 2400 2700 3000 2400 3000 3600 4200 1800 1900 2000 1600 1620 1640 1660 665 660 655 650 645 640 635 630 625 3000 3200 3400 3600 Mn 2p 1/2 C P S B.E. Mn 2p 3/2 Mn2p 4 Fresh 1 2 3 180 240 300 360 300 450 600 200 400 600 800 200 300 400 60 55 50 45 350 400 450 Mn3p C P S B.E. LiOH/Li 2 O 2 Li 2 CO 3 Li1s+Mn3p 4 Fresh 1 2 3 Li 2 O Figure 4.18: XPS spectra for a) O1s, b) C1s, c) Mn2p and d) Li1s+Mn3p of the fresh electrode, Li3MnO4-1, Li3MnO4-2, Li3MnO4-3, and Li3MnO4-4. ~ 85 ~ Raman spectroscopy was carried out on all electrodes to study the Mn chemical environment and to get better insight about the bonding in the amorphous material (Fig.4.19). The deeper penetration depth of Raman spectroscopy in comparison with XPS allowed obtaining infor- mation about Li3MnO4 vibrations in the fresh electrode. A sharp peak was observed at 756 cm-1 related to Mn-O stretching (A1 mode) of the [MnO4] 3- tetrahedron [111, 112]. The car- bon matrix gave rise to signals characteristic of the D and G bands located around 1350-1360 cm-1 and 1580-1590 cm-1, respectively [113]. Li3MnO4-1 and Li3MnO4-4 showed a broad peak centred at 620 cm-1, typical of Mn-O stretching in Oh [91] coordination indicating that the coordination surrounding the Mn ion changed during the cycling. In Li3MnO4-2 and Li3MnO4-3 spectra the Mn signal was not detectable; this is in agreement with XPS analysis and confirmed the presence of a thick SEI formed during discharge. Considering the Raman analysis there are strong indications that the amorphous compound, formed from decomposi- tion of Li3MnO4, is amorphous MnO2. 300 600 900 1200 1500 1800 300 600 900 1200 1500 1800 756 D G 4 Fresh 1 2 3 620 490 490 620 In te n s it y ( a .u .) Raman shift (cm -1 ) Figure 4.19: Raman spectra of the fresh electrode, Li3MnO4-1, Li3MnO4-2, Li3MnO4-3, and Li3MnO4-4. ~ 86 ~ The overall characterization carried out on fresh and cycled electrode allowed us to propose a mechanism for the electrochemical degradation of Li3MnO4 (Fig.4.20). Figure 4.20: Degradation mechanism of Li3MnO4 Starting from the fresh electrode, XRD and Raman spectroscopy confirmed the Li3MnO4 phase. The presence of a very thin layer of LiOH and MnO2 on the surface of the particles, detectable only by XPS, can be explained with the high reactivity of Mn5+. In fact, as it was reported previously [103], the detection of Mn5+ by XPS should be possible in the bare pow- der. However, in this case, the electrode preparation in organic solvents and the absorption of water molecules from the atmosphere could have caused the reduction of Mn5+ to Mn4+ and the formation of LiOH on the surface. When the material was initially charged (point 1), approximately 0.5 Li equivalents could be extracted from the structure (100Ah/kg) with the characteristic plateau at 3.8 V (Eq.1). The extraction of Li ions should theoretically lead to an oxidation of the Mn5+ ions; instead sur- prisingly Mn4+ was detected by XPS with a concomitant destruction of the crystal structure (lower intensity in XRD) and the detection of Li2CO3. Our interpretation of these results is that the amorphization caused by Li removal gave rise to an extraction of oxygen species, as already known for Mn-based cathode materials [30, 114]. Simultaneously Mn ions were re- duced from Mn5+ to Mn4+ and the electrolyte was oxidized to form Li2CO3, as already ob- ~ 87 ~ served before [115]. The manganese has now oxidation state (IV); except a few very rare ex- amples of materials with Mn4+ in tetrahedral coordination, the most stable coordination in this oxidation state is the octahedral one [68]. In fact, the coordination changed from Td to the most stable Oh as verified by Raman spectroscopy. In the following discharge (point 2) around 1 Li equivalent was inserted in the structure (180 Ah/kg) causing the reduction of the remain- ing manganese Mn5+ to Mn4+and Mn3+ (not visible in the XPS). This resulted in the almost complete collapse of the structure and probably further oxygen species were released. The oxygen species formed during these processes can directly react with lithium ions or can be electrochemically reduced to Li2O2 and Li2O [39, 116] creating a thick inorganic SEI on the surface of the particles, as observed by XPS. This SEI hinders the detection of Mn by XPS and Raman spectroscopy in the discharged state. When the material was discharged initially (point 3), Li ions were inserted into the structure for a total of 1.5 Li equivalents (230Ah/kg) at 1.5 V. The same behaviour was observed as de- scribed above: amorphization of the material and SEI formation. In the following charge (point 4) a much higher capacity was recorded (330Ah/kg) in comparison with the previous discharge. This was due to the electrochemical oxidation of Li2O2 and Li2O composing the SEI which left back the “naked” particles of amorphous MnO2 as confirmed by Raman and XPS spectroscopy. Amorphization during the first cycle occurred without apparent change in the particle mor- phology as we discussed above. However, SEM/EDX analysis was also carried out on the separator and lithium anode after cycling, more precisely at point 3 and 4. In all analysed zones no Mn was detected leading to the conclusion that a massive dissolution is not the cause of amorphization. Further characterization was carried out after 30 cycles to study possible modifications related to cycling. In Fig. 4.21a the XRD pattern of Li3MnO4 is shown after 30 cycles. As explained above, the compound formed by degradation of Li3MnO4 during the first cycle was amor- phous MnO2 and this is the compound which was cycled in the following cycles. As expected for amorphous manganese oxide, the cycling did not influence the short range order and the compound remained amorphous during cycling [117]. It is well known that Mn dissolution af- fects the performance of Mn-based battery materials [59], but in our case the particle mor- phology was not strongly affected as can be seen in Fig.4.21b, in the allocated amount of cy- cles. ~ 88 ~ Figure 4.21: (a) XRD pattern of Li3MnO4 after 30 cycles (b) SEM micrographs of Li3MnO4 after 30 cycles. From this detailed post mortem study we could confirm that the fast capacity fading recorded in the previous chapter was due to the electrochemical conversion of Li3MnO4 to amorphous MnO2 by electrochemical degradation already during the first cycle. The low cycling stability is the well-known behaviour of amorphous manganese oxide [118] formed during the first cy- cle. ~ 89 ~ 4.4 Vanadium incorporation in Li3MnO4 In the previous chapter, a post-mortem analysis of SSR- Li3MnO4 revealed the conversion of this material to a-MnO2 during cycling. For this reason, our work focused now on the attempt to improve the structural stability of Li3MnO4 to obtain better capacity retention. Substitution of the electroactive metal ion in cathode materials is a well-known approach to obtain structural stabilization. For example, in layered LiNiO2 substitution of Ni 3+ by Co3+ increased the cation ordering allowing for a higher reversible capacity [20, 25] . In LiMn2O4, substitution of Mn3+ by Ni3+ or Co3+ lowered the Mn3+/Mn4+ ratio, decreased Mn dissolution, suppressed the transition to tetragonal phase, and increased cycling stability [25]. Considering Li3MnO4, Saint et al.[67] proposed to substitute Mn 5+ with V5+ or P5+ ions, this could led to the exchange of MnO4 n− oxyanions with the much more stable VO4 n− or PO4 n− ox- yanions, with the aim to increase the structural stability and lowering the capacity fading. Taking into account the very low formation and decomposition temperature of Li3MnO4, the incorporation of V5+ or P5+ ions by classic solid state reaction techniques is not a very promis- ing approach due to the very low ion diffusion coefficient below 200°C (decomposition tem- perature of Li3MnO4). Therefore, it is preferable to mix the metal ions by solution techniques. This can be easily performed using the FD process proposed in the previous subchapter for the synthesis of pristine Li3MnO4. The FD process has the potential to incorporate metal ions by facilitating the phase mixing procedure and hence contribute to the stabilization of the cy- cling behaviour. Vanadium ions were chosen for the incorporation because vanadium can the- oretically fulfil a double function: 1) it can act as stabilizer for the structure as described above, and it can act as electroactive metal ion increasing capacity or voltage. Considering that Li3MnO4 and Li3VO4 have the same crystal structure and that the ionic radii of Mn5+ (Td) and V 5+(Td) are very similar (0.330 Å and 0.355 Å, respectively) the creation of a solid solution Li3Mn1-xVxO4 is theoretically possible. For this reason our first attempt was to synthesize a solid solution with high vanadium content (50 at%). The synthesis route described for the pristine material was now applied for the solid-solution synthesis between vanadium and manganese. Ammonium metavanadate (NH4VO3) was se- lected as vanadium source because vanadium is in (V) oxidation state and the compound is soluble in water. ~ 90 ~ The XRD pattern of the precursor powder after FD is shown in Figure 4.22. The main peaks were indexed as LiMnO4•3H2O phase. Peaks of KMnO4 (dark square) were also present as impurity due to incomplete exchange of potassium ions with lithium ions. These peaks were superimposed to a weak hump present between 10° and 40°, indication of a secondary phase. 10 20 30 40 50 60 70 80 (2 0 2 ) (2 0 0 ) (2 1 3 ) (3 3 0 ) (3 2 1 ) (2 2 0 ) (3 1 1 ) (2 1 2 ) (1 0 3 ) (2 0 3 ) (2 1 1 ) (2 1 0 ) (0 0 2 ) (2 0 1 ) (1 1 0 ) (1 0 1 ) KMnO 4 In te n s it y ( a .u .) 2 (degree) (1 0 0 ) Figure 4.22: XDR pattern of the FD powder. Li3MnO4 and Li3VO4 show almost identical XRD patterns, therefore the solid solution be- tween 𝑀𝑛𝑂4 3−and 𝑉𝑂4 3−should have similar peak positions as the parent compounds. The XRD pattern of the final compound labelled as FDR-Li3Mn0.5V0.5O4 is shown in Fig.4.23. 10 20 30 40 50 60 70 80 In te n s it y ( a .u .) 2 (degree) (0 0 2 ) (0 2 0 )(2 1 0 ) (1 1 1 ) (0 0 1 ) (1 0 1 ) (1 1 0 ) Figure 4.23: XDR pattern of the FDR- Li3Mn0.5V0.5O4 powder. ~ 91 ~ Two clear contributions were present in the pattern: sharp peaks having low intensity plus very broad peaks. The sharp peaks could be assigned to both phases (Li3MnO4 and/or Li3VO4) but the broad peaks (indicated with arrows), were an indication of a phase separation during the freeze drying process. To understand if the formation of the solid solution was successful, and to clarify the nature of the broad peaks, pristine FDR-Li3MnO4 and FDR-Li3VO4 were synthesized by the FD pro- cess and subjected to the same temperature treatment of FDR-Li3Mn0.5V0.5O4. The XRD pat- terns obtained from these powders were successively compared (Fig.4.24). FDR-Li3MnO4 (Fig.4.24a) showed a crystalline structure with reflection assigned to orthorhombic crystal system (Chapter 4 subchapter 4.2). FDR-Li3VO4 instead showed very broad peaks indication for a nano-crystalline material with crystallite size lower than 10 nm [119, 120]; FDR- Li3Mn0.5V0.5O4 seemed to be constituted by the superimposition of these two different phases: the crystalline FDR-Li3MnO4 and the nano-crystalline FDR-Li3VO4. 10 20 30 40 50 60 70 80 10 20 30 40 50 60 70 80 FDR - Li 3 MnO 4 (a) (c) (b) FDR - Li 3 Mn 0.5 V 0.5 O 4 FDR - Li 3 VO 4 In te n s it y ( a .u .) 2(deg) Figure 4.24: XDR pattern of (a) FDR-Li3MnO4,(b) FDR- Li3Mn0.5V0.5O4 (c) FDR-Li3VO4 powder. ~ 92 ~ The interpretation of the XRD measurements (Fig.4.22, 4.23, 4.24) is that during the sublima- tion of the water in the FD process, a phase separation between LiMnO4 and Li3VO4 occurred. Li3VO4 was made of very small crystallite size and originated the weak hump (Fig.4.22). Dur- ing the subsequent heating step, the low synthesis temperature allowed only the transfor- mation from LiMnO4 to Li3MnO4 to take place, leaving unchanged the nano-crystalline Li3VO4 phase which is now more visible in XRD due to the lower intensity of the Li3MnO4 peaks. The entire process can be summarized by the Eq.4.4: (1 − 𝑥)𝐿𝑖𝑀𝑛𝑂4(𝑎𝑞) + (2 + 𝑥)𝐿𝑖𝑂𝐻(𝑎𝑞) + 𝑥𝑁𝐻4𝑉𝑂3(𝑎𝑞) 𝑭𝑫 → 𝐿𝑖𝑀𝑛𝑂4(𝑠) + 2𝐿𝑖𝑂𝐻(𝑠) + 𝑥𝐿𝑖3𝑉𝑂4(𝑠) ∆ →𝐿𝑖3𝑀𝑛𝑂4 + 𝑥𝐿𝑖3𝑉𝑂4 Eq.4.6 This explanation was confirmed by HT-XRD carried out on FDR- Li3VO4 (Fig.4.25). In fact, the crystallization of the vanadate phase out of the nanocrystalline precursor started around 400°C and was complete around 600°C. At this temperature, the manganate phase was al- ready formed and underwent decomposition (TGA Chapter 4 subchapter 4.2) 10 20 30 40 50 60 70 80 700°C 100°C 2 (degree) RT Figure 4.25: High Temperature HT-XDR pattern of FDR-Li3VO4 powder. 𝑂2 ~ 93 ~ Morphology and chemical composition of the sample were studied by SEM/EDX. FDR- Li3Mn0.5V0.5O4 was composed of micron-sized particles and porous sheets, a probable result of the water sublimation during the FD process (Figure 4.26). The distribution of Mn and V was homogeneous in all analyzed zones with an average composition of 47.9 at% of vanadi- um and 52.1 at% of manganese which was in good agreement with the nominal composition. Figure 4.26: SEM micrographs of FDR- Li3Mn0.5V0.5O4 Galvanostatic measurements were carried out to study the electrochemical properties of FDR- Li3Mn0.5V0.5O4. In Fig.4.27a and Fig.4.27b the voltage profiles of the 1 st and the 2nd cycle us- ing different cut-off voltages during charge are shown. 0 10 20 30 40 50 60 70 80 90 100 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 FDR-Li 3 Mn 0.5 V 0.5 O 4 1.5-4.2V FDR-Li 3 Mn 0.5 V 0.5 O 4 1.5-4.8V FDR-Li 3 Mn 0.5 V 0.5 O 4 1.5-5.2V P o te n ti a l v s L i+ /L i (V o lt s ) Capacity (Ah/kg) 1 st discharge 1 st charge (a) ~ 94 ~ 0 10 20 30 40 50 60 70 80 90 100 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 P o te n ti a l v s L i+ /L i (V o lt s ) Capacity (Ah/kg) FDR-Li 3 Mn 0.5 V 0.5 O 4 1.5-4.2V FDR-Li 3 Mn 0.5 V 0.5 O 4 1.5-4.8V FDR-Li 3 Mn 0.5 V 0.5 O 4 1.5-5.2V2 nd charge 2 nd discharge (b) Figure 4.27: Discharge and charge profile for the (a) 1st cycle and (b) 2nd cycle between 1.5-4.2 V (black), 1.5- 4.8 V (red), and 1.5-5.2 V (blue) at 50A/kg of FDR- Li3Mn0.5V0.5O4. The cells were discharged initially. Taking into account the same voltage range used to analyze the pristine material (1.5-4.2 V, Fig.4.9 and Fig.4.10), a huge capacity decrease during charge and discharge was observed for the vanadium containing compound in comparison with the pristine FD compound. Extending the analyzed voltage range to higher potentials, an increase in charge capacity was observed above 4.5 V. Even though the material was cycled at high voltages during the 1st cycle, the discharge capacity during the 2nd cycle was still very low (< 40 Ah/Kg). For this reason, the additional capacity obtained above 4.5 V during the first charge can be associated to an irre- versible removal of oxygen species. Li3MnO4 is a cathode material active between 4.2 V and 1.5 V while Li3VO4 is known to be an anode material active between 1.5 V and 0.1 V (capacity of 350Ah/kg) [46]. Taking into account the very low capacity recorded at high voltages it is likely that in FDR- Li3Mn0.5V0.5O4 only the vanadate phase is the active phase. To extract more information about the redox potential of this material and to confirm our pre- vious hypothesis, FDR-Li3Mn0.5V0.5O4 was also tested at low voltages (Fig.4.28) and com- pared with pristine FDR-Li3MnO4 and FDR-Li3VO4. The discharge profile of FDR- Li3Mn0.5V0.5O4 showed two main contributions to the total capacity: 1) a plateau between 1.2 V and 0.2 V contributing about 350 Ah/kg and 2) a plateau between 0.2 V and 0.1 V with a total capacity of about 1000 Ah/kg. Considering that the Mn redox reaction should occur be- tween 3.0 V and 1.5 V and this feature was not present in the voltage profile of FDR- Li3Mn0.5V0.5O4, the first contribution can be assigned to the intercalation of Li-ions into the ~ 95 ~ Li3VO4 phase, and the second contribution was due to over-reduction of the manganese phase (probable formation of MnO) with irreversible conversion reactions and SEI formation [121]. 0 100 200 300 900 1000 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 FDR-Li 3 MnO 4 FDR-Li 3 Mn 0.5 V 0.5 O 4 FDR - Li 3 VO 4 Capacity (Ah/kg) P o te n ti a l v s L i+ /L i (V o lt s ) Figure 4.28: 1st discharge profile at 50A/kg of FDR- Li3MnO4 (red), FDR- Li3Mn0.5V0.5O4 (magenta), SSR- Li3VO4 (blue). Cells were discharged initially. To further prove that the disappearance of the Mn redox activity at high voltages in FDR- Li3Mn0.5V0.5O4 was a direct consequence of the FD process, galvanostatic measurements were carried out on a 50/50 wt.% ground mixture between SSR- Li3MnO4 and SSR-Li3VO4 (Fig. 4.29). In this case, indeed, manganese activity occurred between 3.0 V and 1.5 V, as shown by the arrow. 0 100 200 300 600 700 800 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 SSR - Li 3 MnO 4 Li 3 MnO 4 + Li 3 VO 4 50/50 wt% SSR - Li 3 VO 4 Capacity (Ah/kg) P o te n ti a l v s L i+ /L i (V o lt s ) Figure 4.29: 1st discharge profile at 50A/kg of SSR- Li3MnO4 (black), SSR-Li3MnO4+SSR-Li3VO4 50/50 wt.% (purple), SSR-Li3VO4 (blue). Cells were discharged initially. ~ 96 ~ The overall characterization carried out on FDR- Li3Mn0.5V0.5O4 showed that the attempt of creating a solid solution with a vanadium content of 50 at% by FD synthesis led to a bi-phase material composed of a crystalline Li3MnO4 phase and a nano-crystalline Li3VO4 phase; However, the disappearance of the Mn contribution at high voltages was a clear indication that a morphological shielding of the Mn phase was present (i.e. coverage of Li3MnO4 parti- cles with Li3VO4). Taking into account the results shown above, a smaller amount of vanadium (5 wt%) was then used for the solid solution reaction. The synthesis was carried out as previously explained (Eq.1 and Eq.2). In Fig.4.30 the XRD patterns of pristine FDR-Li3MnO4 and FDR- Li3Mn0.95V0.05O4 is shown. No differences were observed in the peak positions; however a weak hump was clearly visible between 20° and 25°. This originated probably from the nano- crystalline Li3VO4 phase formed during the FD process, as shown above for 50 at% V. 10 20 30 40 50 60 70 80 FDR-Li 3 MnO 4 FDR-Li 3 Mn 0.95 V 0.05 O 4 In te n s it y ( a .u .) 2(deg) Figure 4.30: XDR pattern of FDR-Li3MnO4 (red) and FDR- Li3Mn0.95V0.05O4 (green) powder. Le Bail fitting was carried out to study the influence of vanadium content on Li3MnO4 lattice parameters. A small contraction was observed along the a axis, compensated by an expansion along the b and c axes (Table 4.3). The vanadium, if incorporated in the structure, should lead to lattice expansion. Considering the increase in the background and the significant change in the b axis, the assumption was that the incorporation of vanadium in the Li3MnO4 structure occurred only partially (< 5 at.%). ~ 97 ~ Table 4.3: Le bail fitting results. The main difference between the two compounds resided in the electrochemical properties. 0 50 100 150 200 250 1.5 2.0 2.5 3.0 3.5 4.0 FDR - Li 3 MnO 4 FDR - Li 3 Mn 0.95 V 0.05 O 4 P o te n ti a l v s L i+ /L i (V o lt s ) Capacity (Ah/kg) 1 st charge 1 st discharge Figure 4.31: 1st cycle between 1.5-4.2 V at 50A/kg of FDR- Li3MnO4 (red), FDR-Li3Mn0.95V0.05O4 (green). The vanadium containing compound showed a 1st discharge capacity of 45Ah/kg, in compari- son with 185Ah/kg of the pristine Li3MnO4 (Fig.4.31). Even though the vanadium amount was very small, the influence on the discharge capacity was relevant. In conclusion, in both attempts of vanadium doping (50 at% and 5 at%) the final result on the electrochemical properties was a significant capacity decrease. This was probable due to the formation of a nano-crystalline Li3VO4 phase which interacting with Li3MnO4, shield the Mn activity at higher voltages. Samples Rwp (%) Rexp (%) χ2 a (Å) b (Å) c (Å) Crystallite size (nm) FDR- Li3MnO4 2.71 2.40 1.27 6.3074(2) 5.4212(2) 4.9306(2) 35.5(1) FDR- Li3Mn0.95V0.05O4 2.25 1.73 1.69 6.3042(3) 5.4316(3) 4.9320(2) 20.7(1) ~ 98 ~ Chapter 5 5. Manganese in square pyramidal coordination: h- LiMnBO3 5.1 Introduction Lithium metal oxides containing polyanion groups (𝑋𝑂𝑚) 𝑛−, such as phosphate (𝑃𝑂4) 3− [122], silicate (𝑆𝑖𝑂4) 3− [56], borates (𝐵𝑂3) 3− [123], are very attractive candidates as cathode materials for Li-ion batteries . Even though these units are not electro-active, they introduce several advantages especially as enhanced stability at higher cycling voltages [57] and voltage tuning due to the inductive effect of the polyanion group on the metal cation [124]. Among them, borates LiMBO3 (M=Fe,Mn,Co) could be good alternatives to phosphates due to the lower weight of the (𝐵𝑂3) 3− group. In particular, LiMnBO3 is a very promising cath- ode material because of its high theoretical capacity of 220 Ah/kg [69]. LiMnBO3 has two polymorphs: hexagonal [125] and monoclinic [126] with theoretical cell voltage of 4.1 V and 3.7 V vs Li+/Li, respectively [69]. The hexagonal polymorph has a [𝑀𝑛𝐵𝑂3]𝑛 𝑛−framework with MnO5 square pyramids and BO3 units. MnO5 square pyramids shared two opposite edges of its square base with two adjacent pyramids forming chains along the [001] direction. The BO3 groups link three chains via cor- ner sharing. This framework creates channels along the [001] directions where Li atoms oc- cupy tetrahedral sites. LiO4 tetrahedra share two corners forming chains along [001] (Fig.5.1a) [125]. The monoclinic polymorph has a [𝑀𝑛𝐵𝑂3]𝑛 𝑛−framework with MnO5 trigonal bypiramids and BO3 units. MnO5 trigonal bipyramids shares two edges with adjacent bipyramids forming a single chain along the [-101] direction. The BO3 groups link three chains via corner sharing. Li atoms reside in tetrahedral coordination [126] (Fig. 5.1b). ~ 99 ~ Figure 5.1: Crystal structure of (a) hexagonal and (b) monoclinic polymorph (adapted from [57]). Similar to phosphates [127], lithium transition metal borates are considered to have intrinsi- cally low ionic and electronic conductivity, which is believed to be the cause of the poor elec- trochemical performance. For this reason the synthesis of nanosized materials with carbon coating for increased electronic conductivity are compulsory to obtain decent capacities. Up to date, LiMnBO3 is mainly fabricated through solid state reaction at 500°C to obtain the monoclinic phase and at 800-850°C to obtain the hexagonal phase [69] under inert or reduc- ing atmosphere. Two-step calcination processes were usually required together with interme- diate ball-milling to reduce the particle size after the first calcination process. Recently LiMnBO3 was also synthesized by sol-gel [57] and microwave synthesis [128]. The carbon- coated LiMnBO3 was usually fabricated via the calcination of the mixture of LiMnBO3 and carbon sources (i.e sucrose) under inert atmosphere to create a LiMnBO3/C composite [129]. At present, crystalline monoclinic m-LiMnBO3 was the most studied phase reaching a capaci- ty of 150-160 Ah/kg for about 50 cycles. Crystalline hexagonal h-LiMnBO3 instead showed very small capacity of about 50 Ah/kg (in both cases with carbon coating). h-LiMnBO3 was synthesized as nano-crystalline material and it showed much higher capacity around 120 Ah/kg [57]. As shown in Chapter 4 for Li3MnO4, FD synthesis was a technique which allowed obtaining nanocrystalline materials and lowering the reaction temperature due to better reagents mixing at nanoscopic level. In this chapter it will be described, how the use of this synthesis route en- abled the synthesis of nanocrystalline FDR-LiMnBO3 and FDR-LiMnBO3/rGO composite. Structural and morphological electrochemical characterization was carried out and electro- chemical measurements were performed to study both materials as cathodes for Li-ion batter- ies. ~ 100 ~ 5.2 FD synthesis of LiMnBO3 and LiMnBO3/rGO composite One of the main challenges in the synthesis of nanocrystalline materials by FD route was to find suitable precursors with similar and as low as possible decomposition temperatures. This opened up the possibility to low temperatures reactions and in consequence to smaller crystal- lite size. For the synthesis of LiMnBO3, lithium citrate tetrahydrate Li3(Cit)•4H2O as lithium source, manganese acetate tetrahydrate Mn(CH3COO)2•4H2O as manganese source and boric acid H3BO3 as boron source were selected after a careful review of available precursors and their thermogravimetric data. Li3(Cit)•4H2O and Mn(CH3COO)2•4H2O decompose around 250-320°C and H3BO3 around 200°C [130-132]. The expected reaction temperature was around 350°C-400°C. The XRD patterns of the precursors FD mixture with and without GO after sublimation of the water are shown in Fig.5.2. 10 20 30 40 50 60 70 80 precursors FD mixture precursors FD mixture + GO In te n s it y ( a .u .) 2(deg) H 3 BO 3 Figure 5.2: XRD patterns of the precursor FD mixture without GO (black), and with GO (red). The FD mixture containing only the precursors showed two peaks assigned to H3BO3 triclinic phase (JCPDS-PDF 00-030-0199). The reflections of the Li and Mn-containing precursors were not visible indicating a good mixing at atomic scale. The FD mixture of the precursors with GO showed an amorphous XRD pattern with a broad hump around 25°. This was an in- dication for an even better precursors mixing in comparison to the mixture without GO. ~ 101 ~ In Fig.5.3 the XRD patterns of FDR-LiMnBO3 and FDR-LiMnBO3/rGO obtained after heat- ing treatment of the FD mixture were shown. 10 20 30 40 50 60 70 80 (3 3 0 ) (1 4 0 ) (2 2 0 ) (2 2 1 ) (3 0 0 ) (1 0 1 ) (1 1 0 ) FDR-LiMnBO 3 FDR-LiMnBO 3 /rGO In te n s it y ( a .u .) 2(deg) MnO (1 1 1 ) Figure 5.3: XRD pattern of FDR-LiMnBO3 (black) and FDR-LiMnBO3/rGO (red) For both compounds, the diffraction peaks could be indexed with the hexagonal LiMnBO3 structure (JCPDS-PDF 00-053-0371). The small impurity peak at 40.5° in FDR-LiMnBO3 was attributed to MnO as decomposition product of manganese acetate in reductive atmos- phere [131]. The peak broadening is an indication that both materials were composed of nano- sized crystallites. Williamson-Hall calculation was carried out to quantify their dimension. Results showed an average crystallite size of 10-15 nm for both compounds. Preferred orien- tation of the crystallite enhanced the peak intensity of the (111) and (300) peaks at 2θ=35.9 and 38.1 [57]. In contrast to the standard solid state route [58] which required 800°C to obtain the hexagonal phase, h-LiMnBO3 was obtained at 300°C and 400°C by FD route with and without GO, re- spectively. The stabilization of the hexagonal phase at low temperature could be due to the good mixing of the reagents by FD synthesis as already shown before by sol-gel synthesis [57]. The lower temperature obtained using GO was probably due to the additional heat origi- nated by the exothermic reduction reaction from GO to rGO in reducing atmosphere [133]. ~ 102 ~ SEM analysis was carried out to study the morphology of the samples. Micrographs are shown in Fig.5.44. FDR-LiMnBO3 (Fig.5.4a e Fig.5.4b) was composed of 50-80 nm round- shaped primary particles which formed agglomerates of around 500 nm. In FDR- LiMnBO3/rGO (Fig.5.4c) the particles were embedded into graphene sheets. Figure 5.4: SEM micrographs of (a) FDR-LiMnBO3, (b) higher magnification FDR-LiMnBO3, and (c) FDR- LiMnBO3/rGO. TEM high resolution micrographs are shown in Fig.5.5. FDR-LiMnBO3 was composed of su- perimposed crystalline domains in the order of 10 nm in agreement with XRD crystallite size calculations. In FDR-LiMnBO3/rGO these crystalline domains have similar dimension and were interconnected with rGO sheets. ~ 103 ~ Figure 5.5: TEM micrographs of (a) FDR-LiMnBO3, (b) FDR-LiMnBO3/rGO. The physicochemical characterization carried out showed that the approach to create an in- situ nanocrystalline composite material was successful. The good mixing of the precursors with GO was confirmed by XRD. The subsequent simultaneous crystallization of LiMnBO3 and reduction of GO to rGO allowed creating a conductive network between the metal oxide particles and the rGO sheets as shown by electron microscopy. Composites with rGO are usually made with a three steps process: 1) synthesis of the metal oxide 2) ball milling with GO and 3) reduction of GO to rGO. This is a very energy and time consuming method. In our case the composite material was obtained in only one-step method. Galvanostatic measurements were carried out to study the electrochemical performance of the synthesized materials and to evaluate the possible differences between them. In Fig.5.6, the 1st cycle voltage profiles of FDR-LiMnBO3 and FDR-LiMnBO3/rGO are shown. Both materials showed similar voltage shape, with no defined intercalation voltage plateau. During the first cycle FDR-LiMnBO3 showed a charge and discharge capacity of 100 Ah/kg and 70 Ah/kg, respectively. These low capacities were probably the result of the low electronic and ionic conductivity of the material which hinders the lithium (de)intercalation. The synthesis of FDR-LiMnBO3/rGO composite allowed creating a better conductive network and as conse- quence a charge capacity of 290Ah/kg and a discharge capacity of 120Ah/kg were obtained. Obviously, since the theoretical capacity of the material is 220Ah/kg, side reactions (i.e de- composition due to instability of the charged phase, oxidation of the electrolyte) were occur- ring during the initial charge which contributed to the total charge capacity. ~ 104 ~ 0 40 80 120 160 200 240 280 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 3.8 4.0 4.2 4.4 4.6 0 40 80 120 160 200 240 280 Capacity (Ah/kg) P o te n ti a l v s L i+ /L i (V ) FDR-LiMnBO 3 FDR-LiMnBO 3 /rGO Figure 5.6: Charge and discharge profile for the 1st cycle between 4.6 V and 2.0 V at 10A/ kg of FDR-LiMnBO3 (black) and FDR-LiMnBO3/rGO (red). The voltage profile of FDR-LiMnBO3/rGO for the 1 st, 2nd and 10th cycle is shown in Fig.5.7. 0 40 80 120 160 200 240 280 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 3.8 4.0 4.2 4.4 4.6 0 40 80 120 160 200 240 280 1 st cycle 2 nd cycle 10 th cycle P o te n ti a l v s L i+ /L i (V ) Capacity (Ah/kg) Figure 5.7: 1st, 2nd and 10th cycle between 4.6 V and 2.0 V at 10A/ kg of FDR-LiMnBO3/rGO. The lithium (de)intercalation is reversible during the following cycles and the amorphous-like voltage shape was recorded also for the 2nd and 10th cycle. The columbic efficiency increased in the following cycles indicating a lower extent of side reactions; however, the capacity de- creased very fast during cycling. The discharge capacity vs cycle number is plotted in Fig.5.8. ~ 105 ~ The initial increase in discharge capacity recorded for FDR-LiMnBO3/rGO vanished after a few cycles. 0 5 10 15 20 25 30 0 20 40 60 80 100 120 140 FDR-LiMnBO 3 FDR-LiMnBO 3 /rGO D is c h a rg e c a p a c it y ( A h /k g ) Cycle number Figure 5.8: Discharge capacity vs cycle number of FDR-LiMnBO3 (red) and FDR-LiMnBO3/rGO (black) at 10Ah/kg. One of the reasons of this strong capacity fading could be associated with amorphization of the material. XRD patterns were recorded after 30 cycles (Fig. 5 .9). Complete amorphization occurred for FDR-LiMnBO3/rGO, and only two peaks with very low intensity were present between 35° and 40° for FDR-LiMnBO3. 10 20 30 40 50 60 70 80 FDR-LiMnBO 3 after 30 cycles FDR-LiMnBO 3 /rGO after 30 cycles In te n s it y ( a .u .) 2(deg) Figure 5.9. XRD patterns of FDR-LiMnBO3 (black) and FDR-LiMnBO3/rGO (red) after 30 cycles. ~ 106 ~ The overall characterization carried out on FDR-LiMnBO3 and FDR-LiMnBO3/rGO showed that the FD route can be applied also to borates compounds. Nanocrystalline materials can be obtained by selection of the right precursors. Looking at LiMnBO3, a conductive network is necessary to obtain good capacities. We have shown that using a one-step synthesis route is possible to obtain a composite material between LiMnBO3 and rGO. An improvement of 70% in capacity was obtained for FDR-LiMnBO3/rGO in comparison to bare FDR-LiMnBO3 in the first discharge. However, a strong capacity fading is still one of the main problems regard- ing this material. Doping with metal cations such as Co2+ or Ni2+ and single particle carbon coating could lead to better structural stability and higher capacity retention. ~ 107 ~ Concluding remarks The topic of this doctoral thesis was the study of manganese-based cathode materials for Li- ion batteries. Globally, the following results have been achieved: 1) Ca2MnO4 was studied for the first time as cathode materials. A calcium extraction by acid treatment was designed to functionalize the surface of the particles and activate it for Li inter- calation. The amount of lithium inserted in the acid-treated samples was a function of the amount of calcium extracted. On the basis of the extensive characterization carried out, a new bi-functional crystalline-amorphous structure was proposed. It was composed of an inner Ca2MnO4 crystalline bulk core whose aim was to improve the stability, and an outer Ca 2+ con- taining MnO2•H2O amorphous shell which gave rise to the electrochemical response. A con- siderable improvement of the capacity retention was recorded. 2) The freeze drying (FD) technique was applied for the first time to synthesize Li3MnO4. The material synthesized by this synthesis route was very well characterized and compared with Li3MnO4 synthesized by a classical solid state reaction. The FD material showed modification in the micro- and nanostructure (particles and crystallite size) which were responsible for the higher capacity recorded during the initial cycles. However, the strong capacity fading over cycling was a trigger for a deeper study on the degradation mechanism. A post-mort analysis was carried out concluding that the material converts to amorphous MnO2. 3) Preliminary studies were carried out on LiMnBO3. The FD technique was applied for the synthesis of LiMnBO3 for the first time as well. A nano-crystalline LiMnBO3 and a LiMnBO3/rGO composite were obtained and characterized. The recorded capacities were very close to the state of the art. Manganese, as it was shown in this thesis, has a very rich chemistry due to its many oxidation states and its many coordination geometries. However, there are two main drawbacks of Mn- based cathode materials that were tried to be addressed in this thesis: capacity and stability. Taking into account the materials analysed and the techniques used in this work, a lot can still be done to develop better materials and improve their properties. First of all, the new concept developed for Ca2MnO4 can be eventually applied to stabilize other amorphous cathode and anode materials. Secondly, as we have seen in the previous chapters, the possibility to exploit redox couple with manganese oxidation state above Mn4+ can occur only in Td coordination. In this case, a cationic substitution of Mn 5+ by V5+ or P5+ with other synthesis techniques (es: USC in O2) could lead to Li3MnO4 stabilization. Anionic ~ 108 ~ substitution could also be an option, the substitution of O2- by N3- or the incorporation of F- could lead to the formation of compounds with Mn in high oxidation states. Then, the selec- tion of different precursors for the FD synthesis of LiMnBO3 could allow obtaining an higher reaction temperature and bigger crystallite size that in conjunction with cationic substitution of Mn2+ by Co2+ or Ni2+ could stabilize this material. In conclusion, this thesis which was focused on almost unexplored materials for battery appli- cations could be considered as pioneering research and an attempt to go beyond the common families of cathode materials. Hopefully, the results presented in this work will motivate also other researchers around the globe to follow this path and start studying new families of cath- ode materials for Li-ion batteries. This would lead to the development of low cost and envi- ronmental friendly materials with higher electrochemical performance. ~ 109 ~ References [1] A.M. Omer, Energy, environment and sustainable development, Renewable and Sustainable Energy Reviews, 12 (2008) 2265-2300. [2] REN21, Renewables 2014 10 Year Report : the first decade, in, 2014. [3] E.E. Paul Denholm, Brendan Kirby, and Michael Milligan, The Role of Energy Storage with Renewable Electricity Generation, in, 2010. [4] Wikipedia - The Free Encyclopedia, in. [5] I.E. Agency, Global EV Outlook: Understanding the Electric Vehicle Landscape to 2020, in, 2013. [6] M.Z. Jacobson, M.A. Delucchi, G. Bazouin, Z.A.F. Bauer, C.C. Heavey, E. Fisher, S.B. Morris, D.J.Y. Piekutowski, T.A. Vencill, T.W. Yeskoo, 100% clean and renewable wind, water, and sunlight (WWS) all-sector energy roadmaps for the 50 United States, Energy Environ. Sci., (2015). [7] L.R.F. Allen J. Bard, ELECTROCHEMICAL METHODS Fundamentals and Applications, in: I. JOHN WILEY & SONS (Ed.), 2001. [8] P.F. Christine Lefrou, Jean-Claude Poignet, Electrochemistry: The Basics, With Examples, in: Springer (Ed.), 2012. [9] M. Armand, J.M. Tarascon, Building better batteries, Nature, 451 (2008) 652-657. [10] B.J. Landi, M.J. Ganter, C.D. Cress, R.A. DiLeo, R.P. Raffaelle, Carbon nanotubes for lithium ion batteries, Energy Environ. Sci., 2 (2009) 638-654. [11] G. Gritzner, G. Kreysa, G.S. Wilson, D. Landolt, V.M.M. Lobo, W. Plieth, M. Sluytersrehbach, K. Tokuda, C.P. Andrieux, Y.A. Chizmadzhev, B.E. Conway, J. Koryta, O.A. Petrii, D. Pletcher, M.J. Weaver, A.J. Arvia, T. Biegler, H.D. Hurwitz, D. Pavlov, G. Horanyi, S.K. Rangarajan, E. Gileadi, S. Trasatti, T. Watanabe, G.A. Wright, W. Paik, C. Gutierrez, D. Simonsson, M.L. Berkem, A.K. Covington, D. Drazic, NOMENCLATURE, SYMBOLS AND DEFINITIONS IN ELECTROCHEMICAL ENGINEERING, Pure Appl. Chem., 65 (1993) 1009-1020. [12] G.I. Allen J. Bard, Fritz Scholz, Electrochemical Dictionary, in, Springer, 2012. [13] G.N. Lewis, F.G. Keyes, THE POTENTIAL OF THE LITHIUM ELECTRODE, J. Am. Chem. Soc., 35 (1913) 340-344. [14] K. Brandt, Historical development of secondary lithium batteries, Solid State Ion., 69 (1994) 173- 183. [15] E. Peled, Film forming reaction at the lithium/electrolyte interface, J. Power Sources, 9 (1983) 253-266. [16] D. Aurbach, E. Zinigrad, Y. Cohen, H. Teller, A short review of failure mechanisms of lithium metal and lithiated graphite anodes in liquid electrolyte solutions, Solid State Ion., 148 (2002) 405- 416. [17] J.M. Tarascon, M. Armand, Issues and challenges facing rechargeable lithium batteries, Nature, 414 (2001) 359-367. [18] S. Basu, C. Zeller, P.J. Flanders, C.D. Fuerst, W.D. Johnson, J.E. Fischer, Synthesis and properties of lithium-graphite intercalation compounds, Materials Science and Engineering, 38 (1979) 275-283. [19] D. Guerard, A. Herold, Intercalation of lithium into graphite and other carbons, Carbon, 13 (1975) 337-345. [20] M.S. Whittingham, Lithium batteries and cathode materials, Chem. Rev., 104 (2004) 4271-4301. [21] B. Scrosati, Lithium Rocking Chair Batteries: An Old Concept?, J. Electrochem. Soc., 139 (1992) 2776-2781. [22] M. Winter, R.J. Brodd, What are batteries, fuel cells, and supercapacitors?, Chem. Rev., 104 (2004) 4245-4269. [23] B. Xu, D.N. Qian, Z.Y. Wang, Y.S.L. Meng, Recent progress in cathode materials research for advanced lithium ion batteries, Materials Science & Engineering R-Reports, 73 (2012) 51-65. ~ 110 ~ [24] M. Winter, J.O. Besenhard, M.E. Spahr, P. Novak, Insertion electrode materials for rechargeable lithium batteries, Adv. Mater., 10 (1998) 725-763. [25] Lithium Batteries: Science and Technology., in: G.P. Gholam-Abbas Nazri (Ed.), Springer, 2009. [26] T.J. PATEY, Oxide Nanoparticles for Electrodes in Lithium-Ion Batteries, in: Dipl. Verf. -Ing. ETH Zürich, ETH Zurich, Zurich, 2009. [27] K. Mizushima, P.C. Jones, P.J. Wiseman, J.B. Goodenough, LIXCOO2 "(OLESS-THANXLESS-THAN- OR-EQUAL-TO1) - A NEW CATHODE MATERIAL FOR BATTERIES OF HIGH-ENERGY DENSITY, Mater. Res. Bull., 15 (1980) 783-789. [28] E. Antolini, LiCoO2: formation, structure, lithium and oxygen nonstoichiometry, electrochemical behaviour and transport properties, Solid State Ion., 170 (2004) 159-171. [29] I. Belharouak, Y.K. Sun, J. Liu, K. Amine, Li(Ni1/3Co1/3Mn1/3)O2 as a suitable cathode for high power applications, J. Power Sources, 123 (2003) 247-252. [30] N. Yabuuchi, K. Yoshii, S.-T. Myung, I. Nakai, S. Komaba, Detailed Studies of a High-Capacity Electrode Material for Rechargeable Batteries, Li2MnO3-LiCo1/3Ni1/3Mn1/3O2, J. Am. Chem. Soc., 133 (2011) 4404-4419. [31] C.S. Johnson, N. Li, C. Lefief, J.T. Vaughey, M.M. Thackeray, Synthesis, Characterization and Electrochemistry of Lithium Battery Electrodes: xLi(2)MnO(3)center dot(1- x)LiMn0.333Ni0.333Co0.333O2 (0 <= x <= 0.7), Chem. Mat., 20 (2008) 6095-6106. [32] Z. Zhu, L. Zhu, Synthesis of layered cathode material 0.5Li2MnO3·0.5LiMn1/3Ni1/3Co1/3O2 by an improved co-precipitation method for lithium-ion battery, J. Power Sources, 256 (2014) 178-182. [33] S. Birgisson, K.M.O. Jensen, T.L. Christiansen, J.F. von Bulow, B.B. Iversen, In situ powder X-ray diffraction study of the hydro-thermal formation of LiMn2O4 nanocrystallites, Dalton Transactions, 43 (2014) 15075-15084. [34] Batteries for Sustainability, in: R.J. Brodd (Ed.), Springer, 2013. [35] M.M. Thackeray, W.I.F. David, P.G. Bruce, J.B. Goodenough, Lithium insertion into manganese spinels, Mater. Res. Bull., 18 (1983) 461-472. [36] H.C. Le, N.D. Nguyen, S. Brutti, B. Scrosati, Synthesis, characterization and electrochemical properties of 4.8 V LiNi0.5Mn1.5O4 cathode material in lithium-ion batteries, Electrochim. Acta, 55 (2010) 5110-5116. [37] N. Amdouni, K. Zaghib, F. Gendron, A. Mauger, C.M. Julien, Structure and insertion properties of disordered and ordered LiNi0.5Mn1.5O4 spinels prepared by wet chemistry, Ionics, 12 (2006) 117- 126. [38] Nanotechnology for Lithium-Ion Batteries, in: I.D. Yaser Abu-Lebdeh (Ed.), Springer, 2013. [39] V. Etacheri, R. Marom, R. Elazari, G. Salitra, D. Aurbach, Challenges in the development of advanced Li-ion batteries: a review, Energy Environ. Sci., 4 (2011) 3243-3262. [40] R. Marom, S.F. Amalraj, N. Leifer, D. Jacob, D. Aurbach, A review of advanced and practical lithium battery materials, J. Mater. Chem., 21 (2011) 9938-9954. [41] A.K. Padhi, K.S. Nanjundaswamy, J.B. Goodenough, Phospho-olivines as positive-electrode materials for rechargeable lithium batteries, J. Electrochem. Soc., 144 (1997) 1188-1194. [42] V. Aravindan, J. Gnanaraj, Y.S. Lee, S. Madhavi, LiMnPO4 - A next generation cathode material for lithium-ion batteries, J. Mater. Chem. A, 1 (2013) 3518-3539. [43] J.B. Goodenough, Y. Kim, Challenges for Rechargeable Li Batteries, Chem. Mat., 22 (2010) 587- 603. [44] C.K. Chan, H. Peng, G. Liu, K. McIlwrath, X.F. Zhang, R.A. Huggins, Y. Cui, High-performance lithium battery anodes using silicon nanowires, Nat Nano, 3 (2008) 31-35. [45] L. Kavan, J. Prochazka, T.M. Spitler, M. Kalbac, M.T. Zukalova, T. Drezen, M. Gratzel, Li insertion into Li4Ti5O12 p(Spinel) - Charge capability vs. particle size in thin-film electrodes, J. Electrochem. Soc., 150 (2003) A1000-A1007. [46] H. Li, X. Liu, T. Zhai, D. Li, H. Zhou, Li3VO4: A Promising Insertion Anode Material for Lithium-Ion Batteries, Adv. Energy Mater., 3 (2013) 428-432. [47] A. Lewandowski, A. Świderska-Mocek, Ionic liquids as electrolytes for Li-ion batteries—An overview of electrochemical studies, J. Power Sources, 194 (2009) 601-609. ~ 111 ~ [48] P. Knauth, Inorganic solid Li ion conductors: An overview, Solid State Ion., 180 (2009) 911-916. [49] A.H. Whitehead, M. Schreiber, Current Collectors for Positive Electrodes of Lithium-Based Batteries, J. Electrochem. Soc., 152 (2005) A2105-A2113. [50] S.-T. Myung, Y. Hitoshi, Y.-K. Sun, Electrochemical behavior and passivation of current collectors in lithium-ion batteries, J. Mater. Chem., 21 (2011) 9891-9911. [51] S.S. Zhang, A review on the separators of liquid electrolyte Li-ion batteries, J. Power Sources, 164 (2007) 351-364. [52] Y. Idemoto, T. Mochizuki, K. Ui, N. Koura, Properties, crystal structure, and performance of o- LiMnO2 as cathode material for Li secondary batteries, J. Electrochem. Soc., 153 (2006) A418-A424. [53] H. Xia, Z.T. Luo, J.P. Xie, Nanostructured LiMn2O4 and their composites as high-performance cathodes for lithium-ion batteries, Progress in Natural Science-Materials International, 22 (2012) 572- 584. [54] Q. Zhang, J. Mei, X. Wang, J. Guo, F. Tang, W. Lu, Facile synthesis of spherical spinel LiMn2O4 nanoparticles via solution combustion synthesis by controlling calcinating temperature, J. Alloy. Compd., 617 (2014) 326-331. [55] S.-M. Oh, S.W. Oh, S.-T. Myung, S.-M. Lee, Y.-K. Sun, The effects of calcination temperature on the electrochemical performance of LiMnPO4 prepared by ultrasonic spray pyrolysis, J. Alloy. Compd., 506 (2010) 372-376. [56] R.J. Gummow, Y. He, Recent progress in the development of Li2MnSiO4 cathode materials, J. Power Sources, 253 (2014) 315-331. [57] S. Afyon, D. Kundu, F. Krumeich, R. Nesper, Nano LiMnBO3, a high-capacity cathode material for Li-ion batteries, J. Power Sources, 224 (2013) 145-151. [58] L. Chen, Y.M. Zhao, X.N. An, J.M. Liu, Y.Z. Dong, Y.H. Chen, Q. Kuang, Structure and electrochemical properties of LiMnBO3 as a new cathode material for lithium-ion batteries, J. Alloy. Compd., 494 (2010) 415-419. [59] D.H. Jang, Y.J. Shin, S.M. Oh, Dissolution of spinel oxides and capacity losses in 4V Li/LixMn2O4 coils, J. Electrochem. Soc., 143 (1996) 2204-2211. [60] A. Yamada, Lattice Instability in Li(LixMn2−x)O4, J. Solid State Chem., 122 (1996) 160-165. [61] X. Li, Y. Xu, C. Wang, Suppression of Jahn–Teller distortion of spinel LiMn2O4 cathode, J. Alloy. Compd., 479 (2009) 310-313. [62] K.Y. Chung, C.W. Ryu, K.B. Kim, Onset mechanism of jahn-teller distortion in 4 V LiMn2O 4 and its suppression by LiM0.05Mn1.95O 4(M = Co, Ni) coating, J. Electrochem. Soc., 152 (2005) A791-A795. [63] R.G. BURNS, MINERALOGICAL APPLICATIONS OF CRYSTAL FIELD THEORY, in, CAMBRIDGE UNIVERSITY PRESS, 1993. [64] E.A.K. James E. Huheey, Richard L. Keiter, Inorganic Chemistry, in, Harper Collins College Publisher. [65] Q. Liu, Y.X. Li, Z.L. Hu, D.L. Mao, C.K. Chang, F.Q. Huang, One-step hydrothermal routine for pure- phased orthorhombic LiMnO2 for Li ion battery application, Electrochim. Acta, 53 (2008) 7298-7302. [66] F. Cheng, J. Zhao, W. Song, C. Li, H. Ma, J. Chen, P. Shen, Facile Controlled Synthesis of MnO2 Nanostructures of Novel Shapes and Their Application in Batteries, Inorg. Chem., 45 (2006) 2038- 2044. [67] J.A. Saint, M.M. Doeff, J. Reed, Synthesis and electrochemistry of Li3MnO4: Mn in the +5 oxidation state, J. Power Sources, 172 (2007) 189-197. [68] D.M. Sherman, THE ELECTRONIC-STRUCTURES OF MANGANESE OXIDE MINERALS, Am. Miner., 69 (1984) 788-799. [69] J.C. Kim, C.J. Moore, B. Kang, G. Hautier, A. Jain, G. Ceder, Synthesis and Electrochemical Properties of Monoclinic LiMnBO3 as a Li Intercalation Material, J. Electrochem. Soc., 158 (2011) A309-A315. [70] A. Sasmal, S. Saha, C.J. Gomez-Garcia, C. Desplanches, E. Garribba, A. Bauza, A. Frontera, R. Scott, R.J. Butcher, S. Mitra, Reversible switching of the electronic ground state in a pentacoordinated Cu(ii) complex, Chemical Communications, 49 (2013) 7806-7808. ~ 112 ~ [71] R. Robert, D. Logvinovich, M.H. Aguirre, S.G. Ebbinghaus, L. Bocher, P. Tomes, A. Weidenkaff, Crystal structure, morphology and physical properties of LaCo1-xTixO3 +/-delta perovskites prepared by a citric acid assisted soft chemistry synthesis, Acta Materialia, 58 (2010) 680-691. [72] I.D. Fawcett, J.E. Sunstrom, M. Greenblatt, M. Croft, K.V. Ramanujachary, Structure, magnetism, and properties of Ruddlesden-Popper calcium manganates prepared from citrate gels, Chem. Mat., 10 (1998) 3643-3651. [73] E. Krupicka, A. Reller, A. Weidenkaff, Morphology of nanoscaled LaMO3-particles (M=Mn, Fe, Co, Ni) derived by citrate precursors in aqueous and alcoholic solvents, Crystal Engineering, 5 (2002) 195- 202. [74] L. Bocher, M.H. Aguirre, R. Robert, D. Logvinovich, S. Bakardjieva, J. Hejtmanek, A. Weidenkaff, High-temperature stability, structure and thermoelectric properties of CaMn1-xNbxO3 phases, Acta Materialia, 57 (2009) 5667-5680. [75] H.Y. Lee, J.B. Goodenough, Supercapacitor behavior with KCl electrolyte, J. Solid State Chem., 144 (1999) 220-223. [76] F. Leroux, L.F. Nazar, 3-volt manganese dioxide: the amorphous alternative, Solid State Ion., 100 (1997) 103-113. [77] A.A. Andriiko, A.Y. Shpak, Y.O. Andriyko, J.R. Garcia, S.A. Khainakov, N.Y. Vlasenko, Formation of spinel structured compounds in the lithium permanganate thermal decomposition, J. Solid State Electrochem., 16 (2012) 1993-1998. [78] E.E. McCabe, C. Greaves, Synthesis and structural and magnetic characterization of mixed manganese-copper n=1 Ruddlesden-Popper phases, Chem. Mat., 18 (2006) 5774-5781. [79] K.H. Lee, S.W. Kim, H. Ohta, K. Koumoto, Thermoelectric properties of layered perovskite-type (Sr(1-x)Ca(x))(3)(Ti(1-y)Nb(y))(2)O(7), J. Appl. Phys., 101 (2007). [80] M.D. Carvalho, R.P. Borges, A.V. Girao, M.M. Cruz, M.E.M. Jorge, G. Bonfait, P. Dluzewski, M. Godinho, Electroni doping of Ca4Mn3O10 induced by vanadium substitution, Chem. Mat., 17 (2005) 4852-4857. [81] K. Toda, S. Kurita, M. Sato, New layered perovskite compounds, LiLaTiO4 and LiEuTiO4, Journal of the Ceramic Society of Japan, 104 (1996) 140-142. [82] K. Toda, M. Takahashi, T. Teranishi, Z.G. Ye, M. Sato, Y. Hinatsu, Synthesis and structure determination of reduced tantalates, Li2LaTa2O7, Li2Ca2Ta3O10 and Na2Ca2Ta3O10, with a layered perovskite structure, J. Mater. Chem., 9 (1999) 799-803. [83] C.R. Michel, R. Amigo, N. Casan-Pastor, Dramatic change in magnetic properties of manganates Ca(2-x)Ln(x)MnO(4) by low-temperature electrochemical oxidation in fused nitrates, Chem. Mat., 11 (1999) 195-197. [84] S. Nishimoto, M. Matsuda, M. Miyake, Novel protonated and hydrated n=1 Ruddlesden-Popper phases, HxNa1-xLaTiO4 center dot yH(2)O, formed by ion-exchange/intercalation reaction, J. Solid State Chem., 178 (2005) 811-818. [85] S. Lin, Z. Huang, Y. Sun, X. Wang, R. Tang, S. Ma, X. Yang, Synthesis of Ca 1+xMn xO 1+3x(x = 1, 2, 3) compounds with the Ruddlesden-Popper structures and extraction of their interlayer Ca 2+ ions, Journal of Beijing Normal University (Natural Science), 45 (2009) 370-374. [86] D.S. Alfaruq, M.H. Aguirre, E.H. Otal, S. Populoh, L. Karvonen, S. Yoon, Y. Lu, G. Deng, S.G. Ebbinghaus, A. Weidenkaff, Crystal growth and thermoelectric properties of CaMn0.98Nb0.02O3- delta, J. Cryst. Growth, 377 (2013) 170-177. [87] A. Nemudry, P. Weiss, I. Gainutdinov, V. Boldyrev, R. Schollhorn, Room temperature electrochemical redox reactions of the defect perovskite SrFeO2.5+x, Chem. Mat., 10 (1998) 2403- 2411. [88] L. Karvonen, S.H. Yoon, P. Hug, H. Yamauchi, A. Weidenkaff, M. Karppinen, The n=3 member of the SrCoO(3n-1)/n series of layered oxygen-defect perovskites, Mater. Res. Bull., 46 (2011) 1340- 1345. [89] K.R. Poeppelmeier, M.E. Leonowicz, J.M. Longo, CAMNO2.5 AND CA2MNO3.5 - NEW OXYGEN- DEFECT PEROVSKITE-TYPE OXIDES, J. Solid State Chem., 44 (1982) 89-98. ~ 113 ~ [90] R.N. Reddy, R.G. Reddy, Synthesis and electrochemical characterization of amorphous MnO2 electrochemical capacitor electrode material, J. Power Sources, 132 (2004) 315-320. [91] K. Ozawa, Lithium Ion Rechargeable Batteries: Materials, Technology, and New Applications, WILEY-VCH Verlag GmbH & Co, 2009. [92] C.H. Jiang, E. Hosono, H.S. Zhou, Nanomaterials for lithium ion batteries, Nano Today, 1 (2006) 28-33. [93] M. Galceran, M.C. Pujol, M. Aguilo, F. Diaz, Sol-gel modified Pechini method for obtaining nanocrystalline KRE(WO4)(2) (RE = Gd and Yb), J. Sol-Gel Sci. Technol., 42 (2007) 79-88. [94] R. Baddour-Hadjean, J.P. Pereira-Ramos, Raman Microspectrometry Applied to the Study of Electrode Materials for Lithium Batteries, Chem. Rev., 110 (2010) 1278-1319. [95] G. Meyer, R. Hoppe, THERMAL-BEHAVIOR OF LI3MNO4 .1. ALPHA-LI3MNO4 AND BETA- LI3MNO4, Zeitschrift Fur Anorganische Und Allgemeine Chemie, 424 (1976) 249-256. [96] O.A. Shlyakhtin, Y.S. Yoon, S.H. Choi, Y.J. Oh, Freeze drying synthesis of LiNi0.5Mn0.5O2 cathode materials, Electrochim. Acta, 50 (2004) 505-509. [97] G. Sudant, E. Baudrin, B. Dunn, J.M. Tarascon, Synthesis and electrochemical properties of vanadium oxide aerogels prepared by a freeze-drying process, J. Electrochem. Soc., 151 (2004) A666- A671. [98] Y.Q. Qiao, X.L. Wang, Y.J. Mai, X.H. Xia, J. Zhang, C.D. Gu, J.P. Tu, Freeze-drying synthesis of Li3V2(PO4)3/C cathode material for lithium-ion batteries, J. Alloy. Compd., 536 (2012) 132-137. [99] O.A. Shlyakhtin, Y.J. Oh, Inorganic cryogels for energy saving and conversion, J. Electroceram., 23 (2009) 452-461. [100] T. Le Mercier, J. Gaubicher, E. Bermejo, Y. Chabre, M. Quarton, Morphology and electrochemical behavior of an ultrafine LiMn2O4 powder obtained by a new route, from freeze- dried precursors, J. Mater. Chem., 9 (1999) 567-570. [101] H.S. Zhou, D.L. Li, M. Hibino, I. Honma, A self-ordered, crystalline-glass, mesoporous nanocomposite for use as a lithium-based storage device with both high power and high energy densities, Angew. Chem.-Int. Edit., 44 (2005) 797-802. [102] M. Zhao, Y. Fu, N. Xu, G.R. Li, M.T. Wu, X.P. Gao, High performance LiMnPO4/C prepared by a crystallite size control method, J. Mater. Chem. A, 2 (2014) 15070-15077. [103] W.P. Kilroy, S. Dallek, J. Zaykoski, Synthesis and characterization of metastable Li-Mn-O spinels from Mn(V), J. Power Sources, 105 (2002) 75-81. [104] C.N.R. Rao, D.D. Sarma, S. Vasudevan, M.S. Hegde, STUDY OF TRANSITION-METAL OXIDES BY PHOTOELECTRON-SPECTROSCOPY, Proc. R. Soc. London Ser. A-Math. Phys. Eng. Sci., 367 (1979) 239- 252. [105] M. Oku, X-RAY PHOTOELECTRON-SPECTRA OF KMNO4 AND K2MNO4 FRACTURED IN-SITU, J. Electron Spectrosc. Relat. Phenom., 74 (1995) 135-148. [106] S. Malmgren, K. Ciosek, M. Hahlin, T. Gustafsson, M. Gorgoi, H. Rensmo, K. Edström, Comparing anode and cathode electrode/electrolyte interface composition and morphology using soft and hard X-ray photoelectron spectroscopy, Electrochim. Acta, 97 (2013) 23-32. [107] K.P.C. Yao, D.G. Kwabi, R.A. Quinlan, A.N. Mansour, A. Grimaud, Y.-L. Lee, Y.-C. Lu, Y. Shao- Horn, Thermal Stability of Li2O2 and Li2O for Li-Air Batteries: In Situ XRD and XPS Studies, J. Electrochem. Soc., 160 (2013) A824-A831. [108] P. Verma, P. Maire, P. Novák, A review of the features and analyses of the solid electrolyte interphase in Li-ion batteries, Electrochim. Acta, 55 (2010) 6332-6341. [109] Y. Shao, G. Yin, J. Zhang, Y. Gao, Comparative investigation of the resistance to electrochemical oxidation of carbon black and carbon nanotubes in aqueous sulfuric acid solution, Electrochim. Acta, 51 (2006) 5853-5857. [110] K. Edström, T. Gustafsson, J.O. Thomas, The cathode–electrolyte interface in the Li-ion battery, Electrochim. Acta, 50 (2004) 397-403. [111] Gonzalez.F, W.P. Griffith, TRANSITION-METAL TETRA-OXO-COMPLEXES AND THEIR VIBRATIONAL-SPECTRA, J. Chem. Soc.-Dalton Trans., (1972) 1416-&. ~ 114 ~ [112] R.J. Smith, Y.R. Shen, K.L. Bray, The effect of pressure on vibrational modes in Li3PO4, J. Phys.- Condes. Matter, 14 (2002) 461-469. [113] T. Jawhari, A. Roid, J. Casado, Raman spectroscopic characterization of some commercially available carbon black materials, Carbon, 33 (1995) 1561-1565. [114] D.Y.W. Yu, K. Yanagida, Y. Kato, H. Nakamura, Electrochemical Activities in Li2MnO3, J. Electrochem. Soc., 156 (2009) A417-A424. [115] S.S. Zhang, K. Xu, T.R. Jow, Understanding formation of solid electrolyte interface film on LiMn2O4 electrode, J. Electrochem. Soc., 149 (2002) A1521-A1526. [116] Z.Q. Peng, S.A. Freunberger, L.J. Hardwick, Y.H. Chen, V. Giordani, F. Barde, P. Novak, D. Graham, J.M. Tarascon, P.G. Bruce, Oxygen Reactions in a Non-Aqueous Li+ Electrolyte, Angew. Chem.-Int. Edit., 50 (2011) 6351-6355. [117] X. Jun John, Y. Hui, J. Gaurav, Y. Jingsi, Amorphous manganese oxide remains amorphous upon lithium intercalation and cycling, Electrochemistry Communications, 6 (2004) 892-897. [118] J.J. Xu, S. Passerini, B.B. Owens, W.H. Smyrl, Processing and properties of amorphous manganese dioxide formed by sol-gel procedures, in: G.A. Nazri, C. Julien, A. Rougier (Eds.) Solid State Ionics V, Materials Research Society, Warrendale, 1999, pp. 119-124. [119] L. Wu, D. Bresser, D. Buchholz, S. Passerini, Nanocrystalline TiO2(B) as Anode Material for Sodium-Ion Batteries, J. Electrochem. Soc., 162 (2015) A3052-A3058. [120] J.A. Chang, M. Vithal, I.C. Baek, S.I. Seok, Morphological and phase evolution of TiO2 nanocrystals prepared from peroxotitanate complex aqueous solution: Influence of acetic acid, J. Solid State Chem., 182 (2009) 749-756. [121] J. Liu, Q. Pan, MnO/C Nanocomposites as High Capacity Anode Materials for Li-Ion Batteries, Electrochem. Solid-State Lett., 13 (2010) A139-A142. [122] W.-J. Zhang, Structure and performance of LiFePO4 cathode materials: A review, J. Power Sources, 196 (2011) 2962-2970. [123] K.J. Lee, D.W. Kim, H.S. Hong, Preparation and Characterizations of Lithium Iron Borate Nano- sized Powders via Aerosol and Thermal Process, Curr. Nanosci., 10 (2014) 168-170. [124] B.C. Melot, J.M. Tarascon, Design and Preparation of Materials for Advanced Electrochemical Storage, Accounts Chem. Res., 46 (2013) 1226-1238. [125] V. Legagneur, Y. An, A. Mosbah, R. Portal, A.L. La Salle, A. Verbaere, D. Guyomard, Y. Piffard, LiMBO3 (M = Mn, Fe, Co): synthesis, crystal structure and lithium deinsertion/insertion properties, Solid State Ion., 139 (2001) 37-46. [126] O.S. Bondareva, M.A. Simonov, Y.K. Egorovtismenko, N.V. Belov, CRYSTAL-STRUCTURE OF LIZN BO3 AND LIMN BO3, Kristallografiya, 23 (1978) 487-490. [127] J. Molenda, A. Kulka, A. Milewska, W. Zajac, K. Swierczek, Structural, Transport and Electrochemical Properties of LiFePO4 Substituted in Lithium and Iron Sublattices (Al, Zr, W, Mn, Co and Ni), Materials, 6 (2013) 1656-1687. [128] T. Ma, A. Muslim, Z. Su, Microwave synthesis and electrochemical properties of lithium manganese borate as cathode for lithium ion batteries, J. Power Sources, 282 (2015) 95-99. [129] Y.S. Lee, H. Lee, Improved lithium storage capacities of LiMnBO3/C via simple high-energy milling, Mater. Lett., 132 (2014) 401-404. [130] G.E. Tobon-Zapata, E.G. Ferrer, S.B. Etcheverry, E.J. Baran, Thermal behaviour of pharmacologically active lithium compounds, J. Therm. Anal., 61 (2000) 29-35. [131] E.M. Diefallah, KINETIC-ANALYSIS OF THERMAL-DECOMPOSITION REACTIONS .6. THERMAL- DECOMPOSITION OF MANGANESE(II) ACETATE TETRAHYDRATE, Thermochim. Acta, 202 (1992) 1-16. [132] F. Sevim, F. Demir, M. Bilen, H. Okur, Kinetic analysis of thermal decomposition of boric acid from thermogravimetric data, Korean J. Chem. Eng., 23 (2006) 736-740. [133] Z.L. Wang, D. Xu, Y. Huang, Z. Wu, L.M. Wang, X.B. Zhang, Facile, mild and fast thermal- decomposition reduction of graphene oxide in air and its application in high-performance lithium batteries, Chemical Communications, 48 (2012) 976-978. ~ 115 ~ Additional declaration Chapter 2, chapter 3 and subchapter 4.1, 4.2 and 4.3 are based on the following published pa- pers: Y. Surace, M. Simoes, J. Eilertsen, L. Karvonen, S. Pokrant, A. Weidenkaff, Functionalization of Ca2MnO4-delta by controlled calcium extraction: Activation for electrochemical Li intercalation, Solid State Ion., 266 (2014) 36-43. Y. Surace, M. Simões, L. Karvonen, S. Yoon, S. Pokrant, A. Weidenkaff, Freeze drying synthesis of Li3MnO4 cathode material for Li-ion batteries: A physico-electrochemical study, J. Alloy. Compd., 644 (2015) 297-303. Y. Surace, M. Simoes, J. Eilertsen, L. Karvonen, S. Pokrant, A. Weidenkaff, Activation of nano-Ca2MnO4 for electrochemical lithium intercalation, MRS Spring 2015 proceeding, Vol 1805/2015. And on the following submitted paper: Y. Surace, M. Simões, S. Pokrant, A. Weidenkaff, Capacity fading in Li3MnO4: a post- mortem analysis. Dübendorf, 13/11/2015 Yuri Surace ~ 116 ~ Curriculum Vitae Yuri Surace Nationality: Italian Date of birth: 31.07.1988 Education 09.2012-10.2015 Ph.D candidate Swiss Federal Labs for Materials Science and Technology (EMPA) (CH) and University of Stuttgart (Germany) Thesis title: Manganese-based cathode materials for Li-ion batteries Thesis supervisor: Prof. Dr. Anke Weidenkaff (University of Stuttgart) and Dr. Simone Pokrant (EMPA) 10.2009-05.2012 Master degree in Materials Science and Engineering University of Calabria (Italy) with a final vote 110/110 with honors. Thesis title: New luminescent platinum (II) complexes with tetradentate ligands. Thesis supervisor: Prof. Dr. Mauro Ghedini and Dr. Nicolas Godbert (University of Calabria), Dr. J.A.Gareth Williams (University of Durham) 10.2006-12.2009 Bachelor degree in Materials Science University of Calabria (Italy) with a final vote 110/110 with honors. Thesis title: New electrochromic materials with cyclopentadienyl struc- ture Thesis supervisor: Prof. Dr. G.Chidichimo and Dr.Lucia Veltri (Univer- sity of Calabria) 09.2001-07.2006 High School Science Diploma Liceo scientifico “Michele Guerrisi” (Italy) with a final vote 94/100. ~ 117 ~ Publications Y. Surace, M. Simoes, J. Eilertsen, L. Karvonen, S. Pokrant, A. Weidenkaff, Functionalization of Ca2MnO4-delta by controlled calcium extraction: Activation for electrochemical Li intercalation, Solid State Ion., 266 (2014) 36-43. Y. Surace, M. Simões, L. Karvonen, S. Yoon, S. Pokrant, A. Weidenkaff, Freeze drying synthesis of Li3MnO4 cathode material for Li-ion batteries: A physico-electrochemical study, J. Alloy. Compd., 644 (2015) 297-303. Y. Surace, M. Simoes, J. Eilertsen, L. Karvonen, S. Pokrant, A. Weidenkaff, Activation of nano-Ca2MnO4 for electrochemical lithium intercalation, MRS Spring 2015 proceeding, Vol 1805/2015. Y. Surace, M. Simões, S. Pokrant, A. Weidenkaff, Capacity fading in Li3MnO4: a post- mortem analysis, submitted M. Simões, Y. Surace, S. Yoon, C. Battaglia, S. Pokrant, A. Weidenkaff, Hydrothermal vanadium manganese oxides: Anode and cathode materials for lithium-ion batteries, J. Power Sources, 291 (2015) 66-74. J. Eilertsen, Y. Surace, S. Balog, L. Sagarna, G. Rogl, J. Horky, M. Trottmann, P. Rogl, M.A. Subramanian, A. Weidenkaff, From Occupied Voids to Nanoprecipitates: Synthesis of Skutterudite Nanocomposites in situ, Zeitschrift für anorganische und allgemeine Chemie, 641 (2015) 1495-1502. Contributions to conferences Materials Research Society (MRS) Spring meeting, San Francisco (CA), United States, April 6-10, 2015 – Poster contribution 65th Annual Meeting of International Society of Electrochemistry (ISE), Lausanne, Switzer- land, August 31-September 5, 2014 – Oral contribution Gordon Research Conference: Batteries, Ventura (CA), United States, March 9-14, 2014 – Poster Contribution European Material Research Society (E-MRS) Spring Meeting, Strasburg, France, May 27- 31, 2013 – Poster Contribution Workshop Powder Diffraction School (Modern Synchrotron Methods), PSI (Paul Scherrer institute), Vil- ligen , Switzerland, July 1-4, 2014