Borate based cathode materials for lithium- and sodium-ion batteries
Boron, the fifth element of the periodic table, doesn’t occur in its elemental form in nature, only as its oxide form. More than 200 borate minerals are found in earth’s crust, so called borates, but only a few of them are of commercial interest: borax, kernite, ulexite and colemanite (Table 4).
Starting in the 19th century, borates based glasses have been studied for their optical properties and later on for alkali ion conduction.128,129 Nevertheless crystalline borate based materials were mainly investigated because of its non-linear optical properties and as host structures for light emitting phosphors. Since the discovery of β-BaB2O4 and LiB3O5 in the mid 1980’s (which are up to now still the most frequently used borate based non-linear optical properties (NLO) materials) an increased interest in the scientific community has led to the discovery of a large variety of borate based compounds.,132 It is believed that these optical properties derive from the unique crystal- and electronic resulting from the small boron atoms inside an oxide matrix.
Lithium transition metal borates
The first transition metal borate compounds LiMBO3 were investigated since the late 1990’s for their optical and structural properties (M = Mg, Co, Cd, Zn),135–137 it was first in 2001 when Legagneur et al. introduced the family of lithium 3d-metal borates LiMBO3 (M = Fe, Mn, Co) as potential cathode materials for Li-ion batteries, with theoretical capacities exceeding 200 mAh/g. The three compounds were prepared by classical ceramic synthesis at high temperatures.138 The crystal structures for M = Mn, Fe, Co crystallizing in a monoclinic space group are shown in Figure 31, and are built up of edge sharing [MO5] trigonalbipyramids running along the c-axis. These columns are connected through planar BO3- groups and split lithium sites in tetrahedral coordination. In general the three compounds differ only in the size of the trigonal-bipyramid depending on the transition metal ion. Only for M = Mn a hexagonal polymorph is reported (Figure 31d) which can be obtained from the monoclinic form if heated above 550°C.
Search for new pyroborate B2O5 based compounds
It is nowadays established that by modifying the polyanion of a cathode material, elevated redox potentials versus Li/ Na can be achieved. Following this direction, we tried in this thesis to increase the potential of borate based materials, since for instance LiFeBO3 shows a voltage of 2.8 V which is rather low compared to other polyanionic cathode materials. To increase the redox potential condensed polyanions can be introduced, as previous recalled in switching from LiFePO4 to Li2FeP2O7.
Thus, for Fe-based materials, the expected potential should be well above 2.8 V vs. Li+/Li0 if going from BO3 to B2O5 or even B2O5F.
We have followed this strategy and our results are reported in this chapter which first gives a general overview about the reactivity of borates and possible synthetic routes for crystalline borate materials and then focuses on the experimental description of our first synthetic approaches, theoretically leading to pyroborate cathode materials.
A few synthesis considerations
Given by the nature of boric acid H3BO3 (mainly used precursor for borates) which contains one boron atom connected to three hydroxyl groups, borate based compounds can be synthesized through different synthetic methods. It is known that H3BO3 can easily undergo dehydration/ polymerization reactions leading to unpredictable crystal structures. As schematically shown in Figure 36, a complete- or a partial dehydration of H3BO3 can be realized, leading to new polyanions that can be viewed as a polymerization of the ortho borate group BO3. These different synthetic routes can involve more or less all techniques used in solid state/ inorganic chemistry including high temperature ceramic-, or low temperature mechanochemical or solvothermal methods. Some of these approaches will exemplary be described in the next sections, whether successful or not, however with the aim to possibly provide ideas for future synthetic work.
Structural and electrochemical properties of a lithium copper pyroborate Li6CuB4O10
The synthesis of Li6CuB4O10 was first reported in 2006 by Pan et al.,171 but already mentioned in Sparta’s PhD thesis in 2003. The synthesis of this phase was done by the authors through mixing stoichiometric amounts of Li2CO3, CuO and H3BO3 and annealing the mixture at 590°C for 2-3 days, denoted as α-Li6CuB4O10 afterwards.173 To obtain a better crystallized product our synthesis consisted in mixing stoichiometric amounts of the same precursors with mortar and pestle, followed by a decomposition of H3BO3 at 450°C for 4 h and annealing the mixture for 48 h at 650°C. A royal blue powder of pure α-Li6CuB4O10 was obtained, showing a particle size from 20 to 50 μm with no signs of impurities (Figure 38a). Later in 2013, Kuratieva et al. isolated a crystal of another Li6CuB4O10 polymorph denoted as β-Li6CuB4O10, in exploring the Li2O-CuO-B2O3 ternary system.172 Through the course of this thesis we found that bulk samples of β-Li6CuB4O10 can be obtained from α- Li6CuB4O10 if the latter is annealed for at least 9 days at 500°C in air with a particle size distribution from 10 to 50 μm,.
Structure and polymorphism
Two structural models were reported for Li6CuB4O10, the first one by Pan et al. (α- polymorph) and the second one by Kuratieva et al. (β-polymorph) both crystallizing in a triclinic unit cell. Their structural models were originally derived from single crystal diffraction and reported without any structural or synthetic relationship between them. For α- Li6CuB4O10 the XRD pattern can be indexed using Pan’s triclinic cell, with the exception of weak peaks which could not be indexed suggesting a superstructure. This superstructural peaks (Figure 40, inset) were already mentioned in Sparta’s PhD thesis and explained in a structural model including a tripled unit cell. This structural model fits perfectly the recorded XRD pattern (Figure 40a) with lattice parameters listed in Table 7. Regarding β-Li6CuB4O10 the XRD pattern could be proper fitted using the triclinic model described by Kuratieva et al. (Figure 40b) with lattice parameters shown in Table 8.
Synthesis, structural and electrochemical properties of sodium transition metal pentaborates Na3MB5O10 (M = Fe, Co)
In the previous chapter we have demonstrated through the synthesis of Li6CuB4O10 the feasibility to obtain high potentials in borate compounds. Therefore at the same time we show the difficulty in making new pyroborates. This was an impetus to explore other borate anions, hence our interest for the family of sodium metal pentaborates Na3MB5O10. Solid state synthesis and structures have been reported to prepare these phases with M = Mg, Ca, Zn.
Their preparation consists briefly in mixing stoichiometric amounts of Na2CO3, M(II)carbonate/nitrate and H3BO3 followed by annealing of the mixture at temperatures between 650-700°C in air for several days.191,192 Our first attempts by replacing the M(II)- carbonate/nitrate through Co/Fe(II)-oxalate and heating the mixture under argon to avoid oxidation of the transition metal were not successful. Using NaOH instead of Na2CO3, led to the formation of Na3FeB5O10, but was again not successful in the case of Co. Na3CoB5O10 could solely be obtained if Co(OH)2 was used as precursor. If the oxalate was used, we always end up with Co metal due to the easy reduction of Co2+ in Co-oxalate if heated under argon. Since the precursors are prone to react with ambient atmosphere, all steps of the synthesis were carried out in an argon filled glove box or under argon flow in a tube furnace. All chemicals were stored in argon, sodium hydroxide NaOH and boric acid H3BO3 were dried prior to be used at 200°C for 4 h in vacuum, and at 55°C for 24 h in air respectively. In general, stoichiometric amounts of NaOH, Fe(C2O4)∙2H2O or Co(OH)2 and H3BO3 were mixed together for 15 min in argon using a SPEX milling apparatus. We observed that the mechanochemical process causes most likely a reaction between the precursors which solidified within the milling jar, yielding in an amorphous intermediate product. The resulting solid was then reground with a mortar and pestle, and heated with a rate of 10°C/min up to 700°C for 1 h under argon flow. Once the synthesis is finished, the formed product Na3FeB5O10 was immediately transferred to the glove box for further use, since we observed degradation in ambient atmosphere. At the opposite, Na3CoB5O10 is stable in air. Both compounds possess particle sizes ranging from 1 to 50 μm as show in Figure 62 with grey and blue color for M=Fe and Co respectively.
Table of contents :
Table of content
1.1 Batteries for electrochemical energy storage
1.2 Lithium- and sodium ion batteries
1.2.1 Lithium ion batteries
1.2.2 Sodium ion batteries
1.3 Cathode materials for lithium- and sodium –ion batteries
1.3.1 Lithium- and sodium transition metal oxides
1.3.2 Polyanionic cathode materials and the inductive effect
1.3.3 Conversion type cathode materials
1.4 Borate based cathode materials for lithium- and sodium-ion batteries
1.4.1 Why borate based materials?
1.4.2 Lithium transition metal borates
1.5 Motivation and aim of the thesis
2 Search for new pyroborate B2O5 based compounds
2.1 A few synthesis considerations
2.2 Ceramic synthesis
2.3 Low temperature synthesis
3 Structural and electrochemical properties of a lithium copper pyroborate Li6CuB4O10
3.1 Structure and polymorphism
3.2 Electrochemical characterization
3.2.1 Activity versus lithium
3.2.2 Ionic conductivity
4 Synthesis, structural and electrochemical properties of sodium transition metal pentaborates Na3MB5O10 (M = Fe, Co)
4.1 Synthesis and structure
4.2 Electrochemical characterization
4.2.1 Activity versus sodium
4.2.2 Ionic conductivity
5 Study of the electrochemical driven reaction mechanism of bismuth oxyborate Bi4B2O9 versus lithium
5.1 Synthesis and structure
5.2 Electrochemical characterization
6 General conclusion
7.1 Electrochemical characterization
7.1.1 Galvanostatic techniques
7.1.2 Potentiostatic techniques
7.1.3 Electrochemical impedance spectroscopy
7.1.4 Direct current polarization measurements
7.2 Structural characterization
7.2.1 Laboratory XRD
7.2.2 Synchrotron XRD
7.3 Other physical characterization
7.3.1 Thermal analysis
7.3.2 EPR spectroscopy
7.3.3 Mössbauer spectroscopy
7.3.4 Scanning- and transmission electron microscopy
7.3.5 Density functional theory calculations
7.3.6 Bond valence energy landscape calculations