Lithium-rich Phase Research Articles

The Lithium Flash

We present a scenario to explain the lithium-rich phase which occurs on the red giant branch at the so-called bump in the luminosity function. The high transport coefficients required to enhance the surface lithium abundance are obtained in the framework of rotation-induced mixing thanks to the impulse of the important nuclear energy released in a lithium burning shell. Under certain conditions a lithium flash is triggered off. The enhanced mass loss rate due to the temporary increase of the stellar luminosity naturally accounts for a dust shell formation.

Oxygen Nonstoichiometry in Li–Mn–O Spinel Oxides: A Powder Neutron Diffraction Study

Neutron powder diffraction has been carried out on two series of Li–Mn–O samples with spinel structure, which have been shown to lose oxygen with respect to initial stoichiometric spinels: (i) Li1.05Mn2O4quenched from 800 or 925°C, (ii) “Li1.33Mn1.67O4” (nominal) treated with ammonia gas at 200°C. The structural refinements unambiguously show that both 925°C quenching and low-temperature ammonia treatment induce oxygen vacancies in the spinel lattice. However, the mechanism of oxygen loss is markedly different between quenchings at 800 and 925°C. While the latter actually corresponds to the introduction of anionic vacancies at constant cation composition, the former is due to a rearrangement of the Li/Mn ratio in the spinel phase and the formation of an additional lithium-rich phase Li2MnO3. The presence of vacancies induce a significant increase in atomic displacement parameters of oxygen.

Raman scattering studies of lithium-intercalated NiPS 3

We report on the vibrational properties of lithium-intercalated NiPS 3 microcrystalline samples. Raman scattering spectra of Li x NiPS 3 (0 ≤ x ≤ 1.16) compounds have been recorded as a function of temperature in the wavenumber range 50–700 cm −1. The new peak located at 313 cm −1 is interpreted as a stretching mode due to vibration of lithium atoms located in the van der Waals gap against the nearest neighbouring sulphur atoms. Results are analyzed in terms of stretching vibrations of LiS 6 entities. It is concluded that lithium ions are more likely to occupy octahedral sites in the lithium-rich phase. Using a simple model of lattice dynamics, the coupling force constants are determined and compared with those in the pristine material.

Vibrational studies of lithium-intercalated SnS 2

We report on the vibrational properties of lithium-intercalated SnS 2 single crystal. Raman scattering and infrared absorption spectra have been recorded as a function of temperature in the frequency range 50–600 cm −1. The new bands are interpreted such as modes due to vibrations of lithium atoms located in the van der Waals gap against the nearest neighbouring chalcogen atoms. Results are analyzed in terms of stretching vibrations of LiS 4 and LiS 6 entities. From a comparison of the spectra, it is concluded that lithium ions are more likely to occupy tetrahedral sites in the lithium-rich phase. Using a simple model of lattice dynamics, the coupling force constants are determined and compared with those of the pristine material.

Vibrational studies of lithium-intercalated SnSySe2−y (0≤y≤2) layered compounds

We report the vibrational properties of lithium-intercalated SnSySe2−y (0≤y≤2) layered compounds. Infrared absorption spectra have been recorded as a function of temperature in the frequency range 50–600 cm−1. The new bands are interpreted as vibrational modes of lithium atoms against the nearest-neighbouring-chalcogen atoms located in the van der Waals gap. Results are analyzed in terms of stretching vibrations of Li-S4 and Li-S6 entities. From a comparison of the spectra, it is concluded that lithium ions are more likely to occupy tetrahedral sites in the lithium-rich phases. Using a simple model of lattice dynamics, the coupling force constants are determined and compared with those of the pristine materials.

Sol-gel processing of lithium disilicate

Lithium silicate powders were prepared by several sol-gel routes. Starting solutions contained equimolar amounts of lithium and silicon, but single-phase lithium disilicate (Li2Si2O5) formed only when local stoichiometry was maintained through gelation and drying. Gels prepared from solutions containing LiNO3, tetraethylorthosilicate (TEOS), water and ethanol were visibly homogeneous, but on drying the local stoichiometry was upset by the recrystallization of LiNO3. Consequently, a lithium-rich phase (Li2SiO3) was the first to crystallize on heating with a lesser amount of Li2Si2O5 forming at a higher temperature. Solutions containing TEOS and lithium methoxide formed a precipitate when combined with a water/ethanol hydrolysis solution. The precipitate dissolved before gelation, but the resulting powders crystallized into a mixture of Li2SiO3 and Li2Si2O5. The relative amount of Li2Si2O5 could be increased by adding HNO3 to the hydrolysis solution and using lower water contents. Precipitation was avoided by partially hydrolysing TEOS before adding the lithium alkoxide; these powders crystallized directly into Li2Si2O5 after heating at 550 °C. Gel-derived powders prepared using an Li-Si methoxyethoxide solution also crystallized directly into Li2Si2O5.

Comparative IR and Raman studies of various amorphous MoS 3 and Li xMoS 3 phases

Amorphous MoS 3 samples (a-MoS 3) prepared either by thermal decomposition of ammonium tetrathiomolybdate (the “classic” route) or by reaction of molybdenum halides with hexamethyldisilathiane (the “organic” route) are studied using IR and Raman techniques. The vibrational patterns of both amorphous compounds are quite similar and, in agreement with previous studies, the occureences of (S 2) 2− pair anions, of distorted prismatic coordinations for Mo 5+ ions and of MoMo single bonds are evidenced. We thus conclude the existence of a chain-like anisotropic structure similar to that known for the two-dimensional NbS 3 compound, so that the formal charge state is a-Mo 5+ (S 2) 1 2 2− (S 2−) 2 . IR data and partial Raman results for several Li xMoS 3 intercalated phases (0 < x ⪕ 4.5) show that on lithiation there is first a strengthening of the MoS bonds and then a reduction of the (S 2) 2− pairs. We conclude that lithium ions are primarily accommodated in “tetrahedral” sites in between the chains up to x ≈ 1.0 for classic a-MoS 3 and up to x ≈ 2.0 for organic a-MoS 3 and that additional lithium ions occupy other “tetrahedral” and “octahedral” sites in the van der Waals gaps. The different behaviours of the twp amorphous compounds are thus evidenced for lithium-rich phases; these results are related to electrochemical data showing that three Li + ions (per molybdenum) are reversibly intercalated in “organic” a-MoS 3 samples.

Li 21Si 5, a Zintl phase as well as a Hume-Rothery phase

Li 21Si 5 is the most lithium-rich phase in the binaryLiSi system and substitutes the so-called Li 22Si 5. This was proved by a careful X-ray structure analysis ( F4¯3m-T 2 d; a = 1871.0pm; Z = 16; R = 0.052; 397hkl ). The reduced Li content is due to ordered vacancies in the huge 6 × 6 × 6 superstructure of bcc ( cF416). This gives rise to the formation of M 26 clusters typical for γ-brass structures. Two different cluster types, namely Li 22Si 4 and Li 20Si 6, set up a complex linkage of diamond, zinc blende, and NaCl type partial structures, in which short Si Si distances are avoided. Li 21Si 5 is on one hand a representative consistent only of main group elements of the Hume-Rothery phases for structural and electronic reasons (VEC=20.5/13= 1.58). On the other hand the two M 26 clusters fulfill electronically an extended Zintl phase formalism, according to [Li 22Si 4 4+ and [Li 20Si 6 4−. The latter anion follows classical valence rules (Si 4− formation), while the former one carries a cage orbital in addition to the filled Si states, which is delocalized over the Li sites. The two quantum mechanically stable units act as donor-acceptor pairs, which is also reflected in the experimental valence electron density. In a new topological description the whole structure can be understood as being built of M 14 cluster nuclei and a continuous periodic curved surface formed by the network of all remaining Li atoms.

Infrared study of lithium intercalated phases in the Li xFeS 2 system (0 ⩽ x ⩽ 2). Characterization of a new iron disulfide

Infrared spectra (550−200cm −1) of the Li 2FeS 2 phase, of several chemically and electrochemically oxydized Li xFeS 2 samples with x=1.64, 0.98, 0.72, 0.28, 0.14 and ∼0.0, and of some reintercalated phases with x=0.69, 0.93 and 1.50 have been recorded and analyzed in terms of stretching vibrations of LiS 6 or LiS4 entities, of S 2 −2 anionic pairs and of Fe 3+ −S bonds. From a comparison of the spectra, it is concluded that lithium ions occupy “octahedral” and “tetrahedral” sites in the lithium rich phases, and that they are mainly localized in “octahedral” positions for all the other phases (0 ⩽ × < 1). In highly oxydized samples, the formation of S 2 −2 pairs is evidenced and new optical modes probably due to Fe 3+S bonds are also observed in the lithium free compound. Finally, spectra of the oxydized samples are in agreement with the oxydation of Fe 2+ in Fe 3+ in a first step and then of S −2 ions into S 2 −2 in a second one. Different mechanisms of oxydation and reduction are discussed and the existence of a new FeS 2 described by the developped formula Fe 3+ (S 2 −2) 1 2 S −2 is found.

Chemical diffusion in intermediate phases in the lithium-tin system

The compositional variation of the chemical diffusion coefficient in the six intermediate phases LiSn, Li 7Sn 3, Li 5Sn 2, Li 13Sn 5, Li 7Sn 2, and Li 22Sn 5 of the lithium-tin system at 415°C has been measured. Among these intermediate phases, the phase Li 13Sn 5 has the highest chemical diffusion coefficient, varying with composition from 5.01 × 10 −5 to 7.59 × 10 −4 cm 2/sec at that temperature. Combining this information with coulometric titration curves (emf versus composition), the self-diffusion coefficient of lithium has also been determined in the various intermetallic phases as a function of composition under the assumption that the tin atoms do not move appreciably compared with the lithium atoms. The lithium self-diffusion coefficient in the phase LiSn is lower than those in the more lithium-rich phases by one order of magnitude. This result is discussed in terms of the difference between the crystal structures of LiSn and the other lithium-rich phases in the lithium-tin system.