Thermal treatments under a very wide range of oxygen pressures were used to probe the composition and defect nature of a lithium-overstoichiometric “Li x 0 CoO2” (x 0 > 1) sample using X-ray powder diffraction, 7Li NMR, and electrochemical tests. It was found that at 900 °C, under atmospheric and elevated oxygen pressures, the lithium-overstoichiometric sample gradually transformed to stoichiometric LiCoO2 by losing excess lithium in the form of Li2O. In addition, it was shown that the defect associated with Co2+ and oxygen deficiency as reported by Gorshkov et al. and Karelina et al. had a different NMR signature than that present in Li-overstoichiometric samples. Therefore, it is believed that oxygen vacancies are present but Co2+ ions are not present in Li x 0 CoO2 (x 0 > 1). This leads to a formula [Li]interslab[CoIII 1 - 3tCo3+(IS) 2tLit]slab[O2 - t], involving an intermediate spin configuration for 2t Co3+ ions in a square-based pyramidal site. This new model was supported by the NMR and magnetic data of the lithium-overstoichiometric sample and its deintercalated compounds. The effect of the defect on the end-of-discharge voltage profile during cycling was also discussed.

A large variety of pure p-Ni(OH), and cobalt-substituted nickel hydroxides Ni,_,Co(OH), of well-controlled chemical composition, particle size, and morphology have been prepared and characterized by x-ray, transmission electron microscopy, electrochemical, and Raman spectrometry studies. The hydroxide particle size and its Co-substituted content were found to govern the Ni-electrode capacity. Once these parameters were controlled, p-nickel hydroxide specimens able to reversibly exchange more than one electron per metal atom (Ni + Co) were routinely prepared. p-Ni,_1Co(OH), Raman spectra consists of five lines. The intensities of the ones located at 3605 and 515 cm' were found to change as a function of the particle size and the Co content. Indeed, the line at 3605 cm', ascribed to adsorbed water, appears and grows when the particle size decreases (e.g., when the surface/volume ratio increases), while the line located at 515 cm1 is enhanced by the presence of coprecipitated cobalt. A direct eorrelation between the intensity of the 515 cm' Raman line and the electrochemical capacity of the nickel hydroxide sample is found. Raman spectroscopy can then be used as a powerful nondestructive tool to differentiate "high capacity" from low capacity nickel hydroxide samples.

A thorough study was made concerning the formation of LiNiO2 by a low‐temperature ion‐exchange reaction under hydrothermal conditions. LiNiO2 with a low degree of cationic mixing and good electrochemical performance was prepared either from pure β‐NiOOH, γ‐NiOOH, or a mixture of both where the γ phase was the majority. The synthesized product false(LT‐LiNiO2false) presents electrochemical properties similar to LiNiO2 prepared by conventional powder synthesis false(HT‐LiNiO2false) in spite of the difference in Brunauer, Emmett, and Teller (BET) surfaces, around 20 m2/g for LT‐LiNiO2 and 0.7 m2/g for HT‐LiNiO2 . The thermal and moisture stabilities of LT‐LiNiO2 were studied and turned out to be much lower than those of HT‐LiNiO2 . This moisture instability was found (i) to be intrinsic to LiNiO2 , since ground HT‐LiNiO2 samples presented a similar behavior, (ii) to be enhanced by decreasing the LiNiO2 particle size, and (iii) to be much larger than that of LiCoO2 . In view of electrochemical applications, cobalt‐substituted products should be used in order to ensure the stability against both reduction‐delithiation and hydrolysis.