Search

The nature of Li1+αV3O8·nH2O gel-like precipitate and corresponding xerogels has been studied. Unlike V2O5 gel, the Li1+αV3O8 gels are diphasic materials made of intimately mixed solid and liquid phases having very close Li/V stoichiometry. At room temperature, the solid component contains presumably an hewettite type structure, whereas the liquid part crystallizes into lithiated decavanadic acids. Upon drying at 90 °C, both components lead to hydrated Li1+αV3O8·nH2O hewettite-like compounds with different interlayer spacing but with the same water content. By coupling pH measurements, X-ray diffraction, and liquid and magic-angle spinning ⁵¹V NMR, a new mechanism of formation of the Li1+αV3O8 gel-like precipitate via solubilized vanadic species is proposed.

The synthesis and characterization of a polycrystalline sample of the n = 1 Ruddlesden−Popper phase La2Sr2LiRuO8 are reported. X-ray and neutron diffraction show that the six-coordinate sites are occupied by 1:1 ordered xy layers of Li:Ru, and the nine-coordinate sites are occupied by 1:1 ordered layers of La:Sr; a low concentration (7.4%, 9.8% respectively) of anti-site defects was detected on each sublattice. These layers stack in an ordered sequence along z, resulting in an orthorhombic unit cell with a = 5.46001(1), b = 5.45803(1), and c = 25.5410(2) Å (space group Imm2) at 290 K. The structure is confirmed by high-resolution transmission electron microscopy, although the presence of stacking faults is revealed. ⁷Li MAS NMR spectra are consistent with the diffraction data and show that the anti-site defects are not randomly distributed but occur in domains. The sample undergoes a transition to an antiferromagnetic phase at 85 K, with an ordered magnetic moment of 1.58(7) μB per Ru⁵⁺ aligned along [001]. The spins associated with the anti-site Ru cations freeze as a spin glass.

The conversion that occurs from layered LiMnO2 to spinel on electrochemical cycling has been studied by neutron diffraction and NMR. Neutron diffraction results indicate that the tetrahedral sites in the Li layers that share faces with octahedral sites in the transition metal layers are occupied even following the first charge to 4.6 V. NMR results are consistent with the conversion from the monoclinic, Jahn−Teller distorted, to the rhombohedral, layered phase on charging. On subsequent discharging, clear evidence for the monoclinic phase is seen by NMR indicating the presence of Jahn−Teller distorted domains in the material at 3.5 V and below. The fraction of monoclinic phase decreases gradually as a function of cycle number and disappears by 35 charge−discharge cycles. Diffraction patterns obtained as a function of cycle number were refined with a structural model that included both a layered phase (with octahedral and tetrahedral site occupancy) and a spinel phase, with the fraction of spinel increasing from 0.12 (5 cycles) to 0.93 following 92 cycles. Both diffraction and NMR results indicate that the spinel phase that nucleates from the layered material is stoichiometric and does not contain Li occupancy on the Mn sites. An additional site is seen by NMR which reaches a maximum in cycles 25−50, which is assigned to tetrahedrally coordinated Li in a local environment intermediate between that of the spinel and the layered phase. The observation of this site by NMR is associated with two characteristic peaks in the incremental capacity plot at 3.75 and 3.9 V on charge and discharge. The data indicate that the mechanism for Li insertion and removal into these local environments is complex and involves simultaneous structural rearrangements. The intermediate environment decreases in concentration on subsequent cycling as the concentration of the spinel phase continues to grow.

⁷Li and ⁵¹V MAS NMR spectra acquired during the electrochemical cycling of the layered battery material Li1+xV3O8 have been used to follow the local structural and electronic changes that occur during the different stages of intercalation/deintercalation. Several resonances are observed using ⁵¹V MAS NMR and an attempt to assign them to the different vanadium sites on the basis of the differences in the width of their sidebands manifolds is made. Both the ⁵¹V and ⁷Li NMR spectra show multiple vanadium and lithium local environments, and the spectra cannot be explained by using a simple model, based on the number of crystallographically distinct vanadium sites. Deintercalated samples with x close to 0 (prepared at potentials of 4.2 V vs Li) give ⁵¹V spectra with very poor resolution. On lithium intercalation, the ⁵¹V NMR resonances sharpen and shift to higher frequencies; three sharp resonances along with two broader resonances are clearly resolved for the samples prepared at potentials of 3.4 and 3.0 V (x = 0.3 and 0.5, respectively). This behavior is consistent with solid solution behavior in this potential range. Three lithium sites are observed which are assigned to the octahedral site and two tetrahedral sites between the lithium layers. The phase transition that is seen in the electrochemical data, but not with structural probes (such as X-ray diffraction), and that occurs close to 2.8 V (to form Li2V3O8), results in a decrease in the resolution of the ⁵¹V resonances, and the observation of new ⁷Li resonances at negative frequencies with much shorter relaxation times. Variable temperature experiments confirm that the shifts and short relaxation times are due to a transferred hyperfine interaction. Based on the ⁷Li and ⁵¹V spectra, the phase transition is ascribed to the formation of localized V⁴⁺ and V⁵⁺ ions. Further intercalation to form Li4V3O8 results in the complete loss of the ⁵¹V resonances and larger ⁷Li hyperfine shifts, due to the presence of unpaired (d¹) electrons.