Odile Bohnke

Designing fast oxide-ion conductors based on La2Mo2O9

The ability of solid oxides to conduct oxide ions has been known for more than a century, and fast oxide-ion conductors (or oxide electrolytes) are now being used for applications ranging from oxide fuel cells to oxygen pumping devices. To be technologically viable, these oxide electrolytes must exhibit high oxide-ion mobility at low operating temperatures. Because of the size and interaction of oxygen ions with the cationic network, high mobility can only be achieved with classes of materials with suitable structural features. So far, high mobility has been observed in only a small number of structural families, such as fluorite, perovskites, intergrowth perovskite/Bi2O2 layers and pyrochlores. Here we report a family of solid oxides based on the parent compound La2Mo 2O9 (with a different crystal structure from all known oxide electrolytes) which exhibits fast oxide-ion conducting properties. Like other ionic conductors, this material undergoes a structural transition around 580 °C resulting in an increase of conduction by almost two orders of magnitude. Its conductivity is about 6 × 10 -2 S cm-1 at 800 °C, which is comparable to that of stabilized zirconia, the most widely used oxide electrolyte. The structural similarity of La2Mo2O9 with β-SnWO 4 (ref. 14) suggests a structural model for the origin of the oxide-ion conduction. More generally, substitution of a cation that has a lone pair of electrons by a different cation that does not have a lone pair—and which has a higher oxidation state—could be used as an original way to design other oxide-ion conductors.

Mechanism of ionic conduction and electrochemical intercalation of lithium into the perovskite lanthanum lithium titanate

Abstract The ionic conductivity and electrochemical intercalation properties of La 2 3−x Li 3 x TiO 3 solid solutions (for 0.07 ≤ x ≤0.13) have been studied. These compounds present a perovskite-type structure (ABO 3 ) with cation deficiency at the A-sites. The purely ionic conductivity was confirmed and the mechanism of ionic conduction investigated using impedance spectroscopy techniques. We find that the temperature dependence of conductivity can be modelized by a Vogel-Tamman-Fulcher (VTF)-type relationship. In these materials, where the high ionic conductivity may originate from the presence of vacancies in the A-sites of the perovskite structure, the VTF behavior would suggest a mechanism of conduction involving the tilting of the TiO 6 octahedra. The lithium intercalation was also investigated in LiClO 4 (M)-PC electrolyte using galvanostatic discharge and charge at very low rates (one Li/250 and /1500 h) in order to approach the equilibrium. It was shown that the lithium intercalation leads to the presence of a plateau around 1.5 V/Li in the discharge curve, it is partly reversible and the capacity of the electrode is not very high. A maximum lithium uptake of 0.15 was found. The diffusion coefficient of lithium in the intercalated material was determined by impedance spectroscopy at room temperature and found to range from 10 −8 cm 2 s −1 to 10 −9 cm 2 s −1 as intercalation proceeds. Since the experimental impedance spectroscopy data performed at room temperature follow a Warburg behavior at low frequency, the intercalation seems to proceed in a single-phase process although a plateau is observable in the discharge curve.

Conductivity measurements on nasicon and nasicon-modified materials

Conductivity measurements have been performed on the solid solution series, Na3Zr2−x/4Si2−xP1+xO12, for 0<x<2. These materials are Zr-deficient nasicon. A monoclinic to rhombohedral phase transition occurs around x=0.333. The variation of the ionic conductivity as a function of the composition x is investigated in the temperature range 150–600 K. It is shown that the bulk ionic conductivity decreases slightly from 6×10−4 S cm−1 for the Zr non-deficient nasicon (x=0) to 6×10−5 S cm−1 for the Zr highest-deficient material (x=1.667). The presence of the above mentioned structural change does not affect the ionic conductivity. The mobility of the Na+ ions in these materials seems to be mostly influenced by the size of the bottlenecks through which the ions have to pass. The variations of both the ionic conductivity and the activation energy of the ionic motion process in the bulk of the material can be explained by structural considerations.

Electrochromism in WO3 thin films. I. LiClO4-propylene carbonate-water electrolytes

Abstract The performances of electrochromic cells containing evaporated amorphous WO 3 thin films as electrochromic material in 1M LiClO 4 -propylene carbonate-water electrolytes are presented. Much attention has been paid to some parameters such as the thickness of the layer, the overpotential applied to WO 3 electrode during the electrochemical coloration and the amount of water contained in the electrolyte (from 50 ppm to 10% in weight). Simultaneous electrical and optical in situ measurements have been carried out to study electrochromism. The optical data were stored into a microprocessor and restituted after treatment. The method used here gave us the possibility to rapidly test electrochromic materials.

Anomalies in Li+ ion dynamics observed by impedance spectroscopy and 7Li NMR in the perovskite fast ion conductor (Li3xLa2/3-x□1/3-2x)TiO3

Abstract A previous study, performed on the fast ionic conductor Li 3 x La 2/3- x TiO 3 by 7 Li and 6 Li Nuclear Magnetic Resonance (NMR), has shown that Li + ions undergo two different motions: a fast motion ( τ c ≈10 −9 s) inside the A-cage of the perovskite structure and a slower one ( τ c ≈10 −6 s) from one A-cage to a next vacant one. Furthermore, a change of these two motion mechanisms is observed around 200 K. Apart from NMR, impedance spectroscopy may also afford information on the ionic motion mechanism. Lithium motion in Li 3 x La 2/3- x TiO 3 is then studied by impedance spectroscopy in the 1 Hz–10 MHz frequency range and in the 140–500 K temperature range. The results obtained by these two techniques, i.e. 7 Li NMR and impedance spectroscopy, are then compared in the 140–270 K temperature range. As observed in NMR, the dc conductivity shows a change in the mechanism of ionic motion around 200 K. Apart from the dc plateau, the real part of conductivity ( σ ′) displays a dispersive behavior at high frequencies. Plotting the ac data in terms of impedance and modulus reveals the presence, in the mechanism of conduction, of both a nonlocalized conduction (long-range motion of the mobile ions) and a localized one (dipolar relaxation). According to these experimental observations, an equivalent electrical circuit is proposed, taking into account the physical processes assumed to be present when a small electrical signal is applied to the oxide. Both dipole polarization and long-range motion of the mobile ions are included in the electrical circuit of the conductive pathways. A complex nonlinear least squares fitting procedure (CNLS) is used to fit this electrical model to the experimental conductivity vs. frequency response ( σ ′ and σ ″). This procedure shows that all the parameters linked to the conductive pathways undergo a sudden change around 200 K, suggesting that a change in the ionic motion mechanism occurs at this temperature. This result is discussed in relation to both the crystallographic structure of the ionic conductor and the results previously obtained by 7 Li NMR.

Modeling Li-ion conductivity in fast ionic conductor La2/3−xLi3xTiO3

Abstract Monophasic samples of fast ionic conductors La2/3−xLi3xTiO3 (LLTO), with x varying from 0.06 to 0.13, are prepared by solid-state reaction. The total dc-conductivity is measured by complex impedance spectroscopy in the 10 MHz –1 Hz frequency range. Considering only the resulting location of oxygen atoms and employing bond valence equations, the conduction geometry and dc ionic conductivity are modeled. An averaged pathway for the Li+ conduction is proposed in this paper, assuming that the time-averaged position for Li+ is the geometrical centre of the A-cage. The saddle point of this pathway (Vumax) can be related to the activation energy for the ionic jump. Moreover, in order to model dc-ionic conductivity, not only activation energy, but also number of carriers and site occupancies have been considered. We propose three possibilities for the Li+ location in the structure in order to predict bulk conductivity in LLTO phase. Experimental evidence allows the exclusion of one of the three possibilities, while the other two are both in agreement with experimental values.

A distribution of activation energies for the local and long-range ionic motion is consistent with the disordered structure of the perovskite Li3xLa2/3-xTiO3

Abstract 7 Li nuclear magnetic resonance spin–lattice relaxation time T1 versus temperature is reported in the 150 K–900 K temperature range on lithium lanthanum titanate fast ionic conductors. Because of the presence of disorder in the distribution of the lanthanum ions in the crystalline structure of this oxide and consequently in the conduction pathways of the lithium ions we propose to explain the strong asymmetry shown by these T1 versus 1/T curves by assuming independent ionic hops over a distribution of activation energies for the thermally activated Li+ ion hops. According to this assumed model the spin–lattice relaxation times T1 and the DC conductivity are fitted consistently in the 200–600 K and 300–400 K temperature ranges respectively. For both lower and higher temperatures a departure of the experimental data from the model is observed and explained. The use of this model to fit both T1 and DC conductivity data ruled out the possibility that different forms of the distribution would lead to a reasonable representation of T1. The physical meaning of the obtained parameters is discussed in accordance with the structure of the compounds.

Electrochemical intercalation of lithium into the ramsdellite-type structure of Li2Ti3O7

The chemical and electrochemical intercalation of lithium ions into the ramsdellite-type structure of Li2Ti3O7 has been carried out by chemical reduction with n-BuLi or by galvanostatic discharge of an electrochemical cell. A maximum lithium uptake has been found to be of 0.60 and 1 per Li2Ti3O7 respectively at room temperature. During this intercalation process the ramsdellite network is preserved as shown either by electrochemical galvanostatic discharge or by X-ray powder diffraction analysis. However it is clearly shown by X-ray diffraction analysis and by electrochemical spectroscopy that two intercalation sites of very close energy are involved in this process. The potential difference between the two peaks observed in the voltammogram is 350 ± 20 mV. The variations of both the charge transfer resistance and the diffusion coefficient observed by a.c. impedance spectroscopy as intercalation proceeds can be explained by the presence of these two different intercalation sites.

An electrochemical quartz crystal microbalance study of lithium insertion into thin films of tungsten trioxide I. Modeling of the ionic insertion mechanism

A theoretical description of the mechanism of lithium insertion into amorphous thin films of tungsten trioxide (a-WO3) prepared by thermal vacuum evaporation of WO3 powder is presented. The model developed is based on the experimental results obtained by chronoamperometry and ac impedance spectroscopy associated with electrochemical quartz crystal microbalance (EQCM). The electrode mass change and the current flowing through the electrochemical cell during cathodic polarization are simulta neously recorded. As expected, it can be observed that the insertion process is associated with a gain of mass of the inserted electrode at long times (t > 1 s). On the other hand at short times (t < 1 s) a net loss of mass is observed. The existence of a non faradaic surface process involving the expulsion of anions from the electrode surface and occuring simultaneously with the faradaic insertion of non-solvated lithium ions into the oxide layer is discussed.

Ionic conductivity of crystalline and amorphous Na3Al2(PO4)2F3

Abstract The ionic conductivity of Na 3 Al 2 (PO 4 ) 2 F 3 (space group I4/mmm) has been studied by impedance spectroscopy in the temperature range 25–650°C for both the crystalline and the amorphous compounds. The dc conductivity is mainly due to a long-range Na + motion in the cavities formed by the arrangement of the [Al 2 O 8 F 3 ] bioctahedra and the [PO 4 ] tetrahedra of the material. According to the structure this motion would be bidimensional. The frequency dependence of σ ′ shows that, for both materials, the Na + motion is somehow correlated and that the increase of conductivity at high frequency may be ascribed to the relaxation of the neighbourhood of the mobile species after hops occurred. On the other hand, the frequency dependence of σ ″, for the amorphous compound, reveals at low frequency a dipolar relaxation which may be caused by the motion of small units of polyhedra present in this disordered material. This relaxation is thermally activated and does not appear in the crystalline compound which presents a long-range ordering.