Structure and Stoichiometry in Lithium Inserted Metal Oxides

Abstract

The stoichiometries of lithium inserted transition metal oxides synthesised at ambient temperature are strongly dependent on the structure of the host compound. Recent chemical and structural studies of insertion compounds are discussed in this paper as examples of the structural characteristics critical in the determination of lithium stoichiometries.

The recent interest in materials which can incorporate lithium atoms into their crystal structures at ambient temperature has been stimulated by their potential use as the positive electrode in rechargeable (secondary) batteries.⁽¹⁾ Lithium is the lightest and most electropositive metal, and has an ionic radius which facilities its incorporation into many crystal structures without severe distortion or bond breaking. The secondary cells generally consist of a lithium metal anode, a liquid or polymer electrolyte which passes lithium ions, and a cathode material which acts as a host for the Li on discharge of the cell. Most commonly, Li reacts with solid compounds in a displacement type reaction, e.g.:

$$yLi + M{O_x} \to \frac{y}{2}L{i_2}O + M{O_{x - \frac{y}{2}}}$$

(1)

These reactions involve extensive bond breaking and structural reorganization, and are therefore not useful in a rechargeable cell. For some transition metal compounds, however, Li is incorporated into the structure on cell discharge to form a single compound:

$$Liy + M{O_x} \to L{i_y}M{O_x}$$

(1)

with an accompanying chemical reduction of the host metal atom. This reaction will be easily reversed if the structural change on Li accommodation is not severe; that generally means M-O bonds in the host are not broken. For such a reaction to be fast enough to form the basis of a practical electrochemical cell, the diffusion coefficient of Li in MO_x must be unusually large at ambient temperature. In its most general form, the reaction of this type is known as an insertion reaction, if the host compound has a layered structure, the more specific term intercalation is used.

The EMF of an electrochemical cell as a function of discharge is a sensitive measure of the stoichiometry of the chemical phases produced during the insertion process. The cell potential is given by:

$${V_x} = \frac{{RT}}{F}\ln \frac{{{\mu _x}}}{{{\mu _{Li}}}}$$

(1)

where x is the lithium stoichiometry, R, T and F are gas constant, temperature, and Faraday constant, μ _x is the chemical potential of Li in the host at stoichiometry x and μ _Li is the chemical potential of Li in the anode. The chemical potential of Li in the host depends on the reducibility of the host transition metal and is sensitive to the energy of the site Li occupies: Li in octahedral and tetrahedral sites in the same host will have different chemical potentials. Thus, if Li_xMO_y is a solid solution, V_x changes continuously with x, whereas if a 2 phase region occurs during discharge (in a particular range of x, 2 phases with different fixed x values are changing in proportion only) then V_x does not change over that range.

Insertion compounds form a broad class of materials in which both line phases and wide ranges of Li non-stoichiometry have been reported.⁽²⁾ Common stoichiometrics are in the range between 0.01 and 2.0 Li/ host metal. Aside from their technological use, insertion compounds are of great interest to the solid state chemist as they most often form metastable phases which cannot be synthesized by conventional routes. Such compounds may be synthesised either electrochemically, through discharge of Li/electrolyte/insertion compound cell, or chemically, by reaction of the host with a nonaqueous reagent which contains Li at a fixed chemical potential. Such chemical synthesis routes often make it possible to synthesize larger quantities of the pure insertion compound which are of interest for physical study, once the appropriate Li chemical potential for favorable reaction has been determined. Due to the significant volume changes associated with the insertion process, the compounds are generally available only in powder form.

Chalcogenides which have been studied extensively for lithium insertion reactions have had primarily layer type structures: the inserted lithium is accommodated in octahedral or tetrahedral sites between MX_n slabs otherwise loosely bonded to each other through chalcogen-chalcogen Van der Waals interactions.⁽²⁾ In oxides, however, layer compounds are generally unstable, and most reversible Li insertion reactions have involved transition metal oxide hosts with close packed oxygen arrays, or crystallographic shear structures. There are four general structural requirements for such compounds: an interconnected network of sites in the structure allowing the fast transport of the Li ions, sites of appropriate geometry in large numbers which allow the inserted Li to be accommodated in favorable coordination polyhedra, stability of the host structure against bond bending, and strong host M-O bonds that will not break on reaction with lithium. Structural factors play a major role in determining the stoichiometrics of the lithium insertion compounds, defining not only limiting and line phase stoichiometrics, but also the ranges of composition within which solid solution phases form. Many electrochemical and structural studies have been performed on metal oxide insertion compounds in recent years. The results of those studies can be used to determine the important structural influences on stoichiometry in this class of compounds. I will use the results of studies on two classes of compounds, close packed structures, and crystallographic shear structures, to discuss and illustrate those structural influences.