Conversion-type electrodes based on transition metal oxides are promising anode materials for the next-generation lithium-ion batteries (LIB) due to their high theoretic specific capacities, compared to the commercially used graphite anodes (372 mAh/g). Cobalt oxide based anodes exhibit theoretic specific capacities of up to 890 mAh/g (Co₃O₄). However, conversion materials need to overcome several drawbacks, including significant capacity losses after the first discharge, relatively low cycling stabilities, and a large potential hysteresis between charge and discharge.

According to Larcher et al.^[1], the charge / discharge reactions at very low current densities can be written as:

10 Li⁺ + Co₃O₄ + 10 e^- ⇌ Li₂O + 3 CoO + 8 Li⁺ + 8 e^- ⇌ 3 Co + 4 Li₂O. (1)

For higher discharge current densities, the formation of the Li_xCo₃O₄ metastable intermediate phase, was reported^[1].

x Li⁺ + Co₃O₄ + x e^- → Li_xCo₃O₄. (2)

Since the intermediate reactions taking place during the conversion mechanism are only partially understood, a combined experimental, modelling and simulation approach can be used to clarify the thermodynamics of the first discharge of Co₃O₄-based conversion anodes.

Phase formation and open circuit voltages during lithiation / delithiation can be calculated using CALPHAD-based models and simulations of the multi-component systems. However, reliable thermodynamic and phase diagram data are essential pre-requisites to be able to model the Gibbs free energies of the phases for generation of self-consistent thermodynamic descriptions. Therefore, key thermochemical and phase diagram investigations were performed in this work to clarify inconsistencies in the existing literature data. Further, electrochemical tests were conducted to validate the results of the simulations as well as to generate new data for the thermodynamic modelling.

First, the enthalpy of reduction of Co₃O₄ to CoO was determined using high-temperature oxide melt and transposed temperature drop calorimetry. This reaction is of particular importance because it is an intermediate step in the conversion reaction (see equation (1)). Secondly, since reliable heat capacity data are needed to extrapolate the thermodynamic descriptions to temperatures relevant for battery applications, the heat capacity of Co₃O₄ was measured in the temperature range of 200 to 1150 K using differential scanning calorimetry. The experimental data were then compared to calculations performed using an existing thermodynamic description of the Co-O system^[2] to assess its reliability for extrapolations to higher order systems.

A thermodynamic dataset of the Li-Co-O material system was developed. The results obtained from the simulation of the titration of LiCoO₂ and CoO versus Li using the thermodynamic description were checked with EMF and phase diagram data at 673 K available in literature^[3]. Furthermore, electrochemical cycling tests, galvanostatic intermittent titration technique (GITT) experiments and entropy measurements using coin cells were performed in our lab to verify extrapolations of the thermodynamic calculations to room temperature. In order to reproduce the experimentally measured quasi-open circuit voltage profiles, a thermodynamic description of the metastable Li_xCo₃O₄ phase (see equation 2) was included into the thermodynamic Li-Co-O dataset.

[1] D. Larcher, G. Sudant, J.-B. Leriche, Y. Chabre, J.-M. Tarascon, J. Electrochem. Soc. 2002, 149, A234.

[2] M. Chen, B. Hallstedt, L. J. Gauckler, J. Phase Equilib. 2003, 24, 212.

[3] N. A. Godshall, I. D. Raistrick, R. A. Huggins, Mat. Res. Bull. 1980, 15, 561.