Co3O4 has shown acceptable electrochemical properties as the anode material of Li secondary batteries. In detail, its capacity reached about 700 mAh/g, twice as high as graphite, and 93.4% of initial capacity was retained after 100 cycles. EIS (electrochemical impedance spectroscopy) analyses revealed that after the 1st cycle, the insertion or extraction of Li ions in Co3O4 can occur homogeneously and reversibly (randless-like behavior, homogeneous mixture: Co + Li2O (in the state of Li insertion), Co3O4 (in the state of Li extraction)). In fact, the coulombic efficiency of Co3O4 was almost 100% except for the 1st cycle. According to P. Poizot's research on several kinds of transition metal oxides, such as Co3O4, CoO, NiO, etc., a small Li2O particle size and catalytic activity of the transition metal are expected to decrease the binding energy of Li2O tremendously. As a result, Li2O should be easy to decompose, and transition metal oxides should be able to charge or discharge reversibly by formation or decomposition of Li2O. However, this assumption has never been confirmed by experimental results. In our results, the CV (cyclic voltammogram) of a Li2O-Co mixture shows much larger oxidation and reduction peaks than that of Li2O. Based on XRD analyses, oxidation and reduction in the CV of Li2O correspond, respectively, to the decomposition and formation of Li2O, it can be asserted that Co addition to Li2O facilitates decomposition and formation processes in Li2O and that the catalytic effect of the transition metal must be one of the main causes that make Li2O form or decompose repeatedly. (c) 2005 Elsevier Ltd. All fights reserved.