First Principles and Experimental Study of Surface Redox Reactions in Li2MnO3

The development of rechargeable lithium-ion batteries (LIB) has progressed rapidly to meet the demand for consumer electronic devices such as cellular phones and laptop computers, and the recent demands for larger energy storage applications (xEV and grid storage) require further development of LIB or alternative energy storage technology. Among possible alternative cathode materials compared to the currently used material LiCoO₂, Li-rich compounds xLi₂MnO₃·(1-x)LiMO₂ (M = Ni, Co, Mn) are the most promising candidate with higher capacity, lower cost and absence of toxic element. Within the two-component composite compounds, Li₂MnO₃ plays a central role in increasing the capacity of the compounds and independently possess a high theoretical capacity of 460 mAh/g (if the entire lithium is utilized). Although the Li₂MnO₃ is not electrochemically active between 2.0V and 4.4V, it is active to extract lithium ions over 4.5V. During the initial delithiation process, Li₂MnO₃ is known to transform into layered LiMnO₂structure, and the subsequent charge/discharge cycling induces a gradual phase transformation to the spinel structure, resulting in lowering voltage plateau and a large irreversible capacity accompanied by oxygen loss. To overcome these capacity loss problems, diverse approaches have been made to improve electrochemical properties of the cathode materials: adjusting the relative ratio of different transition metal ions, coating layers on the particles, and synthesizing at various temperature to control specific surface area.

Even though the bulk properties of cathode materials have been studied by many research groups, the surface properties of cathode have seldom been researched so far. Specifically, theoretical study on cathode surface has been rarely performed even though major reactions and chemical transformations occur near the surface and interfaces between different phases. Recently, several research groups highlighted and reported experimental observations of a phase transformation from the layered structure to spinel phase at the particle surface upon the first charge cycle. However, it is still unclear and controversial for theoretical and fundamental understanding on the mechanism of phase transformations. Without the basic understanding based on detailed atomic scale theoretical analysis, many problems of LIB materials will take extreme efforts and time to resolve the problems.

In the present paper, we established a fundamental understanding on the origin of phase transformation at the surface through first principles study of interface models. The theoretical findings suggest unusual redox activities of the surface layers of Li₂MnO₃which is confirmed by detailed experimental study. These theoretical and experimental findings provide basic understanding to improve performance degradation by phase transformation.

For the experimental study, coin cells are made with Li₂MnO₃ cathode and Li reference electrode, and variation of voltage between 2.0 V and various limiting voltage is applied to the electrochemical test with a constant current at room temperature and under 0.05 C-rate during 10 cycles. There is a small amount of redox capacity (4 mAh/g) at lower voltage as shown in Fig. 1a inset, and Figs. 1d-f show 10 mAh/g capacity during subsequent 10 cycles with lower limiting voltage of 4.1-4.3 V. Fig. 1a shows voltage plateau at 4.6 V extracting Li in the bulk Li₂MnO₃ with accompanying phase change to layered LiMnO₂with lower voltage during subsequent cycles.

For the modeling study, in order to elucidate the low voltage redox mechanism observed in Fig. 1, we examined the Li intercalation potential at the surface using the model shown in Fig. 2. The model interface contains stoichiometric ratio of constituents (Li : Mn :O = 2 : 1 : 3) split into two regions of a semi-infinite bulk (SIB), which is inactive at low voltage (4-4.5 V), and a interface layer (IFL) exposed to vaccum, which is calculated to be active at low voltage (~4 V).

Finally, detailed electronic structures of the interface model show different oxidation state of surface Mn atoms compared to the fully oxidized bulk Mn⁴⁺ within the bulk. Since the high activation potential of 4.6 V of bulk Mn is known to be related to electron extraction from oxygen atoms coordinating Mn⁴⁺, any lower Mn oxidation state on the surface would reduce the delithiation potential comparable to those of layered LiMnO₂ and spinel (3-4 V). Furthermore, delithiation potential at the surface show strong dependence on the Li location in Mn-layer (4.6 V) and Li-layer (4V) at the surface of Li₂MnO₃. The preferential delithiation of Li from the surface Li-layer would facilitate the Mn atom migration to the Li-layer causing spinel phase transformation at the surface. These underlying understanding mechanisms of the surface will provide a conceptual basis to develop diverse approached to suppress phase transitions in the cathode materials in LIB, which is our current modeling and experimental research topics.

This work was supported by the Industrial Strategic technology development program(10041589) funded by the Ministry of Knowledge Economy(MKE, Korea)