Analysis of SEI on Graphite Anode in Li Ion Cell Using LiMn2O4 Spinel Cathode

Introduction

We have been developing large capacity Li ion batteries for backing up of telecommunication apparatuses. Among the many types of Li ion batteries, The LiMn₂O₄ is convenient for replacing conventional lead-acid batteries, which have been widely used in such backup systems. This is because the LiMn₂O₄ has a voltage plateau, which is close to that of lead-acid batteries, due to the combination with a graphite anode. On the other hand, long calendar life of more than 10 years at full charge is also required for backup. Mn²⁺ ion dissolution from the LiMn₂O₄ cathode is well known, and we have already reported that calendar life is greatly affected by the concentration of dissolved Mn²⁺ ions [1]. We have postulated the degradation mechanism by charge exchange between Mn²⁺and intercalated Li [1]; however, the relationships between the amount of Mn deposited on the anode and that of Li consumption, which are calculated from capacity reduction, are unclear.

In this study, we analyzed the solid-electrolyte-interface (SEI) on the anode to investigate the role of Mn in the degradation of Li ion batteries for backup.

 

Experimental

We used the 18650-type cylindrical Li-ion cell (capacity: about 0.7 Ah) [2]. The cathode was a Mg-doped manganese-spinel and the anode was carbon-coated graphite. We used an ethylene carbonate and dimethyl carbonate based electrolyte containing phosphazene retardant.

Sample cells were charged at four voltages (4.05, 4.10, 4.15, and 4.20 V), and the environmental temperature during float charging was maintained at 50°C. The cells were discharged periodically at a current of 0.2 C (at 25°C) to evaluate their residual capacity.

Degraded cells were disassembled and six different points of the anode were analyzed. The Mn content was analyzed using an inductively coupled plasma (ICP) method. We also used Li⁷-nuclear magnetic resonance(NMR) to identify the species on the anode. A LiCl solution of 1M was used as the standard sample.

Results

Figure 1 shows the lifetime of each floating charge voltage. Lifetime is defined as the duration at which the discharged capacity remains under 70% initial capacity. We found that lifetime decreased greatly between 4.15 and 4.2 V, presumably due to the phenomenon of deterioration caused by electrolyte decomposition since this test was conducted at 50°C. The lifetime at 4.05 V was slightly shorter compared with those at 4.10 and 4.15 V. The reason for this is uncertain.

More than 103 times the amount of Li was observed on the degraded anode compared with that of Mg by using the ICP method. This result shows that the simple charge exchange mechanism between Mn²⁺ and intercalated Li will be rejected. Figure 2 shows the Li⁷-NMR spectrum of the anode. Two peaks were identified. One was the broad peak of which the chemical shift was -12.2 ppm. The other was that of which the chemical shift was 2.4 ppm. The area of the former was about 98%, and the latter was identified inorganic Li compound. The former may be an organic Li compound, although the chemical species remains unclear. It is known that SEI forms compact inorganic compounds, such as LiF, and organic Li compounds. The results from this study suggest that the formation of organic Li compounds is the main cause of Li consumption and that Mg deposition may accelerate organic Li compound formation.

References

(1)     T. Tsujikawa, K. Yabuta, T. Matsushita, M. Arakawa, K. Hayashi, J. Electrochem. Soc., 158, 322 (2011).

(2)     T. Tsujikawa, K. Yabuta, T. Matsushita, T. Matsushima, K. Hayashi, and M. Arakawa, J. Power Sources, 189, pp. 429-434 (2009).