Lithium-ion batteries (LiBs) are currently growing in popularity for high output applications such as power sources for electric vehicles, etc. However, it is known that the present composite electrodes of LiBs show relatively lower capacity under high rate charge/discharge [1]. One of the causes of the capacity loss is inhomogeneous electrochemical reaction in composite electrodes. Therefore, it is important to figure out where and how the reaction progresses in composite electrodes during charge/discharge. Several studies have reported that the reaction in a composite electrode proceeds from the side facing a counter electrode [2]. This suggests that the slow ionic transport in a liquid electrolyte causes the formation of the reaction distribution. In order to evaluate reaction distribution due to the slow ionic transport in a liquid electrolyte, we have proposed to use a model electrode with a laminated structure (Fig. 1) [3]. In this model electrode, the reaction distribution is formed in the in-plane direction if the formation of reaction distribution is controlled by the ionic transport. In this study, we performed operando observation of the reaction distribution in this model electrode by using two-dimensional X-ray absorption spectroscopy (2D-XAS). Combining with numerical simulations, we tried to quantitatively understand the influence of the slow ionic transport in a liquid electrolyte on the formation of reaction distribution.  

EXPERIMENTAL

The model electrodes were fabricated by mixing LiCoO₂ (LCO) powders (Nichia Co., Japan), acetylene black (Denka-Black, Denki Kagaku Kogyo Kabushiki Kaisha, Japan), and organic binder PVDF (KF Polymer L#1120, Kureha Co., Japan) with the weight ratio of 80:10:10 and covering the top surface of the composites by kapton films. The model electrodes had the dimension of 10 × 10 mm and about 40 µm thickness. The porosity was around 45 %. Electrochemical cells were fabricated with the model electrode, a lithium metal foil, EC-EMC with 1 moll⁻¹ of LiPF₆(EC:EMC = 3 : 7 in volume, Kishida Chemical Co., Ltd., Japan), and an polymer separator (Cell-Gurd #2400, Polypore International, Inc., U.S.).

The operando 2D-XAS measurements were carried out at the synchrotron radiation beam line BL28XU in SPring-8, Japan. The spatial and time resolution of the 2D-XAS measurements were 6.5μm and 1.5 min, respectively. The reaction distribution in the area of 300 × 1400 µm near the edge of the model electrodes were evaluated during a charge process of the electrochemical cells. The charge rate was 7 mAcm^-2for the cross-section of the model electrode, and the cut off voltage was 4.3 V.

RESULTS AND DISCUSSION

Figure 2 shows the charge curve of the electrochemical cell. The charged capacity was about 18 mAhg^-1. The reaction area can be roughly estimated from the charged capacity, assuming that the reaction took place from all four sides of the model electrode and the reaction area was uniformly fully charged (state of charge (SOC) ≈ 60 %). The reaction area estimated by this way was 280 μm. Figure 3 shows the reaction distribution in the model electrode at time 0, 20, 40, 60, and 69 min (cut off time) during a charge process. Blue area indicates the SOC of LCO is low, while the red area indicates the SOC is high. The charged area propagated from the edge to the inner part. This result suggested that the formation of the reaction distribution was caused by the slow ionic transport in liquid electrolyte. At the end of the charge, the edge of the model electrode was deeply charged (SOC ≈ 60 %) while the part over than 400 μm from the edge was little charged (SOC ≈ 0 %). The reaction area was in reasonable agreement with the estimated one from the observed capacity. Thus, it can be said that the reaction distribution in the electrode was strongly related to the capacity loss. In the presentation, we will make a more quantitative discussion on the relation between the formation of reaction distribution and the ionic transport in liquid electrolyte, comparing the results between the experiments and the numerical simulation.

REFERENCES

[1] J.W. Fergus, J.Power Sources, 195(2010)939.

[2] J.Liu et al., J. Phys. Chem. Lett., 1(2010)2120.

[3] T. Nakamura et al., Solid State Ionics, 262 (2014) 66.

ACKNOWLEDGEMENT

This work was supported by the RISING project from NEDO of Japan.