Influence of Surface Area of Mesopore Volume on Discharge Characteristics for Lithium-Air Secondary Battery

Abstract

Introduction Lithium-air secondary batteries have been expected to the next generation secondary battery due to extremely high theoretical energy density. The carbon material is generally used as a positive electrode material because of low cost and high conductivity. However, the overpotential is high because Li2O2 is deposited on the positive carbon electrode during discharge process and then blocks the diffusion path for lithium ion, oxygen, and electrolyte. Therefore, the structure of carbon material affects the electrochemical performance of lithium-air secondary battery. The relationship between the mesopore sutructure of the carbon material and the electrochemical performance is not clear. In this study, we prepared the various carbon electrodes using the porous carbons (CNovel) having larger surface area and pore volume and the conventional carbon material (acetylene black (AB)), and compared their electrochemical performances. Experimental CNovel MH (MH), CNovel P(2)010 (P(2)010), CNovel P(4)050 (P(4)050), and AB were used as the carbon materials. The carbon materials and polyvinylidene fluoride were mixed in an N-methyl-2-pyrrolidone solution. The obtained slurry was coated on the carbon paper and dried in vacuum at 80 oC overnight. After drying, the loading of the carbon material on the carbon paper was about 0.5 mg cm-2. The carbon electrodes were transferred into an argon-filled glovebox. The CR2032 coin type cell was assembled in the glovebox under argon atmosphere by using the obtained carbon electrodes (16 mm in diameter) as a positive electrode, lithium metal foil (16 mm in diameter) as a negative electrode, 1 mol dm-3 LiCF3SO3 dissolved in tetraethylene glycol dimethyl ether (TEGDME) as an electrolyte, and glass fiber filter (Whatman GF/A) as a separator. The separator was saturated with 1 mol dm-3 LiCF3SO3/TEGDME. For the electrochemical measurements, the discharge-charge tests were carried out in dry air atmosphere at 25 oC in the 2.0-4.5 V voltage range at the current density of 100 mA g-1. The specific capacity was calculated by the loading of the carbon material. Results and Discussion Fromthe pore size distribution of various carbon materials, the modal pore size of MH and P(2)010 was about 4 nm, while that of P(4)050 was about 50 nm. From the nitrogen adsorption-desorption isotherms, the order of the surface area of mesopore (mesopore range: 2-50 nm) was as follows: P(2)010 > MH > P(4)050 > AB. From the first discharge curves of various carbon electrodes (current density: 200 mA g-1), the discharge plateau voltages of P(2)010, MH, P(4)050, and AB electrodes were 2.69, 2.67, 2.68, and 2.58 V, respectively. Fig. 1 shows the relationship between the pore structure (the surface area of mesopore and the mesopore volume) of carbon materials and the first discharge capacity. The first discharge capacities of P(2)010, MH, P(4)050, and AB electrodes are 4231, 3652, 3552, and 2484 mAh g-1, respectively. The CNovel electrodes show higher first discharge voltage and capacity than the AB electrode. From Fig. 1a, the first discharge capacity increases with increasing the surface area of mesopore. The first discharge capacity is more strongly related to the surface area of mesopore than the mesopore volume. It was suggested that the increase in the thickness of the Li2O2 layer enhanced the overpotential. Therefore, the discharge reaction was cut off at the certain thickness of the Li2O2 layer. Based on these results, it was found out that the surface area of mesopore of the carbon material is more important parameter than volume for lithium-air secondary battery, since the thickness of the Li2O2 layer directly affects overpotential. References 1) J. Xiao et al., J. Electrochem. Soc., 157,A487-A492 (2010). 2) L. Wang et al., J. Power Sources, 234,8-15 (2013). 3) P. Guan et al., Electrochim. Acta, 129,326 (2014). 4) I. Landa-Medrano et al., J. Electrochem. Soc., 162,A3126-3132 (2015). Acknowledgments This research was partially supported by ALCA-SPRING, JST. Figure 1