Understanding Why Sodium Intercalated Graphite Is Unstable

Sodium-ion batteries offer a potential solution to the problem of limited lithium resources, although developing compatible, high-performance cathode, anode, and electrolyte materials for such devices presents a more difficult challenge. In conventional lithium ion batteries, graphite is used as the anode, but it has been known for more than 40 years from experiment that, unlike the other alkali metals (Li, K, Rb, and Cs), intercalation of Na into graphite is thermodynamically unfavorable, so that modification of the anode material or addition of solvating species is needed if intercalation is to occur. First-principles calculations by Nobuhara et al. [2] and Wang et al. [3] showed that even at low Na densities NaC x compounds are energetically unstable. However, the underlying mechanisms responsible for the endothermic formation energies are still uncertain. To investigate this phenomenon in more detail we performed first-principles calculations of the intercalation behavior of alkali metals into graphite. A van der Waals correction was added to the standard density functional theory methods to better reproduce the laminated structures of graphite intercalated compounds. Calculated formation energies of intercalated systems XC6 (X = Li, Na, K, Rb, or Cs) are shown in Fig. 1(a). As the atomic number decreases from Cs to Na, the magnitude of the formation energy decreases, becoming positive (endothermic) in the case of Na. This tendency can be explained in terms of ionic bonding energies, which decrease as the atomic number of the alkali metal decreases because electronegativity increases as the alkali metal group is ascended. The reason for the negative formation energy in the case of LiC6 can be understood by examining the electron density and density of states, plotted in Fig. 1(b), which show greater covalent bonding between Li and C than the other alkali metals. References [1] K. Kim, et al., Energy Environ. Sci., 8, 2963 (2015). [2] K. Nobuhara, H. Nakayama, M. Nose, S. Nakanishi, and H. Iba, J. Power Sources, 243, 585 (2013). [3] Z. Wang, S. M. Selbach, and T. Grande, RSC Adv., 4, 4069 (2014). Figure 1