In this report, we show how controlling the size and ionic content of 1D diffusion channels in tunnel manganese oxides affects the diffusion of charge-carrying Li⁺ and Na⁺ ions. Manganese oxides are of significant interest for Li-ion and Na-ion batteries (LIBs and SIBs, respectively) due to their low cost, environmental friendliness, and high electrochemical activity. Specifically, tunnel manganese oxides are attractive candidates for intercalation-based batteries due to their one-dimensional (1D) diffusion channels built from corner and edge sharing MnO₆ octahedra arranged around stabilizing cations and water molecules. Hydrothermal synthesis methods are used to synthesize tunnel manganese oxides with high aspect ratio nanowire morphologies. By controlling the time, temperature, and precursors during hydrothermal synthesis, the size, shape, and ionic content of the tunnels can be carefully controlled.

Here, we synthesize three tunnel manganese oxides with different size 1D diffusion channels and ionic content and compare their rate performance in LIBs and SIBs with apparent Li⁺ ion and Na⁺ ion diffusion coefficients calculated from the galvanostatic intermittent titration technique (GITT). GITT consists of a repeated sequence of a current pulse to allow a set amount of charge to titrate into the cell, followed by a relaxation period to allow the electrochemical cell to return to equilibrium (Weppner and Huggins, J. Electrochem. Soc., 1977, 124, 1569). From the changes in voltage during each step, the apparent diffusion coefficient of the charge-carrying ion can be calculated. The three different tunnel manganese oxide phases studied are α-MnO₂, consisting of 2 octahedra by 2 octahedra tunnels stabilized by K⁺ ions, todorokite MnO₂, consisting of 3xn (n = 3, 4, 5, 6, 7) octahedra tunnels stabilized by Mg²⁺ ions, and a newly reported phase 2xn-MnO₂, which consists of an ordered array of 2x2, 2x3, and 2x4 octahedra tunnels stabilized by Na⁺ ions.

The apparent Li⁺ and Na⁺ diffusion coefficients are shown in Figure 1. In LIBs, todorokite MnO₂ demonstrates an initial Li⁺ ion diffusion coefficient of 1e-7 cm² s^-1, an order of magnitude higher than α-MnO₂ and 2xn-MnO₂. This result indicates the diffusion channels in todorokite MnO₂, which are larger than in the other two phases studied, are advantageous for Li⁺ ion diffusion. This is also reflected in the superior rate performance demonstrated by todorokite MnO₂ in LIBs. In SIBs, 2xn-MnO₂ demonstrates an initial Na⁺ ion diffusion coefficient of nearly 1e-7 cm² s^-1, one order of magnitude higher than α-MnO₂ and nearly two orders of magnitude higher than todorokite MnO₂. This superior Na⁺ ion diffusion coefficient is supported by the greater capacity demonstrated at higher current rates by the 2xn-MnO₂. Thus, despite containing smaller structural tunnels than the tunnels in todorokite MnO₂, the 2xn-MnO₂ demonstrates more facile Na⁺ ion diffusion. This observation is attributed to the fact that the 2xn-MnO₂ structure is stabilized by Na⁺ ions that create more well-defined diffusion channels and insertion sites favored by Na⁺ ions. This is further supported by the Li⁺ diffusion coefficient for 2xn-MnO₂, which is an order of magnitude lower than the Na⁺ ion diffusion coefficient. Thus, we find that the mobility of charge-carrying ions is not only determined by the size of the 1D diffusion channels in tunnel manganese oxides, but also by the ionic content stabilizing the channels. In summary, by utilizing the GITT, we show how 1D diffusion channel size and ionic content affects the diffusion coefficient of charge-carrying ions in tunnel structured materials.