Na is a probable alternative for Li in rechargeable batteries due to its lower cost and higher abundance. Na, however, has a larger ionic radius and smaller ionisation potential compared to Li. Given these shortcomings, to design a Na ion based cathode material comparable in performance to Li, all mechanisms that may increase the capacity and voltage beyond that of the current limits should be carefully fine-tuned [1]. Here, through comprehensive density functional calculations, we screened the full range of ternary hexagonal layered compounds with ilmenite structure (space group 158) that consist of Na, O and 4d or 3d transition elements (TM) in search for candidates with a possibly higher operating voltage. The investigated TMs were V, Cr, Mn, Fe, Co, Ni, Nb, Mo, Tc, Ru, Rh, Pd and Ag.

In the calculations, we treated the exchange-correlation interaction of the 3d TM containing compounds with hybrid HSE03 functional to account for the strong correlation effects. For 4d TM cases, on the other hand, we found that the GGA+U formalism sufficed for describing the band structure accurately. We, furthermore, found that a U_eff value of 2 eV could particularly reproduce the electrochemical potential of Na extraction in Na_xRuO₃ compound [2], which we then utilised for the rest of 4d TM containing compounds. Figures 1(a) and (b) show the structure of the sodiated compound Na₁TMO₃ (R^-3) and the desodiated compound Na_0.5TMO₃ (P^-31m) respectively.

In the ilmenite type Na₁TMO₃, due to the R^-3 symmetry, O ions are coordinated by two TM and two Na ions and two vacant sites. In this case, O under-coordination elevates the unhybridised or orphaned portion of O 2p states closer to the Fermi level. High O electronic population near the Fermi level facilitates greater O participation in redox over a wide range of Na concentrations. Since the potential associated with O redox is higher than that of TM redox, this mechanism also offers a new opportunity to achieve high voltage cathodes. Accordingly, our most notable prediction was Na₁VO₃ that is likely to have a potential of 5.91 V and is thermodynamically stable with respect to major competing Na_xV_yO_z compounds promising excellent cyclability. Interestingly, the redox mechanism in Na₁VO₃ is entirely borne on O centres as V has a d⁰ configuration. Similarly, Na₁NbO₃ was also found to be another suitable and economic candidate for high voltage and high capacity Na ion batteries with a voltage of 4.94 V which is entirely O borne [2]. We also found that Na₁FeO₃, Na₁MnO₃ and Na₁RuO₃ have a desirable potential of 5.18 V and 4.53 V and 3.72 V respectively. Finally, according to our calculations, large O–O distances in the R^-3 structure during cycling likely inhibit O–O bonding and subsequent O₂ evolution which are otherwise detrimental to the battery application.

Fig. 1. (a) and (b) are the schematic representation of the sodiated ilmenite type Na₁TMO₃ (R^-3 symmetry) and the desodiated Na_0.5TMO₃ (P^-31m symmetry), respectively.

References:

[1] B. M. de Boisse, M. Okubo, A. Yamada et al., Nat. Commun., 7, 11397 (2016).

[2] H. M. N. Assadi, M. Okubo, A. Yamada, Y. Tateyama, J. Mater. Chem. A 6, 3747-3753 (2018).