Stabilized O3‐Type Layered Sodium Oxides with Enhanced Rate Performance and Cycling Stability by Dual‐Site Ti4+/K+ Substitution

Abstract High‐capacity O3‐type layered sodium oxides are considered one of the most promising cathode materials for the next generation of Na‐ion batteries (NIBs). However, these cathodes usually suffer from low high‐rate capacity and poor cycling stability due to structure deformation, native air sensitivity, and interfacial side reactions. Herein, a multi‐site substituted strategy is employed to enhance the stability of O3‐type NaNi0.5Mn0.5O2. Simulations indicate that the Ti substitution decreases the charge density of Ni ions and improves the antioxidative capability of the material. In addition, the synergistic effect of K+ and Ti4+ significantly reduces the formation energy of Na+ vacancy and delivers an ultra‐low lattice strain during the repeated Na+ extraction/insertion. In situ characterizations verify that the complicated phase transformation is mitigated during the charge/discharge process, resulting in greatly improved structure stability. The co‐substituted cathode delivers a high‐rate capacity of 97 mAh g−1 at 5 C and excellent capacity retention of 81% after 400 cycles at 0.5 C. The full cell paired with commercial hard carbon anode also exhibits high capacity and long cycling life. This dual‐ion substitution strategy will provide a universal approach for the new rational design of high‐capacity cathode materials for NIBs.

Air-exposed NaNMO and f) Air-exposed NaK0.01NMTi0.1O.XPS analysis of g) C 1s and h) O 1s.Where NaK0.01NMTi0.1O is displayed at the top and NaNMO is at the bottom.i) FTIR spectra of NaNMO and NaK0.01NMTi0.1Obefore and after the air exposure for 2 days.

Air stability analysis
To investigate the effect of K/Ti co-doping on the air stability of NaNMO, we performed a comprehensive characterization by XRD, SEM, and electrochemical properties to provide information on the structure-interface-property relationship between the fresh and air-exposed samples.As shown in Figure S5a, the XRD pattern of fresh NaNMO showed splitting of the (003) and (104) peaks, which indicates the generation of the monoclinic phase O3'-Na1-xNMO and surface residual bases.After be exposed to air for 2 days, the diffraction peaks of the aged NaNMO become significantly broader and weaker, which demonstrates the reduced crystallinity.
Further, with the appearance of the monoclinic phase, the O3 peak shifts to a lower angle due to the massive spontaneous escape of bulk sodium.The escaped sodium species tend to react with H2O and CO2 in air, and then produce large amounts of residual alkali impurities on the surface. [1]In contrast, the aged NaK0.01NMTi0.1Omaintains the original XRD pattern without any new peaks (Figure S5c), which indicates the successful suppression of the unfavorable bulk Na escape and H2O embedding.The NaNMO provides an initial charge capacity of 115.1 mAh g -1 , which drops dramatically to 92.0 mAh g -1 after 2 days of air exposure (Figure S5b).The large capacity loss of up to 23.1 mAh g -1 implies that spontaneously active lattice sodium escape is significantly present in the air-exposed O3-type material.Note that, suffering from severe spontaneous escape of Na, the initial charging capacity of airexposed NaNMO is much lower than the theoretical capacity, thus delivering an unusually high initial Coulomb efficiency of 124.8%.In contrast, the aged NaK0.01NMTi0.1Ohas a higher initial charging capacity and thus shows an excellent initial coulombic efficiency (102.8%)(Figure S5d).Additionally, the morphological evolution of the air-exposed samples was further investigated via SEM.The aged NaNMO particles were heavily agglomerated and covered with many significant particulate substances on the surface (Figure S5e), which could be attributed to the generation of Na-resistive surface species during exposure to air. [2]Additional evidence of increased air stability was provided by the minimal number of alkaline impurities that were found on the surface of NaK0.01NMTi0.1O(Figure S5f).The above results suggest that the K/Ti co-doping can effectively inhibit the chemical sensitivity of NaNMO to H2O and CO2, which results in more active Na + retained in the structure, less residual alkaline material generated on the surface, and excellent structural and cycling stability (Figure S6).
The specific composition of the surface products of the particles was examined using XPS to further evaluate the impact of K/Ti co-doping on the transition metal valence of NaNMO.The peaks at 289.1, 285.5, and 284.8 eV in the C 1s spectra are attributed to carbonate, amorphous C-O, and C-C, respectively, while the peaks in the O 1s spectra are primarily generated from absorbed OH -, carbonate, and lattice oxygen (Figures S5g, S5h, and S5i). [3]For these fresh samples, the minute amount of alkaline material on the surface may have resulted from inevitable residual Na2CO3 during transferring the material from the furnace to the sealed vial in air (< 1 min) or during natural cooling.It can be observed that the peak intensity of CO3 2-in the NaK0.01NMTi0.1O is much lower than that of the pristine.The characteristic peaks of FTIR around 1433 cm -1 and 880 cm -1 have been attributed to CO3 2-.After 2 days of air exposure, the peak intensities of carbonates of NaNMO and NaK0.01NMTi0.1Oare more robust than those of fresh samples.However, the signal intensity of CO3 2-or HCO3 -species of NaK0.01NMTi0.1Owas significantly lower than that of NaNMO, demonstrating the air stability of NaK0.01NMTi0.1O is better than that of pristine.Furthermore, the intensity variation of the peak at 3400 cm -1 has a similar trend to that of CO3 2-, which was positively correlated with the degree of hydration of the cathode material. [4]Additionally, we detected K + and Ti 4+ , which is in line with our predictions (Figures S7 and S8).
Based on the significant difference in the Fermi energy levels between Ni 2+ /Mn 4+ and Ti 4+ , the introduction of Ti 4+ can effectively increase the valence state of Ni via inhibiting charge localization (the electronic delocalization decreases the number of electrons around Ni). [1,5] As shown in Figure S9, while the peaks of Mn 2p do not shift significantly, the binding energy of Ni 2p shifts to higher values, which verifies that the higher oxidation state of Ni in NaK0.01NMTi0.1O.This high-valence Ni ion facilitates the suppression of spontaneous oxidation reactions, resulting in superior air stability, as clarified in the DFT calculations.

Kinetic analysis
To grasp more information about the kinetic processes, the diffusion behavior of Na + in NaNMO and NaK0.01NMTi0.1Owas analyzed by the GITT.The voltage change during relaxation on the GITT charge-discharge curve indicates the overpotential during the electrochemical reaction. [6]With reference to the GITT curves of both materials, the curve of NaK0.01NMTi0.1O6a] The calculated Na + diffusion coefficients DNa + in the two samples are compared in Figure S19, and it can be observed that the DNa + in NaK0.01NMTi0.1O is significantly larger than the pristine throughout the cycling process, which well demonstrates the fast Na + diffusion and explains the enhanced rate performance of NaK0.01NMTi0.1O.Due to the O3-P3 biphasic reaction, a sharp drop in DNa + occurs between 2.5 and 3 V. [7]          128.8 mAh g -1 (24 mA g -1 , 2.0-4.0V) 81.5 %, 120 mA g -1 , 400 cycles 97.1 mAh g -1 (1200 mA g -1 ) This work High entropy

Figure S3 .
Figure S3.SEM image and corresponding EDS mappings of NaNMO.

Figure S7 .
Figure S7.XPS analysis of a) Ni 2p, b) Mn 2p, and c) K 2p and Ti 2p.Where NaK0.01NMTi0.1O is displayed at the top and NaNMO is at the bottom.

Figure S17 .
Figure S17.Rate performance of hard carbon.

Figure S19 .
Figure S19.a, b) GITT profiles for the charge-discharge process of the initial cycles for NaNMO and NaK0.01NMO, respectively.The Na + diffusion coefficient (DNa + ) values are calculated for NaNMO and NaK0.01NMO at charge c) and discharge d) processes.Individual GITT titration curves of e) NaNMO and f) NaK0.01NMO.Linear fit of τ 1/2 and E during GITT titration of g) NaNMO and h) NaK0.01NMO.

Figure S21 .
Figure S21.Atomic concentrations of the main elements comprising the CEI.

Figure S22 .
Figure S22.Ex-situ XRD patterns of NaNMO during the initial charge-discharge process at 0.1 C in the voltage range of 2.0-4.0V.

Figure S23 .
Figure S23.Ex-situ XRD patterns of NaK0.01NMTi0.1Oduring the initial charge-discharge process at 0.1 C in the voltage range of 2.0-4.0V.

Figure S26 .
Figure S26.DOS comparison of a) O 2p and b) TM 3d.

Table S4 .
Performance comparison of this work with previously reported the O3-NaNi0.5Mn0.5O2prepared by other doping modification methods.