The stability of P2-layered sodium transition metal oxides in ambient atmospheres

Air-stability is one of the most important considerations for the practical application of electrode materials in energy-harvesting/storage devices, ranging from solar cells to rechargeable batteries. The promising P2-layered sodium transition metal oxides (P2-NaxTmO2) often suffer from structural/chemical transformations when contacted with moist air. However, these elaborate transitions and the evaluation rules towards air-stable P2-NaxTmO2 have not yet been clearly elucidated. Herein, taking P2-Na0.67MnO2 and P2-Na0.67Ni0.33Mn0.67O2 as key examples, we unveil the comprehensive structural/chemical degradation mechanisms of P2-NaxTmO2 in different ambient atmospheres by using various microscopic/spectroscopic characterizations and first-principle calculations. The extent of bulk structural/chemical transformation of P2-NaxTmO2 is determined by the amount of extracted Na+, which is mainly compensated by Na+/H+ exchange. By expanding our study to a series of Mn-based oxides, we reveal that the air-stability of P2-NaxTmO2 is highly related to their oxidation features in the first charge process and further propose a practical evaluating rule associated with redox couples for air-stable NaxTmO2 cathodes.

Supplementary Figure 5. NaHCO3 crystals. (a-b) The SEM images of exposed Na0.67MnO2 to RH 93% + CO2 for 3 days. (c) The EDS mapping results, which confirm that the crystals with regular shapes are NaHCO3.
Supplementary Figure 13. The structural and chemical evolution mechanisms upon airexposure. (a) The K-edge XAS of Ni for pristine and exposed Na0.67Ni0.33Mn0.67O2 sample. (b) The K-edge XAS of Mn for pristine, partially hydrated and totally hydrated Na0.67MnO2 samples. (c) The XRD patterns for Na0.67MnO2 samples stored in three different atmospheres for 3 days. (d) The sodium extraction models of double-layer model for calculating the hydration energies.
Supplementary Figure 14. Equivalent circuit. The equivalent circuit used for EIS analysis in this study. The impedance of electrodes can be attributed to Ohmic resistance (Rohm), charge transfer resistance of the anode (RA, CT) and the cathode (RC, CT), the surface resistance of the anode (RA, SEI) and cathode (RC, SEI), and Warburg resistance. Moreover, because the RA, SEI, RA, CT and RC, SEI locate at the high-frequency region, they often coincide and form a semi-circle in the high-frequency region of the Nyquist spectra. 1 Supplementary Figure 16. Temperature-resolved in-situ XRD patterns. Temperatureresolved in-situ XRD patterns for the hydrated Na0.67Zn0.1Mn0.9O2 powder ranging from 25-570 ℃.
Supplementary Figure 20. Surface degradation. The comparison of XRD patterns of pristine and exposed Na0.67Cu0.33Mn0.67O2 samples. The sample was exposed in RH 93% + CO2 for 6 days.
Supplementary Figure 26. The influence of crystallinity on air-stability. (a) The weight loss and (b) the XRD evolutions during the preparation process of Na0.67MnO2. (c) the XRD patterns of the Na0.67MnO2 samples at different calcination stages after the exposure in RH 18% for 3 days.

Supplementary Tables
Supplementary Table 1. Crystal parameters of birnessite phase. Refined crystallographic parameters obtained by Rietveld refinement of birnessite NaxMnO2.
Naf (1)  . The processes marked with a * and b * correspond to the weight losses of the water extraction from the sodium layers and NaHCO3 decomposition processes, respectively, while c * stands for the weight losses of the decomposition of Na2CO3. It should be point out that it is Na + /H + exchange, rather than O2 oxidation dominates the charge-compensation mechanisms of Na + loss on the basis of the XAS results.  It can be clearly observed that the (002) peak of this immersed Na0.67Ni0.33Mn0.67O2 sample shifts to a lower 2-theta degree (Supplementary Figure 7d). We also found that the pH value of the water after soaking Na0.67Ni0.33Mn0.67O2 increases gradually from 7.73 to 11.47 (Supplementary Figure 8a), suggesting the presence of the loss and dissolution of Na + ions when Na0.67Ni0.33Mn0.67O2 is soaked in water. The SEM image of the soaked Na0.67Ni0.33Mn-0.67O2 sample in Supplementary Figure 8b shows that the smooth surface of Na0.67Ni0.33Mn0.67O2 ( Figure 2g) is deformed, indicative of the erosion of water to the Na0.67Ni0.33Mn0.67O2 sample.
In conclusion, the above results suggest that the structural transformation mechanisms of P2-NaxTmO2 in water are similar to that in moisture air, and these changes in water can give new insights into the degradation mechanisms in aqueous batteries.

Protonated phase
As shown in Figure 6, the intensity of XRD patterns in stage b is much stronger than stage c, suggesting the crystallization of the structure at stage b is higher than stage c. We speculate that after the extraction of H2O (70-130 ℃, stage a), the H + ions are extracted from the hydrated sample (stage b), and the good crystallization in stage b is benefited from the proton ions in the dehydrated Na0.67-x-zHxMnO2 sample. As shown in Supplementary Figure 15, no hydration signals (311 ppm) and the obvious dephasing in the 23 Na-1 H REDOR-dephased 23 Na ss-NMR spectra suggest that there is a high possibility of the existence of proton in the Na0.67-x-zHxMnO2 sample in stage b, which confirms our speculation.
The sample in Supplementary Figure 15 was annealed as follow: the hydrated powder sample was heated from room temperature (~ 30 ℃) to 150 ℃ by 3 ℃ per minutes. After annealing the sample at 150 ℃ for 3 minutes, the dehydrated sample was stored at 120 ℃ to keep it from re-hydration.

Supplementary Note 6
Verifying the healing effect of high temperature annealing To verify whether the hydrated materials can be truly fully healed by high temperature.
The totally hydrated Na0.67MnO2 samples were annealed at 500 ℃ and 700 ℃ for 3 h, separately, and the comparison of their XRD patterns with that of pristine Na0.67MnO2 are presented in Supplementary Figure 17. As shown in Supplementary Figure 17a, it can be observed that all three samples show high crystallinity. However, the comparison of (002) peaks in Supplementary Figure 17b shows that 500 ℃-annealed sample exhibit the larger layer spacing than pristine samples, while the location of (002) peaks of pristine and 700 ℃-annealed samples are nearly the same. The above results indicate that the hydrated materials are fully recovered at the high temperature of 700 ℃.

TGA analysis
To ascertain the compositions of various hydrated phases, the TGA profiles of Na0.67Ni0.33Mn0.67O2 samples exposed at different atmospheres have been studied firstly, as shown in Supplementary Figure 18a. As we confirmed in Supplementary Figure 11, the degradation products of Na0.67Ni0.33Mn0.67O2 samples exposed at RH 18% and RH 93% + CO2 atmosphere are Na2CO3 and NaHCO3, respectively. In Supplementary Figure 18a, no obvious mass loss at the temperature range of 25-130 ℃ of Na0.67Ni0.33Mn0.67O2 sample exposed in RH 18% for 15 days can be observed, while the mass loss at the same temperature range of Na0.67Ni0.33Mn0.67O2 sample exposed in RH 93% + CO2 for 15 days is 2.71 %. In addition, the mass loss at 25-130 ℃ is similar to that at 130-900 ℃, coincides well with the ratio of mass losses at the different stages of NaHCO3 decomposition, as will be shown in Supplementary Note 8.
Combing the above results and the fact that there is no H2O intercalated into the Na + layers of Na0.67Ni0.33Mn0.67O2 samples, we can conclude that: 1) the content of absorbed H2O molecules at the particle surface is almost negligible, 2) in the exposed samples, NaHCO3 decomposes at the temperature range of 25-130 ℃ and Na2CO3 decomposes at 130-900 ℃.
Furthermore, according to our in-situ variable-temperature XRD results, the protons are also extracted at 130-900 ℃. Therefore, as shown in Supplementary Figure 19, the mass loss of hydrated phases at 25-130 ℃ and 130-900 ℃ can be identified to the water deintercalation and NaHCO3 decomposition, and deprotonation plus Na2CO3 decomposition, respectively.