Regulating Na⁺ vacancies and stabilizing lattice structure via Al³⁺/O^2- doping for low-cost dry-air-stable NaCl-based solid electrolytes

Material synthesis

NaCl, AlCl₃, and NaOH were purchased from Sinopharm and used without further purification. Na_0.25+xAl_0.25Cl_1-xO_x and Na_1-3yAl_yCl were prepared by the same synthesis method. NaCl and AlCl₃ or NaCl,AlCl₃ and NaOH were milled with a NaCl:AlCl₃ molar ratio of (1-3y):y or with a NaCl:AlCl₃:NaOH molar ratio of 0.25:0.25:x for 30 min. The pre-blended powder mixtures were sintered at 200 °C in a muffle furnace (Kejing, KSL-1400X-A1, China) for 2 h under a high-purity argon (Ar) atmosphere. The sintered products were subsequently subjected to high-energy ball milling in a 100 mL zirconia (ZrO₂) jar under vacuum at 550 rpm for 10 h using a planetary ball mill (Nanda Instruments, QM-3SP2, China). All sample handling and processing operations were conducted in an Ar-filled glovebox (Mikrouna, Super 750, China), where the contents of H₂O and O₂ were strictly controlled below 0.1 ppm to avoid contamination.

The Na₃PS₄ solid electrolyte was fabricated via a conventional solid-state synthesis route. Specifically, sodium sulfide (Na₂S, Sigma-Aldrich) and phosphorus pentasulfide (P₂S₅, 99% purity, Sigma-Aldrich) were weighed according to the stoichiometric ratio and homogeneously mixed by ball milling at 500 rpm for 12 h. The fully mixed powders were vacuum-sealed in a quartz tube and thermally annealed at 280 °C for 3 h to obtain the target electrolyte material.

The Na₂Sn anode material was synthesized through mechanical ball milling of metallic sodium (Na) and tin (Sn) powders (Sinopharm). Initially, the stoichiometric Na and Sn metal mixture was pre-rolled for preliminary blending, then transferred into a stainless-steel milling jar and hermetically sealed under Ar protection. The sealed mixture was ball-milled at 300 rpm for 10 h to yield a uniform Na₂Sn product. An additional 10 h of ball milling was performed if the obtained product exhibited an inhomogeneous morphology and composition.

Characterization

X-ray diffraction (XRD) experiments were performed using a Philips X’Pert powder diffractometer (Rigaku, D/MAX2500VL/PC, Japan) at 45 kV and 40 mA with Cu-K_α radiation (λ = 1.5406 Å). The samples were placed in a zero-background holder and sealed with a Kapton film to avoid air exposure. The data were collected at room temperature with 2θ from 10° to 80°. Rietveld refinement was carried out from XRD data with strong intensity. The following parameters were refined stepwise: (1) scale factor, (2) background using linear interpolation function with 10 coefficients, (3) peak shape using the pseudo-Voigt function, (4) unit cell parameters and fractional atomic coordinates, (5) fractional occupancy and thermal displacement parameters (Uiso). High-resolution synchrotron XRD and total scattering measurements were performed at beamline ID31 of the European Synchrotron Radiation Facility. NIST SRM 660b (NIST, LaB6, America) was used for geometry calibration. Scanning electron microscopy (SEM) images were conducted on a high-resolution field-emission scanning electron microscope (Hitachi, Regulus 8230, Japan). X-ray photoelectron spectroscopy (XPS) measurements were conducted on a Thermo Scientific K-Alpha instrument (Thermo Scientific, ESCALAB Qxi, China).

The ionic conductivity of all solid electrolytes was measured at 30 °C using electrochemical impedance spectroscopy (EIS) in the frequency range from 7 MHz to 1 Hz with a potential perturbation of 50 mV (Bio-Logic, SP-200, France). The activation energy was calculated based on variable-temperature impedance from room temperature up to 70 °C in a microclimate chamber. The as-synthesized solid electrolyte powders were pressed into pellets with a diameter of 12 mm, under a pressure of 370 MPa and then sandwiched between two steel rods for all measurements. The Young's modulus test was carried out on an Atomic Force Microscope (Bruker, Dimension ICON, Germany).

Linear sweep voltammetry (LSV) was employed to evaluate the electrochemical stability of the prepared SSEs. The working cathode was fabricated by blending the SSE powder with 10 wt% vapor-grown carbon fibers (VGCF). The assembled battery configuration adopted NACO_0.2125 as catholyte, the as-prepared Na₃PS₄ as anolyte and Na₂Sn as the anode material. The NACO_0.2125-VGCF cathode (10 mg), NACO_0.2125 (70 mg), Na₃PS₄ (70 mg), and Na₂Sn anode (50 mg) were separately compacted at a pressure of 300 MPa to construct a 12 mm-diameter cylindrical cell. The LSV measurements were performed on a Biologic SP-200 electrochemical workstation at a constant scan rate of 0.1 mV/s, with a voltage testing window ranging from 0 to 5 V.

All-solid-state NaNi_1/3Mn_1/3Ti_1/3O₂||Na_0.4625Al_0.25Cl_0.7875O_0.2125||Na₃PS₄||Na₂Sn battery was fabricated using the following procedure. The composite cathode was made by ball milling the mixture of NaNi_1/3Mn_1/3Ti_1/3O₂, Na_0.4625Al_0.25Cl_0.7875O_0.2125 and VGCF (50:50:5 in weight ratio) at 300 rpm for 30 min. The ASSB was made by co-pressing Na₂Sn anode (50 mg), Na₃PS₄ (70 mg) anolyte, Na_0.4625Al_0.25Cl_0.7875O_0.2125 (70 mg) catholyte and NaNi_1/3Mn_1/3Ti_1/3O₂ composite cathode (12 mg) together in order, and under 300 MPa. Battery tests were carried out on a Neware battery test system (Neware, CT-4008Q-100mA, China).

Synthesis and characterization of Na_0.25+xAl_0.25Cl_1-xO_x

To generate Na⁺ vacancies in the NaCl lattice, partial Na⁺ substitution with high-valence cations was implemented. Considering the economic factor, Al³⁺ was chosen as the dopant, which is the most abundant metal element in nature (Na_1-3yAl_yCl, 0 < y ≤ 0.2). As shown in Supplementary Figure 1, XRD revealed that as the Al content increased, well-resolved diffraction peaks characteristic of the NaAlCl₄ phase appeared and gradually intensified^[24]. This observation demonstrated that Al³⁺ ions could not fit into the NaCl structure. Instead, AlCl₃ reacted with the NaCl matrix and formed NaAlCl₄. This situation can be explained by Pauling's first rule, which states that the cation-to-anion radius ratio (r_c/r_a) is the primary determinant of the cation coordination geometry and polyhedron stability^[40]. For NaCl-type structures belonging to the Fm-3m space group, Na⁺ ions (102 pm) sit in the 4a octahedral sites where a cation is coordinated by six Cl^- anions (181 pm) in a highly symmetrical packing scheme. The mismatch in ionic radius, especially between Al³⁺ (53.5 pm) and Na⁺ (102 pm), hinders effective Na⁺ vacancy creation, leading to low ionic conductivity, which is critical for understanding material limitations. This imbalance induces excessive lattice strain and disrupts the original coordination frameworks. The ionic conductivity of Na_1-3yAl_yCl with impurities was hardly improved, as shown in Supplementary Figure 2.

It is important to note that Al³⁺ can achieve stable sixfold coordination in α-Al₂O₃ primarily due to the influence of O^2- size and electronegativity. The small ionic radius of O^2- (140 pm) and its high electronegativity work together to optimize the cation-to-anion radius ratio for Al³⁺, making this coordination thermodynamically favorable^[41]. The high electronegativity of O^2- exerts a strong polarizing effect, distorting the electron cloud of Al³⁺ and slightly increasing its effective ionic radius, which further stabilizes the sixfold coordination. Based on the theoretical insight, we proposed a dual-anion modification strategy that involves the simultaneous incorporation of Al³⁺ and O^2- into the NaCl lattice. The introduction of O^2- reduces the average ionic radius of the anionic sublattice, which tailors the cation-to-anion radius ratio to the critical range required for producing six-coordinate configurations and stabilizing the doping of Al³⁺ at the 4a site. The dual-anion approach overcame issues inherent with the traditional aliovalent doping method and led to a novel NaCl-type halide SSE, denoted as Na_0.25+xAl_0.25Cl_1-xO_x (NACO_x) (0 ≤ x ≤ 0.225).

Upon the doping of O^2-, NACO_x retains the FCC architecture with the Fm-3m space group, which is isostructural to the parent NaCl lattice [Figure 1A]. XRD characterization [Figure 1B] revealed that as the O^2- doping content increases, the intensity of diffraction peaks corresponding to the NaAlCl₄ impurity phase gradually decreases. This observation confirmed that O^2- doping promotes the incorporation of Al³⁺ into the NaCl lattice. When x ≥ 0.2125, diffraction peaks associated with NaAlCl₄ were eliminated, verifying the formation of a single-phase FCC solid solution.

Figure 1. Structural characterization. (A) Crystal structure of NaCl; (B) XRD patterns of Na_0.25+xAl_0.25Cl_1-xO_x (0 ≤ x ≤ 0.225); (C) Rietveld refinements of Na_0.4625Al_0.25Cl_0.7875O_0.2125; (D) PDF profiles of Na_0.4625Al_0.25Cl_0.7875O_0.2125; (E) ²³Na NMR of Na_0.25+xAl_0.25Cl_1-xO_x (0 ≤ x ≤ 0.225). (F) ²⁷Al NMR of Na_0.4625Al_0.25Cl_0.7875O_0.2125.

XPS was first used to probe the local bonding environment of the phase-pure Na_0.4625Al_0.25Cl_0.7875O_0.2125 (NACO_0.2125) sample [Supplementary Figure 3]. Distinct Na-O and Al-O features were observed, indicating mixed Cl^-/O^2- coordination around Na and Al species^[29,38]. When coupled with complementary XRD data [Figure 1B], these results provided compelling evidence that Al³⁺ and O^2- have been successfully doped into the NaCl lattice. To elucidate the atomic-scale structural features of the optimal single-phase composition, Rietveld refinement was performed on NACO_0.2125 [Figure 1C]. Based on the NaCl framework as a matrix, a partial substitution of Na⁺ and Cl^- by Al³⁺ and O^2- occurred concurrently with the introduction of specific numbers of Na⁺ vacancies. Subsequently, taking into account the large ionic radius difference between Al³⁺ and Cl^-, which drives the formation of [AlCl₄]^-, a part of Al³⁺ was incorporated at the tetrahedral 8c sites to build the final structural model. The lattice constants, atomic coordinates, site occupancies, and isotropic displacement parameters were refined without additional constraints. The fitted curve aligned well with experimental data (Rwp = 4.094%), verifying both the rationality of the established structure and the sample purity, as no NaAlCl₄ impurity phases were observed. The refinement results confirmed a robust FCC framework with a high Na⁺ vacancy concentration of 37.6% at the 4a sites, which constructs a continuous 3D percolation network for ion migration with a calculated Na⁺ hopping distance of 3.98 Å. This unique structural configuration features a high density of interconnected Na⁺ vacancies, which differs markedly from conventional halide SSEs and endows the NACO_x system with enhanced ionic transport kinetics via the vacancy diffusion mechanism^[42,43]. The refined lattice parameter is a = b = c = 5.63 Å, slightly smaller than that of pristine NaCl (5.64 Å), consistent with partial replacement of Na⁺ (102 pm) by the smaller Al³⁺ cation (53.5 pm). A small fraction of Al also occupies the tetrahedral 8c sites, with a refined occupancy of 0.044 [Supplementary Table 1], indicating that Al³⁺ is distributed over both octahedral and tetrahedral environments.

To further resolve the local structure of NACO_0.2125, X-ray pair distribution function (PDF) analysis was performed [Figure 1D and Supplementary Figure 4]. The PDF spectrum of NACO_0.2125 in the high-r region (r > 10 Å) closely matches that of pristine NaCl, indicating that the lattice structure remains intact, which supports confidence in the material's reliability despite Al³⁺/O^2- co-doping^[38]. In contrast, in the low-r region (r < 5 Å), the sensitivity to the short-range atomic coordination environment led to different characteristics and reflects the success of heteroatomic incorporation. Specifically, two prominent peaks centered at 2.83 and 2.40 Å were resolved in the PDF profile of NACO_0.2125, which are unambiguously assigned to the Na-Cl and Na-O bond pairs, respectively. The emergence of the Na-O signature peak provided direct structural evidence for the successful doping of O^2- anions into the halide lattice. Moreover, additional peaks at 2.16 and 1.73 Å correspond to Al-Cl and Al-O bonds in tetrahedral coordination environments, as confirmed by the Rietveld refinement results and corroborating the occupation of the tetrahedral 8c interstitial sites by a fraction of Al³⁺ ions. Further structural analysis identified peaks at 2.67 and 1.90 Å, which are attributed to Al-Cl and Al-O bond pairs associated with Al³⁺ ions residing in the octahedral 4a sites of the NaCl-type lattice. Notably, these bond lengths were slightly longer than the theoretical values for undoped analogs. This phenomenon can be rationalized by the strong polarization induced by the high charge density of Al³⁺ cations. This polarization distorts the electron cloud of adjacent anions, leading to a subtle expansion of the Al-anion coordination sphere^[44,45]. Moreover, the atomic probability distribution derived from the PDF fitting clearly showed that the occupancy probability of Al³⁺ at the octahedral 4a sites is higher than that at the tetrahedral 8c sites. This result provides unambiguous quantitative evidence that over half of the doped Al³⁺ ions are successfully incorporated into the octahedral 4a sites of the NaCl lattice, thereby fully demonstrating the rationality of the defect-engineering design process.

Solid-state nuclear magnetic resonance (ssNMR) spectroscopy provided insights into the chemical coordination environments of Na⁺ and Al³⁺ species^[46], thereby delineating the structural evolution induced by the co-doping of Al³⁺ and O^2-. As illustrated in Figure 1E, the ²³Na NMR spectrum of the oxygen-free counterpart (x = 0) was dominated by a broad resonance spanning from -13 to -17 ppm, a characteristic chemical shift of Na⁺ ions residing in the NaAlCl₄ phase. By contrast, only a faint signal at 7 ppm could be discerned, corresponding to Na⁺ in the pristine NaCl lattice. This spectral feature confirmed the thermodynamic metastability of Al³⁺-doped pristine NaCl, aligning with the XRD results [Figure 1B]. Upon the introduction of O^2- anions, the resonance intensity associated with the NaCl lattice was pronouncedly enhanced, demonstrating the positive impact of oxygen doping on lattice stability. Concomitantly, the signal attributable to the NaAlCl₄ phase gradually diminished until it was completely suppressed at the optimal doping stoichiometry of x = 0.2125. This spectral evolution provided unambiguous evidence for the formation of a single-phase NaCl-type solid solution, in excellent agreement with the phase identification results from XRD [Figure 1B]. Moreover, a distinct new resonance emerged at -11 ppm in the O^2--doped samples, indicative of the formation of a novel Na⁺ coordination environment within the SSE. This signal was attributed to the local environment of Na⁺ adjacent to Al³⁺. With the gradual increase in O^2- content, the higher electronegativity of O^2- relative to Cl^- exerted a stronger electrostatic effect, which induced a progressive downfield shift of the chemical shift^[47,48].

The ²⁷Al NMR spectra further resolved the local coordination environments of Al³⁺ ions [Figure 1F]. Specifically, the sharp resonance centered at 100 ppm is characteristic of tetrahedrally coordinated Al³⁺ in the isolated [AlCl₄]^- anion, while the peak at 92 ppm was assigned to the mixed-anion tetrahedral species [AlCl_4-xO_x]^- formed via partial Cl^-/O^2- substitution. The peak at 78 ppm corresponded to octahedrally coordinated Al³⁺ in the mixed-anion polyhedron [AlCl_6-xO_x]^-, a configuration that was commensurate with the 4a crystallographic sites of the NaCl-type lattice. In addition, three minor peaks at 65, 36, and 4.6 ppm were unambiguously identified as the signatures of [AlO₄]^-, [AlO₅]^-, and [AlO₆]^-, respectively^[49-52]. Quantitative deconvolution analysis of the ²⁷Al NMR spectra revealed that octahedral Al³⁺ species account for 65% of the total Al³⁺ species present in the system. This observation confirmed that the majority of Al³⁺ ions occupy the octahedral 4a sites of the NaCl-type framework, consistent with the design intent of our Al³⁺/O^2- co-doping strategy. Meanwhile, the presence of tetrahedral Al-containing species (e.g., [AlO₄]^-, [AlCl_4-xO_x]^-) indicated that a minor fraction of Al³⁺ ions resides in the tetrahedral 8c interstitial sites of the FCC lattice. This preferential occupation of tetrahedral sites by a portion of Al³⁺ could be attributed to its small ionic radius (53.5 pm), which was better suited to the confined space of the 8c interstitial sites. Collectively, these ²⁷Al NMR results demonstrated that the NACO_0.2125 system does not adopt a simple, stoichiometric NaCl structure, but rather forms a complex substitutional solid solution with Al³⁺ ions distributed across both octahedral 4a and tetrahedral 8c sites. This refined atomic-scale structural model is in excellent agreement with the phase purity and lattice parameter analysis derived from Rietveld refinement and PDF results [Figure 1C and D], thereby providing a consistent structural basis for interpreting the ionic transport behavior of the NACO_0.2125 SSE.

To resolve the Na⁺ transport mechanism within NaCl-type crystalline framework, we constructed Li||NACO_0.2125||Li symmetric cells and subjected them to a constant polarization bias [Figure 2A]^[53]. This experimental design capitalizes on the Li⁺/Na⁺ cation exchange reaction driven by Li⁺ diffusion across the electrolyte. Specifically, the displacement of Na⁺ from its original lattice sites under an external bias directly reveals the spatial distribution of active Na⁺ migration channels. Incoming Li⁺ can replace only Na⁺ ions residing in transport-accessible sites. Notably, ²³Na NMR spectroscopy characterization of the polarized electrolyte [Figure 2B] revealed a pronounced decrease in the resonance intensities corresponding to the NaCl host lattice (7 ppm) and the Na_(Al) local environment (-11 ppm), with the reduction ratio of the Na_(Al) peak (59%) being markedly higher than that of the NaCl peak (18.1%). Such spectral evolution furnished unambiguous evidence for the construction of continuous Na⁺ conduction channels within the electrolyte: Na⁺ ions form a three-dimensional continuous ion transport network across the bulk lattice and exhibit a distinct tendency to migrate via the Na⁺ vacancies adjacent to Al³⁺ sites. This result directly verified the efficacy of the defect engineering strategy proposed in this work.

Figure 2. Transport mechanism analysis. (A) Li||NACO_0.2125||Li cell; (B) ²³Na MAS NMR of polarized and pristine Na_0.4625Al_0.25Cl_0.7875O_0.2125.

Electrochemical performance and cost of Na_0.25+xAl_0.25Cl_1-xO_x

The optimized structural features of NACO_0.2125 (high Na⁺ vacancy concentration and continuous 3D migration network) directly determine its electrochemical performance. Herein, we systematically evaluated the ionic conductivity, electrochemical stability window, and electronic conductivity of the NACO_x series. Both ionic conductivity and phase purity showed a positive correlation with oxygen doping content over a specific range. This trend arose primarily from the progressive suppression of the NaAlCl₄ impurity phase.

At the optimal doping stoichiometry of x = 0.2125, the composition achieved a maximum ionic conductivity of 2.7 × 10^-4 S cm^-1 at 30 °C [Figure 3A and B, Supplementary Figure 5]. This performance peak coincided with the high Na⁺ vacancy concentration of 37.6% at the 4a sites, confirming that the vacancy-mediated diffusion mechanism affects ion transport in this system. In contrast, excessive oxygen doping (x = 0.225) triggered a decline in ionic conductivity to 1.7 × 10^-4 S cm^-1 at 30 °C [Supplementary Figure 5]. This deterioration could be ascribed to structural factors: the Na⁺ vacancy concentration at the 4a sites decreased to 36.3%, thereby reducing the density of mobile charge carriers and raising the migration energy barrier^[54,55]. On the other hand, increasing O^2- doping effectively reduces the content of NaAlCl₄ impurities, thereby improving the ionic conductivity of the NACO_x samples [Supplementary Figure 5]. The optimal electrochemical performance is achieved at x = 0.2125, at which point impurity phases are completely eliminated. Further increasing the O^2- content causes a right shift of the XRD diffraction peaks for NACO_0.225 [Supplementary Figure 6]. Since O^2- possesses a smaller ionic radius than Cl^-, continuous oxygen substitution reduces the unit cell volume. This structural contraction distorts the Na⁺ fast-migration pathways and ultimately deteriorates the ionic conductivity.

Figure 3. Electrochemical performance of NACO_x. (A) Arrhenius conductivity plots of Na_0.25+xAl_0.25Cl_1-xO_x (0.15 ≤ x ≤ 0.225); (B) Ionic conductivities and activation energies of Na_0.25+xAl_0.25Cl_1-xO_x (0.15 ≤ x ≤ 0.225); (C) Linear scanning voltammetry of NACO_0.2125 at 0.1 mV s^-1; (D) DC polarization curve of NACO_0.2125 under 0.2 V.

Electrochemical stability represents a pivotal metric for SSEs. To evaluate the electrochemical stability of NACO_0.2125, LSV measurements were performed on a tailored cell configuration: NACO_0.2125-VGCF||NACO_0.2125||Na₃PS₄||Na₂Sn, where Na₃PS₄ was introduced as an intermediate layer to mitigate direct interfacial reactions between NACO_0.2125 and the Na₂Sn alloy anode. As illustrated in Figure 3C, the LSV profile reveals that NACO_0.2125 exhibits an electrochemical stability window spanning from 1.05-4.24 V (vs. Na₂Sn) at 30 °C. This wide oxidation stability up to 4.24 V surpasses that of most reported sodium-ion halide SSEs (typically 4.0-4.1 V), demonstrating its excellent stability with high-voltage cathode materials. Conversely, NACO_0.2125 shows limited reduction stability, with a noticeable increase in reductive current initiating at 1.05 V. This reductive decomposition behavior could be rationalized by the high electronegativity of Al³⁺ ions within the electrolyte lattice. To further confirm whether NACO_0.2125 inherits the reduction stability inherent to NaCl, symmetric Na₂Sn||NACO_0.2125||Na₂Sn cells were assembled for subsequent performance testing [Supplementary Figure 7]. When cycled at a current density of 0.01 mA cm^-2, the cell exhibits significant voltage polarization, with the overpotential rising to 5 V or higher within 10 h of operation. This pronounced polarization stems from the continuous formation of resistive interfacial layers induced by the reductive decomposition of NACO_0.2125 at the anode-electrolyte interface. In addition to ionic conductivity, NACO_0.2125 exhibited an electronic conductivity as low as 3.8 × 10^-10 S cm^-1 [Figure 3D]. This ultra-low electronic conductivity effectively suppresses the risk of internal short circuits in ASSBs.

Another remarkable advantage of NACO_0.2125 lies in its extremely low cost. It can be synthesized via a simple preparation process using low-cost raw materials, namely NaCl, AlCl₃, and NaOH. Among these starting materials, AlCl₃ stands out for its exceptional economic viability, with a market price of merely 35.6 USD/kg [Supplementary Figure 8]^[56]. In addition, the market prices of NaOH and NaCl are as low as 6.41 and 0.10 USD/kg, respectively, resulting in an overall material cost for NACO_0.2125 of only 25.5 USD/kg^[56]. When benchmarked against other Na-based SSEs reported in the literature, this value represents a remarkable cost superiority. Compared to the expensive Na₃PS₄ sulfide solid electrolytes and other halide solid electrolytes that use rare earth elements^[56], NACO_0.2125 is extremely cheap, even cheaper than the already low-cost NASICON oxide electrolytes [Supplementary Figure 9].

Mechanical properties and dry air stability

Mechanical factors play a crucial role in the industrial fabrication and practical implementation of SSEs. These factors directly affect interfacial contact with electrodes, and long-term cycling durability in ASSBs. For NACO_0.2125, primary particle size analysis reveals a size distribution ranging from approximately 7 to 15 μm. Owing to the intrinsic cohesive nature of halide materials, these fine primary particles tended to aggregate spontaneously into secondary agglomerates with an average size of ~40 μm, as evidenced by morphological characterization [Supplementary Figure 10]. To quantitatively assess the mechanical performance of NACO_0.2125, atomic force microscopy (AFM) based nanoindentation measurements were conducted, with the representative modulus mapping and statistical analysis presented in Figure 4A and B, respectively. The morphological and mechanical data collectively demonstrate that NACO_0.2125 possesses an ultralow Young’s modulus of ~ 2 GPa. This value is one order of magnitude lower than that of conventional inorganic SSEs (e.g., Na₃PS₄ with a Young’s modulus of 20 GPa)^[57,58] and is even comparable to the mechanical compliance of polymer-based SSEs (≤ 2 GPa)^[59]. Such exceptional mechanical softness can be attributed to the formation of Al-O covalent bonds within the NACO_0.2125 lattice. Beyond Al-O mixed covalent-ionic bonds, multiple interrelated structural features synergistically reduce the Young’s modulus of NACO_0.2125. First, the high density of Na⁺ vacancies (37.6%) at octahedral 4a sites breaks the intact ionic bonding network of pristine NaCl, decreasing the total number of interatomic interactions per unit cell and weakening lattice stiffness. Second, mixed Cl^-/O^2- anion substitution introduces widespread local structural disorder and mild lattice distortion, disrupting long-range ordered ionic stacking and further lowering deformation resistance. Together with flexible Al-O bonds with low shear barriers, these three structural factors cooperatively endow NACO_0.2125 with an ultralow Young’s modulus of ~2 GPa, enabling superior densification under compression. Under identical uniaxial pressing pressure, the cross-sectional morphology of NACO_0.2125 pellets exhibited a highly dense and flat microstructure with negligible visible grain boundaries [Supplementary Figure 11A], whereas Na₃PS₄ pellets typically displayed distinct intergranular gaps and rough surfaces [Supplementary Figure 11B]. This enhanced densification capability facilitates intimate interfacial contact between the electrolyte and electrode materials, thereby minimizing interfacial resistance in assembled batteries. In addition, EIS measurements were performed on NACO_0.2125 pellets consolidated at various pressing pressures. As presented in Supplementary Figure 12, the NACO_0.2125 sample pressed at a low pressure of 200 MPa exhibits a moderate total resistance of 381 Ω, merely larger than that of the counterpart fabricated at 400 MPa (330 Ω). Benefiting from its low Young’s modulus (~2 GPa), this electrolyte undergoes homogeneous plastic deformation upon compression and establishes intimate contact with electrode grains, effectively eliminating interfacial voids. These pressure-dependent EIS results solidly verify that the intrinsic mechanical compliance of NACO enables robust interfacial electrochemical properties even under low-pressure cell assembly conditions. Combined with its facile, cost-efficient synthesis protocol, the superior mechanical compliance of NACO_0.2125 renders it a far more industrially viable candidate than conventional halide SSEs.

Figure 4. Chemical stability and mechanical properties. (A) AFM topography image of NACO_0.2125 pellet; (B) Young’s modulus distribution of NACO_0.2125 pellet; (C) Nyquist plots of exposed NACO_0.2125; (D) XRD patterns of NACO_0.2125 and exposed NACO_0.2125 in dry air for 24 h.

Dry air stability is another important factor affecting the practical generation and application of SSEs. It is critical to clarify that the air stability characterization herein is performed under standardized industrial dry-room conditions with a dew point of -40 °C, rather than uncontrolled humid ambient air at room temperature. Mass production of all-solid-state sodium batteries uniformly adopts -40~-60 °C dew dry rooms for electrolyte storage, powder transfer and pellet assembly, making this test condition practically relevant for industrial processing. After 24 h of exposure to an environment with a dew point of -40 °C, NACO_0.2125 powder exhibited a slight impedance rise from 280 to 300 Ω, corresponding to a mere 6.7% reduction in ionic conductivity [Figure 4C and Supplementary Figure 13]. Complementary XRD measurements further validated the structural robustness of NACO_0.2125 after dry air storage [Figure 4D]. In addition, XPS characterization results confirmed the absence of new characteristic peaks in the samples following exposure to the test environment [Supplementary Figure 14], providing further evidence of the excellent dry air stability of NACO_0.2125.

ASSBs using NACO_0.2125

Benefiting from the high ionic conductivity, wide stability window, and excellent mechanical compatibility, NACO_0.2125 is expected to serve as a practical SSE for ASSBs. Full cells with a high-voltage cathode NaNi_1/3Mn_1/3Ti_1/3O₂ and a Na₂Sn anode were assembled and evaluated. The composite cathode formulation consists of NACO_0.2125, NaNi_1/3Mn_1/3Ti_1/3O₂, and VGCF. After thorough mechanical mixing, the active material particles were uniformly dispersed together with NACO_0.2125 and VGCF in the composite cathode matrix [Supplementary Figures 11C and 15]. For the ASSB assembly, Na₃PS₄ (10^-4 S cm^-1, Supplementary Figure 16) was introduced as an interlayer between the Na₂Sn anode [Supplementary Figure 11D] and the NACO_0.2125 electrolyte. The schematic diagram of the ASSB configuration is presented in Figure 5A. Cross-sectional morphological characterization of the assembled ASSB [Figure 5B and Supplementary Figure 17] reveals distinct layered structures with well-defined thicknesses: ~35 μm for the NaNi_1/3Mn_1/3Ti_1/3O₂ composite cathode, ~150 μm for the NACO_0.2125 electrolyte layer, ~175 μm for the Na₃PS₄ interlayer, and 45 μm for the Na₂Sn anode. Importantly, the composite cathode forms a tight interfacial bond with the NACO_0.2125 electrolyte layer, with no visible gaps at the interface. This superior interfacial contact is attributed to the excellent machinability of NACO_0.2125.

Figure 5. All-solid-state battery. (A) Configuration of ASSB with NACO_0.2125 as SSEs; (B) Cross-section SEM for different layers in the ASSB; (C) Charge-discharge profiles at different cycles (2.4~4.2 V); (D) Charge-discharge profiles at different cycles (2.4 V~4.4 V). (E) Cycle performance.

NaNi_1/3Mn_1/3Ti_1/3O₂||NACO_0.2125||Na₃PS₄||Na₂Sn cell delivers stable cycling performance within a voltage window of 2.4~4.2 V, with an initial discharge capacity of 92.9 mAh·g^-1 at 30 °C [Figure 5C]. Even after 100 cycles at 0.3 C, the battery retains a discharge capacity of 82.8 mAh g^-1, corresponding to a capacity retention rate of 89%. EIS measurements reveal that the initial impedance of the fresh full cell is 503 Ω, whereas the impedance surges to 1,100 Ω after 100 cycles [Supplementary Figure 18]. This impedance escalation induces electrochemical polarization, which is primarily responsible for the gradual capacity decay observed during long-term cycling. To elucidate the origin of this impedance growth, a compatibility study was conducted on the NACO_0.2125/Na₃PS₄ bilayer system: 100 mg pellets of NACO_0.2125 and Na₃PS₄ were fabricated separately under a pressure of 400 MPa and then stacked and aged for 72 h in an argon atmosphere. Post-aging impedance characterization shows an increase in the total interfacial impedance of the bilayer, providing direct evidence of chemical incompatibility between NACO_0.2125 and Na₃PS₄ [Supplementary Figure 19]. In addition, XRD analysis of the aged Na₃PS₄ showed the presence of impurity phases, confirming that side reactions occurred at the interface [Supplementary Figure 20]. This interfacial incompatibility is identified as one of the key factors contributing to the performance degradation of the full cell. Owing to the excellent oxidative stability of NACO_0.2125 (4.24 V vs. Na₂Sn), the ASSB based on this material enables stable cycling for 100 cycles at 0.3 C within the voltage window of 2.4~4.4 V, delivering an initial discharge capacity of 105.9 mAh g^-1 and 83% capacity retention [Figure 5D and E]. In addition to cycling stability, the rate capability of the ASSB was evaluated to assess the dynamic ion transport behavior of NACO_0.2125 under different current conditions [Supplementary Figure 21]. At progressively increased current rates of 0.1, 0.2, 0.3, 0.5, and 1 C, the battery delivers discharge capacities of 116.2, 102.4, 88.0, 75.6, and 54.5 mAh g^-1, respectively. On the other hand, in this work, NACO_0.2125 was exposed to dry air for 24 h and then employed as an SSE in ASSBs to investigate the influence of air exposure. As presented in Supplementary Figure 22, the NaNi_1/3Mn_1/3Ti_1/3O₂||NACO_0.2125 (after exposure)||Na₃PS₄||Na₂Sn cell retained 86% of its initial capacity after 100 cycles at 0.3C. These results confirm that air-exposed NACO_0.2125 still exhibits satisfactory electrochemical performance in full cells, verifying its outstanding dry air stability. The battery performance showed the structural stability and high voltage compatibility of the NACO_0.2125-based ASSB system.