The pursuit of advanced energy storage systems has intensified the focus on sodium-ion batteries as a compelling complement to the ubiquitous lithium-ion technology. Sharing similar chemical properties and intercalation mechanisms with lithium, sodium offers significant advantages in terms of natural abundance and cost. However, state-of-the-art sodium-ion batteries predominantly employ flammable organic liquid electrolytes, which pose inherent safety risks and limit the achievable energy density. The transition to all-solid-state batteries, utilizing non-flammable, mechanically robust solid electrolytes, is widely regarded as a pivotal step toward realizing safer and higher-energy-density sodium-based energy storage. The success of this paradigm hinges critically on the development of solid electrolytes that combine high ionic conductivity, excellent electrochemical stability, and facile processability.
Among the various families of sodium-ion conductors, NASICON (Sodium Super Ionic Conductor)-type materials, particularly Na3Zr2Si2PO12 (NZSP), stand out due to their appealing combination of properties. These include a wide electrochemical window (enabling the use of high-voltage cathodes), high mechanical strength, exceptional stability in ambient air, and a three-dimensional framework that facilitates fast Na+ migration. The ionic conductivity of NZSP is highly dependent on its structural phase and microstructural characteristics. While the monoclinic phase (space group C2/c) exhibits superior ionic conductivity, the practical performance of a ceramic electrolyte pellet is predominantly governed by its relative density. High grain boundary resistance, often stemming from residual porosity and poor inter-granular contact, can severely hamper the total ionic conductivity. Consequently, achieving high densification during sintering is paramount.
Traditional ceramic processing for NASICON electrolytes involves mixing pre-synthesized powder with a polymeric binder (e.g., polyvinyl alcohol, PVA) via ball milling. This method often results in inhomogeneous binder distribution and irregular, polydisperse particle morphology in the green body. These factors lead to inefficient particle packing, creating numerous pores that are difficult to eliminate during subsequent sintering, ultimately yielding ceramics with suboptimal density and conductivity. Therefore, innovative powder processing techniques are required to overcome these limitations and unlock the full potential of NZSP electrolytes for solid-state sodium-ion battery applications.
In this work, I introduce and systematically evaluate the spray-drying method as a superior powder processing route for fabricating high-performance Na3Zr2Si2PO12 ceramic electrolytes. This technique enables the simultaneous and uniform coating of ceramic particles with binder and the formation of spherical secondary granules with a narrow, near-Gaussian size distribution. I hypothesize that this optimized granule morphology will lead to superior particle packing in the green compact, which in turn will facilitate the fabrication of highly dense ceramics after sintering. The primary objective is to demonstrate that this process-driven enhancement in microstructure directly translates to improved ionic transport properties, providing a viable and effective strategy for advancing ceramic electrolytes for solid-state sodium-ion batteries without relying solely on compositional doping.
Experimental Methodology
Powder Synthesis
The Na3Zr2Si2PO12 powder was synthesized via a conventional solid-state reaction. Stoichiometric amounts of Na2CO3 (99.99%), ZrO2 (99.99%), SiO2 (99.99%), and NH4H2PO4 (99%) were used as precursors. To compensate for the volatilization of sodium and phosphorus during high-temperature sintering, an excess of 8 wt% Na2CO3 and 15 wt% NH4H2PO4 was added. The raw materials were planetary ball-milled in anhydrous ethanol for 12 hours using zirconia balls. The obtained slurry was dried at 80°C, ground, and sieved through a 100 μm mesh. The resulting precursor powder was first calcined at 400°C for 2 hours to decompose carbonates and ammonium species, followed by a high-temperature treatment. To determine the optimal synthesis temperature, separate batches were sintered between 1000°C and 1150°C for 12 hours in air.
Spray-Drying Granulation Process
The synthesized NZSP powder was subjected to a spray-drying process to create optimized granules. A slurry was prepared by dispersing the NZSP powder in anhydrous ethanol with 2 wt% polyvinyl alcohol (PVA, binder) and 2 wt% polyethylene glycol (PEG, dispersant). The solid content was maintained at 15 wt%. This slurry was ball-milled again for 12 hours to ensure homogeneity. The well-dispersed slurry was then fed into a spray dryer (Yamato ADL311S) with an inlet temperature of 130°C and a feed rate of 5 mL/min. The atomized droplets underwent rapid drying, resulting in spherical, free-flowing granules which were collected via a cyclone separator. For comparison, a control batch of powder was prepared using the conventional method: simply mixing the NZSP powder with 2 wt% PVA in ethanol via ball milling, followed by drying and grinding.
Pellet Fabrication and Sintering
Both the spray-dried granules (SD-GB) and the conventionally mixed powder (GB) were uniaxially pressed into green pellets (12 mm diameter) under 200 MPa. To minimize element loss during sintering, the pellets were embedded in sacrificial powder of the same composition. All pellets were sintered in a muffle furnace at 1250°C for 6 hours with a heating rate of 4°C/min, followed by furnace cooling. The resulting ceramic pellets from the conventional and spray-drying routes are labeled CS-NZSP and SD-CS-NZSP, respectively. The sintered pellets were polished with sandpaper to achieve smooth and parallel surfaces for electrochemical characterization.
Material Characterization
The phase purity and crystal structure of the powders and ceramics were analyzed by X-ray diffraction (XRD, Bruker D8 Advance) using Cu Kα radiation. The microstructures of the granules, green bodies, and fractured surfaces of the sintered pellets were examined using scanning electron microscopy (SEM, Hitachi S-4800). Transmission electron microscopy (TEM, JEOL JEM-2100F) was used to observe the binder coating on individual spray-dried particles. Elemental distribution was analyzed by energy-dispersive X-ray spectroscopy (EDX) coupled with SEM. Particle size distribution was measured using a laser diffraction particle size analyzer.
Electrochemical Characterization
The ionic conductivity of the ceramic pellets was determined by electrochemical impedance spectroscopy (EIS) using an electrochemical workstation (Ametek PARSTAT 3000A-DX). Gold blocking electrodes were deposited on both sides of the pellets via thermal evaporation. EIS measurements were performed in the frequency range of 7 MHz to 0.1 Hz with an AC amplitude of 10 mV. The bulk ($R_b$) and grain boundary ($R_{gb}$) resistances were extracted by fitting the impedance spectra with an equivalent circuit. The total ionic conductivity ($\sigma_t$) was calculated using the formula:
$$\sigma_t = \frac{L}{R_t \cdot A}$$
where $L$ is the pellet thickness, $A$ is the electrode area, and $R_t = R_b + R_{gb}$ is the total resistance. The activation energy ($E_a$) for ion conduction was derived from the Arrhenius equation by measuring conductivity over a temperature range from 25°C to 100°C:
$$\sigma T = A \exp\left(\frac{-E_a}{k_B T}\right)$$
where $A$ is the pre-exponential factor, $k_B$ is the Boltzmann constant, and $T$ is the absolute temperature.
To evaluate the electronic conductivity and the sodium ion transference number, direct current (DC) polarization tests were conducted. A constant DC voltage of 5 V was applied across the Au-blocking-electrode|pellet|Au-blocking-electrode cell, and the current was monitored for 5000 s. The steady-state current ($I_{ss}$) was used to calculate the electronic conductivity ($\sigma_e$):
$$\sigma_e = \frac{L \cdot I_{ss}}{V \cdot A}$$
The ionic transference number ($t_{+}$) was then calculated as:
$$t_{+} = \frac{\sigma_t – \sigma_e}{\sigma_t}$$
The electrochemical stability window was assessed by cyclic voltammetry (CV) at a scan rate of 1 mV s-1 within a voltage range of 0 to 6 V vs. Na/Na+.
Results and Discussion
Phase Formation and Powder Morphology
XRD analysis of powders synthesized at different temperatures confirmed that 1150°C is the optimal temperature for obtaining phase-pure, monoclinic Na3Zr2Si2PO12. Powders sintered at lower temperatures (1000-1100°C) showed impurities such as ZrO2 and Na2ZrSi2O7, along with broader diffraction peaks indicating poorer crystallinity. The powder synthesized at 1150°C exhibited sharp, well-defined peaks matching the standard pattern for monoclinic NZSP (PDF#84-1200), confirming the formation of the desired high-conductivity phase.
The impact of the granulation process on particle morphology was striking. SEM images revealed that the conventionally mixed powder (GB) consisted of irregularly shaped particles with a broad size distribution, including large agglomerates exceeding 20 μm. In stark contrast, the spray-dried powder (SD-GB) was composed of spherical secondary granules with a uniform size predominantly below 5 μm. Laser particle size analysis quantitatively confirmed this observation, showing a much narrower, near-Gaussian size distribution for the spray-dried granules compared to the polydisperse conventional powder. TEM analysis further revealed that the spray-drying process resulted in a thin, uniform coating of PVA binder (approximately 5 nm thick) on the surface of the primary NZSP particles within each granule.
This optimized granule morphology has profound implications for green body formation. The spherical shape and narrow size distribution enable much more efficient particle packing according to principles of granular mechanics. This directly translated to a higher green density, as measured by the Archimedes method. The relative density of the green body increased from 83.0% for the conventionally processed sample (GB) to 89.1% for the spray-dried sample (SD-GB), providing a superior starting point for the final sintering stage.
Microstructure and Densification of Sintered Ceramics
The superior packing of the spray-dried granules manifested dramatically in the microstructure of the sintered ceramics. SEM images of the fracture surface of the CS-NZSP pellet (conventional route) revealed a porous structure with irregularly shaped voids, some exceeding 5 μm in size. These pores act as strong barriers to ion transport, significantly increasing grain boundary resistance. Conversely, the SD-CS-NZSP pellet exhibited a remarkably dense and homogeneous microstructure with well-connected grains and negligible visible porosity. The elimination of large pores indicates that the uniform initial packing facilitated a more complete and homogeneous sintering process.
This microstructural improvement is quantitatively reflected in the final relative density of the sintered pellets. As summarized in Table 1, the SD-CS-NZSP pellet achieved a relative density of 97.5%, significantly higher than the 88.1% attained by the CS-NZSP pellet. This value of 97.5% represents one of the highest densities reported for pure, undoped NZSP ceramics prepared by conventional pressureless sintering, surpassing densities achieved by other advanced sintering techniques like microwave or cold sintering for similar compositions. The enhanced densification is also visually apparent in the higher linear shrinkage observed for the SD-CS-NZSP pellet (~13.7%) compared to the CS-NZSP pellet (~5.5%). EDX elemental mapping on the fracture surface of the dense SD-CS-NZSP ceramic confirmed a homogeneous distribution of Na, Zr, Si, P, and O, with no evidence of element segregation or secondary phase formation at the grain boundaries. The measured atomic ratio of Na to P was approximately 2.98:1, close to the theoretical 3:1, confirming that the initial excess of Na and P precursors effectively compensated for volatilization losses during high-temperature sintering.
| Sintering Method | Composition | Sintering Temp. (°C) | Relative Density (%) | $\sigma_t$ at 25°C (S cm-1) | $E_a$ (eV) | Reference |
|---|---|---|---|---|---|---|
| Cold Sintering | Na3.256Mg0.128Zr1.872Si2PO12 | 140 | 82.9 | 0.41 × 10-4 | – | Ref. [19] |
| Microwave Sintering | Na3Zr2Si2PO12 | 850 | 96.0 | 2.5 × 10-4 | 0.31 | Ref. [23] |
| Conventional Sintering | Na3Zr2Si2PO12 | 1250 | 71.4 – 88.1 | 1.7×10-4 – 6.7×10-4 | 0.34 – 0.41 | Refs. [20,21] & This work |
| Conventional Sintering (O2) | Na3.4Zr1.9Zn0.1Si2.2P0.8O12 | 1250 | 99.5 | 5.27 × 10-3 | 0.285 | Ref. [18] |
| Spray-Drying + Conventional Sintering | Na3Zr2Si2PO12 | 1250 | 97.5 | 6.96 × 10-4 | 0.32 | This work |
Ionic Transport Properties
The EIS spectra of both ceramic pellets at room temperature consisted of a depressed semicircle at high-to-medium frequencies (attributed to the parallel combination of grain boundary resistance and capacitance) and an inclined spike at low frequencies (attributed to the electrode/electrolyte interface). The intercept of the semicircle with the real axis at high frequency corresponds to the bulk resistance ($R_b$), while the diameter of the semicircle corresponds to the grain boundary resistance ($R_{gb}$). The Nyquist plots clearly showed that both $R_b$ and $R_{gb}$ were significantly smaller for the dense SD-CS-NZSP pellet compared to the porous CS-NZSP pellet.
The calculated room-temperature total ionic conductivity ($\sigma_t$) for the SD-CS-NZSP electrolyte was $$6.96 \times 10^{-4} \text{ S cm}^{-1}$$, which is approximately 41% higher than the value of $$4.94 \times 10^{-4} \text{ S cm}^{-1}$$ obtained for the CS-NZSP electrolyte. This enhancement is directly attributable to the drastic reduction in grain boundary resistance due to minimized porosity and improved inter-granular connectivity. The bulk conductivity also showed improvement, likely due to better crystallinity and fewer intra-grain defects in the densely sintered material. The temperature-dependent conductivity followed Arrhenius behavior. The activation energy ($E_a$) for ion migration decreased from 0.34 eV for CS-NZSP to 0.32 eV for SD-CS-NZSP, indicating a slightly lower energy barrier for Na+ hopping in the denser ceramic. At an elevated temperature of 100°C, the conductivity of SD-CS-NZSP reached $$5.24 \times 10^{-3} \text{ S cm}^{-1}$$, demonstrating its suitability for medium-temperature operation.
Electronic Conductivity and Electrochemical Stability
For a solid electrolyte to be effective in a solid-state sodium-ion battery, it must possess negligible electronic conductivity to prevent internal short circuits and self-discharge. The DC polarization curve for the SD-CS-NZSP pellet showed a rapid decay of current from an initial ionic+electronic current to a steady-state plateau dominated solely by electronic conduction. The calculated electronic conductivity ($\sigma_e$) was extremely low, on the order of $$1.23 \times 10^{-7} \text{ S cm}^{-1}$$. Using the values for $\sigma_t$ and $\sigma_e$, the sodium ion transference number ($t_{+}$) was calculated to be greater than 0.999. This confirms that the SD-CS-NZSP ceramic is a practically pure ionic conductor, which is a critical requirement for battery electrolytes.
The electrochemical stability window was assessed by CV. The voltammogram showed a pair of cathodic and anodic peaks near 0 V vs. Na/Na+, corresponding to the deposition and stripping of metallic sodium, respectively. Crucially, no other significant Faradaic current was observed within the wide scanned range of 0 to 6 V. This indicates that the NZSP electrolyte is electrochemically stable against decomposition up to at least 6 V, a voltage window that comfortably accommodates most high-potential cathode materials for sodium-ion batteries. This wide stability window is a key advantage of NASICON-type oxides over many sulfide-based solid electrolytes and is essential for constructing high-voltage, high-energy-density solid-state sodium-ion battery cells.
Conclusion
In this study, I have successfully demonstrated that the spray-drying granulation technique is a highly effective and process-oriented strategy for fabricating high-performance Na3Zr2Si2PO12 ceramic electrolytes. By transforming irregular, polydisperse powder into spherical granules with a uniform binder coating and a narrow particle size distribution, the spray-drying process fundamentally improves the powder’s packing characteristics. This leads to the formation of a highly dense green body, which upon sintering, yields a ceramic electrolyte with a relative density of 97.5%—a significant achievement for undoped NZSP via pressureless sintering.
The microstructural superiority directly translates to enhanced electrochemical performance. The spray-drying derived ceramic electrolyte (SD-CS-NZSP) exhibits a room-temperature ionic conductivity of $$6.96 \times 10^{-4} \text{ S cm}^{-1}$$, which is substantially higher than that of its conventionally processed counterpart. It also possesses an extremely low electronic conductivity (~10-7 S cm-1), a high Na+ transference number (>0.999), and a wide electrochemical stability window up to 6 V vs. Na/Na+.
This work underscores that beyond compositional doping, advanced powder processing is a critical lever for optimizing the properties of ceramic solid electrolytes. The spray-drying method presented here provides a scalable and reproducible route to achieve high densification and conductivity in NASICON-type electrolytes. This development is a meaningful step forward in the quest for viable solid electrolytes for next-generation solid-state sodium-ion batteries. The principles established are not limited to NZSP but are broadly applicable to other ceramic ionic conductors, potentially accelerating the development of various all-solid-state battery systems.