Halogen Doping in Na₃PS₄ Solid Electrolytes for High Performance All-Solid-State Sodium Batteries

Abstract

Sulfide-based solid electrolytes are promising for all-solid-state sodium batteries due to their high ionic conductivity and facile processability, but their practical use is limited by moisture sensitivity and poor interfacial stability. To address these issues, Na_3−xPS_4−xM_x (M = F, Cl, Br, I) electrolytes were first synthesized as a preliminary study to evaluate the effect of halogen doping. Chlorine was identified as the most effective dopant and was therefore selected for a systematic investigation of doping concentration. Na_3−xPS_4−xCl_x (x = 0.1–0.3) electrolytes were prepared by solid-state sintering, and the optimum composition was determined to be Na_2.85PS_3.85Cl_0.15, which achieved a high ionic conductivity of 5.5 × 10⁻⁴ S·cm⁻¹ with a reduced activation energy of 33.3 kJ·mol⁻¹. When employed in TiS₂|Na_2.85PS_3.85Cl_0.15|Na₃Sn full cells, the optimized electrolyte enabled high initial capacity, excellent rate capability, and stable long-term cycling. These results highlight the effectiveness of Cl doping concentration control in enhancing both the intrinsic properties of Na₃PS₄-based electrolytes and the overall electrochemical performance of all-solid-state sodium batteries.

1. Introduction

With the accelerating global transition toward renewable energy, sodium-ion batteries (SIBs) have emerged as promising candidates for large scale energy storage owing to the abundance, low cost, and wide distribution of sodium resources. However, conventional liquid electrolyte SIBs face inherent safety challenges, including electrolyte leakage, flammability, and uncontrolled dendrite growth, that severely hinder their practical deployment. To address these issues, all-solid-state sodium-ion batteries (ASSIBs) employing solid-state electrolytes (SSEs) have attracted extensive attention because of their superior thermal stability, wide electrochemical window, and enhanced interfacial compatibility with sodium metal anodes [1,2,3].

Among various SSEs, sulfide-based electrolytes, particularly Na₃PS₄, are regarded as highly attractive due to their relatively high ionic conductivity at room temperature, favorable mechanical properties, and facile processability. Na₃PS₄ can be synthesized via cost-effective mechanochemical routes under ambient conditions without high temperature sintering, while its soft nature ensures good interfacial contact with electrodes and thus lowers interfacial resistance [4,5,6]. Nevertheless, pristine Na₃PS₄ suffers from several intrinsic drawbacks: (i) high sensitivity to moisture, accompanied by the release of toxic H₂S gas; (ii) insufficient chemical stability against sodium metal, leading to side reactions and dendrite growth; and (iii) limited ionic conductivity compared with oxide-based conductors [7,8,9]. Recent studies have sought to address these issues through multiple strategies, such as surface molecular engineering for improved processing stability [10], polymeric or buffer interphases to suppress interfacial degradation [11], fluorine-rich interphase chemistries to homogenize the electric field and inhibit dendrites [9], and oxysulfide compositions to enhance chemical robustness [12]. Despite these advances, achieving a balance between high ionic conductivity and long-term interfacial stability remains a pressing challenge [13,14,15].

Doping has emerged as a particularly effective strategy for improving SSE performance by introducing lattice distortions, optimizing Na⁺ migration pathways, and tailoring defect chemistry [16,17]. In sulfide-based electrolytes such as Na₃SbS₄, halogen substitution for sulfur has been shown to expand the lattice, lower migration barriers, and thereby enhance ionic conductivity [18]. More recently, systematic investigations have begun to explore halogen doping in Na₃PS₄, with encouraging results: Cl⁻ doping has been reported to stabilize the Na|Na₃PS₄ interface and suppress dendritic growth [19,20,21], while Br⁻ doping can reduce activation energy and improve bulk ionic conductivity [22,23]. However, excessive doping may disrupt the Na₃PS₄ framework and introduce impurity phases, ultimately compromising conductivity [24]. Thus, careful optimization of the dopant type and concentration is essential to simultaneously maximize ionic conductivity and interfacial stability.

In this context, the present study focuses on the design and synthesis of halogen-doped Na₃PS₄ electrolytes (Na_3−xPS_4−xM_x, M = F, Cl, Br, I) via solid-state sintering. Systematic structural and electrochemical investigations are conducted to clarify the effects of halogen type and doping level on lattice stability, Na⁺ migration, and electrochemical performance. Particular attention is devoted to Cl-doped Na₃PS₄, which exhibits the most favorable balance between ionic conductivity and interfacial stability. This work provides mechanistic insights into halogen-induced structural modulation and offers practical guidance for the rational design of high-performance sulfide-based SSEs for next-generation all-solid-state sodium batteries.

2. Material and Experimental Method

2.1. Preparation of Halogen-Doped Na₃PS₄ Electrolytes

Halogen-doped Na₃PS₄ solid electrolytes with the nominal composition Na_3−xPS_4−xM_x (M = F, Cl, Br, I; x = 0.1, 0.15, 0.2, 0.3) were synthesized via a solid-state method. Stoichiometric amounts of Na₂S, P₂S₅, and the corresponding halide sources (NaF, NaCl, NaBr, or NaI) were weighed in an argon filled glovebox (H₂O/O₂ < 0.1 ppm), with a total batch mass of 1.5 g. The mixtures were ground for ~15 min and then were transferred to a sealed ball mill container to mill at 550 rpm for 11 h. The smiled powders were finely ground, pressed into pellets, sealed in evacuated quartz tubes, and sintered at 420 °C for 12 h, followed by furnace cooling to room temperature.

2.2. Cathode Preparation and Cell Assembly

Electrolyte pellets for electrochemical testing were prepared by pressing 100 mg of powder into 12 mm disks under 350 MPa for 3 min to obtain dense pellets of ~0.5 mm thickness. For the composite cathode, TiS₂ (Alfa Aesar, Ward Hill, Haverhill, MA, USA, 99.8%), solid electrolyte and Super P (TIMCAL, Bodil, Switzerland, 99.5%) were mixed at a mass ratio of 4:6:1 and ground in an agate mortar for ~20 min. For full-cell assembly, the composite cathode was spread onto one side of the electrolyte pellet, while a Na₃Sn alloy anode was placed on the opposite side. The stack was pressed at 400 MPa for 3 min to ensure intimate interfacial contact. All procedures were carried out in an argon glovebox (MBraun, Garching, Germany).

2.3. Characterization Methods

The phase structures of the electrolytes were analyzed by X-ray diffraction (XRD, Rigaku SmartLab, Kyoto, Japan, Cu Kα radiation, λ = 1.5406 Å). Raman spectra were collected on a Raman spectrometer (Horiba HR800, Kyoto, Japan) with a 532 nm laser source. Morphology and elemental distribution were examined by scanning electron microscopy (SEM, Zeiss GeminiSEM 300, Baden, Germany) equipped with energy-dispersive X-ray spectroscopy (EDS).

2.4. Electrochemical Measurements

Electrochemical impedance spectroscopy (EIS, ParStat 4000, Princeton, NJ, USA) was carried out in the frequency range of 0.1 Hz–10 MHz with an AC amplitude of 30 mV. Ionic conductivities were measured at 303–373 K, and activation energies were determined from Arrhenius plots according to σ = Aexp(–Ea/RT).

Galvanostatic charge–discharge (GCD, CT-4008, Neware, Shenzhen, China) and cycling tests were carried out in the voltage range of 1.2–2.4 V at various C-rates. Rate capability and long-term cycling stability were also evaluated under different C-rates. Sodium symmetric cells were assembled with Na metal as both electrodes and solid electrolyte pellets as separators, and were cycled at current densities from 0.01 to 0.1 mA cm⁻² in 0.01 mA cm⁻² increments.

3. Results and Discussion

3.1. Structural, Transport, and Interfacial Properties of Halogen-Doped Na₃PS₄

To investigate the effect of halogen doping on Na₃PS₄, four halide ions (F⁻, Cl⁻, Br⁻, I⁻) were introduced by substituting sulfur sites in the lattice. This approach utilizes the differences in ionic radius and electronegativity to tune the local structure and transport channels. The resulting compositions, Na_3−xPS_4−xM_x (M = F, Cl, Br, I; denoted as NPSF, NPSCl, NPSBr, and NPSI), were synthesized under identical conditions for direct comparison.

Figure 1 shows the XRD patterns of the halogen-doped electrolytes. The reveals that the characteristic diffraction peaks of electrolytes doped with four halogens (F, Cl, Br, and I) are consistent with those of pure Na₃PS₄ and closely match the standard NPS phase structure. Notably, despite the incorporation of halogen ions with different electronegativities and ionic radii, no significant shift is observed in the diffraction peak positions of the doped samples. This indicates that, under the present doping conditions, the introduction of halogen atoms does not noticeably affect the lattice parameters or crystal structure of the electrolytes. Accordingly, the four halogen-doped electrolytes, together with pure Na₃PS₄, can be classified as fast ion conductors. The XRD pattern of pure Na₃PS₄, shown by the black line in the figure, is consistent with the diffraction peaks reported in the literature [11]. In addition, several unidentified peaks appear in the patterns of the NaBr- and NaI-doped electrolytes, suggesting the presence of impurity phases introduced during the doping process. By contrast, the NaF- and NaCl-doped electrolytes show no impurity peaks. However, the diffraction peak intensity of the NaF-doped sample is slightly reduced, whereas that of the NaCl-doped sample is enhanced, indicating that the synthesized solid electrolytes exhibit a high degree of crystallinity while maintaining the tetragonal crystal structure of Na₃PS₄.

The ionic transport behavior was examined by EIS (0.1 Hz–10 MHz, 30 mV). As shown in Figure 2a, NPSCl exhibits a nearly linear response with negligible semicircle, suggesting suppressed grain boundary resistance. In contrast, NPSF, NPSBr, and NPSI display pronounced semicircles, indicating significant grain boundary contributions [25]. The calculated room temperature ionic conductivities (Figure 2b) are 7.62 × 10⁻⁵ S·cm⁻¹ (NPSF), 3.16 × 10⁻⁴ S·cm⁻¹ (NPSCl), 4.25 × 10⁻⁵ S·cm⁻¹ (NPSBr), 2.05 × 10⁻⁵ S·cm⁻¹ (NPSI), and 1.43 × 10⁻⁴ S·cm⁻¹ (pristine Na₃PS₄). Thus, Cl doping enhances conductivity by more than twofold compared to the pristine sample, while Br and I doping reduce conductivity by nearly an order of magnitude. Notably, the ionic radius and electronegativity differences lead to varying degrees of lattice distortion. Cl⁻ (1.81 Å) closely matches S²⁻ (1.84 Å), allowing isovalent substitution without disrupting PS₄ tetrahedra. F⁻ is too small (1.33 Å), causing local contraction that hinders Na⁺ migration. Br⁻ (1.96 Å) and I⁻ (2.20 Å) are too large, inducing lattice strain and impurity phases, thereby lowering conductivity [12].

The interfacial stability with sodium metal was evaluated using TiS₂|Na_3−xPS_4−xCl_x|Na₃Sn symmetric cells (Figure 3). Cycling was performed at 0.01–0.1 mA cm⁻² in steps of 0.01 mA cm⁻², with at least five cycles at each current density. At low current densities, all halide-doped electrolytes displayed stable cycling with low polarization voltages. However, when the current density reached 0.04 mA·cm⁻², cells with NPSBr and NPSI exhibited sharp polarization increases to ~3 V and ~2 V, respectively, indicating severe interfacial side reactions with sodium metal. The decomposition products formed at the interface possess poor ionic conductivity, thereby accelerating polarization growth and eventually leading to continuous electrolyte degradation. As a result, both NPSBr and NPSI failed within 30–35 h, reaching the 5 V cut-off voltage. In contrast, the NPSF and NPSCl electrolytes exhibited superior compatibility with sodium. NPSF remained stable up to 0.09 mA·cm⁻² for ~50 h, while NPSCl maintained excellent cycling stability, sustaining 0.1 mA·cm⁻² for more than 100 h with polarization below 1 V. This enhanced stability can be attributed to the formation of a passivating interfacial layer, which suppresses continuous electrolyte decomposition. These observations, consistent with the XRD and EIS analyses, highlight the crucial role of Cl doping in optimizing both ionic conductivity and interfacial stability.

Collectively, the structural, transport, and interfacial analyses identify Cl as the most effective halogen dopant, laying the foundation for a detailed investigation of NaCl-doped electrolytes in the following section.

3.2. Structure and Properties of Cl-Doped Na₃PS₄ Electrolytes

To further clarify the role of chlorine as the most effective halogen dopant, structural and vibrational characterizations of Na_3−xPS_4−xCl_x (x = 0.1, 0.15, 0.2, 0.3) were investigated by X-ray diffraction (XRD) and Raman spectroscopy. As shown in Figure 4a, all samples largely maintain the tetragonal Na₃PS₄ framework, confirming the successful synthesis of sulfide solid electrolytes. When the Cl content exceeds x = 0.2, additional reflections, likely originating from Cl-rich secondary phases, appear near 32° and 52°, accompanied by weakened intensities of the main peaks. At x = 0.3, the reflection near 32° is no longer discernible, suggesting partial structural distortion or the formation of secondary phases, which correlates with the observed reduction in ionic conductivity at higher doping levels. However, the impurity content is too low to be detected (for x = 0.3) [18]. Raman spectra recorded with a 532 nm laser in the 100–600 cm⁻¹ range are presented in Figure 4b. For x ≤ 0.15, the characteristic Raman bands of PS₄³⁻ units at ~410 and ~560 cm⁻¹ remain essentially unchanged apart from a gradual decrease in intensity. At higher doping levels (x ≥ 0.2), additional bands emerge and the ~410 cm⁻¹ band diminishes markedly; at x = 0.3 it becomes nearly invisible. These observations corroborate the XRD results, indicating the onset of structural disorder and the presence of Cl-related secondary phases at elevated doping levels.

The surface morphologies of Na_3−xPS_4−xCl_x electrolytes are presented in Figure 5a–e. All samples exhibit irregularly shaped particles with a broad size distribution. As the Cl content increases to x = 0.15, the particles become larger and exhibit partial agglomeration (Figure 5c). No significant morphological changes are observed at x = 0.2 (Figure 5d). In contrast, the sample with x = 0.3 shows dense block-like aggregates with visible surface pores, which are unfavorable for pellet densification and are expected to increase interfacial resistance (Figure 5e).

The elemental distribution of the optimized Na_2.85PS_3.85Cl_0.15 composition was further examined by EDS mapping (Figure 6). Na, P, S, and Cl are homogeneously distributed across the particles, confirming the uniform incorporation of Cl into the Na₃PS₄ framework after ball milling and sintering.

3.3. Electrochemical Performance of Cl-Doped Na₃PS₄ Electrolytes

The ionic conductivities of Na_3−xPS_4−xCl_x (x = 0, 0.1, 0.15, 0.2, 0.3) were measured by electrochemical impedance spectroscopy (EIS). As shown in Figure 7, the conductivity strongly depends on the Cl content. The highest value of 5.5 × 10⁻⁴ S·cm⁻¹ was obtained at x = 0.15, where a balance between defect formation and structural stability was achieved. At a lower content (x = 0.1), the conductivity moderately increased compared with pristine Na₃PS₄, while higher doping levels (x ≥ 0.2) resulted in a decline due to the emergence of impurity phases and disruption of Na⁺ transport pathways.

Impedance changes of electrolyte materials prepared with different NaCl doping amounts were tested at 303 K, 323 K, 343 K, 363 K, and 373 K. The activation energies of the corresponding solid electrolytes were calculated according to the Arrhenius equation σ = Aexp(−E_a/RT) (where A is the pre-exponential factor, R is the molar gas constant, and T is the thermodynamic temperature), as shown in

Table 1. Activation Energies Corresponding to Na_3−xPS_4−xM_x Series Electrolytes.

	Activation Energy (KJ/mol)
Na3PS4	36.4
Na2.9PS3.9M0.1	34.7
Na2.85PS3.85M0.15	33.3
Na2.8PS3.8M0.2	35.3
Na2.7PS3.7M0.3	39.9

. Figure 8 presents the Arrhenius curves and corresponding activation energies of NPS electrolytes prepared at different temperatures. From the activation energy data shown in Figure 8b, it can be seen that the content of doped elements has a significant impact on the activation energy of the electrolyte: with the increase in the amount of doped elements introduced, the activation energy first decreases and then increases. The lowest activation energy of 33.3 kJ/mol is achieved when x = 0.15; when the doping amount exceeds 0.15, the activation energy gradually increases, reaching the highest value of 39.9 kJ/mol at x = 0.3.

To demonstrate practical applicability, Na_2.85PS_3.85Cl_0.15 was integrated into all-solid-state batteries with TiS₂ composite cathodes and Na₃Sn alloy anodes. A schematic illustration of the pelletized electrolyte and the assembled cell is presented in Figure 9. The thin electrolyte layer (~0.54 mm) and intimate electrode–electrolyte contact provide favorable conditions for stable operation.

The rate performance of TiS₂|Na_2.85PS_3.85Cl_0.15|Na₃Sn cells is shown in Figure 9b,c. At 0.02, 0.05, 0.1, 0.2, and 0.5C, the initial discharge capacities were 201.6, 211.9, 148.4, 125.3, and 63.7 mAh·g⁻¹, respectively. Even at elevated current densities, the voltage profiles remained smooth and reproducible, indicating highly reversible Na⁺ intercalation and deintercalation.

Cycling performance under different rates is summarized in Figure 10. At 0.02C, the cell retained 86% of its capacity after 50 cycles, while at 0.05C it maintained 77.6% after 175 cycles. Although the voltage plateau gradually decreased with extended cycling, the overall retention remained competitive. At 0.1–0.2C, capacities stabilized around 98–97 mAh·g⁻¹ after 100 cycles, whereas at 0.5C the discharge capacity dropped to 64.5 mAh·g⁻¹ after 100 cycles. These results demonstrate that the optimized Cl-doped electrolyte enables stable cycling at low to moderate rates, while at higher rates the growth of interfacial resistance and polarization leads to accelerated capacity fading.

When the rate was increased to 0.5 C, the charge and discharge curves and cycling profiles (Figure 11) showed no distinct voltage plateaus even in the 1st, 2nd, 5th, 10th, and 100th cycles. This behavior contrasts sharply with conventional batteries that exhibit defined plateaus, suggesting anomalies in the reaction kinetics, the phase transition mechanisms of the electrode materials, or the interfacial processes between the electrolyte and electrodes. In other words, stable redox reactions are not established, and substantial resistance arises during Na⁺ intercalation and deintercalation, while the electrolyte may fail to sustain stable interfacial reactions. As a result, the voltage changes continuously and gradually throughout the charge–discharge process. This phenomenon is consistent with the observations reported by Chu et al. [9], implying that the present system shares mechanistic features with previously studied solid-state sodium batteries.

After 100 cycles, the capacity at 0.5 C was only 64.46 mAh·g⁻¹, whereas at 0.1 C and 0.2 C the capacities remained 98.37 and 97.6 mAh·g⁻¹, respectively. The capacity degradation at higher rates is attributed to structural damage and aggravated side reactions. Because ion diffusion in the electrode materials is intrinsically limited, high rate cycling forces a large flux of Na⁺ ions to intercalate or deintercalated within a short time, generating steep concentration gradients at the electrode surface. Ions in some regions cannot diffuse rapidly enough, resulting in polarization and reduced accessible capacity. Meanwhile, enhanced side reactions continuously consume both the solid electrolyte and sodium ions, further accelerating capacity fading.

4. Conclusions

In this work, halogen doping was employed to tailor the properties of Na₃PS₄ solid electrolytes for all-solid-state sodium batteries. Preliminary screening indicated that chlorine is the most effective dopant, and subsequent studies focused on controlling its substitution level.

Systematic investigation of Na_3−xPS_4−xCl_x revealed that the optimal composition, Na_2.85PS_3.85Cl_0.15, achieves an ionic conductivity of 5.5 × 10⁻⁴ S·cm⁻¹ at room temperature, owing to a favorable balance between defect formation and lattice stability. Excessive Cl incorporation was found to introduce secondary phases and impair ion transport. Electrochemical testing further demonstrated that the optimized Cl-doped electrolyte enhances interfacial stability with sodium metal, reduces polarization, and suppresses dendrite formation in symmetric cells. Integration into TiS₂|Na_2.85PS_3.85Cl_0.15|Na₃Sn full cells yielded high capacity, good rate capability and stable cycling. This confirms that controlled Cl doping in Na₃PS₄ effectively enhances ionic conductivity and interfacial stability, offering a viable route to advance sulfide-based solid electrolytes for all-solid-state sodium batteries.