Sintering conditions such as temperature, time, and atmosphere of the sintering process are important factors affecting the structural and electrochemical characteristics of SEs [32]. Therefore, it is important to optimize the sintering conditions for synthesizing the SE. In addition, such optimization is necessary to prevent unwanted side effects, such as Li-loss during the sintering process and optimization of sample composition. As the primary heat treatment, LiTaSiO₅ SE pellets were sintered in a box furnace in air atmosphere for 6 h at temperatures of 900°C to 1100°C, where the temperature was increased in steps of 50°C, to check the optimum synthesis conditions.

X-ray diffraction analysis was performed to determine the structural characteristics of the samples sintered under various sintering temperature conditions. Fig. 1a depicts the XRD patterns of the LTSO SEs sintered at different temperatures ranging from 900°C to 1100°C. The XRD patterns of the sample sintered at 900°C and 950°C did not match the reference pattern (ICSD#39648) [27]. This result was attributed to the failure to bind the oxygen atoms and silicon atoms during synthesis of the LiTaSiO₅ SE under low-temperature conditions. As depicted in Fig. 1a, the diffraction peaks of the pristine LTSO powder at 1000°C or higher belong to a monoclinic phase, confirming that the LTSO structure clearly changes from the trigonal phase (space group: P 1, LiTaO₃) to the monoclinic phase with increasing sintering temperature. This result indicated that the sintering temperature had a significant effect on the crystal structure and phase formation of the LTSO SE. Additionally, a minor impurity phase (LiTaO₃) was detected in all LTSO powders sintered at high temperatures.

Fig. 1b depicts the XRD pattern of the (110) plane of the LTSO SE obtained by sintering at 1000–1100°C. For the pellets sintered at 1000–1050°C, the (110) peak intensities increased. However, for the pellets sintered at 1100°C, the peak intensities and d-spacing values decreased (Fig. 1c). This indicated that LTSO might have undergone lithium evaporation at around 1100°C. For clarity, the lattice parameters of the samples sintered at different temperatures are summarized in Table 1. The lattice parameters and volume of the LTSO SE were higher only at the sintering temperature of 1050°C. Seemingly, the lattice parameters of the sample sintered at 1100°C were marginally lower, possibly because of lithium loss.

Electrochemical impedance spectroscopy (EIS) was performed to study the ionic conductivity of the LiTaSiO₅ SE as a function of the sintering temperature in the range of 900–1100°C. The Cole–Cole plots of all the sintered samples exhibit only one semicircle in the high-to-low-frequency region (Fig. 1d). The semicircle in the graph is related to the total impedance of grains and grain boundaries. Compared to the resistance of the low-temperature sample, the resistance of the high-temperature samples was lower, and it decreased and then increased as the sintering temperature increased. Li evaporation during long-term sintering may have led to an increase in charge transfer resistance [4,33]. The ionic conductivity of the SE samples at RT increased as the sintering temperature increased, and it reached approximately 7.76×10^–8 S cm^–1 for the LTSO SE sintered at 1050°C. However, the ionic conductivity of the sample sintered at 1100°C was lower. The sample sintered at 1050°C exhibited the smallest grain boundary resistance value (2,090,000 Ω) in the EIS curve, and this result was attributed to its higher d-spacing value and lattice parameter, as confirmed by XRD analysis (Fig. 1a–c). Hence, the sintering temperature of 1050°C was selected as a suitable condition for further synthesis of LTSO SEs.

FE-SEM was used to examine the surface morphology of the LTSO SEs pellets. Fig. 2a–d present morphological images of the LTSO pellets prepared by sintering at 900–1100°C. The images in Fig. 2a,b confirm that sintering at 900–950°C was inadequate for particle binding. According to Fig. 2c, at the sintering temperature of 1000°C, the interparticle contact area increased due to neck growth, and a continuous pore channel was developed in the pellets [34–37]. At the sintering temperatures of 1050–1100°C, this continuous pore channel vanished owing to particle fusion, and large pores in the pellets were eliminated because the particle size increased (Fig. 2d) [34–37]. Additionally, cracks were observed in all the pellets sintered at high temperatures. Hence, sintering at temperatures exceeding 1000°C was confirmed to have increased the pellet’s compaction strength and integrity.

The XRD patterns of the LTSO SEs sintered at 1050°C for durations of 6 h to 24 h are presented in Fig. 3. These XRD results exhibit characteristic peaks, and there are no significant differences in the crystallinity of the samples prepared by sintering for 6 h to 18 h (Fig. 3a). Table 2 summarizes the lattice parameters of the samples prepared by sintering for different durations. As the sintering time increased from 6 h to 12 h, the average grain size of the samples increased owing to grain boundary migration [32,38]. In samples sintered for more than 12 h, the average grain size decreased owing to enhanced grain boundary generation [39]. Table 3 summarizes the average grain sizes of the samples sintered for different durations. The peak intensity of the sample sintered for 24 h decreased owing to Li evaporation (Fig. 3b). In addition, LiTaO₃ impurity peaks were identified in the spectra of all the prepared samples owing to inadequate bonding between oxygen and silicon atoms.

Fig. 3d shows the Cole–Cole plot of the LTSO SEs sintered for different durations (6–24 h). In Fig. 3d, a correlation is established between the sintering time and resistance properties of the LTSO SEs. As the temperatures increased, the diameter of the semicircle decreased steadily, and simultaneously, ionic conductivities of the LTSO SEs increased. The total resistance (R_total) of the LTSO SEs decreased as the sintering time increased from 6 to 12 h. The total resistance of the LTSO SEs first decreased from 6 h to 12 h then increased as the holding time reached 18 h, as can be inferred from their microstructural properties and grain size data (Table 3). The samples sintered at 1050°C for 12 h exhibited the highest ionic conductivity (1.78×10^–7 S cm^–1), which was attributed to its lowest total resistance (1,350,000 Ω), highest d-spacing (110 plane) value, and improved crystallinity (Fig. 3b,c). Apart from sintering temperature, sintering time, too, significantly influenced Li⁺ volatilization. As the sintering time increased (> 18 h), the volatilization of Li⁺ in the LTSO SEs was unavoidable, which could have led to a decrease in peak (110) intensity (Fig. 3b). When the holding time was short (12 h), the conductivity of the LTSO SEs increased owing to the large average grain size and low intensity of the LiTaO₃ phase. However, the conductivity of the LTSO SEs decreased as the holding time increased beyond a certain threshold (>12 h), probably because of Li⁺ volatilization and an increase in LiTaO₃ phase intensity (Fig. 3a). Accordingly, prolonged sintering times adversely affected the ionic conductivity of the SEs. Therefore, the sintering time of 12 h was selected as the optimal condition for further processing.

The morphological images presented in Fig. 4 systematically reveal the effect of sintering time on the sintered LTSO SE pellets. Additionally, the porosity of the LTSO pellets decreased significantly as the sintering holding time increased. No major transformation in particle shape occurred with increasing sintering time. Moreover, all the samples exhibited crack formation owing to thermal expansion of particles.

An ICP-OES analysis was performed to check the actual stoichiometry of the Li_1+xTa_1–xZr_xSiO₅ (x = 0–0.3) SEs, and the results are summarized in Table 5. The ICP-OES analysis was carried out after pre-treating the SE samples with aqua regia. For five SE compositions, most of the target values of the synthesis materials and the values obtained through analysis were consistent.

The XRD patterns of the Zr-doped Li_1+xTa_1–xZr_xSiO₅ (x = 0–0.3) powders are presented in Fig. 5a. In the composition range of x = 0–0.2, the diffraction patterns were indexed based on the reference pattern (ICSD#39648) with a small amount of LiTaO₃ impurity phase. For x = 0.1 or higher, the major (110) diffraction plane shifted to a lower degree, meaning that Zr⁴⁺ was successfully incorporated into the crystal structure, and the d-spacing and lattice parameter values increased, which favored fast Li-ion transfer (Fig. 5b,c). Table 4 summarizes the lattice parameters of the prepared Li_1+xTa_1–xZr_xSiO₅ (x = 0–0.3) electrolyte. The diffraction peaks shifted to lower angles owing to the larger ionic radius of Zr⁴⁺ (0.72 Å) than that of Ta⁵⁺ (0.64 Å) [24,30]. However, as the zirconium substitution composition value x increased, the LiTaO₃ peak intensified, which was attributed to destabilization of the binding of oxygen and silicon due to zirconium substitution. Additionally, for x = 0.3, a new ZrSiO₄ peak evolved in the sample [40,41]. Therefore, the zirconium substitution composition range for maintaining the LiTaSiO₅ crystal phase was set to x = 0–0.2. The morphology of the as-synthesized Zr-doped Li_1+xTa_1–xZr_xSiO₅ (x = 0–0.3) SE powders sintered at 1050°C for 12 h in the air is depicted in Fig. 6. zirconium content played a critical role in decreasing cracking and porosity of the pellets. Furthermore, zirconium doping improved contact between the SE particles. This reduced cracking and increased the relative density of the pellets. This phenomenon can be viewed from various perspectives. Solid-state diffusion plays a significant role in particle formation and interparticle bonding, and therefore, diffusion-based bonding has a vital effect on the mechanical and microstructural properties of materials [42–44]. Furthermore, diffusion itself is governed by sintering temperature and time [42–44]. Consequently, the relative density of the Li_1+xTa_1–xZr_xSiO₅ (x = 0–0.3) SEs decreased when the zirconium content was higher than x = 0.1. The large particle size of the SEs was attributed to the migration of boundaries [32]. Furthermore, the large particle size reduced the contact area between the SE particles, leading to lower density [32,42]. High zirconium doping concentrations induced discontinuous grain growth. According to these results, particle size growth was directly proportional to doping concentration. In addition, the ionic conductivities of the Li_1+xTa_1–xZr_xSiO₅ (x = 0–0.3) pellets prepared with different zirconium doping concentrations were determined by performing an EIS analysis.

A Cole–Cole plot of the Li_1+xTa_1–xZr_xSiO₅ (x = 0–0.3) SEs sintered at 1050°C for 12 h is presented in Fig. 5d. As depicted in this figure, a semicircle similar to that mentioned in the preceding sections was observed. For x = 0 and 0.05, the SEs exhibited low Li-ionic conductivities of 1.78×10^–7 and 2.97×10^–7 S cm^–1, respectively, owing to the lower density, d-spacing (110 plane) value, and lattice parameters (Table 4). As the Zr concentration increased, the total resistance value decreased at first and then increased, reaching a minimum at x = 0.10 (R_tot =11,500 Ω). The total ionic conductivity of Li_1+xTa_1–xZr_xSiO₅ (x = 0.1) was 2.01×10^–5 S cm^–1. For comparison, the ionic conductivity properties of the LTSO SEs are summarized in Table 6. The heterovalent substitution of Zr⁴⁺ for Ta⁵⁺ could increase the lithium-ion concentration to accommodate charge compensation. In a similar manner, it may reduce the activation energy and, additionally, induce concerted diffusion to improve Li⁺ ionic conductivity [24,27,28]. Zr doping in the LTSO SEs enhanced their ionic conductivities to up to 10^–5 S cm^–1, which is two times higher than those of the pristine LTSO SEs (1.78×10^–7 S cm^–1). This improvement was ascribed to the reduction of pores and cracks, as well as to the denser structure of the Li_1.1Ta_0.9Zr_0.1SiO₅ SE. Nevertheless, the conductivity of the Li_1+xTa_1–xZr_xSiO₅ (x = 0–0.3) SEs decreased as the doping concentration exceeded a certain value (x > 0.1), which was undoubtedly attributed to an increase in the LiTaO₃ peak intensity, decreasing density, and abnormal particle size growth (Fig. 6) triggered by the generation of a new ZrSiO₄ phase (Fig. 5a).

The ionic conductivities and relative densities of the Li_1+xTa_1–xZr_xSiO₅ (x = 0–0.3) SEs are summarized in Table 6. Compared to the pristine LTSO SEs (relative density of 79.5%), the relative densities of the Zr-doped LTSO SEs were higher (in excess of 90%), which first increased and then decreased with increasing doping concentration. It is well established that zirconium doping of LTSO SEs improves their conductivity and mechanical strength, in addition to reducing their resistance. In addition, it was concluded that sintering temperature (1050°C), time (12 h), and zirconium substitution (x = 0.1) conditions increased the relative density and reduced cracking, which helped improve the ionic conductivities of the doped LTSO SEs by 250 times compared to those of the pristine LTSO SEs. Nevertheless, the ion conductivities of the doped LTSO SEs were lower than the theoretical value of 6.4×10^–3 S cm^–1 calculated through simulation. This was attributed to surface cracks, crystalline collapse, presence of LiTaO₃, and changes in particle size during the sintering process. Therefore, the optimization of process variables such as sintering temperature and time, substitution composition, substitution element changes, pressure required for pellet production, particle size, and density should be explored in future studies.

Fig. 7 shows the Galvanostatic charge-discharge, Cyclic voltammetry, and Charge-discharge performance of the LTZSO SE cells. The working current density of the cell was 0.1 mA cm^–2 for the corresponding plating and stripping process and the duration time was 30 min. Fig. 7a confirms there was no fluctuations during the cell performance and the electrolyte is stable up to 100 cycles. We added liquid electrolyte at the interface to improve the contact between electrolyte the Li metal electrode. The CV diagram (Fig. 7b) shows clear oxidation and reduction peaks corresponding to lithium intercalation and de-intercalation in the NCM811 cathode. After the first cycle, the oxidation and reduction potential of the cell was shifted to lower potential due to the activation process. From Fig. 7c–d, the initial charge and discharge capacity of the cell were 123 and 99 mAh g^–1, respectively. Subsequently after the first cycle, the discharge capacity of the cell was increased due to the activation effect in the electrode/electrolyte interface. Afterwards 15 cycles, the cell provided a discharge capacity of 150 mAh g^–1. Moreover, we are working to improve the ionic conductivity, capacity and cell performance.