The phase compositions of NZSP and Zn-NZSP samples prepared according to Table S2 were analyzed by X-ray diffraction (XRD) (Fig. 2a). The results show that the peaks of all samples are indexed to the NZSP phase (PDF No. 84-1200). However, the extra peaks marked by ★ and ▲ of NZSP sample are indexed to the ZrO2 phase (PDF No. 86-1451) and Na2ZrSi2O7 phase (PDF No. 73-2309), while the pattern of 0.15Zn-NZSP sample displays the peak marked by ⚪ of Na2ZnSiO4 phase (PDF No. 87-1572). These results indicate that the doping of Zn element can reduce the unexpected ZrO2 and Na2ZrSi2O7 phases, however, the amount of dopant needs to be limited. As shown in Fig. 2b and Fig. 2c, the high resolution transmission electron microscope (HRTEM) images the crystalline phases of NZSP and 0.1Zn-NZSP samples, respectively. In addition to the NASICON phase, the HRTEM results also indicate the co-existence of ZrO2, Na2ZrSi2O7 and glass phases in NZSP samples. Because ZnO, as a typical flux, helps to NASICON crystal growth and inhibit the volatilization of Na and P sources, 0.1Zn-NZSP sample only displays the NZSP phase [38, 39]. The scanning electron microscope (SEM) images of NZSP and 0.1Zn-NZSP samples are shown in Fig. S1 and Fig. S2, respectively. The granule size of NZSP is a range of 0.1 − 1.5 µm, while the size of 0.1Zn-NZSP is a range of 0.5 − 3.0 µm, and it suggests that ZnO can be used to increase the granule size. The thermogravimetry (TG)-differential scanning calorimeter (DSC) analysis is shown in Fig. 2d. The red lines are the DSC curves and the blue lines are the TG curves. The pronounced peaks at 109°C, 293°C and 572°C can be assigned to the volatilization of free water, the decomposition of ammonium dihydrogen phosphate and the decomposition of sodium carbonate, respectively. While the peaks at 1080°C corresponding to the crystallization of NZSP. These results indicate that appropriate Zn-doping can reduce the unexpected ZrO2 and Na2ZrSi2O7 phases. Electrochemical impedance spectroscopy (EIS) analyses were used to examine the ion transport performance. The EIS spectra are in the form of a semi-circle with the real axis intercepts occurring at the lower frequency end, this corresponds to the total resistance, including the grain and interfacial resistances. The EIS curves of NZSP, 0.05Zn-NZSP, 0.1Zn-NZSP, 0.15Zn-NZSP samples are shown in Fig. S3. The total conductivities were calculated for the sample optimization (Fig. 2e). The divalent metal Zn can replace Zr into ZrO6 octahedron to modify the sodium transport channel [40, 41], thus 0.1Zn-NZSP sample displays the highest ion conductivity of 0.32 mS cm− 1, while that of NZSP sample is 0.048 mS cm− 1. These results prove that ZnO is a good dopant to suppress the ZrO2 and Na2ZrSi2O7 phases, thus improving the ion conductivity.
A schematic illustration of the synthesis route for the Zn-NZSP/ZIF-8/PVDF-HFP SSEs is shown in Fig. 3a. Firstly, the raw materials with strictly stoichiometry were mixed to obtain the precursor, the precursor was transformed into Zn-NZSP after annealed in an air atmosphere. Secondly, the ZIF-8 film grow up along with the ZnO6 octahedron on the surface of Zn-NZSP granules via heterogeneous nucleation and growth of ZIF-8, forming the Zn-NZSP/ZIF-8 intermediate. Finally, the Zn-NZSP/ZIF-8 intermediates were evenly dispersed in PVDF-HFP solution, the Zn-NZSP/ZIF-8/PVDF-HFP samples were obtained after consolidation and drying. The ratios of Zn-NZSP and ZIF-8 were used to optimize ZIF-8 film (Table S3). As-obtained samples were probed by XRD and SEM. The thin ZIF-8 film on the surface of Zn-NZSP can’t be probed by XRD, so only the pure NASICON phase is displayed (Fig. S4). Notably, the SEM image in Fig. S5 exhibits the rough surface compared with the image in Fig. S2, suggesting the ZIF-8 surface coating on the Zn-NZSP granules. The XRD (Fig. S4) and SEM (Fig. S6) results indicate the formation of SSEs with good interface contact. The EIS curves of all samples were obtained at different temperatures (Fig. S7). As-obtained total sodium ionic conductivities in Fig. 3b indicate that C sample possesses the highest conductivity of 2.2 mS cm–1 at 25 oC, and the sodium ion conductivities of C sample at 30 to 80 oC are the highest among all of samples (Fig. 3c). Remarkably, the SSE films of NZSP/PVDF-HFP, NZSP/ZIF-8 and ZIF-8/PVDF-HFP can’t be formed through such technology.
The ZIF-8 structure on the NZSP surface was probed by the TEM and EDS mapping. In Fig. 4a, the low magnification image displays some granules with the size range of 0.5 − 2 µm, it is consistent with the SEM result (Fig. S2). The enlarged image shown in Fig. 4b displays that the d-spacing of 0.452 nm corresponds to the (020) plane of NZSP around by the amorphous structure. The clear and orderly diffraction pattern in the inset of Fig. 4b indicates the single crystal feature. Note that the position marked by the red dashed line reveals the good connected interface between NZSP and ZIF-8. In Fig. 4c, the elemental mappings of Na, Zr, Si, P and O exhibit the phase interface between NZSP and ZIF-8, while the mappings of Zn and N are uniformly distributed, indicating the uniformity of ZIF-8 coating layer. Therefore, above TEM and EDS mapping results prove that the good interface was constructed between the NZSP and the ZIF-8 layer. Furthermore, XPS was employed to identify the interface contact between the NZSP, ZIF-8 and PVDF-HFP. The peak located at 285.0 eV was used for the spectrum calibration. The survey spectra of the Zn-NZSP/ZIF-8 and Zn-NZSP/ZIF-8/PVDF-HFP samples evidence the presence of Na, Zr, Zn, Si, P, N (Fig. S8) and Na, Zr, Zn, Si, P, N, C, F (Fig. S9), respectively. In addition, the peak locations of Na, Zr, Zn, Si, P in the high − resolution XPS spectra indicate the bonding consistency. As shown in Fig. 4d, compared to the N 1s peak at 399.2 eV of ZIF-8 sample, the N 1s peak at 399.1 eV of 0.1Zn-NZSP/ZIF-8 confirm the formation of good interface. While, the N 1s peak at 400.0 eV of Zn-NZSP/ZIF-8/PVDF-HFP sample indicate the bonding combination of N in ZIF-8 layer was increased. Similarly, the F 1s peak (686.6 eV) of PVDF-HFP moved in the direction of large bonding energy (688.3 eV) in Fig. 4e, indicating the bonding combination of F in PVDF-HFP layer was increased. Therefore, the XPS results confirm that the formation of compatibility interface in Zn-NZSP/ZIF-8/PVDF-HFP bridging structure[42, 43], which are attributed to both rigid (inorganic) and flexible (organic) properties of ZIF-8 [44, 45].
In order to highlight the bridging effect of ZIF-8/PVDF-HFP, we compare the structure and performance of the NZSP, 0.1Zn-NZSP and 0.1Zn-NZSP/ZIF-8/PVDF-HFP samples. The schematic illustration of such sample structures is shown in Fig. 5a. The pores, cracks and ZrO2 impurity in the NZSP sample (Fig. 5b) hindered the sodium ion transport. After Zn-doping, the cracks and ZrO2 impurity were avoided due to the assistive melting effect of ZnO, but the size range of pores increased (Fig. 5c) due to the foaming caused by oxygen production. The performance enhancement (Fig. 2e) after Zn-doping indicates the avoiding of cracks and ZrO2 impurity is significant. Furthermore, the formation of Zn-NZSP/ZIF-8/PVDF-HFP avoided the pore structure (Fig. 5c) evidenced by the SEM results in Fig. 5d, delivering the improved sodium ion transport performance (Fig. 3).
To explore the reasons for performance improvement, the EIS curves of NZSP, 0.1Zn-NZSP and C samples at 30 − 80 oC (Fig. S10) were obtained to calculate the sodium ion conductivities of these samples for the inner-granular, intergranular and total. Please see the sodium ion conductivities of NZSP and 0.1Zn-NZSP samples at different temperatures in Fig. 6a and Fig. 6b. The crossover phenomenon is very obvious, especially before the granule bridging, the total conductivity is significantly affected by temperature. Differently, the sodium ion conductivities of C sample are smoothly enhanced with the temperature increasing (Fig. 6c). This pattern indicates that the regular consistency of inner-granular, intergranular and total conductivities with the temperature variation is conducive to obtaining high performance. In Fig. S11, the curves of 1000/T − lnɑ were used to obtain the activation energies, the results are compared in Fig. 6d. The minimum difference in activation energy between grains and boundaries determines the minimum value of total activation energy (0.25 eV). It is the result of bridging effect of ZIF-8/PVDF-HFP among the NASICON granules, resulting in a homogeneous ion transport.
For proof-of-application, the battery performances of the Zn-NZSP/ZIF-8/PVDF-HFP SSEs were evaluated via the symmetrical Na|SSEs|Na cells, the charge-discharge were tested at 1 mA cm− 2. As shown in Fig. 7a, the charge-discharge performances of all samples are compared, the cell with C sample displays the smooth charge-discharge voltages of ± 0.15 V, such optimize result is consistent with the optimized ionic conductivity. Furthermore, the optimal sample was used to compare with the commercial glass fiber SSE and polypropylene SSE (Fig. 7b), exhibiting sufficient stability due to the excellent interface connectivity via the bridging structure [46, 47]. In Fig. 7c, the long cycle test displays good stability for more than 1500 h, reflecting the good structural stability. Moreover, the voltage window of the best sample is 6.5 V (Fig. S12), indicating its stability.