Preparing 3D Perovskite Li0.33La0.557TiO3 Nanotubes Framework Via Facile Coaxial Electro‐Spinning Towards Reinforced Solid Polymer Electrolyte

It is of significance to construct continuous multiphase percolation channels with fast lithium‐ion pathway in hybrid solid electrolytes. 3D ceramic nanostructure frameworks have attracted great attention in this field. Herein, the three‐dimensional perovskite Li0.33La0.557TiO3 nanotubes framework (3D‐LLTO‐NT) is fabricated via a facile coaxial electro‐spinning process followed by a calcination process at 800 °C. The hybrid polymer electrolyte of 3D‐LLTO‐NT framework and poly (ethylene carbonate) (3D‐LLTO‐NT@PEC) shows improved ionic conductivity of 1.73 × 10−4 S cm−1 at ambient temperature, higher lithium‐ion transference number (tLi+) of 0.78 and electrochemical stability window up to 5.0 V vs Li/Li+. The all‐solid‐state cell of LiFePO4/3D‐LLTO‐NT@PEC/Li delivers a high specific capacity of 140.2 mAh g−1 at 0.1 C at ambient temperature. This outstanding performance is attributed to the 3D ceramic nanotubes frameworks which provide fast lithium ion transfer pathway and stable interfaces.

thermostability and wide electrochemical stability window. [26,27] 2D and 3D nanostructured LLTO framework were recently reported for the reinforced ionic conductivity of composite solid electrolytes with satisfying solid battery performance. [28] In previous work, the openstructured LLTO nanotubes (NTs) were fabricated as ion-conductive fillers in composite solid electrolytes by a gradient electro-spinning method. Since the gradient electro-spinning method only uses one channel with one precursor solution, the prepared hollow nanofibers often have an uncontrollable hollow structure. In contrast, coaxial electro-spinning uses two sorts of precursor solutions in separate inner and outer channels. [29,30] Consequently, it can precisely control the shell-core structure of the coaxial fibers. Therefore, coaxial electrospinning is an easier and effective strategy for the precise preparation of coaxial nanofibers and hollow nanotubes.
In this work, the 3D perovskite LLTO nanotubes framework (3D-LLTO-NT) was fabricated via a facile coaxial electro-spinning process following with calcination process. The precisely prepared hollow LLTO nanowires have a controllable hollow structure. The nanotubes framework (3D-LLTO-NT) was combined with poly (ethylene carbonate) (PEC) and bis (trifluoromethane) sulfonamide lithium (LiTFSI) to prepare the hybrid polymer electrolyte (3D-LLTO-NT@PEC). The obtained 3D-LLTO-NT@PEC polymer electrolyte showed a remarkable enhancement in the ion conductivity, higher Li-ion transfer number and satisfying solid state battery performance at ambient temperature, as shown in Table S1, Supporting Information.

Results and Discussion
The coaxial electro-spinning process was illustrated in Figure 1. The coaxial electro-spinning was verified to be a simple and effective strategy for the precise preparation of coaxial nanofibers and hollow nanotubes. [31,32] The coaxial electro-spinning uses two kinds of precursor solutions in separate inner and outer channels with varied individual feeding rates. Therefore, it can precisely control the core/shell structures and morphologies of the coaxial fibers. In this work, the uniform coaxial core/shell nanotubes were prepared via coaxial electro-spinning by precisely adjusting the parameters of the electrostatic spinning process, i.e., the viscosity of the precursor solution, the applied voltage ( Figure S1, Supporting Information), and the feeding rate ratio of the core/shell feed solution ( Figure S2, Supporting Information). Figure S1, Supporting Information, showed SEM images and TEM images of as-spun nanofibers and the calcined three-dimensional perovskite Li 0.33 La 0.557 TiO 3 nanotubes framework (3D-LLTO-NT) prepared in a spinning voltage range of 12-18 kV. As shown, the microstructures of the as-spun nanofibers were the same in nature regardless of the spinning voltage. Only the diameter of the as-spun nanofibers changed with the increase of spinning voltage. The flow rate of the core precursor solution was 0.4 mL h À1 . In addition, the shell precursor solution was set at various flow rate range from 0.4-0.8 mL h À1 . The wall thickness of the nanotubes increased with the increase of the flow rate of the shell solution ( Figure S2, Supporting Information), which demonstrated the wall thickness of the 3D-LLTO-NT could be easily adjusted by the flow rates of shell and core solutions. In order to obtain 3D-LLTO-NT with fine hollow structures, the applied voltage of 15 kV and the feeding rate ratio of the core/shell feed solution of 1:1 were used in this research. The viscosity of the precursor solution was regulated by the concentration of polyvinyl pyrrolidone (PVP). The concentrations of PVP in core/ shell solutions were 21 wt.% and 10.5 wt.%, respectively. The pristine fibers were solid and the core/shell structures were not noticeable. They were uniform with an average diameter of about 900 nm. After calcined at 800°C for 2 h, the core/shell hollow structures were observed as shown in Figure 2b,c,e,f. It was confirmed in Figure 2b,e that coaxial nanofibers had turned into nanotube structures. After calcination, the PVP was burned out and only the shells LLTO remained. The average diameter of shell's inner holes was about 600 nm and the thickness of the shell was about 125 nm. In Figure 2c,f, the closing SEM and TEM observation showed that the shell of 3D-LLTO-NT was made up of aggregation of LLTO nanoparticles. BET measurement results showed that the specific surface area of hollow LLTO nanotube and LLTO nanorod was 11.425 and 11.295 m 2 g À1 , respectively. The unobvious difference was due to the characteristic structure of LLTO, which was composed of nanosphere. However, the hollow LLTO nanotubes showed a lower compaction density (0.39 g cm À3 ) than solid LLTO nanorod (0.75 g cm À3 ). In other words, in the case of the same mass, hollow LLTO nanotubes showed larger contact area than nanorod, increasing the lithium ion number in the interface between polymer electrolyte and inorganic solid-state electrolyte.
In our previous study, solid LLTO nanofibers with mass loading of 5 wt.% was enough to achieve a maximum lithium ionic conductivity. [25] Herein, the mass loading of 3D-LLTO-NT in current polymer composite electrolytes remained at 5 wt.% as well. The 3D-LLTO-NT filled PEC composite electrolyte film was fabricated with the solution casting. 3D-LLTO-NT displayed a better mechanical property compared to the solid LLTO nanofibers of the same diameter size. They were not damaged or broken into non-uniform particle aggregate fragments in the PEC/LiTFSI suspension under 500 W of ultrasonic dispersion or 700 rpm of mechanical stirring and remained tubular. This was consistent with the previous reports by others that the hollow-nanotubes, e.g., wood or metal tubes were robust against pressure than the solid one at the similar length and weight. The reason was probably due to that the hollownanotubes structure was lighter and the stress distribution of hollownanotubes structure was more effective than the solid ones. [33] Figure 3a showed the optical image of composite electrolyte membrane composed of 3D-LLTO-NT framework with PEC and 50 wt.% LiTFSI(3D-LLTO-NT@PEC). The membrane was flexible and robust. Figure 3b presented the SEM image of the membrane. The 3D-LLTO-NT was well compatible in PEC polymer matrix, demonstrating a good mechanical strength. Figure 3c showed the XRD patterns of 3D-LLTO-NT, PEC, PEC/ LiTFSI and 3D-LLTO-NT@PEC. The XRD patterns of 3D-LLTO-NT were well indexed to tetragonal phase structure of the Li 0.33 La 0.557 TiO 3 (JCPDS:01-87-0935). No characteristic diffraction peak of LiTFSI was observed in the XRD spectrum of PEC/LiTFSI electrolyte, implying that the LiTFSI had been solvated by PEC completely. In the XRD pattern of the 3D-LLTO-NT@PEC membrane, the peaks located at 32.7°, 46.9°a nd 58.3°were assigned to the 3D-LLTO-NT. The relative intensity of the peak of PEC did not change and no peak for LiTFSI was found, indicating that 3D-LLTO-NT addition had no adverse effect on the morphology of PEC and LiTFSI. Figure 3d revealed that the glass transition temperature (T g ) of PEC, PEC/LiTFSI and the 3D-LLTO-NT@PEC membrane. The T g value of PEC/LiTFSI was À13.59°C. While the T g value of 3D-LLTO-NT@PEC was reduced to À19.73°C. The T g shifted to a lower temperature indicated that the segments movement of the PEC was enhanced by the 3D-LLTO-NT addition. It is known that PEC has -OH end group and C=O group. There exist hydrogen bonding interactions between the -OH Energy Environ. Mater. 2023, 6, e12636 2 of 7 and C=O that will hinder the movement of the chain segments. 3D-LLTO-NT addition weakened the hydrogen bond reaction by separating the PEC chains and the segments movement become faster. As a result, the decreased T g was observed. This would be beneficial for the lithium-ion conduction.
The EIS curves of the SS/3D-LLTO-NT@PEC/SS at 25°C were shown in Figure 4a. The ionic conductivity of the 3D-LLTO-NT@PEC composite electrolyte was 1.73 × 10 À4 S cm À1 , twice the ionic conductivity of the 3D-LLTO-NT-free electrolyte (PEC/LiTFSI, 8.88 × 10 À5 S cm À1 ) at 25°C. When the temperature increased to 65°C, the ionic conductivity of the 3D-LLTO-NT@PEC composite electrolyte reached 1.57 × 10 À3 S cm À1 , while the ionic conductivity of the PEC/LiTFSI electrolyte was 7.78 × 10 À4 S cm À1 (seen in Figure 4b). The activated energy of ionic conduction of electrolytes was estimated using the Arrhenius equation (1): where T is the absolute temperature, E a is the active energy, and A is a pre-exponential factor. After the fitting analysis for the obtained results in Figure 4b, the corresponding activation energies of 3D-LLTO-NT@PEC is 0.21 eV, which was closed to that of PEC/LiTFSI electrolyte (0.20 eV). The hollow nanotube structures morphology will provide more lithium-ion transport paths both on the inner and outer surfaces of the nanotubes. Therefore, the addition of 3D-LLTO-NT, active ceramic fillers with nanotube morphology, can improves the ions conductivity in the electrolyte. Wide electrochemical stability window is one of the most important factors for the composite electrolytes in solid-state batteries. Figure 4c showed the LSV curve of PEC/LiTFSI and 3D-LLTO-NT@PEC electrolyte. The electrochemical stability window of the pristine PEC/LiTFSI was 4.5 V versus Li/Li + . The 3D-LLTO-NT@PEC exhibited a more stable electrochemical stability with the window up to 5.0 V, which has greater potential to match high-voltage lithium battery applications. The decomposition of polymer usually started from the terminated group. The terminated hydroxyl group will accelerate the oxidation decomposition of PEC. The addition of Lewis basic LLTO can decrease the content of -OH in PEC, improving the oxidative resistance of electrolyte. In addition, the wide potential window of LLTO (6.2 V vs Li/ Li + ) will also broaden the oxidation decomposition potential of the designed composited electrolyte. [34] Moreover, the hydrophilic  inorganic 3D-LLTO-NT can absorb the trace moisture in the PEC matrix avoiding the oxidation reaction on the electrodes. [35] The mechanical properties are also very important for the solid electrolytes. [36] Figure S3, Supporting Information, depicted the tensile strength of the PEC/LiTFSI electrolyte and 3D-LLTO-NT@PEC composite electrolyte. The PEC/LiTFSI membrane was sticky and soft with the tensile strength of 1.19 MPa and elongation at breaking of 1168%. The membrane was too weak to be self-standing. After adding 3D-LLTO-NT frameworks into PEC, the tensile strength of 3D-LLTO-NT@PEC increased to 1.83 MPa and elongation at breaking increased to 532%. The obtained electrolyte coupled hardness with softness: PEC acted as soft component; while, 3D-LLTO-NT acted as hard component. The specific design endowed the composited electrolyte with an excellent mechanical property. Therefore, adding 3D-LLTO-NT frameworks into PEC made the membrane more robust and be a selfstanding membrane.
For high performance polymer electrolytes in solid-state Li batteries, higher ionic conductivity and Li ion transference number (t Li  Figure 4d, the t Li + value of 3D-LLTO-NT@PEC electrolyte is 0.78, much higher than the average level (0.2-0.5) of the polymer electrolytes. [37,38] The high lithium-ion transfer number was not only contributed by the nanotubes (3D-LLTO-NT), but also by the high lithium salt content in the polymer matrix. Unlike PEO-based electrolytes, the PEC-based composite electrolytes showed a rather high t Li + with the addition of high lithium salt content. In the PEC-based electrolyte system, most anions of lithium salts existed as aggregates at high concentrations. [11] The interaction between C=O groups and Li + in the aggregate state was weakened by contacts between Li + and anions. This coordination structure allowed the electrolytes compositions prepared with higher concentration LiTFSI to obtain high t Li

+
. On the basis of PEC-based electrolyte with high lithium salt content, the addition of conductive nanotube fillers, 3D-LLTO-NT, could further increase t Li + . The 3D-LLTO-NT was quasi one-dimensional fast ionic conductor, could promote the migration of Li + in the electrolyte film. Second, the enhanced cationic conduction was also believed to be due to the large active interface between the polymer and the nanotube surface. Moreover, the addition of 3D-LLTO-NT adsorbed anions, which could restrain their migration and promote the movement of Li + . This special 3D lithium-ion transport channel synergistically enabled fast Li + transport and high t Li Polymer electrolytes usually have a good interfacial compatibility with lithium metal electrode. To verify this conclusion, a Li/electrolyte/ Li symmetrical cell was assembled and examined by galvanostatic   Figure S4, Supporting Information). The decreased polarization potential was probably due to the activation of interface between electrolyte and lithium electrode. Unsurprisingly, no short circuit was found in 280 h. The lithium ion-conducted LLTO in PEC based electrolyte can disperse lithium ion flux, avoiding the local aggregation of lithium ion. In addition, the rigid LLTO can suppress the growth of lithium dendrite.
Although 3D-LLTO-NT@PEC electrolyte possessed a high oxidation decomposition potential of 5 V, the active electrode would induce the oxidation decomposition of electrolyte. For example, singlet oxygen and free radicals can cause the chemical degradation of electrolytes. [39] Hence, LiFePO 4 was used as the active material, avoiding the battery deterioration caused by the inductive effect of cathode. The galvanostatic charge/discharge profile of the solid lithium batteries were shown in Figure 5a. The discharging capacity of the LiFePO 4 /Li cell using 3D-LLTO-NT@PEC electrolyte at room temperature reached to 140.2 mAh g À1 , which was higher than that of PEC/LiTFSI (132.5 mAh g À1 ) at 0.1 C. This was due to high conductivity of 3D-LLTO-NT@PEC electrolyte. Additionally, LiFePO 4 /PEC/LiTFSI/Li battery showed an unusual behavior cycling curves below 3 V due to the decomposition of PEC on lithium metal. The discharge curves of LiFePO 4 /PEC/LiTFSI/Li gradually got back to normal after five cycles, which was maybe due to that the decomposed PEC participated in the formation of SEI on lithium metal electrode ( Figure S5, Supporting Information).The induce of 3D-LLTO-NT would limit this reaction at the interface by physical obstruction. It can be seen in Figure 5b that the LiFePO 4 /Li cell using 3D-LLTO-NT@PEC electrolyte achieved a Coulombic efficiency of 99% after 200 cycles 0.1 C at room temperature. So, the obtained 3D-LLTO-NT@PEC polymer electrolyte showed a stable charging/discharging performance and satisfying cycling performance in solid state LiFePO 4 / Li battery at ambient temperature.

Conclusion
In summary, the active ceramic 3D-LLTO-NT frameworks have been successfully fabricated by coaxial electro-spinning process. It is worth noting that a remarkable improvement in the electrochemical performance and tensile strength is achieved in solid PEC/LiTFSI polymer electrolyte. The 3D-LLTO-NT@PEC shows an ionic conductivity of 1.73 × 10 À4 S cm À1 at room temperature and has a surprisingly high t Li + value of 0.78. The Li + transference number is much higher than that of typical electrolytes (0.2-0.5). The 3D-LLTO-NT@PEC composite electrolyte also exhibits a stable electrochemical window up to 5.0 V versus Li/Li + . The solidstate battery LiFePO 4 /3D-LLTO-NT@PEC/Li delivered a specific capacity of 140.2 mAh g À1 at 0.1 C. The 3D ceramic nanotube framework provides a fast Li + transfer pathway and a stable interface, resulting in this excellent performance of the composite electrolyte.
The coaxial spinneret (NanoNc Co. Ltd., Seoul, Korea) was made up of two stainless-steel needles, the inner needle had an outer diameter of 0.51 mm (25gauge) and the outer needle had an inner diameter of PVP 6 mm (18-gauge). The injecting rate of both precursor solutions was at 0.4 mL h À1 . The voltage between the spinneret and the collector was 15 kV and the tip-to-collector distance was 15 cm. The as-spun fibers were collected on the aluminum foils. The as-spun fibers were dried in a vacuum oven at 60°C more than 24 h, and then was transferred to a muffle furnace and calcined at 800°C for 2 h in air with a heating rate of 5°C min À1 to obtain the 3D-LLTO-NT.
Preparation of composite electrolytes: The mixture of PEC (M n = 1.5 × 10 5 , M w /M n = 2.7 from Gel Permeation Chromatography) and bis (trifluoromethane) sulfonamide lithium (LiTFSI, DoDoChem) with a mass ratio of 1:1 was dissolved in acetonitrile (CH 3 CN, Sigma-Aldrich) under stirring for 12 h to give a transparent solution. The 3D-LLTO-NT was added to above solution with 5 wt.% fraction (based on the PEC) and then sonicated for 10 min. The resulting slurry was cast onto a poly(tetrafluoroethylene) (PTFE) plate and heated in an oven filled with N 2 at 60°C for 2 h to remove the solvent, finally dried in a vacuum oven at 60°C for 12 h to obtain the composite electrolyte membranes. In this process, methylene chloride and ethylene carbonate will volatilize during the drying process. LiTFSI was purchased from DoDoChem with a battery grade. Moreover, the obtained composited electrolyte with acetonitrile was vacuum-dried at 60°C for 12 h. The electrolyte can be regarded as solvent free. The average thickness of the electrolytes is about 250-400 μm.
Samples characterization: The morphology of the 3D-LLTO-NT and 3D-LLTO-NT@PEC composite electrolyte membranes were observed by scanning electron microscope (SEM, SU8010, Hitachi, Japan) at 3-15 kV and transmission electron microscope (TEM, JEM-2100HR). X-ray diffraction (XRD) measurements was performed with X-ray diffractometer (X'Pert Powder PANalytical, Netherlands) with CuKa radiation (1.5418Å) at a scan rate (2θ) of 5°C min À1 . The glass transition temperature (T g ) of the samples were evaluated by a differential scanning calorimeter analyzer (DSC Q20, TA, USA) from À50°C and 50°C at a heating rate of 10°C min À1 under N 2 atmosphere.
Electrochemical impedance spectroscopy (EIS) was performed by a multichannel electrochemical workstation (VSP-300, Bio Logic, France) at various temperatures over the frequency range from 500 kHz to 100 mHz with an amplitude of 10 mV at the open-circuit voltage (OCV). The ionic conductivity was measured after sandwiching polymer composite electrolyte between two stainless-steel (SS) blocking electrodes. The linear sweep voltammetry (LSV) was recorded at a scan rate of 1 mV s À1 from 2.0 to 6.0 V. The electrochemical window of the samples was examined after sandwiching them in a model having the structure SS/electrolyte/Li. The tensile test was measured using Universal Testing Machine (SANS TAS-10, Suns, China) according to standard GB 13022-1991. [40] The EIS and chronoamperometry (CA) of the Li/electrolyte/Li symmetrical cells were performed by a multichannel electrochemical workstation (VSP-300, Bio Logic, France) to determine the lithium-ion transference number (t Li+ ) of the composite samples. The values of the t Li+ is estimated from equation 2: [41,42] where I 0 is the initial current, I s represents the steady current. R 0 and R s are the impedance at the initial and steady states, respectively. ΔV is the applied DC polarization potential voltage (10 mV). The solid lithium batteries (coin type 2032) were assembled with the composite electrolyte, lithium foil anode and composite cathode. The composite cathode slurry was prepared by blending LiFePO 4 (Hefei Kejing Material Technology Co., China), acetylene black (Hefei Kejing Material Technology Co., China) and the binder LA-132 (Chengdu Indigo Power Sources Co., China) at a ratio of weight of 80:10:10 in acetonitrile and was cast on the Al current collector. The mass loading of active material LiFePO 4 was about 1.10 mg cm À2 . The galvanostatic charging/ discharging properties and cycling performance of the battery was performed at battery testers CT3001A system (Wuhan LAND Electronic Co., China) at 25°C with 0.1 C rates within a voltage range of 2.5-4.0 V. The battery was kept at 50°C for 2 h and at 25°C for another 10 h before the testing.