Design of a polymer electrolyte membrane for enhanced zinc anode stability in reversible aqueous zinc-ion batteries

INTRODUCTION

The heightened prevalence of environmental pollution and escalating energy consumption have spurred the dynamic advancement of clean and renewable energy technologies. However, the intermittent nature and inherent variability of wind and solar energy have posed obstacles to their seamless integration into electrical infrastructure. Consequently, the storage and conversion of energy necessitate secondary batteries endowed with superior performance characteristics to ensure the consistent delivery of clean power. Among the widely discussed secondary batteries, zinc-ion batteries (ZIBs), utilizing a blend of pure water and zinc salts as electrolytes, have emerged as a promising contender for large-scale energy storage and mobile power applications. This is attributed to their notable features of being low-cost, high-energy, highly safe, and environmentally friendly^[1-6]. However, the utilization of ZIBs is significantly impeded by the instability of zinc metal anodes during the plating/stripping process^[7], primarily attributable to zinc dendrites, zinc corrosion, and side reactions occurring on the anode^[8-10]. Zinc dendrites, characterized by a high Young’s modulus, proliferate with cycling, introducing dead zinc metal that extends to the cathode and induces short circuits in the battery. Conversely, pure zinc metal is prone to corrosion when exposed to salt solutions, leading to its degradation and the accumulation of byproducts on the anode surface. This circumstance severely hampers the zinc plating/stripping process and diminishes the cycling lifespan of ZIBs. To address these challenges and enhance the reversibility of the anode in ZIBs, numerous effective strategies have been proposed. These include the incorporation of additives into the electrolyte^[5,11,12], the fabrication of semi/all-solid-state electrolytes^[13-17], the construction of layered protection around the anodes^[10,18,19], the design of intricate anode structures^[20-24], and the utilization of polymer electrolytes^[25-27]. Among these approaches, the utilization of polymer electrolytes stands out as the simplest and most effective method to enhance the electrochemical performance of ZIBs. Polymer electrolytes offer numerous advantages^[28,29], including higher electrolyte-confinement capability^[30], superior uniform distribution of ions in solution, easily controlled thickness, broader operating temperatures, and better interface compatibility^[28] compared to aqueous electrolytes. Furthermore, owing to their cross-linked structure and functional groups, the electrochemical properties of the polymer can be readily optimized through the design of unique structures, incorporation of functional groups, and combination with other materials^[29,31].

In order to mitigate zinc dendrite formation and enhance ion conductivity in ZIBs using optimized polymer electrolytes, researchers have proposed and designed a series of polymer electrolytes. These include poly (ethylene oxide)^[32], xanthan gum^[33], gelatin^[34], polyacrylamide^[35], carboxymethyl cellulose sodium^[36], and polyvinyl alcohol^[37,38]. It is noteworthy that the aforementioned polymer electrolytes have demonstrated excellent electrochemical performance with significant implications. However, the preparation processes are complicated and often yield numerous byproducts. Therefore, there is an urgent need to develop high-performance ZIBs by utilizing polymer electrolytes based on sustainable materials through easily implementable methods.

Cellulose, a renewable biomaterial composed of nanofibers made of hexose units, possesses exceptional hydrophilicity and moisture retention due to the abundance of hydroxyl groups along its lengthy chains. Additionally, it boasts advantages such as low cost, sustainability, biodegradability, chemical stability, and widespread availability as a natural polymer, rendering it a promising candidate for high-performance electrolytes in ZIB applications^[39-43]. However, the pristine cellulose-based electrolytes in ZIBs exhibit poor performance and low capacity during cycling, primarily due to their inability to prevent the formation of zinc dendrites on the anode and the dissolution of cathode materials.

In this study, a polymer electrolyte membrane (TAC) was synthesized by incorporating tannic acid (TA) into cellulose, facilitating the formation of ether bonds and introducing numerous phenolic hydroxyl groups. Compared to pristine cellulose, TAC exhibits enhanced liquid retention capacity, effectively mitigating cathode material dissolution and non-reactive parts corrosion of zinc foil. Moreover, the lower binding energy (E_b) of phenolic hydroxyl groups and zinc ions in TAC, compared to pristine cellulose, significantly inhibits zinc dendrite growth on the anode, resulting in reduced desolvation energy and accelerated uniform zinc ion deposition. These findings were corroborated through thermodynamic calculations and Zn-MnO₂ battery test. Notably, a battery employing 1% TAC demonstrated impressive performance metrics, maintaining a high discharge capacity retention of 83.1% after 1,000 cycles at a current density of 5 C, coupled with exceptional reversible ability (99.9% coulombic efficiency) and prolonged lifespan exceeding 2,300 h at a current density of 0.5 mA cm^-2. This innovative design not only provides effective strategies for fabricating high-performance ZIBs but also holds promise for broader applications in secondary batteries. Hence, as depicted in Figure 1A, TAC was synthesized by introducing TA into the cellulose solution. In brief, a homogeneous cellulose solution was blended with TA and reacted in pure water for approximately 30 min. The reaction between cellulose and TA resulted in the formation of a polymer electrolyte membrane, which was subsequently vacuum-filtered to produce a film with abundant phenolic hydroxyl groups distributed on the surface of the fibers. Figure 1B illustrates the schematic diagram for protecting the Zn anode. It is noteworthy that due to the stronger attractive interaction between the phenolic hydroxyl groups in TAC and Zn²⁺, the flux of zinc ions was effectively and uniformly regulated. Consequently, in the absence of TA, the battery produced vertical zinc dendrites after extended cycling. Conversely, in the presence of TA, zinc deposition and stripping occurred uniformly, resulting in horizontally stacked zinc layers without zinc dendrites.

Figure 1. Schematic of (A) the fabrication process of TAC; (B) the protection diagram for Zn anode.

EXPERIMENTAL

Preparation of TAC: X% (X = 0, 0.5, 1, 3) TAC was prepared as follows: 2 g bacterial cellulose (solid content = 70%, Guilin Qihong Technology Co. Ltd) was dispersed in 100 mL pure water through magnetic stirring for 30 min; then, X g (X = 0, 0.5, 1, 3) TA (98%, Macklin) was dissolved in abovementioned mixture and stirred for 30 min, respectively, which formed TAC films by vacuum filtration. TAC films were dried at 80 °C until constant weight. According to the mass of the TA (X = 0, 0.5, 1, 3), the samples was named as X% (X = 0, 0.5, 1, 3) TAC, respectively.

MnO₂/carbon nanotube (CNT) material was synthesized referencing a previous work with a slight modification^[44]; the detail of experiment was as follows: 0.15 g multi-walled CNT (diameter: 5-15 nm, length: 10-30 μm, Nanjing XFNANO Materials Tech Co., Ltd) with hydroxylation was dispersed in 150 mL pure water and 2.03 g MnSO₄ (AR grade, Aladdin) was added into the above solution stirring for 30 min. 1.27 g KMnO₄ (AR grade, Alfa) was dispersed into 80 mL pure water stirring for 30 min and added into the above solution slowly stirring for 10 min. Then, the mixture solution was transferred to an autoclave heating for 12 h at 120 °C and obtained a dark brown powder through filtering and drying.

For preparing cathode electrode, MnO₂/CNT powder, super P carbon black (TIMCAL), and poly(vinylidene fluoride) binder(Arkema) (weight ratio of 7:2:1) were mixed in N-Methylpyrrolidone solvent (99.9%, Aladdin), stirred for 1 h, and coated on a piece of single-walled CNTs paper (Nanjing XFNANO Materials Tech Co., Ltd). Then it was dried in a vacuum at 70 °C and cut into electrodes with a diameter of 9 mm and a mass loading of 0.6-1.5 mg MnO₂. Zinc foil (purity > 99.98%, Alfa) was polished using sandpaper and used as anode.

Characterization: The surface morphology of electrode and TAC materials was revealed using a scanning electron microscope (SEM) (Zeiss Sigma 300). The functional groups and chemical states of TAC materials were investigated by X-ray photoelectron spectroscopy (XPS) (Thermo Scientific K-Alpha) and Fourier transform infrared spectroscopy (FTIR) spectra (Thermo Scientific Nicolet iS5). An electronic universal testing machine (MTS CMT6103) and optical contact angle measuring system (KRUSS DSA100) were utilized to evaluate the mechanical properties and wettability of TAC materials. The thermal curves were performed by simultaneous thermal analyzer (NETZSCH STA 449F5) at 10 °C min^-1 under nitrogen atmosphere.

Electrochemical measurements: Cyclic voltammetry (CV) curves, linear sweeping voltammetry (LSV), and electrochemical impedance spectroscopy (EIS) were tested using an Admiral potentiostats (Squidstat Plus). The CV and LSV were conducted using polished Zn foil and obtained MnO₂/CNT material corresponding to the counter electrode, the reference electrode, and the work electrode, respectively. Electrochemical properties of Zn-MnO₂ batteries (voltage range of 0.8-1.85 V) and Zn||Zn batteries were tested using CR2032-type coin cells and 2001A battery testing system at 25 °C. Before the X% TAC materials were used as polymer electrolyte membranes, a small amount of 2 M ZnSO₄ mixed with 0.1 M MnSO₄, 2 M Zn(CF₃SO₃)₂ mixed with 0.1 M Mn(CF₃SO₃)₂, 2 M Zn(OAc)₂, or 2 M ZnCl₂ was added into the X% TAC materials. Additionally, except where otherwise stated, all the electrochemical results are gathered using 2 M ZnSO₄ mixed with 0.1 M MnSO₄ as the electrolyte.

The liquid retention capacity of TAC materials under 5 MPa was determined by

where M₀ is the weight of TAC before soaked in the mixture solution of 2 M ZnSO₄ + 0.1 M MnSO₄. M₁ is the weight of TAC after being soaked in the mixture solution under 5 MPa.

The test of liquid retention rates with 5 MPa: First, 0%, 0.5%, 1%, and 3% TAC were cut into samples with 19 mm diameter and soaked in the liquid electrolyte with 2 M ZnSO₄ + 0.1 M MnSO₄ for 6 h. The weights of abovementioned polymer electrolytes were recorded and named as M₀ and M₁, where M₀ is the weight of TAC before soaked in the mixture solution of 2 M ZnSO₄ + 0.1 M MnSO₄. M₁ is the weight of TAC after being soaked in the mixture solution under pressure of 5 MPa. The hollow diameter size of bibulous paper is larger than 19 mm (by 1-2 mm) avoiding the polymer electrolyte contact with bibulous directly.

DFT calculation: The first principal calculation was conducted in the computational chemistry package of Gaussian 16 program^[45] based on the density functional theory (DFT). The optimized structures without virtual frequency were acquired with the B3LYP functional^[46]. The basis set 6-311+G (d, p)^[47] with D3 (BJ)^[48] was utilized for C, H and O atoms while the SDD basis set^[49,50] with an effective core potential (ECP) was employed for Zn atoms. The impact of the water solvent environment on all the systems was taken into consideration on the basis of the IEFPCM model^[51] for structure optimization. The Multiwfn^[52-55] and VMD^[56] program were utilized in the calculation and analysis of electrostatic potential (ESP) on molecule surface. The binding energies of all systems were calculated under the M06-2X-D3/def2-TZVP^[57-59] level and the ECP was also used for Zn atoms. The solvation model density (SMD) model^[60] was taken into consideration in the calculation of binding energies. The E_b for system was defined according to

where E_total, E₂, and E₁ stand for the energy for the interaction system, the different substrate materials (0% TAC, 1% TAC), and Zn²⁺, respectively.

RESULTS AND DISCUSSION

Optical images illustrating TAC concentrations of 0%, 0.5%, 1%, and 3% are shown in Supplementary Figure 1A-D. Relative to the optically clear 0% TAC, the transparency of X% TAC (where X = 0.5, 1, 3) decreased proportionally with increasing TA content. Scanning electron microscopy (SEM) was employed to investigate the microstructure of TAC, revealing a three-dimensional porous network comprising randomly oriented nanofibers of varying diameters, as depicted in Figure 2A-D. While 0% TAC showed numerous irregular pores, the introduction of TA resulted in composite TAC structures featuring well-defined pores and a dense fiber arrangement. Notably, 1% TAC exhibited fewer pores and tightly interconnected fibers compared to both 0.5% and 3% TAC. This characteristic is advantageous for suppressing the formation of zinc dendrites during the plating/stripping process^[61]. XPS was utilized to probe the chemical bonds present on the surface of TAC. Analysis via XPS revealed the presence of various functional groups such as C-C, C-OH, and C-O-C, identified by their respective binding energies of 284.8, 286.5, and 288 eV^[62,63], as illustrated in Figure 2E.

Figure 2. SEM images of (A) 0% TAC, (B) 0.5% TAC, (C) 1% TAC, and (D) 3% TAC; (E) XPS of C 1s of the TAC samples; (F) FTIR spectra of TAC samples; (G) Galvanostatic Zn stripping and plating in a Zn||Zn cell using TAC under different current density (0.1, 0.2, 0.5, 1, and 2 mA cm^-2) with 1 mAh cm^-2 capacity limitation; Potential profiles of Zn/Zn cell under (H) 0.1, (I) 0.2, (J) 0.5, and (K) 1 mA cm^-2 for a specified time range with a capacity limitation of 1 mAh cm^-2.

Upon the addition of TA, there was an observable increase in the C-O-C content of TAC, suggesting a reaction between TA and the oxygen groups within the cellulose matrix [Supplementary Table 1]. Furthermore, comparison of the infrared spectroscopy (IR) spectrum of 0% TAC with those of 0.5%, 1%, and 3% TAC revealed prominent absorption peaks around ~1,700 and 1,600 cm^-1, indicative of C=O and aromatic skeletal vibrations^[64,65]. This observation suggests the introduction of TA onto the cellulose surface, as shown in Figure 2F. Interestingly, despite having different TA contact angles (measured at 40.7°, 68.9°, 46.5°, and 35.4°, for 0%, 0.5%, 1%, and 3% TA, respectively, as illustrated in Supplementary Figure 2, all TAC samples demonstrated comparable levels of hydrophilic performance. Supplementary Figure 3A depicts the test results regarding the liquid retention capacity of TAC materials at concentrations of 0%, 0.5%, 1%, and 3% under a pressure of 5 MPa. It is evident that the liquid retention capacity increases with higher TA content, indicating that TA-modified cellulose exhibits enhanced capacity in retaining liquid electrolytes, particularly under elevated pressure conditions, as illustrated in Supplementary Figure 3B. This heightened liquid retention capability serves to shield the zinc foil anode from corrosion by the liquid electrolyte. Notably, during charging and discharging at a current density of 5 C for approximately 36 h, the rear surface of the zinc foil in batteries using 1% TAC appeared notably smoother compared to those employing 0% TAC. The latter exhibited pronounced corrosion with numerous perforations in the non-reactive region, as evidenced in Supplementary Figure 4. To investigate the electrochemical characteristics of the polymer electrolyte membrane containing 0%, 0.5%, 1%, and 3% TAC, as depicted in Figure 2G, galvanostatic zinc plating/stripping processes were conducted within a Zn||Zn cell across various current densities ranging from 0.1 to 2 mA cm^-2, with a consistent capacity of 1 mAh cm^-2. The Zn||Zn cell incorporating 1% TAC displayed overpotential values of 32.81, 36.00, 46.33, and 55.94 mV at current densities of 0.1, 0.2, 0.5 and 1 mA cm^-2, respectively. This outcome suggests that 1% TAC could serve effectively as a polymer electrolyte membrane, enhancing the electrochemical performance of the zinc anode during the plating/stripping process, as delineated in Figure 2H-K. To comprehensively assess the electrochemical attributes of TAC materials, both 0% and 1% TAC were employed as polymer electrolyte membranes in Zn-MnO₂ batteries. Supplementary Figure 5A and B shows the structure and morphology of the positive electrode materials, characterized by small-sized particles (~200 nm) comprising nanotubes intermixed with MnO₂ nanosheets. Additionally, Figure 3A and B illustrates smoother, flatter surface with more uniform and denser fiber structures in 1% TAC compared to 0% TAC. These features are advantageous for mitigating zinc dendrite growth towards the positive electrode, thus averting internal short circuits within the battery^[66]. The introduction of 1% TA resulted in the formation of interwoven tortuous nanosized pores within denser fibers, a notable improvement over the 0% TAC shown in Figure 3C and D. This structural modification led to a reduction in pore size. Exothermic data, as shown in Supplementary Figure 5C, and collected via differential scanning calorimetry (DSC), revealed specific peaks at 249, 362, 346 and 352 °C corresponding to pristine cellulose, TA, TA/pristine cellulose composite, and 1% TAC, respectively. Compared to the TA/pristine cellulose composite, the exothermic peak of 1% TAC exhibited reduced intensity, lower density, and narrower peak width, indicative of a reaction between TA and cellulose^[67,68].

Figure 3. SEM images of (A) 10 μm and (B) 200 nm of 1% TAC, (C) 10 μm and (D) 200 nm of 0% TAC; SEM images of Zn anode in Zn||Zn symmetry cells after 700 h plating/stripping cycles in ZnSO₄ using (E) 10 μm and (F) 500 nm of 1% TAC, (G) 10 μm and (H) 500 nm of 0% TAC at 0.5 mA cm^-2; (I) Galvanostatic Zn stripping and plating in a Zn||Zn cell using TAC with 2 M ZnSO₄ as the electrolyte under current density of 0.5 mA cm^-2 with 1 mAh cm^-2 capacity limitation; Potential profiles of Zn||Zn cell plating/stripping cycles using 1% TAC at (J) the first, (K) the 100th, (L) 200th, (M) 400th, (N) 500th with 1 mAh cm^-2 capacity limitation.

The electrochemical stability of 0% and 1% TAC was investigated through LSV curves, employing a limited voltage of 2.5 V and a scan rate of 0.1 mV s^-1, as shown in Supplementary Figure 5D. It was observed that the current profiles of 0% and 1% TAC in Zn-MnO₂ batteries remained stable below 1.54 and 1.90 V, respectively. This suggests that 1% TAC exhibits superior stability, particularly at higher voltage. Furthermore, the stability and electrochemical performances of TAC materials and both positive and negative electrodes were assessed using EIS and CV, as illustrated in Supplementary Figure 6A-D. Notably, in the context of Zn-MnO₂ batteries, 1% TAC demonstrated enhanced electrochemical stability and interfacial compatibility compared to 0% TAC. To explore the protective role of TAC on the Zn anode, Zinc foils were employed as electrodes in Zn||Zn batteries, utilizing both 0% and 1% TAC with a 2 M ZnSO₄ electrolyte. These batteries underwent plating and stripping cycles for 700 h at a current density of 0.5 mA cm^-2. The zinc foils displayed notable differences in anode surface morphology, clearly observed in the SEM images shown in Figure 3E-H. In comparison to the pristine zinc foil [Supplementary Figure 7], the anode foil from the battery that used 0% TAC exhibited a coarse surface with abundant flake-like zinc dendrites, as shown in Figure 3G. Furthermore, these dendrites featured cracking and irregular protuberances, leading to discontinuous morphology as shown in Figure 3H, indicating potential short-circuiting and premature battery failure due to dendrite growth with cycling. In contrast, that utilizing 1% TAC resulted a relatively smooth surface with densely packed nanostructures, as shown in Figure 3E and F, after approximately 700 h of plating/stripping process. To explore the cyclability of zinc electrodes containing different concentration of TAC mixed with ZnSO₄ at a current density of 0.5 mA cm^-2, electrochemical performance assessments were carried out using Zn||Zn symmetric cells. Results depicted in Figure 3I illustrate that after 700 h of cycling, the polarization voltages of cells utilizing 1% TAC were notably lower compared to those employing 0% TAC. This discrepancy was attributed to the side reaction occurring on the surface of the zinc foil in 0% TAC-containing cells, leading to the formation and accumulation of zinc dendrites with the increased cycles, as shown in Figure 3G. Upon cycling for 700 h at a limited capacity of 1 mAh cm^-2, the battery using 0% TAC showed an erratic polarization voltage behavior and eventual short circuits. Conversely, the battery using 1% TAC exhibited comparatively stable polarization voltage profiles and sustained functionality for over 2,300 h at a current density of 0.5 mA cm^-2. To monitor the evolution of polarization voltages over cycling, potential profiles of Zn||Zn cells utilizing 1% TAC were recorded at a limited capacity of 1 mAh cm^-2. Notably, the potentials observed after the first, 100th, 200th, 400th, and 500th cycles were 38.40, 39.80, 35.62, 45.76, and 42.96 mV, respectively [Figure 3J-N], indicating sustained stability of the zinc foil anode during the plating/stripping processes. Furthermore, to elucidate the intrinsic advantages of 1% TAC over 0% TAC, the influence of cellulosic hydroxyl and phenolic hydroxyl groups on the solvation structure of zinc ions was investigated via DFT calculations, as shown in Supplementary Figure 8. Comparative analysis between 1% TAC and 0% TAC is presented in Figure 4. Figure 4A shows the typical electrochemical reactions of Zn²⁺ in the batteries that used 0% TAC or 1% TAC as a polymer electrolyte membrane on the interface of the zinc foil anode. Zn²⁺ interacted with cellulose modified with or without TA. As shown in Supplementary Figure 9, the E_b of 0% TAC-Zn²⁺ and 1% TAC-Zn²⁺ was -0.2898 and -0.5713 eV, respectively. It was established that the participation of phenolic hydroxyl from TAC is conducive to the desolvation and deposition of zinc ions on the zinc foil anode^[69,70]. Figure 4B displayed energy gap between the highest occupied molecular orbital (HOMO) and the lowest unoccupied molecular orbital (LUMO) of 1% TAC and 0% TAC. Obviously, 1% TAC exhibited a lower energy gap than 0% TAC, which implied better electrical conductivity of 1% TAC than 0% TAC^[71,72], in accordance with the abovementioned conclusion from the summary energies. Figure 4C and D showed the ESP distribution on the Van der Waals surface of 1% TAC and 0% TAC, respectively. In comparison to larger pink and white regions with positive ESP and reactive sites of hydroxyl groups from 0% TAC, there were larger light blue and gentian blue regions with negative ESP of phenolic hydroxyl from 1% TAC, which provides more abundant active sites to induce the deposition and stripping of Zn^2+[70,73].

Figure 4. (A) Schematic of Zn²⁺ plating/stripping process by using 0% TAC or 1% TAC; (B) LUMO and HOMO gap comparison between 1% TAC and 0% TAC; electrostatic potential (ESP) distribution on the Van der Waals surface of (C) 0% TAC and (D) 1% TAC.

Hence, the maxima ESP value of 1% TAC with Zn²⁺ was reduced to 395.1561 kcal mol^-1, which was lower than that of 0% TAC with Zn²⁺ (440.0305 kcal mol^-1), as shown in Supplementary Figure 10. Therefore, in comparison with 0% TAC, the easier interaction and stronger attraction interaction between 1% TAC and Zn²⁺ indicated that the desolvation process of Zn²⁺ with 1% TAC was easier than that with 0% TAC. Less energy was consumed during the process when zinc ions left the electrolyte to move to the interface of the anode electrode, which made the reaction process easier and boosted the uniform distribution of zinc ions on the zinc foil electrode. As a result of the weaker attraction interaction between 0% TAC and Zn²⁺, zinc ions were nonuniformly distributed on the surface of the anode electrode and a large number of dendrites were found, as shown in Figure 4A. On the other hand, 1% TAC possessed abundant hydroxyl and π-π stacking that exhibited high interaction with H₂O molecules, which ensured excellent liquid electrolyte retention of the 1% TAC at 5 MPa. In addition, in order to assess the compatibility of the TAC with different electrolytes, galvanostatic Zn stripping and plating in a Zn||Zn cell using TAC was conducted in various electrolytes, as shown in Figure 5. After long cycles of the Zn foil with 1% TAC in Zn(CF₃SO₃)₂ as shown in Figure 5A, abundant regular bulks composed of stacked flakes can be observed. while in Figure 5B from the battery using 0% TAC, amounts of vertically stacked flakes interlaced each other and a mass of tiny particles dispersed on the surface, indicating zinc dendrites and other byproducts from side reactions during the cycles. Similarly, from the Zn anode with 1% TAC in Zn(OAc)₂ shown in Figure 5C, horizontally stacked flakes in small size interweaved and stayed close together; on the contrary from 0% TAC in Figure 5D, massive vertical bulks connected with each other, which is the same as the one from ZnCl₂ in Figure 5F. In Figure 5E from 1% TAC in ZnCl₂, irregular flakes horizontally stacked and formed multi-hierarchy structure. The above results demonstrated that 1% TAC induced uniform distribution of Zn ions, which effectively inhibited the growth of zinc dendrites.

Figure 5. SEM images of Zn anode in Zn|| Zn cells after long cycles in different electrolytes: (A) 1% TAC and (B) 0% TAC in Zn(CF₃SO₃)₂, (C) 1% TAC and (D) 0% TAC in Zn(OAc)₂, and (E) 1% TAC and (F) 0% TAC in ZnCl₂; Galvanostatic Zn stripping and plating in a Zn||Zn cell using TAC with 1 mAh cm^-2 capacity limitation in the electrolyte of (G) Zn(CF₃SO₃)₂ under current density of 0.5 mA cm^-2, and (H) Zn(OAc)₂ under current density of 2 mA cm^-2.

Hence, the cycle performance of 1% TAC is prior to that of 0% TAC, which was displayed in Figure 5G for Zn(CF₃SO₃)₂ and Figure 5H for Zn(OAc)₂. Concretely, when in Zn(OAc)₂ at 2 mA cm^-2, 1% TAC nearly stably cycles for 1,200 h while the battery using 0% TAC shot-circuited within 100 h; when used in Zn(CF₃SO₃)₂ at 0.5 mA cm^-2, 1% TAC consistently steadily work for 2,800 h without degradation while 0% TAC failed in short time. Besides, when utilized in ZnCl₂ at 0.5 mA cm^-2 as shown in Supplementary Figure 11, 1% TAC performed steadily for 420 h; inversely, the battery using 0% TAC failed within 75 h. In short, 1% TAC with outstanding performances and compatibility could be used as a polymer electrolyte membrane in different electrolytes for ZIBs.

In addition, the CV curves of the Zn-MnO₂ batteries that used 0% TAC or 1% TAC were investigated, as shown in Figure 6A. Both CV profiles presented similar redox reaction pairs, indicating that 1% TAC did not influence the intercalation/deintercalation process of Zn²⁺. To study the electrochemical properties of the TAC, CNT paper with MnO₂, 0% TAC or 1% TAC, and zinc foil, were used as the positive electrode, polymer electrolyte membrane, and negative electrode assembled in Zn-MnO₂ batteries, respectively. To investigate the adaptability of 1% TAC, the electrochemical performance of the assembled battery was investigated in ZnSO₄, as shown in Figure 6B-D and Zn(CF₃SO₃)₂, as shown in Figure 6E-G, as the matrix of the aqueous electrolyte, respectively. As shown in Figure 6B, the battery using 1% TAC displayed a higher rate capability than that using 0% TAC. This was because 1% TAC possessed abundant phenolic hydroxyls, which modified the solvation environment of Zn²⁺ ions, weakened the solvation-free energy, and ultimately lowered the reaction barrier for the deposition and insertion of zinc ions^[74]. The Zn-MnO₂ battery using 1% TAC with ZnSO₄ exhibited excellent discharge capacities of 288.1, 254.7, 221.5, 202.0, 188.3, and 178.1 mAh g^-1, which were higher than those used the 0% TAC at current densities of 0.5, 1, 2, 3, 4, and 5 C. This was due to the lower solvation-free energy and the faster redox reactions of the positive material. Moreover, the coulomb efficiencies of 1% TAC and 0% TAC approached 99.9% at various current densities. When the assembled batteries were operated at a current density of 1 C, the battery using 1% TAC with ZnSO₄ exhibited a higher discharge capacity retention of 92.60% and a higher capacity of 33.33% after 100 cycles in comparison to that using 0% TAC [Figure 6C]. When the current rate was increased to 5 C, the cyclability of the Zn-MnO₂ battery was investigated using 1% TAC with ZnSO₄, as shown in Figure 6D. The first-cycle discharge capacity of the battery using 1% TAC was 178.4 mAh g^-1 which thereafter changed to 148.3 mAh g^-1 after 1,000 cycles. Thus, this battery exhibited a high retention capacity of 83.1% at a large current density of 5 C. The coulombic efficiency of the battery using 1% TAC reached 99.9% during 1,000 cycles, which was attributed to the strong inhibitory effect of the 1% TAC on the side reaction and the formation of zinc dendrites. When evaluating the applicability of 1% TAC in different electrolytes, the assembled battery using 1% TAC with Zn(CF₃SO₃)₂ exhibited a high discharge capacity of 265.2 mAh g^-1 at a rate of 0.5 C, as shown in Figure 6E. When the current rate was raised to 1, 2, 3, 4, and 5 C, discharge capacities of 231.1, 192.4, 171.6, 153.5, and 137.4 mAh g^-1, could be maintained, respectively. However, the Zn-MnO₂ battery using 0% TAC exhibited discharge capacities of 163.1, 141.4, 118.7, 106.6, 98.2, and 92.1 mAh g^-1 at current densities of 0.5, 1, 2, 3, 4, and 5 C, respectively, confirming that modifying cellulose with TA could increase the capacity of batteries significantly. The cyclability of Zn-MnO₂ batteries using 0% TAC and 1% TAC were investigated at 1 C for 100 cycles; after 100 cycles, the charge capacity of 1% TAC and 0% TAC with Zn(CF₃SO₃)₂ were 174.2 and 131.3 mAh g^-1, respectively. Moreover, the discharge capacity of the Zn-MnO₂ battery using 0% TAC with Zn (CF₃SO₃)₂ exhibited sharp fluctuations at a current density of 1 C within the first 100 cycles as shown in Figure 6F, which was attributed to the side reactions resulting from the reaction-free electrolyte. To investigate the lifespan of the battery using 1% TAC with Zn(CF₃SO₃)₂, the battery was tested at a current density of 5 C for 1,000 cycles, as shown in Figure 6G. The discharge capacity retention maintained at 73.7% and the coulombic efficiency approached 99.9% after 1000 cycles. Thus, the battery using 1% TAC with ZnSO₄, or Zn(CF₃SO₃)₂ exhibited high capability rates and long-life cyclability, which were ascribed to the outstanding adaptability of 1% TAC to different liquid electrolyte components.

Figure 6. (A) CV curves of 0% TAC and 1% TAC at a scan rate of 1 mV s^-1; discharge capacity and the corresponding coulombic efficiency of the battery using ZnSO₄ at (B) different current densities, (C) 1 C, (D) 5 C and using 2 M Zn(CF₃SO₃)₂ as the electrolyte at (E) different current densities, (F) 1 C, (G) 5 C, respectively.

CONCLUSIONS

By regulating cellulose with TA, we obtained a polymer electrolyte membrane of 1% TAC, which possessed a three-dimensional porous network and dense structure with abundant phenolic hydroxyls. Additionally, the 1% TAC delivered a high liquid retention rate at 5 MPa and protected both the front and back surfaces of the zinc foil from corrosion by the electrolyte. Supported by the excellent electrochemical properties mentioned above and simulation calculations, this strategy has proven to be an efficient way to reduce the hydroxyl groups on the surface of cellulose and increase the quantity of phenolic hydroxyl groups, which suppressed the reaction of the polymer electrolyte and Zn²⁺, and subsequently demonstrated a lower desolvation energy during the plating/stripping process of Zn²⁺ as well as a lower overpotential. Remarkably, these advantages were beneficial to the uniform distribution of Zn²⁺ on the surface of the anode, the alleviated zinc dendrite, and the prolonged life of the Zn-MnO₂ battery. This novel method has been demonstrated to be an effective strategy for the design and fabrication of high-performance polymer electrolyte membranes for ZIBs, and offers a viable path to assemble other batteries with excellent performance.