In the context of the global energy transition, the development of renewable energy sources has made large-scale, highly safe, and long-duration energy storage battery systems an urgent necessity. Traditional fossil fuels are non-renewable and cause significant environmental damage, while renewable sources such as wind, solar, and tidal energy are intermittent and fluctuating. Therefore, efficient and reliable energy storage battery technologies are essential to stabilize the grid and support the widespread adoption of clean energy. Among many electrochemical storage technologies, the zinc-iron flow energy storage battery has received considerable attention because of its low electrolyte cost, high safety, environmental friendliness, and promising energy density. This technology is considered an excellent candidate for distributed energy storage battery systems and grid-scale applications.
A zinc-iron flow energy storage battery operates by utilizing the redox reactions between zinc species and iron-based complexes. In the alkaline system, the negative electrode employs the deposition and dissolution of metallic zinc, while the positive electrode relies on the ferrocyanide/ferricyanide redox couple. This system offers a high operating voltage and a high theoretical energy density. However, the growth of zinc dendrites during repeated charge-discharge cycles remains one of the most critical problems. Zinc dendrites can penetrate the separator causing internal short circuits, which ultimately destroys the battery. Therefore, the separator membrane is a key component that must simultaneously block zinc dendrites, suppress crossover of active species, and efficiently transport carrier ions.
In my research, I focused on the design and preparation of a composite membrane for a zinc-iron flow energy storage battery. The commercial perfluorosulfonic acid membrane is widely used because of its high ionic conductivity and excellent chemical stability, but it suffers from high cost, high swelling, and insufficient mechanical strength. In particular, it is easily pierced by zinc dendrites during long-term operation. Therefore, I proposed to incorporate a rigid zeolite material, ZSM-5, into a Nafion matrix. The rigid zeolite framework can mechanically reinforce the membrane and provide ion-selective channels, while the Nafion matrix ensures proton or sodium ion transport. However, pristine ZSM-5 has sub-nanometer micropores that limit ion transport. To overcome this limitation, I treated ZSM-5 crystals with a mixed alkaline etchant containing TPAOH and NaAlO2, producing hollow, hierarchical porous ZSM-5 particles with improved diffusion characteristics and lower Si/Al ratios. These modified particles were then used as functional fillers in Nafion composite membranes.
Experimental Details
I synthesized the parent ZSM-5 zeolite with a nominal Si/Al ratio of 50 by a hydrothermal method. The molar composition of the synthesis gel was TEOS:TPAOH:NaAlO2:H2O = 10:1:0.2:600. After stirring at room temperature for 12 hours, the gel was transferred to a Teflon-lined stainless-steel autoclave and heated at 135 °C for 36 hours under rotation. The resulting white powder was washed by repeated centrifugation and dried overnight, followed by calcination at 550 °C for 6 hours to remove the organic structure-directing agent. The parent sample was denoted as Z.
For the alkaline etching treatment, I used TPAOH and NaAlO2 as dual etching agents. In a typical preparation, TPAOH solution was diluted with deionized water to obtain the desired concentration, and a certain amount of NaAlO2 was added together with the calcined ZSM-5 powder. After ultrasonic dispersion for 1 hour, the suspension was transferred to an autoclave and heated at 170 °C for 6 hours. The etched products were thoroughly washed, dried, and calcined at 550 °C for 6 hours. I systematically varied TPAOH concentration, NaAlO2 amount, etching temperature, and etching time to investigate the synergistic etching behavior. The names and conditions of the representative etched samples are listed in Table 1.
| Sample | TPAOH (mol/L) | NaAlO2 (g) | Etching temp. (°C) | Etching time (h) | Yield (%) |
|---|---|---|---|---|---|
| Z | — | — | — | — | — |
| 0.75M-Z | 0.75 | 0 | 170 | 6 | 15 |
| Z-0.1 | 0 | 0.1 | 170 | 6 | 25 |
| 0.65M-Z-0.1 | 0.65 | 0.1 | 170 | 6 | 15 |
| 0.75M-Z-0.1 | 0.75 | 0.1 | 170 | 6 | 10 |
| 0.8M-Z-0.26 | 0.8 | 0.26 | 170 | 6 | 6 |
| 0.85M-Z-0.46 | 0.85 | 0.46 | 170 | 6 | 5 |
| 0.75M-Z-0.05 | 0.75 | 0.05 | 170 | 6 | 11 |
| 0.75M-Z-0.16 | 0.75 | 0.16 | 170 | 6 | 10 |
| 0.75M-Z-0.27 | 0.75 | 0.27 | 170 | 6 | 8 |
| 0.75M-Z-0.38 | 0.75 | 0.38 | 170 | 6 | 7 |
To prepare the composite membranes, I used a solution-casting method. A certain amount of etched ZSM-5 powder was dispersed into a mixture containing 3 mL of Nafion dispersion, 0.5 mL of N,N-dimethylformamide, and 0.5 mL of N,N-dimethylacetamide. The suspension was alternately sonicated and stirred for 15 minutes each, repeating four times to obtain a homogeneous dispersion. The dispersion was then poured into a clean glass Petri dish, covered lightly with aluminum foil, and dried at 80 °C for 24 hours. After drying, the membrane was aged at 160 °C for 2 hours, cooled to room temperature, and peeled off. The membrane containing 0.75M-Z-0.16 particles was denoted M-X, where X is the filler mass in milligrams. For the lower-Si/Al sample 0.75M-Z-0.27, the membranes were denoted M-X-0.27. A pure Nafion membrane was prepared under the same conditions without adding any zeolite and denoted M-0.
The water uptake (WU) and swelling ratio (SR) were measured by soaking membrane samples in deionized water for 24 hours, removing surface water, and comparing the wet and dry masses and lengths. The equations are:
$$WU = \frac{W_{\mathrm{wet}} – W_{\mathrm{dry}}}{W_{\mathrm{wet}}} \times 100\%$$
$$SR = \frac{L_{\mathrm{wet}} – L_{\mathrm{dry}}}{L_{\mathrm{wet}}} \times 100\%$$
Mechanical properties including tensile strength, elongation at break, and puncture strength were tested with a universal electronic testing machine at a tensile speed of 5 mm/min. Ion permeability was measured using an H-type diffusion cell. The left compartment contained 50 mL of 0.4 M K4Fe(CN)6 in 3 M NaOH solution, while the right compartment contained 50 mL of 0.4 M K2SO4 in 3 M NaOH solution. The membrane was sealed between the two compartments. At regular intervals, 4 mL of solution was removed from the right side and replaced with fresh K2SO4/NaOH solution. The concentration of \(\mathrm{Fe(CN)_6^{4-}}\) was determined by UV-visible spectroscopy, and the permeability \(P\) was calculated using Fick’s law:
$$\frac{dC_R(t)}{dt} = \frac{P A}{V_R L} \left( C_L – C_R(t) \right)$$
where \(V_R\) is the volume of the right compartment, \(A\) is the effective membrane area, \(L\) is the membrane thickness, \(C_L\) is the initial concentration on the left side, and \(C_R(t)\) is the concentration on the right side at time \(t\). The area resistance and ionic conductivity of the membranes were measured by electrochemical impedance spectroscopy in 0.5 M NaOH solution using a CHI660E electrochemical workstation. The frequency range was 1000 Hz to 1000 kHz with a voltage amplitude of 5 mV. The area resistance \(R_A\) and conductivity \(\sigma\) were obtained as:
$$R_A = R \times A$$
$$\sigma = \frac{L}{R_A}$$
To evaluate the practical performance of the composite membranes in an energy storage battery system, I assembled single cells using 0.6 M K4Fe(CN)6 in 5 M NaOH as the positive electrolyte and 0.3 M Na2Zn(OH)4 in 5 M NaOH as the negative electrolyte. The effective area of the membrane was 1.6 cm × 1.6 cm. Charge-discharge tests were performed at current densities from 40 to 120 mA/cm² with a cutoff voltage of 0.4 V. The coulombic efficiency (CE), voltage efficiency (VE), and energy efficiency (EE) were calculated as:
$$CE = \frac{\text{discharge capacity}}{\text{charge capacity}} \times 100\%$$
$$EE = \frac{\text{discharge energy}}{\text{charge energy}} \times 100\%$$
$$VE = \frac{EE}{CE} \times 100\%$$
Results and Discussion: Etching of ZSM-5
I first examined the morphology of the parent ZSM-5 crystals using SEM and TEM. The parent crystals displayed a brick-like shape with a length of approximately 2.5 μm along the c-axis. I observed a discontinuous amorphous layer on the crystal surface, which was identified as unreacted gel. Elemental mapping showed that aluminum was concentrated near the outer region of the parent crystals. This amorphous skin layer played an important role during the subsequent etching process.
When I etched the parent ZSM-5 with TPAOH alone, the internal part of the crystal dissolved only partially, forming an X-shaped cavity rather than a regular hollow structure. When I etched with NaAlO2 alone, the dissolution occurred mainly inside the crystal, while the external morphology remained almost unchanged. These observations clearly demonstrated that TPAOH and NaAlO2 act synergistically rather than independently as etchants for constructing hollow ZSM-5 crystals.
By using the dual etchants, I successfully prepared hollow “nanobox” ZSM-5 crystals with regular internal cavities. The formation mechanism follows the “dissolution-recrystallization” process. In a basic TPAOH solution, OH⁻ anions penetrate into the zeolite framework and dissolve the silicon-rich internal region, while TPA⁺ cations adsorb on the external surface and direct the recrystallization of a dense shell. NaAlO2 plays a dual role: it increases the alkalinity through hydrolysis and simultaneously supplies aluminate species that protect the external shell from excessive dissolution. The hydrolysis of aluminate can be expressed as:
$$\mathrm{AlO_2^- + 2H_2O \rightleftharpoons Al(OH)_4^-}$$
This balanced dissolution and recrystallization process is essential for generating a well-defined hollow morphology while preserving the MFI framework. I found that the TPAOH concentration had a strong effect on the extent of internal dissolution. As the TPAOH concentration increased from 0.65 M to 0.75 M, the internal cavity gradually enlarged. However, at 0.8 M, the crystals became severely etched, and at 0.85 M they completely dissolved under the same NaAlO2 content. To recover intact hollow crystals at high TPAOH concentrations, I increased the NaAlO2 amount. For example, at 0.8 M TPAOH, adding 0.26 g NaAlO2 produced complete hollow shells, while at 0.85 M TPAOH, 0.46 g NaAlO2 was required. These observations indicate that NaAlO2 acts as a protective agent against over-etching.
In the lower TPAOH concentration range from 0.65 M to 0.75 M, increasing the NaAlO2 amount enhanced the internal dissolution. The hollow volume became larger and the shell thickness decreased as the NaAlO2 amount increased from 0.05 g to 0.38 g. The amorphous skin layer crystallized during etching, and at higher NaAlO2 contents, almost no residual amorphous layer remained. This behavior shows that NaAlO2 can also promote etching in a low-alkalinity environment. Therefore, the synergistic balance between TPAOH and NaAlO2 determines the final crystal structure.
I also investigated the influence of etching temperature and time. When the etching temperature was reduced to 150 °C or 160 °C, the crystals were over-etched and partially broken. At 180 °C, regular hollow structures with intact shells were obtained, and the crystal size increased to about 3 μm, indicating enhanced recrystallization at higher temperatures. Extending the etching time from 6 h to 12 h or 18 h also improved the recrystallization of the outer layer. In particular, after 18 h of etching, the crystals exhibited a more uniform hollow structure and a larger crystal size of about 4 μm. These results confirm that higher temperature and longer etching time promote the recrystallization of the outer shell.
X-ray diffraction confirmed that all etched samples retained the characteristic MFI structure. No new crystalline phases were observed. The relative peak intensities decreased after etching because of the partial dissolution of the internal framework, but the topological structure was preserved. The N₂ adsorption-desorption isotherms showed a significant increase in mesoporosity after etching. The parent ZSM-5 had a mesopore volume of about 0.15 cm³/g, whereas all etched samples showed mesopore volumes in the range of 0.33 to 0.68 cm³/g. The pore size distributions were centered around 30 nm. This hierarchical pore system is beneficial for ion transport in an energy storage battery membrane because it reduces diffusion path lengths and provides more accessible channels.
| Sample | Smicro (m²/g) | SBET (m²/g) | Vmicro (cm³/g) | Vmeso (cm³/g) | Vtotal (cm³/g) |
|---|---|---|---|---|---|
| Z | 191 | 368 | 0.10 | 0.15 | 0.25 |
| 0.65M-Z-0.1 | 202 | 359 | 0.11 | 0.33 | 0.44 |
| 0.75M-Z-0.1 | 229 | 411 | 0.12 | 0.37 | 0.49 |
| 0.75M-Z-0.05 | 220 | 396 | 0.12 | 0.46 | 0.58 |
| 0.75M-Z-0.16 | 211 | 412 | 0.11 | 0.57 | 0.68 |
| 0.75M-Z-0.27 | 227 | 392 | 0.12 | 0.53 | 0.65 |
| 0.75M-Z-0.38 | 209 | 416 | 0.11 | 0.68 | 0.79 |
ICP-OES measurements revealed that the Si/Al ratio of the parent ZSM-5 was about 55, close to the nominal value of 50. After etching, the Si/Al ratio decreased significantly to 11–15 depending on the amount of NaAlO2. This decrease is attributed to the dissolution of silicon-rich internal regions and the incorporation of additional aluminum from the etching solution. The lower Si/Al ratio is beneficial for ion conductivity because aluminum sites provide negative charge centers that can promote cation transport. Solid-state \(^{27}\mathrm{Al}\) NMR spectroscopy showed a single peak at approximately 59 ppm, corresponding to tetrahedrally coordinated framework aluminum. This result confirms that aluminum remained in the framework positions after etching and was uniformly distributed in the zeolite structure.
| Sample | Si/Al |
|---|---|
| Z | 55 |
| Z-0.1 | 34 |
| 0.75M-Z | 20 |
| 0.75M-Z-0.05 | 15 |
| 0.75M-Z-0.16 | 14 |
| 0.75M-Z-0.27 | 12 |
| 0.75M-Z-0.38 | 11 |
Fourier-transform infrared spectroscopy showed characteristic absorption bands related to silanol groups after etching. The band near 3630 cm⁻¹ was assigned to external silanol groups, while the bands near 3441 cm⁻¹ and 1630 cm⁻¹ were assigned to the stretching and bending vibrations of silanol groups, respectively. The presence of these silanol groups increases the hydrophilicity of the zeolite surface, which can be advantageous for water uptake and ion transport in a composite membrane for an energy storage battery.
Composite Membranes Containing 0.75M-Z-0.16
Because the etched ZSM-5 sample 0.75M-Z-0.16 exhibited a high mesopore volume of 0.57 cm³/g, a low Si/Al ratio of 14, and a well-defined hollow nanobox morphology, I selected it as the filler for preparing a series of Nafion composite membranes. I varied the filler content from 1 mg to 7 mg and denoted the samples M-1, M-2.5, M-4, M-5.5, and M-7. The pure recast Nafion membrane was denoted M-0. SEM images of the composite membranes showed smooth surfaces without obvious defects. The etched ZSM-5 particles settled uniformly near the bottom of the membrane during solvent evaporation, forming a dense zeolite-rich layer. As the filler content increased, the thickness of this bottom layer increased, and some aggregates began to appear at higher loadings.
The water uptake of the composite membranes was higher than that of pristine Nafion. The pure Nafion membrane absorbed about 12% water, while the composite membranes absorbed between 15% and 19% depending on the filler content. This improvement is attributed to the hydrophilic silanol groups and abundant mesopores of the hollow ZSM-5 particles. The mesopores can act as water reservoirs and provide additional pathways for ion transport. Importantly, the swelling ratio of the composite membranes was lower than that of pure Nafion. The rigid zeolite framework restricted the expansion of the polymer chains and maintained dimensional stability. For example, the swelling ratio of M-4 was about 48% lower than that of M-0. This balance between high water uptake and low swelling is desirable for a practical energy storage battery membrane because it ensures both high ion conductivity and good mechanical stability.
Mechanical testing showed that the tensile strength of the composite membranes was significantly enhanced compared with the pure Nafion membrane. The tensile strength of M-0 was 10.14 MPa, whereas M-4 reached 18.46 MPa, representing an improvement of 82.1%. Interestingly, the tensile strength first increased with increasing filler content and then decreased at higher loadings. This trend is caused by particle agglomeration at high filler contents, which reduces the efficiency of stress transfer between the zeolite particles and the polymer matrix. The elongation at break decreased from 95% for M-0 to about 60% for M-7. This decrease is a direct consequence of the rigid zeolite framework, which restricts the mobility of polymer chains.
Furthermore, the puncture strength of the composite membranes increased steadily with increasing ZSM-5 content. The puncture strength of M-0 was 3.02 N, while that of M-7 was 4.70 N, an improvement of 55.6%. The rigid hollow ZSM-5 particles form a strong physical barrier against zinc dendrite penetration. In a zinc-iron flow energy storage battery, zinc dendrites often grow through the membrane and cause short circuits. The enhanced puncture strength is therefore crucial for improving the safety and cycle life of the energy storage battery.
I also measured the Fe(CN)6⁴⁻ permeability of the membranes. The composite membranes exhibited lower ion permeability compared with M-0. The MFI zeolite channel structure and the tortuous diffusion path created by the zeolite particles effectively hindered the crossover of ferrocyanide ions. As the filler content increased, the permeability decreased gradually, although the decrement became smaller at higher loadings. The lower ion permeability is beneficial for reducing self-discharge and capacity loss in the energy storage battery.
Electrochemical impedance measurements showed that the area resistance of the composite membranes was lower than that of M-0. The conductivity of M-0 was 1.26 mS/cm, while M-4 exhibited a conductivity of 1.80 mS/cm, an improvement of 42.9%. This improvement is attributed to the lower Si/Al ratio of the etched ZSM-5. The aluminum framework sites provide additional charge carriers and hydrophilic sites, promoting ion migration within the membrane. However, when the filler content was too high, the conductivity decreased because of particle aggregation and the disruption of continuous ion-conducting pathways. Therefore, an optimal filler content exists for maximizing the electrochemical performance of the composite membrane.
| Membrane | ZSM-5 sample | Filler mass (mg) |
|---|---|---|
| M-0 | — | 0 |
| M-1 | 0.75M-Z-0.16 | 1 |
| M-2.5 | 0.75M-Z-0.16 | 2.5 |
| M-4 | 0.75M-Z-0.16 | 4 |
| M-5.5 | 0.75M-Z-0.16 | 5.5 |
| M-7 | 0.75M-Z-0.16 | 7 |
Single-cell tests were conducted by assembling the membranes into an alkaline zinc-iron flow energy storage battery. The coulombic efficiency, voltage efficiency, and energy efficiency were measured at current densities of 40, 60, 80, 100, and 120 mA/cm². All composite membranes showed higher CE and EE than pure Nafion. The CE increased with increasing current density because the shorter charge-discharge time reduced the amount of side reactions. The VE decreased with increasing current density because of higher ohmic and concentration polarization. Among all the composite membranes, M-4 exhibited the best overall performance. At a current density of 80 mA/cm², M-4 achieved a CE of about 98%, a VE of about 87.6%, and an EE of 85.85%. In contrast, the pure Nafion membrane achieved an EE of only 74.54%. This significant improvement in energy efficiency demonstrates that the incorporation of hollow hierarchical ZSM-5 into Nafion is an effective strategy for enhancing the performance of an energy storage battery membrane.
Long-cycle testing of M-4 was carried out at 80 mA/cm² for more than 50 cycles. The battery assembled with M-4 maintained stable charge-discharge performance without noticeable capacity decay. The rigid ZSM-5 layer effectively blocked zinc dendrites, preventing them from penetrating the membrane and causing short circuits. As a result, the voltage efficiency remained stable, and the discharge capacity retention was improved. This long-term stability is essential for the practical application of the composite membrane in a large-scale energy storage battery system.
Composite Membranes Containing 0.75M-Z-0.27
I further investigated the influence of the zeolite structure on membrane performance by using another etched ZSM-5 sample, 0.75M-Z-0.27, which has a slightly smaller mesopore volume of 0.53 cm³/g and a lower Si/Al ratio of 12. The lower Si/Al ratio provides more aluminum sites and thus potentially higher ion exchange capacity. However, the lower Si/Al ratio also weakens the framework stability because the Al–O bond is weaker than the Si–O bond. I prepared composite membranes M-1-0.27, M-2.5-0.27, and M-4-0.27 by using 1 mg, 2.5 mg, and 4 mg of 0.75M-Z-0.27, respectively.
SEM analysis confirmed that the 0.75M-Z-0.27 particles were uniformly embedded in the Nafion membrane with no obvious defects. The water uptake of these membranes was slightly higher than that of the analogous membranes containing 0.75M-Z-0.16 because of the greater number of hydrophilic aluminum sites. However, the swelling ratio also increased because the lower Si/Al ratio reduced the rigidity of the zeolite framework. Consequently, the dimensional stability was somewhat compromised compared with the membranes containing 0.75M-Z-0.16.
Mechanical testing showed that the tensile strengths of M-1-0.27, M-2.5-0.27, and M-4-0.27 were lower than those of M-1, M-2.5, and M-4, respectively. The tensile strength of M-2.5-0.27 was 15.96 MPa, which was still 57.5% higher than that of pure Nafion. The puncture strength of the 0.75M-Z-0.27 composite membranes was also slightly lower than that of the 0.75M-Z-0.16 counterparts, but all values remained significantly higher than that of M-0. The lower mechanical strength is consistent with the lower proportion of strong Si–O bonds and the thinner shell of the 0.75M-Z-0.27 particles. Nevertheless, these membranes still provide adequate mechanical reinforcement for an energy storage battery separator.
The Fe(CN)6⁴⁻ permeability of the 0.75M-Z-0.27 composite membranes was lower than that of the corresponding 0.75M-Z-0.16 membranes at the same filler loading. The smaller mesopores of 0.75M-Z-0.27 created narrower transport channels, which further hindered the crossover of large ferrocyanide ions. This reduced permeability enhances the coulombic efficiency by suppressing self-discharge and cross-contamination. However, the lower Si/Al ratio also caused a larger increase in permeability over time because the framework was more susceptible to chemical degradation in the alkaline environment. Therefore, a trade-off exists between initial selectivity and long-term stability.
The conductivity of the 0.75M-Z-0.27 membranes followed a bell-shaped curve with filler content. M-2.5-0.27 exhibited the highest conductivity of 1.80 mS/cm, which is the same as the best value obtained for M-4. The optimum filler content shifted from 4 mg for 0.75M-Z-0.16 to 2.5 mg for 0.75M-Z-0.27. This shift occurs because the lower Si/Al ratio of 0.75M-Z-0.27 provides more aluminum sites, so a smaller amount of filler is needed to reach the optimum ionic conductivity. At higher loadings, excessive aggregation and the possible disintegration of the less stable framework increase the resistance and decrease the conductivity.
Single-cell tests of the 0.75M-Z-0.27 composite membranes were performed at 80 mA/cm². M-2.5-0.27 achieved a CE of 98.05%, a VE of 84.13%, and an EE of 84.13%. Although this EE is slightly lower than that of M-4 (85.85%), it is still much higher than that of pure Nafion. The coulombic efficiency of the 0.75M-Z-0.27 membranes was slightly higher than that of the 0.75M-Z-0.16 membranes because of the lower Fe(CN)6⁴⁻ permeability. However, the voltage efficiency was reduced, especially for M-1-0.27 and M-4-0.27, due to their lower conductivity and less stable mechanical properties. These results indicate that when the Si/Al ratio of the ZSM-5 filler is lower, the optimal doping content must be reduced in order to balance the positive and negative effects.
| Membrane | Tensile strength (MPa) | Puncture strength (N) | Conductivity (mS/cm) | Area resistance (Ω cm²) | EE at 80 mA/cm² (%) |
|---|---|---|---|---|---|
| M-0 | 10.14 | 3.02 | 1.26 | 4.42 | 74.54 |
| M-4 | 18.46 | 4.50 | 1.80 | 3.06 | 85.85 |
| M-2.5-0.27 | 15.96 | 3.98 | 1.80 | 3.08 | 82.49 |
Conclusion
In this work, I successfully prepared hollow hierarchical porous ZSM-5 zeolite crystals by synergistic etching with TPAOH and NaAlO2. The etching process followed a dissolution-recrystallization mechanism, and the balance between the two etchants was essential for obtaining intact nanobox structures. The etched ZSM-5 samples retained the MFI framework while exhibiting higher mesoporosity, lower Si/Al ratios, and more abundant silanol groups. These structural features are highly favorable for ion transport in an energy storage battery membrane.
I then fabricated ZSM-5/Nafion composite membranes by solution casting. The composite membranes exhibited improved water uptake, reduced swelling, enhanced tensile strength, and significantly higher puncture resistance compared with pure Nafion. The hollow ZSM-5 particles effectively suppressed the crossover of Fe(CN)6⁴⁻ ions and simultaneously improved the ionic conductivity by providing additional aluminum-based charge transport sites. For the sample 0.75M-Z-0.16, the optimal filler content was 4 mg, corresponding to the M-4 membrane. This membrane increased the tensile strength by 82.1%, the puncture strength by 49%, and the conductivity by 42.9% relative to pure Nafion. In an alkaline zinc-iron flow energy storage battery, M-4 achieved an energy efficiency of 85.85% at 80 mA/cm² and maintained stable performance for more than 50 cycles.
When I used the lower-Si/Al sample 0.75M-Z-0.27, the optimal filler content decreased to 2.5 mg. The M-2.5-0.27 membrane exhibited good electrochemical performance with an EE of 82.49% at 80 mA/cm², but its mechanical stability was slightly lower than that of M-4. These findings demonstrate that the structure of the zeolite filler, especially its mesoporosity and Si/Al ratio, must be carefully matched with the doping content to optimize the overall membrane performance.
Overall, the incorporation of hierarchically porous hollow ZSM-5 into Nafion is a promising strategy for developing high-performance, mechanically robust membranes for zinc-iron flow energy storage battery systems. The rigid zeolite framework provides a physical barrier against zinc dendrites, while the hollow channels and low Si/Al ratio promote efficient ion transport. Future work should focus on further optimizing the orientation of zeolite particles within the membrane, reducing the membrane cost, and scaling up the preparation process for practical large-scale energy storage battery applications. By continuing to improve the membrane performance, I believe that the zinc-iron flow energy storage battery will become an even more attractive technology for safe, low-cost, and long-duration energy storage.