Flow battery diaphragm of 3D printing column array pore channel and preparation method of flow battery diaphragm

Abstract

The invention provides a digital light processing 3D printing column array pore channel ceramic membrane for zinc-based flow batteries (ZFBs), and belongs to the technical field of flow batteries. According to the flow battery diaphragm disclosed by the invention, the 3D printed ceramic membrane is used as a substrate, and the ultrathin sub-hydrophilic polyelectrolyte selection layer is prepared on the surface of the ceramic membrane, so that rapid ion transmission is realized, and zinc deposition is effectively regulated and controlled. Due to the existence of the 3D printed column array pore channels, the ion transmission path is effectively shortened, the membrane contact resistance is reduced, rich three-dimensional space in the pore channels can be used for zinc deposition and electrolyte storage, and the area capacity of the battery is effectively improved. And the mechanical firmness of the ceramic can resist structural damage caused by zinc dendrites, so that the battery can stably run for a long time.

Description

Technical Field

[0001] The invention relates to a liquid flow battery diaphragm with 3D printed column array channels and a preparation method thereof, belonging to the technical field of liquid flow batteries. Background Art

[0002] Zinc-based flow batteries with metallic zinc as the negative electrode have attracted extensive attention in the field of grid-scale energy storage due to their advantages of low cost, good safety, and long cycle life. Unfortunately, the generation of zinc dendrites during long-term charge and discharge cycles and the limited area capacity restrict the large-scale development of zinc-based flow batteries.

[0003] Membrane is one of the key materials to solve the above problems, and it plays an important role in the efficient transmission of conductive ions and resistance to zinc dendrites. Inorganic ceramic membranes have the advantages of high mechanical strength, good chemical stability, large flux, and no environmental pollution. They are widely used in water treatment, food processing, and environmental protection. Previous studies have attempted to use traditional planar ceramic membranes in zinc-based flow batteries, which have excellent stability against zinc dendrites. However, the challenges posed by reducing the thickness of ceramic membranes and the inherent planar surface morphology of planar ceramic membranes limit the progress of ion conductivity and zinc deposition capabilities. How to further reduce mass transfer resistance and optimize the spatial structure of zinc deposition remains a key challenge. Summary of the invention

[0004] In view of the above-mentioned technical problems, the present invention provides a 3D printed column array pore ceramic membrane and a preparation method thereof.

[0005] The technical concept adopted by the present invention is: using digital light processing (DLP) three-dimensional printing technology, an innovative column array pore ceramic membrane for ZFBs is proposed. On the one hand, due to the presence of the pores, the ion transmission path is effectively shortened, and the conductive ion K + and Na + The ion flux increased by about 150%. In addition, the transmembrane contact resistance was significantly reduced by about 60%. On the other hand, the array channels increased the contact area between the membrane and the electrode, while also acting as a reservoir for the electrolyte, thereby increasing the concentration of zincate ions at the interface and providing ample space for zinc deposition. At the same time, the ultra-high mechanical strength of the ceramic membrane prevents the formation of zinc dendrites, thereby contributing to the uniform deposition of zinc. The above ideas are used to achieve stable operation of high-performance zinc-based flow batteries.

[0006] Technical solution:

[0007] A liquid flow battery diaphragm is composed of a 3D printed ceramic membrane with array channels as a substrate and a polymer electrolyte layer on the surface.

[0008] The ceramic membrane is prepared by digital light processing (DLP) 3D printing technology; the thickness of the 3D printed ceramic membrane substrate is 0.5-2mm, and the ceramic membrane substrate consists of two parts: a lower support layer and an upper separation layer; the lower support layer includes a three-dimensional printed column array channel with a pore size of 800-1200nm and a depth of 0.5-2mm; the upper separation layer has a thickness of 10-40um and a pore size of 80-120nm.

[0009] The polymer in the polymer electrolyte layer is selected from one or a mixture of sulfonated polyethersulfone, sulfonated polyetheretherketone, sulfonated polyimide, polytetrafluoroethylene, polyvinylidene fluoride, polybenzimidazole, and chitosan; the polymer is preferably a mixture of one or two of polybenzimidazole and sulfonated polyetheretherketone, and the content of the polymer per unit surface area of ​​the ceramic membrane is 0.1-3 mg cm-2.

[0010] A method for preparing a flow battery diaphragm, characterized in that it comprises the following steps:

[0011] Step 1, dissolving the polymer in an organic solvent as a casting solution;

[0012] Step 2: Spin the casting solution onto the substrate to form a film, and then dry it.

[0013] In the step 1, the organic solvent is selected from one or more of N,N-dimethylacetamide, N,N-dimethylformamide, N-methylpyrrolidone, tetrahydrofuran or dimethyl sulfoxide.

[0014] In the step 1, the mass ratio of the polymer in the casting solution is 1-20%; the spin coating speed is controlled to be 500-4000 r / min, and the spin coating time is 10-120 s.

[0015] In the step 2, the drying process adopts forced air drying or vacuum drying, the drying temperature is 60-120° C., and the drying time is 12-48 hours.

[0016] Application of flow battery separators in the preparation of zinc-based flow batteries.

[0017] A method for predicting zinc deposition on a diaphragm during operation of a zinc-based flow battery by means of physical field simulation comprises the following steps:

[0018] Step 1, setting a ceramic membrane model with a pore structure in COMSOL software, and setting a polymer electrolyte layer on the surface of the ceramic membrane, and a diaphragm composed of the ceramic membrane and the polymer electrolyte is located in an electrolyte;

[0019] Step 2, set the electric field potential and the chemical reaction formula during charging:

[0020] Step 3: Use the tertiary current distribution module to introduce the Nernst-Planck equation to describe the ion transport process and solve the polymer electrolyte concentration; use the Butler-Volmer equation to describe the electrode kinetic reaction and combine it with Faraday's law to solve the ion flux;

[0021] Step 4: After the charging reaction, the amount of zinc produced at each position on the diaphragm is obtained by measuring the deformation of the membrane surface.

[0022] In the step 2, the electric field potential is set to 10-30 mV.

[0023] Beneficial Effects

[0024] The hydrophilic polymer electrolyte is spin-coated on the surface of the 3D printed ceramic membrane. The 3D printed pores can effectively shorten the ion transmission path, reducing the membrane contact resistance by about 60%. The rich hydrogen bond network of the polymer accelerates the transmission of conductive ions, and the rich three-dimensional space in the pores can be used for zinc deposition and electrolyte storage, effectively increasing the area capacity of the battery. In addition, the mechanical robustness of the ceramic of up to 1.46Gpa can resist structural damage caused by zinc dendrites, which allows the battery to remain stable during long-term operation, providing a new strategy for the design and preparation of a new generation of high-performance ion conductive membranes. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 : Digital photo of the prepared 3D printed array pore ceramic membrane.

[0026] Figure 2 :SEM images of 3D printed array pore ceramic membrane. a. Upper surface of membrane, b. Cross section of membrane, c. Lower surface.

[0027] Figure 3 : AFM images of films: a.3DPC, b.3DPC-S-1.8, c.3DPC-S-3.0.

[0028] Figure 4 :SEM and X-ray photoelectron spectroscopy (XPS) of the membrane. a. SEM image of the surface of 3D printed ceramic support, b. SEM image of the surface of 3DPC-S-1.8 membrane, c. XPS of the surface of 3DPC-S-1.8 membrane, d. SEM image of the cross section of 3D printed ceramic support, e. SEM image of the cross section of 3DPC-S-1.8 membrane, f. In-situ etching XPS of the surface of 3DPC-S-1.8 membrane.

[0029] Figure 5 :a.Fe(CN) 6 4-a. Permeability of 3DPC-S-1.8 and 3DPC, b. Low-field nuclear magnetic resonance (LF-NMR) analysis results of 3DPC-S-1.8 and 3DPC, c. Small-angle X-ray scattering (SAXS) of 3DPC-S-1.8, 3DPC and SPEEK.

[0030] Figure 6 :a. Permeability of hydroxide ions through different membranes; COMSOL simulation, b. 3DPC-S-1.8 surface resistance simulation, c. Total K + , Na + Flux simulation.

[0031] Figure 7 :Battery performance test. a. Variable current performance of ZFBs of 3DPC-S-1.8, b. 80mA cm -2 , 28 mAh cm -2 Long cycle test, c.80mA cm -2 , 280mAh cm -2 Long cycle test, d. Comparison of the performance of ZFBs with different membranes in terms of cycle life and areal capacity, e. Voltage-time distribution of ZFBs of 3DPC-S-1.8, f and g at 80 mA cm -2 Below, SEM images of Zn morphology in ZFBs assembled with Nafion212 and 3DPC-S-1.8 at the end of charging.

[0032] Figure 8 : a. ZFBs assembled with 3DPC-S-1.8 at 80 mA cm -2 , 280mAh cm -2 a. Digital photograph of the film at the end of charging, b. Digital photograph of the carbon felt morphology, c. SEM image and EDS on the carbon felt; COMSOL physical simulation, d. Aggregation distribution of Zn(OH)42-, e. The process of zinc deposition in the 3D printed holes within a certain period of time.

[0033] Fig. 9 :a.3DPC-S-1.8 and carbon felt at 80mA cm -2 , 280mAh cm -2 Lower micro-CT scanning image of ZFBs at the end of charging, b, c, d, e, f, and g are top views of the two-dimensional slices at the positions marked in a, respectively.

[0034] Fig.10 : Model establishment and potential field distribution diagram.

[0035] Fig.11 : Comparison of average pore resistance and ion flux.

[0036] Fig.12 : Zincate ion distribution and zinc deposition simulation.

[0037] Fig.13 : Pore size distribution diagram of 3D printed ceramic membrane. DETAILED DESCRIPTION

[0038] Example 1

[0039] Preparation of 3D printed ceramic substrates

[0040] A 3D printed column array hole ceramic membrane was produced by digital light processing (DLP) 3D printing technology. First, a mixing and defoaming machine was used to premix the photocurable resin, dispersant and photoinitiator solution in a mass ratio of 40:8:1. Then, the ceramic particles were added to the premix and mixed thoroughly. The mass ratio of alumina powder to photocurable resin was controlled at 3:1. Then, the above slurry was DLP 3D printed using a 3D printer equipped with a UV LED light source. Finally, the top ceramic membrane layer was further prepared by dip coating, and the sample was sintered at a certain temperature to obtain a double-layer ceramic membrane. The ceramic membrane was used as a substrate, such as Figure 1 and Figure 2 As shown, the substrate consists of two parts: a lower support layer and an upper separation layer. The 3D printed ceramic membrane substrate has a thickness of 1.2 mm and consists of two parts: the lower support layer has a pore size of about 1000 nm and contains a 3D printed column array channel with a diameter of 2.5 cm and a depth of about 1 mm at the bottom; the upper separation layer has a thickness of about 20 μm and a pore size of about 100 nm.

[0041] Example 2

[0042] Preparation of Sulfonated Polyetheretherketone (SPEEK)

[0043] The polyetheretherketone resin (PEEK) was placed in a vacuum oven at 80°C for drying. According to a certain ratio, 98% concentrated sulfuric acid was measured and added to the reactor. When the temperature rose to 50°C, PEEK was added and stirred. The reaction time was set to 5h. The resulting mixture was then poured into ice water and stirred with a glass rod to form fibrous SPEEK. The SPEEK was ultrasonically washed to neutrality and placed in an oven at 60°C for drying to obtain a light yellow SPEEK.

[0044] Example 3

[0045] Preparation of SPEEK / 3D printed ceramic membrane

[0046] Before dissolving SPEEK, dry SPEEK in a moisture analyzer at 120°C for 12 minutes; weigh 0.5g of the dried SPEEK and dissolve it in 10ml of DMF solution, and place it on a shaker for 24 hours until SPEEK is completely dissolved. Fix the 3D printed ceramic membrane on a spin coater, and use a pipette to transfer 1ml of the prepared solution twice to the surface of the ceramic sheet. The speed of the spin coater is 3000r / min. After the spin coating is completed, place the composite membrane in an 80°C oven and dry it for 24 hours for use. The composite membrane is marked as 3DPC-SX, where X is the content of the polymer per unit area in mg cm -2 , from 0 to 3 mg cm -2 .

[0047] Characterization of materials

[0048] As shown by atomic force microscopy (AFM) results ( Figure 3 ), compared with the pure ceramic film, the surface roughness of the film gradually decreases with the increase of spin coating amount. The surface and cross-sectional morphology of the film were characterized by field emission scanning electron microscopy (FE-SEM). Figure 4 As shown in a and b, compared with the pure ceramic film, a clear coating is observed on the film surface after spin coating of the polymer. Figure 4 D and Figure 4 The cross-sectional image of e also observed that SPEEK had slightly penetrated into the pores and was tightly bonded to the ceramic substrate. To explore the thickness of the polymer layer, we confirmed the formation of a 3DPC-S-1.8 film with a thickness of less than 10 nm by using surface Al and S element XPS (detection depth less than 10 nm). Figure 4 c). Figure 4 As shown in f, the in-situ etching XPS results show that the S element content gradually decreases with increasing etching depth, which indicates that the surface SPEEK penetrates into the pores, which is consistent with the results observed by SEM characterization.

[0049] Ion permeability test

[0050] Fe(CN) 6 4- Permeability through a membrane.

[0051] Solution on the left: 0.4 mol L -1 K 4 Fe(CN) 6 +3.0mol L -1 NaOH, right solution: 0.4 mol L -1 MgSO4+3.0mol L -1NaOH. The right solution was used as a balance solution to balance the ionic strength and minimize the osmotic pressure effect. During the experiment, the solutions in both half cells were vigorously stirred with a magnetic stirrer to avoid the influence of concentration polarization. At regular intervals, 4 mL of sample solution was collected from the right solution and then 4 mL of balance solution was added. K in the sample solution 4 Fe(CN) 6 The concentration was measured by UV-visible spectrometer and calculated according to the following formula

[0052]

[0053] Among them, V B (L) is the volume of the right half of the battery. B (t)(mol / L) is the Fe(CN) in the right half of the cell 6 4- Functional relationship between concentration and time t, C A is Fe(CN) in the left half of the cell 6 4- Concentration. A(cm 2 ) and L (cm) are the effective membrane area and membrane thickness, respectively. P is Fe(CN) 6 4- The penetration rate.

[0054] OH - Permeability test.

[0055] The solution on the left is 50mL 5mol L -1 NaOH solution, the solution on the right is 50mL deionized water. The liquid is circulated by a peristaltic pump to simulate the battery working environment, and a pH meter is used to measure the concentration of hydroxide ions in the solution on the right at different diffusion times. The calculation formula is as follows:

[0056] pH = 14 + lg (c (OH - ))

[0057] The results of active ion permeation test show that the Fe(CN) 6 4- The ion permeation rate is much lower than that of 3DPC membrane ( Figure 5 a), which may be due to the sieving effect formed by the infiltration of the polymer and the ceramic pores. To further explore the size of the ion channels of 3DPC-S-1.8, we used pore water as a probe and low-frequency nuclear magnetic resonance (LF-NMR) to explore its pore structure. Figure 5As shown in b, the transverse relaxation time T represents different types of pores in the membrane, and the peak area represents the distribution of the pores. Compared with 3DPC, after the introduction of the polymer, the peaks of 3DPC-S-1.8 in regions II and III shifted significantly to the left, and a new peak appeared in region I, indicating that the number of mesopores and macropores decreased, and micropores (<10nm) were generated. This is because the polymer fills the pores and refines the pore structure, thereby improving the ion screening ability of the membrane.

[0058] Excessive swelling of the polymer membrane in the electrolyte will reduce the ion selectivity of the membrane. To this end, we used small-angle X-ray scattering (SAXS) to analyze the membrane in a completely wet state ( Figure 5 c). SPEEK-H calculated according to the Bragg formula 2 Bragg spacing d of O (5.86nm) and 3DPC-S-1.8-H 2 The Bragg spacing d of O (3.45 nm) proves that the polymer infiltration into the pores effectively suppresses the swelling behavior of the membrane, further ensuring the higher ion selectivity of 3DPC-S-1.8.

[0059] At the same time, we found in the test that the ionic conductivity of the 3DPC membrane is as high as 137mS cm -1 , which is much better than the traditional planar ceramic membrane (73mS cm -1 ) and commercial Nafion 212 (11.6 mS cm -1 ) and SPEEK membrane (8.2mS cm -1 ), even after spin coating of the polymer layer, it can reach 37.2 mS cm -1 ( Figure 6 a). OH - This is further supported by the consistency of the ion diffusion data ( Figure 6 b). This indicates that the uniform 3D printed column array pore structure of 3DPC-S-1.8 may shorten the ion transport path, and the large pore structure inside the ceramic also contributes to the rapid migration of ions.

[0060] To demonstrate the feasibility of the prepared 3DPC-S-1.8, ZFBs were assembled for battery testing.

[0061] The battery assembly and performance test method is: the carbon felt electrode is sandwiched between two electrode plates separated by a gasket and a membrane to assemble a homemade ZFBs single cell, and then the effective area of ​​the membrane is captured and calculated by software (Image). The positive electrode electrolyte contains 5 mol L -1 NaOH and 0.6 mol L -1 K 4 Fe(CN) 6 3H2 O, the negative electrode electrolyte contains 5 mol L -1 NaOH and 0.3 mol L -1 ZnO, the electrolyte circulates through a peristaltic pump. The charge and discharge performance at different current densities was tested using Neware. During the charging process, the constant capacity was maintained by controlling the charging time, and the discharge process ended with a cut-off voltage of 0.1V. Variable current and long cycle tests were performed during the charging process. In the variable current and long cycle tests, the volume of the positive and negative electrolytes was 10mL, and at high area capacity (280mAhcm -2 ) test, the volume of the positive and negative electrolytes was 100 mL.

[0062] Benefiting from the advantages brought by 3DPC-S-1.8, ZFBs maintain high performance in a wide current density range. Figure 7 As shown in a (40-200mA cm -2 ), at 80 mA cm -2 At a current density of 1.5 Å, the ZFBs using 3DPC-S-1.8 still showed stable performance for nearly 1200 h of cycling, with an average Coulombic efficiency of 99.5% and an average energy efficiency of 83.2% ( Figure 7 b). The stability is much better than commercial Nafion 212 and SPEEK pure membrane.

[0063] Analyze the voltage curve during battery cycling, such as Figure 7 As shown in e, 3DPC-S-1.8 remained stable during operation, while commercial Nafion212 showed abnormal fluctuations at 70h, which was caused by micro-short circuits in the battery. We disassembled the ZFBs assembled with different separators at the end of long cycle charging and analyzed the morphology of the metal zinc deposited on the anode and the membrane surface. As shown in the figure, the battery using Nafion 212 as a separator exhibited a sharp and disordered zinc dendrite morphology, while the morphology of the metal zinc in the battery using the composite membrane was smooth and dense without dendrites under the same conditions ( Figure 7 f and Figure 7 g).

[0064] For ZFBs batteries, limited area capacity is another major factor restricting their development. To further evaluate the electrochemical performance of 3DPC-S-1.8, the -2 Current density, 280 mAh cm -2Thanks to the excellent strength of the 3DPC-S-1.8 membrane and the deposition space in the 3D printed channels, the battery using 3DPC-S-1.8 can maintain an average coulombic efficiency (CE) of 99% and an energy efficiency (EE) of 79% for nearly 1000 hours ( Figure 7 c). The performance of the 3DPC-S-1.8 membrane is superior to most of the membranes recently reported for ZFBs with different areal capacities ( Figure 7 d).

[0065] Control of zinc deposition by pillar array channels

[0066] To gain a deeper understanding of Zn deposition during the entire charging process, we disassembled the ZFBs at the end of charging at high surface area and characterized Zn deposition on carbon felts as well as ceramic membranes. Figure 8 As shown in a and b, when the regular 3D array columnar channel side faces the zinc side, after the charging is completed, zinc deposition is distributed in large quantities in the carbon felt and the 3D array columnar channel. SEM and EDS further confirm this result ( Figure 8 c).

[0067] Micro-CT technology was further used as a non-destructive testing tool to obtain the distribution information of zinc deposition inside the film and carbon felt at the end of high-volume charging. The XLAB-2000 micro-computed tomography (micro-CT) instrument was used to characterize the three-dimensional structure and morphology of the sample. Test voltage: 100KV; current: 150μA; resolution: 14.89μm.

[0068] like Fig. 9 As shown in Figure 2, zinc deposition gradually increases from the membrane surface to the interface between the membrane and carbon felt, and the deposition tends to be distributed in the pores ( Fig. 9 From the interface to the bottom of the carbon felt, the deposition is less affected by the pores and tends to be evenly distributed ( Fig. 9 fg). This fully demonstrates that the ample space in the pores of the 3DPC-S-1.8 membrane array and the large pores between the ceramic particles provide abundant deposition sites for zinc and reduce concentration polarization. At the same time, the ultra-high mechanical properties of the ceramic help to resist zinc dendrites. With the synergistic effect of these factors, stable operation and high capacity are achieved.

[0069] COMSOL Simulation

[0070] In order to realize computer simulation to explore the influence of 3D pores on ion transport and the regulatory effect on zinc deposition, an alkaline zinc-iron flow battery (ZFBs) model assembled by 3DPC-S-1.8 membrane was established to simulate the distribution of ions on the membrane surface and in the array column pores as well as the electrochemical reaction process, and the solution was performed using COMSOL Multiphysics 6.2 software.

[0071] The present invention adopts a sequence type of physical field controlled grid, selects ultrafine unit size, and encrypts the pore surface grid to ensure the calculation accuracy here, which is more in line with the actual ion transport and zinc deposition process. The main contribution items are dilute species transfer (tds), current (ec) and dynamic grid (ale).

[0072] In the established geometric model, the length is set to 5.1mm and the height is 3.2mm. The surface is separated by a SPEEK membrane (3DPC-S-1.8). The separation layer on the ceramic membrane surface is set to 20um. There are 1mm wide and 1mm deep grooves in the membrane to represent the 3D printed channels. The channel area and the gap between the channels are marked as area I and II ( Fig.10 a). The electric field potential distribution is as follows Fig.10 As shown in b and c, the movement direction of cations is opposite to the voltage direction, while that of anions is the same.

[0073] The charging chemical reaction is set in the model:

[0074]

[0075] During the charging process, Zn(OH) 4 2- Transformed into single substance Zn, the concentration at the membrane interface gradually decreased, and the Zn(OH) 4 2- Under the action of electric field and flow field, it begins to migrate to the interface to make up for the reduction of ions, while the discharge process is the opposite. Zn(OH) 4 2- The diffusion coefficient is 1.2×10 -6 cm 2 / s. OH-, K + 、Na + 、Fe(CN) 6 4- and Fe(CN) 6 3- The diffusion coefficients are 5.3×10 -5 , 1.9×10 -5 , 1.3×10 -5 , 7.8×10 -6 and 8.2×10 -6 cm 2 / s.

[0076] First, select the domain and set the initial concentration values ​​of the upper and lower parts of the model. + and K + and Zn(OH) 4 2-The initial concentration was set to 5 mol L -1 ,2.4mol L -1 and 0 mol L -1 , set all domain ranges, specifically 1-9 regions. + and K + and Zn(OH) 4 2- The initial concentration was set to 5 mol L -1 ,2.4mol L -1 and 0.3 mol L -1 , set the domain. Then, set the potential, select manual for the upper potential boundary, set the value to 9, and the potential V 0 The lower potential is set to 20mV, the lower potential is grounded, and the value is set to 2. In this model, the Brinkman equation is used to ensure continuity and momentum conservation. The Nernst-Planck equation is introduced using the tertiary current distribution module to describe the ion transport process, and the Butler-Volmer equation is used to describe the electrode kinetic reaction. The change in zinc deposition is reflected by the deformation of the membrane surface, which is calculated as follows: Zn(OH) 4 2- Ion flux*Zn(OH) 4 2- Molar concentration / Zn(OH) 4 2- density.

[0077] like Fig.11 As shown in a, the simulation results show that the average resistance of the pores (0.94Ω·cm 2 ) is much smaller than the average resistance between the channels (2.41Ω·cm 2 ). Conductive ion K + 、Na + These results are also confirmed by the fact that the total flux through the pores is much higher than the total flux between the pores ( Fig.11 b). The unique 3D array columnar pore structure in 3DPC-S-1.8 facilitates the rapid and selective transport of charge-carrying ions. Fig.12 As shown in a, under the action of electric field and flow field, the zincate ion concentration in region I is much higher than that in region II. This means that the zinc plating rate in region I is much higher than that in region II, and zinc deposition tends to occur in region I. During the battery charging process, as time increases (T = 10s-1000s), a large amount of metal zinc is deposited in the internal pores of region I, increasing the area capacity of the battery ( Fig.12 b). The simulation results are consistent with the actual characterization test results, which proves the reliability of the simulation method.

Claims

1. A flow battery separator, characterized in that: The diaphragm is a 3D printed ceramic membrane with array channels as a substrate, and a polymer electrolyte layer is provided on the surface.

2. The flow battery membrane according to claim 1, characterized in that: The ceramic membrane substrate is prepared by digital light processing (DLP) 3D printing technology; the thickness of the 3D printed ceramic membrane substrate is 0.5-2mm, and the ceramic membrane substrate consists of two parts Composition: a lower support layer and an upper separation layer; the lower support layer includes a three-dimensional printed column array channel with a pore size of 800-1200nm and a depth of 0.5-2mm; the upper separation layer has a thickness of 10-40um and a pore size of 80-120nm.

3. The flow battery membrane according to claim 1, characterized in that: The polymer in the polymer electrolyte layer is selected from one or a mixture of sulfonated polyethersulfone, sulfonated polyetheretherketone, sulfonated polyimide, polytetrafluoroethylene, polyvinylidene fluoride, polybenzimidazole, and chitosan; the polymer is preferably a mixture of one or two of polybenzimidazole and sulfonated polyetheretherketone, and the content of the polymer per unit surface area of ​​the ceramic membrane is 0.1-3 mg cm -2 .

4. A method for preparing a flow battery diaphragm, characterized in that: The steps include: Step 1, dissolving the polymer in an organic solvent as a casting solution; Step 2: Spin the casting solution onto the substrate to form a film, and then dry it.

5. The preparation method according to claim 4, characterized in that: In the step 1, the organic solvent is selected from one or more of N,N-dimethylacetamide, N,N-dimethylformamide, N-methylpyrrolidone, tetrahydrofuran or dimethyl sulfoxide.

6. The preparation method according to claim 4, characterized in that: In the step 1, the mass ratio of the polymer in the casting solution is 1-20%; the spin coating speed is controlled to be 500-4000 r / min, and the spin coating time is 10-120 s.

7. The preparation method according to claim 4, characterized in that: In step 2, the drying process adopts forced air drying or vacuum drying, the drying temperature is 60-120° C., and the drying time is 12-48 hours.

8. Use of the liquid flow battery separator according to claim 1 in the preparation of zinc-based liquid flow batteries.

9. A method for predicting zinc deposition on a diaphragm during operation of a zinc-based flow battery by means of physical field simulation, comprising the following steps: Step 1: In COMSOL software, a ceramic membrane model with a pore structure is set, and a polymer electrolyte layer is set on the surface of the ceramic membrane. The diaphragm composed of the ceramic membrane and the polymer electrolyte is located in a In electrolyte; Step 2, set the electric field potential and the chemical reaction formula during charging: Step 3: Use the tertiary current distribution module to introduce the Nernst-Planck equation to describe the ion transport process and solve the polymer electrolyte concentration; use the Butler-Volmer equation to describe the electrode kinetic reaction and combine it with Faraday's law to solve the ion flux; Step 4: After the charging reaction, the amount of zinc produced at each position on the diaphragm is obtained by measuring the deformation of the membrane surface.

10. The method for predicting diaphragm zinc deposition by physical field simulation according to claim 9, characterized in that: In the step 2, the electric field potential is set to 10-30 mV.

Images