Spiral-potato–like nanostructure COF enabling nearly unidirectional zinc deposition for Ah-level pouch cells Authors: Fengxue Duan, Qi Li, Xiaohui Wu, Xinyi Chen, Zixiong Zeng, Tianyu Wu, Haifu Zhang, Yuting Chen, Wenhai Feng, Yifa Chen, Ya-Qian Lan Published: 2026 ABSTRACT A major challenge in realizing ampere hour (Ah)–level aqueous zinc (Zn) ion pouch batteries (Ah-level AZIBs) is the requirements for high Zn utilization rate coupled with even Zn nucleation/growth. Here, we report a spiral-potato–like covalent organic framework (SP-COF) nanostructure to construct isotropic solid electrolyte interphase for Ah-level AZIBs. Compared with conventional COF nanostructures, the unique spiral-potato–like SP-COF combined with affinity sites and porosity enables isotropic interfacial regulation to achieve nearly unidirectional Zn deposition following the theoretical three-dimensional nucleation curve. As a result, symmetric cells based on SP-COF/Zn electrode display ultrastable plating/stripping over 2800 hours at 10 milliamperes per square centimeter. The successfully assembled Ah-level triple stacked layer pouch cell (SP-COF/Zn||I2, 1.2 Ah) demonstrates stable cycling over 200 cycles with an ultrahigh Zn utilization rate of 40%. The pouch cell achieves an unprecedented energy density of 156 watt-hours per kilogram. These findings highlight the strategic fine-tuning of COF-based battery architectures to be a promising route for high energy density Ah-level AZIBs. INTRODUCTION The domestic aqueous zinc ion batteries (AZIBs) industry is now at the pilot-scale stage, and the development of Ah-level cells with stable performance and reproducible fabrication is essential for their practical implementation (1). With advantages including large capacity, intrinsic safety, and low cost, Ah-level AZIBs hold considerable potential for large-scale renewable energy storage systems (2–5). The main challenges for Ah-level AZIBs mainly include the following three points: (i) Nonuniform Zn plating in large-area pouch cells: Under deep discharge conditions, uneven current distribution induces localized overdeposition, forming dendrites/dead Zn and causing rapid capacity fade and safety risks. (ii) Mass-transport limitation and increased polarization in thick electrodes: High areal capacity/thick cathodes lengthen ion diffusion paths, leading to stronger concentration polarization and poorer rate capability. (iii) Exacerbated side reactions and hydrogen evolution issues: Hydrogen evolution, corrosion, and passivation by-products accumulate more readily and consume electrolyte, increasing impedance, promoting pouch swelling, and accelerating performance decay (6–10). To conquer these challenges of AZIBs, the central issue is achieving precise, uniform Zn plating/stripping under high zinc utilization rate, which is essential for maintaining high performance. A key strategy for stabilizing the zinc anode is the construction of solid electrolyte interphase (SEI) on the zinc surface, which has been widely recognized as an effective method for regulating Zn2+ flux, suppressing side reactions and stabilizing the interfacial electric field distribution (11–13). However, traditional SEI materials are mostly based on nanostructures like nanoplates, nanocubes, corn-cob–like nanostructures, or array structures (Fig. 1A and fig. S1) (14–20). Due to the lack in isotropic distribution of SEI microstructures, traditional SEI materials generally fail to achieve nearly unidirectional Zn deposition. In particular, for the conditions under high Zn utilization rate (i.e., practical high areal capacities), Zn deposition becomes increasingly anisotropic and prone to dendrite growth, thereby undermining interfacial stability and falling short of the stringent requirements of Ah-level AZIBs. In view of these limitations, an ideal SEI should feature ordered and isotropic arrays with abundant active sites to enable nearly unidirectional Zn deposition, high energy density, as well as low cost under high Zn utilization rate conditions, entailing ‌customized nanostructure engineering within latent functional materials. Advantages of an isotropic spiral-potato–like COF SEI over traditional SEI layer in AZIBs. (A) Conventional SEI morphology design. (B) Spiral-potato–like COF SEI layer. Covalent organic frameworks (COFs) have emerged as a highly promising platform for SEI materials owing to their tunable functionalities, pore architectures, nanostructural design, etc. These merits make COFs particularly attractive for the rational design of advanced SEI layers in AZIBs (21–23): (i) COFs, through the integration of zincophilic groups, enable precise control over Zn2+ flux and interfacial electric field, which promotes uniform Zn deposition and reduces the nucleation barrier (24). (ii) The tunable pore sizes of COFs can be tailored to match Zn2+ transport pathways, ensuring homogeneous Zn2+ flux and mitigating localized current density (25). (iii) The nanostructural engineering offers an effective strategy to regulate Zn plating/stripping behaviors and enhance cycling stability under depth of discharge. In recent years, several COF-based SEI, such as fluorinated COF (26), zincophilic heterometallic cluster COF (27), and copper phthalocyanine COF (28) have been demonstrated to regulate Zn2+ flux and suppress dendrite growth. In general, COFs reported to date are mostly based on simple nanostructures (e.g., nanosheets, nanospheres, nanotubes, and nanowires) (29). Only a few cases of specialized nanostructures like COFs based arrays have been reported, yet they are still limited to array construction strategies (e.g., postsynthetic coating and covalent grafting) and isotropic bottleneck in SEI layer, hardly meeting the potential demand of isotropic SEI layer for Ah-level AZIBs (30, 31). By leveraging the structural tunability, high specific surface area, and uniformly distributed pore channels of COFs, customizing COFs into unique nanostructures for isotropic SEI layer might provide a promising route toward Ah-level AZIBs with high Zn utilization rate (32, 33). Herein, we propose a spiral-potato–like COF (SP-COF) nanostructure to construct isotropic SEI layer for Ah-level AZIBs (Fig. 1B). As a result, benefiting from its unique spiral-potato–like nanostructure and abundant zincophilic sites, the SP-COF enables regulated zinc nucleation and growth, inducing a nearly unidirectional Zn deposition behavior. Such interfacial modulation effectively stabilizes Zn plating/stripping, as evidenced by the prolonged cycling stability of SP-COF/Zn electrode in symmetric cells under high depth of discharge. Furthermore, the feasibility of SP-COF/Zn is extended to Ah-level triple stacked layer pouch cells, which exhibits stable cycling behavior and high energy density. This work demonstrates a promising pathway by rationally designing COF-based interfacial architectures toward high energy density AZIBs. RESULTS Structural and characterization The SP-COF was constructed from 5,10,15,20-tetrakis(4-aminophenyl)porphyrin-zinc (Zn-TAPP) and 4,7-bis[3,5-di(4′-formylphenyl)phenyl]benzothiazole (BTTDA-CHO) via Schiff-base condensation in a mixed 1,2-dichlorobenzene (o-DCB) and n-butanol (n-BuOH) solvent system (Fig. 2A). The structure of SP-COF was confirmed by powder x-ray diffraction (PXRD) analysis combined with theoretical structural simulations with Materials Studio software package, revealing an AA stacking model (Fig. 2B and fig. S2). The PXRD pattern manifests diffraction peak at ~4.4°, which corresponds to (0 0 2) crystal plane with refined unit cell parameters (a = 31.7753 Å, b = 50.5619 Å, and c = 4.0093 Å) and low reliability factors (Rp = 3.52% and Rwp = 2.21%), thereby validating the accuracy and reliability of the calculated model (table S1). Moreover, SP-COF features one-dimensional (1D) pore channel extending along the c axis with calculated pore diameter of ~1.7 nm and interlayer distance of ~3.4 Å (fig. S3). The presence of imine linkages in SP-COF was confirmed by Fourier transform infrared (FTIR) spectroscopy and 13C cross-polarization magic-angle spinning nuclear magnetic resonance (13C NMR) tests. In FTIR spectrum, the disappearance of peaks for two monomers at 3200 to 3500 cm−1 (N─H) and 1695 cm−1 (C═O), together with the emergence of a characteristic C═N peak at 1625 cm−1 (fig. S4), demonstrates the formation of SP-COF. Meanwhile, the distinct signal of the imine moieties at 158 parts per million in the 13C NMR spectrum further confirms the successful synthesis of SP-COF (Fig. 2D). The high-resolution x-ray photoelectron spectroscopy measurement shows that the binding energy values of 1021.8 and 1045.8 eV are assigned to Zn 2p3/2 and Zn 2p1/2, respectively, whereas 390.9, 284.5, and 166.7 eV correspond to N 1s, C 1s, and S 2p, respectively (fig. S5). The porosity and specific surface area of the SP-COF were determined by N2 sorption measurement at 77 K. The Brunauer-Emmett-Teller surface area (SBET) of SP-COF is calculated to be 1023 m2 g−1, and its adsorption-desorption isotherms display a reversible profile (Fig. 2C). The pore size distribution curve exhibits a centered pore width at ~1.7 nm, matching well with the simulated structural result. Besides, thermogravimetric analysis indicates that SP-COF exhibits thermal stability at about 200°C, followed by a pronounced decomposition at higher temperatures (fig. S6). The chemical stability of SP-COF was evaluated by immersing it in 2 M ZnSO4, n-BuOH, mesitylene (MES), CH2Cl2, acetonitrile (MeCN), and ethanol (EtOH) for 2 weeks. Notably, the PXRD patterns remain unchanged after the stability tests, indicating high chemical stability of SP-COF (fig. S7). Structure and characterizations of SP-COF. (A) The structure of SP-COF. (B) Powder x-ray diffraction (PXRD) patterns of SP-COF. a.u., arbitrary units. (C) N2 sorption curves tested at 77 K (inset is the pore size distribution profile). STP, standard temperature and pressure. (D) 13C NMR spectrum. ppm, parts per million. (E) High-angle annular dark-field (HAADF)–scanning transmission electron microscopy image. (F) Elemental mapping images. Scanning electron microscopy (SEM) observations show that SP-COF has a spiral-potato–like morphology with an average length of ~0.80 μm (Fig. 3B and fig. S8). In detail, the SP-COF consists of a nanospindle-like core decorated with surface-embedded nanoplates. The central diameter of SP-COF is ~0.23 μm, whereas the nanoplates have a thickness of ~20 nm. Transmission electron microscopy (TEM) image shows good agreement with the SEM results (Fig. 3C). High-resolution–TEM displays abundant and well-defined lattice fringes for the SP-COF. The Fourier transform image reveals an interplanar spacing of 2.0 nm, which can be indexed to the (0 0 2) crystallographic plane, indicating its good crystallinity (fig. S9). In addition, elemental mapping images show that Zn, C, S, and N elements are uniformly distributed throughout the entire SP-COF (Fig. 2, E and F). SEM and TEM monitoring of COF morphology. (A) Schematic illustration of the polymerization of Zn-TAPP and BTTDA-CHO under various solvent. (B) SEM image of SP-COF. (C) TEM image of SP-COF. (D) SEM image of CC-COF. (E) TEM image of CC-COF. (F) SEM image of NP-COF. (G) TEM image of NP-COF. (H) SEM image of NC-COF. (I) TEM image of NC-COF. (J) SEM images of SP-COF at different time. (K) SEM images of CC-COF at different time. h, hours. Morphology characterization of COFs By tailoring the solvent composition during the solvothermal synthesis, COFs with distinct morphologies, including spiral-potato–like nanostructure (SP-COF), corn-cob–like nanostructure (CC-COF), nanoplate (NP-COF), and nanocube (NC-COF) morphologies can be selectively obtained, underscoring the critical role of solvent-driven effects in directing COFs growth and assembly (Fig. 3A). Specifically, based on the solvent system (o-DCB/n-BuOH) used to obtain SP-COF, CC-COF was prepared by replacing the solvent with MES/methanol (MeOH). In addition, NP-COF and NC-COF were synthesized in n-BuOH and MES/n-BuOH solvent systems based on the similar procedures of SP-COF, respectively. The obtained morphology of CC-COF, NP-COF, and NC-COF was also thoroughly confirmed by using SEM and TEM techniques (Fig. 3, D to I). The PXRD and FTIR results of CC-COF, NP-COF, and NC-COF are similar to that of SP-COF (figs. S10 and S11). Moreover, compared with SP-COF, N2 sorption results indicate lower SBET values for CC-COF (920 m2 g−1), NP-COF (401 m2 g−1), and NC-COF (599 m2 g−1), which might be ascribed to the unique spiral-potato–like morphology that has more exposed surface area (fig. S12). To visualize the structural evolution process, the SP-COF and CC-COF were selected as representative samples. We tracked the solvothermal reaction products at different reaction time (6, 12, 48, and 72 hours) by SEM and TEM tests. For SP-COF, a nanospindle-like morphology (length, ~0.7 μm; and middle diameter, ~200 nm) was formed at 6 hours (Fig. 3J). Extending the reaction time to 12 hours led to the embedded growth of nanoplates on the middle area of the nanospindle-like nanostructure. From 12 to 48 hours, the nanoplates progressively extended toward the edge of the nanospindle-like nanostructure. At 72 hours, the entire nanospindle-like nanostructure was completely covered with nanoplates, forming the final spiral-potato–like nanostructure (Fig. 3J and fig. S13). During this process, the increasing intensity of the main diffraction peak at ~4.4° in the PXRD patterns of SP-COF indicates improved crystallinity (fig. S14). Additionally, the stepwise morphology evolution of CC-COF has also been visualized through SEM tests, showcasing its transformation during the synthesis process (Fig. 3K). Similar to SP-COF, CC-COF exhibits a nanospindle-like morphology at 6 hours. As the reaction proceeds to 12 hours, the nanospindle-like nanostructure undergoes gradual volumetric expansion, accompanied by the nucleation of nanoparticles on the surface. When the reaction was prolonged to 48 and 72 hours, the nanoparticles on the surface gradually increase in size to lastly form the corn-cob–like nanostructure. The PXRD results also show that the intensity of the main peak at ~4.4° increases with prolonged reaction time (fig. S15). Analysis of zinc deposition behavior on COFs/Zn electrodes Based on the obtained four morphologies by tuning the solvent system, the Zn deposition behaviors of them were further investigated (Fig. 4A). The detailed preparation method of the COFs/Zn electrodes has been provided in the Supplementary Materials. Among them, the unique spiral-potato–like nanostructure of SP-COF is expected to induce Zn nucleation and growth along its ordered array, thereby promoting uniform Zn plating. In contrast, the corn-cob–like nanostructure, nanoplate and nanotube structures lack isotropic interface. As a result, Zn deposition on these nanostructures evolves from initially sparse nuclei to agglomerated bulk clusters, which, in turn, triggers nonuniform Zn growth (Fig. 4A). In addition, the surface and cross-sectional morphologies before and after cycling (fig. S16, A to D), together with PXRD analysis, confirm that the SP-COF layer remains intact and firmly attached to the Zn substrate after 200 cycles, demonstrating its excellent structural stability (fig. S16E). Meanwhile, in addition to SP-COF (fig. S17A), we also characterized the surface and cross section of other COF-based electrodes with different morphology, such as CC-COF/Zn (fig. S17B), NP-COF/Zn (fig. S17C), and NC-COF/Zn (fig. S17D) before and after cycling. The results show that the COFs can be stably maintained after cycling. To verify the location of zinc deposition, we deposited Zn onto SP-COF/Cu electrode. Due to the electronic insulating nature (fig. S18A) and Zn2+-conducting property (fig. S18B) of SP-COF, electrons are blocked while Zn2+ ions can migrate through the coating. Therefore, Zn preferentially deposits at the bottom of the SP-COF, as confirmed by the EDS mapping of the cycled SP-COF/Cu electrode (fig. S19). To verify the above hypothesis, an analysis of zinc deposition capacity on different electrodes was conducted. As shown in Fig. 4C, the SP-COF/Zn electrode delivers the highest Zn deposition capacity, reaching up to 58 mAh cm−2. In comparison, the CC-COF/Zn, NC-COF/Zn, and NP-COF/Zn electrodes exhibit reduced maximum deposition capacities of 44, 37, and 29 mAh cm−2, respectively. SEM images of the cycled SP-COF/Zn electrode reveal nearly unidirectional stripe-like features (Fig. 4B). In contrast, CC-COF/Zn, NC-COF/Zn, and NP-COF/Zn electrodes exhibit disordered crystal orientations after cycling. This pronounced performance disparity can be primarily attributed to the unique spiral-potato–like nanostructure of SP-COF, which acts as an effective dendrite blocker and guides nearly unidirectional zinc deposition, thereby enabling evenly dense zinc metal deposition. Furthermore, Scharifker-Hills normalization fitted by chronoamperometry results (34) reveals evident differences in the nucleation and growth modes among the various electrodes (Fig. 4D). Specifically, SP-COF/Zn closely follows a 3D nucleation behavior, implying that additional nucleation sites continuously emerge while preexisting nuclei continue to grow. In comparison, CC-COF/Zn exhibits a weaker capability to regulate the nucleation process than SP-COF/Zn. In addition, NC-COF/Zn and NP-COF/Zn electrodes lead to pronounced current fluctuations (at t < tm), indicative of an unstable nucleation stage. For bare Zn, when t < tm, rapid Zn nucleation dominates and occupies the available nucleation sites. As t exceeds tm, these active sites become quickly covered by Zn deposits, leading to rapid exhaustion of active sites and consequently deviated nucleation curve from 3D nucleation curve. Zinc deposition behaviors of COFs/Zn and bare Zn electrodes. (A) Schematic illustration of the Zn electrodeposition process on SP-COF/Zn, CC-COF/Zn, NP-COF/Zn, and NC-COF/Zn electrodes. (B) SEM images of the cycled SP-COF/Zn, CC-COF/Zn, NP-COF/Zn, and NC-COF/Zn electrodes after removing COF coatings. (C) Limiting deposition capacity on SP-COF/Zn, CC-COF/Zn, NP-COF/Zn, NC-COF/Zn, and bare Zn electrodes. (D) Dimensionless current transients fitted to 3D nucleation models on SP-COF/Zn, CC-COF/Zn, NP-COF/Zn NC-COF/Zn, and bare Zn electrodes. (E) In situ optical microscope images of Zn deposition on SP-COF/Zn and bare Zn electrodes at different time with current density of 10 mA cm−2 (top view). Based on the superior nucleation behavior of SP-COF/Zn over CC-COF/Zn, NC-COF/Zn, and NP-COF/Zn electrodes, the SP-COF/Zn electrode was selected as the primary sample for subsequent investigations on Zn coatings in AZIBs. We also monitored the polarization response of symmetric cells to further elucidate the mass transport behavior. As shown in fig. S20, the SP-COF/Zn cell exhibits a delayed diffusion controlled current peak and a higher limiting current under steady state, indicating faster mass transport. Meanwhile, the activation energy obtained by fitting the impedance values at different temperatures shows that the SP-COF/Zn cell has a much lower activation energy (16.2 kJ mol−1) compared with bare Zn (29.2 kJ mol−1) (fig. S21). Meanwhile, we conducted in situ observations of the SP-COF/Zn and bare Zn electrode with 20 times enlarged electrolyte volume to monitor the time-dependent zinc electrodeposition behaviors via optical microscope and SEM tests (Figs. 4E and 5, H and I, and fig. S22). The SP-COF/Zn electrode consistently exhibits uniform and dense zinc deposition, whereas bare Zn shows apparent zinc protrusions. Electrochemical performances and mechanism study of SP-COF/Zn electrodes in AZIBs. (A) Long-term cycling performances of asymmetric cells at 5 mA cm−2 and 1 mAh cm−2. CE, coulombic efficiency. (B) Cycling performances of symmetric cells at depth of discharge of 80%. (C) PXRD patterns of Zn plating/stripping at different cycles on SP-COF/Zn at 10 mA cm−2. a.u., arbitrary units. (D) Calculated interfacial adsorption energies of SP-COF/Zn. (E) Zn2+ migration energy barrier in route 1 and route 2. (F) 3D AFM images of cycled bare Zn (left) and SP-COF/Zn (right) electrodes. (G) COMSOL simulated electric field distributions for SP-COF/Zn and bare Zn electrodes. SEM images of cycled bare Zn (H) and SP-COF/Zn (I) electrodes in expanded cells (2 cm by 2 cm) at 10 mA cm−2. Electrochemical performances of COFs/Zn electrodes in AZIBs Given the superior zinc deposition behavior enabled by SP-COF, we further investigated its cycling stability in symmetric and asymmetric cells. At 5 mA cm−2, the bare Cu electrode in asymmetric cell remains stable for only 250 cycles, whereas the SP-COF/Cu electrode delivers a significantly prolonged lifespan of up to 3400 cycles with an average CE of 99.80% (Fig. 5A). Compared with bare Cu, the SP-COF/Cu electrode maintains a low voltage hysteresis of 73 mV, indicating reduced internal resistance and effectively suppressed by-product formation (fig. S23). Even at a higher current density of 10 mA cm−2, the SP-COF/Cu electrode exhibits a much longer lifespan of up to 1400 cycles, outperforming the bare Cu electrode (fig. S24). In symmetric cells operated at 5 mA cm−2, the SP-COF/Zn cell remains stable for about 2800 hours, whereas the bare Zn cell fails after only 100 hours (fig. S25). Under a higher current density of 10 mA cm−2, it still operates stably for ~2400 hours (fig. S26). Notably, even under a depth of discharge of 80 and 40%, the SP-COF/Zn electrode maintains stable cycling for over 200 and 400 hours, respectively, highlighting its strong capability for dense Zn plating (Fig. 5B and fig. S27). The rate performance of the symmetric cells at current densities ranging from 1 to 50 mA cm−2 implies that the SP-COF/Zn electrode exhibits smaller and more stable polarization voltages (fig. S28). To elucidate the roles of zincophilic thiazole and Zn-TAPP units, TDA-COF and TAPP-COF were prepared through the condensation of 3,3′,5,5′-tetrakis(4′-formylphenyl)tertriphenyl (TDA-CHO; a thiazole-free ligand) with Zn-TAPP and BTTDA-CHO with TAPP (a Zn2+-free ligand), respectively. The successful syntheses of the TDA-COF and TAPP-COF were confirmed by PXRD tests (fig. S29A). In contrast, the symmetric cell performances of TAPP-COF/Zn and TDA-COF/Zn were compared at a current density of 5 mA cm−2, and their lifetime was both shorter than that of SP-COF/Zn cell, verifying the vital roles of zincophilic thiazole and Zn-TAPP units in enhancing the battery performance (fig. S29B). To further understand the Zn deposition behavior responsible for the superior cycling stability, the postcycling morphology of the electrodes was examined. SEM images reveal that the surface of the SP-COF/Zn electrode exhibits uniformly aligned stripe-like patterns after 120 cycles, whereas the bare Zn surface shows disordered zinc deposition orientations (fig. S30). These observations are further corroborated by atomic force microscopy (AFM) measurements (Fig. 5F and fig. S31). PXRD results indicate that, as the number of cycles increases, the exposure of the Zn (1 0 1) crystal plane becomes increasingly pronounced (Fig. 5C). To elucidate the origin of this crystallographic orientation, we calculated the interfacial binding energies between SP-COF and different Zn crystal planes. As shown in Fig. 5D and fig. S32, the interfacial binding energy between SP-COF and the Zn (1 0 1) plane is significantly higher than those of the Zn (0 0 2) and Zn (1 0 0) planes. This result indicates a thermodynamically more stable interaction between SP-COF and Zn (1 0 1) plane, which favors the preferential exposure of the (1 0 1) facet during Zn deposition. To more intuitively investigate the Zn migration pathways, theoretical calculations reveal a significantly lower energy barrier for Zn2+ migration through nanochannel transport (route 1) than interlayer migration (route 2) (Fig. 5E and fig. S33). This result may arise from the zincophilic thiazole groups on the pore walls of the SP-COF that promote uniform Zn2+ flux. Electric field simulations further reveal that the SP-COF layer homogenizes the interfacial electric field, thereby reducing the electric field hotspots that can trigger dendrite growth (Fig. 5G). In contrast, bare Zn exhibits a highly nonuniform electric field distribution with pronounced localized intensification at surface protrusions, which results in uneven Zn2+ flux and accelerated dendrite nucleation and growth. Electrochemical performances of Zn||I2 cells To evaluate the practical applicability of SP-COF, its performance was examined in Zn||I2 full cells. The cyclic voltammetry curves of the SP-COF/Zn||I2 cell exhibit higher response currents, confirming the accelerated reaction kinetics (fig. S34). Rate performance tests from 1 to 50 mA cm−2 demonstrate the robustness of SP-COF, during which its capacity can be fully recovered when the current density returns to its initial value after rate cycling (fig. S35). Under long-term galvanostatic cycling, the bare Zn cell suffers from internal failure shortly after 1900 cycles, whereas the SP-COF/Zn||I2 cell maintains 100% capacity retention over 10,000 cycles (Fig. 6A). An anode-free SP-COF/Cu||ZnI2 cell delivers an energy density of up to 270 Wh kg−1 (based on the active materials of both anode and cathode) and operates for 260 cycles without capacity decay (Fig. 6B). In contrast, the corresponding bare Cu cell experiences a rapid capacity drop (41% capacity retention). Electrochemical performances of SP-COF/Zn||I2 and Zn||I2 full cells. (A) Long-term cycling performances of full cells at 5 A g−1. (B) Cycling performances of full cells at 1 A g−1 with an anode-free structure. (C) Schematic illustration of a triple-layer stacked SP-COF/Zn||I2 pouch cell. (D) Key parameters of SP-COF/Zn||I2 pouch cell. (E) Comparison of SP-COF/Zn||I2 full cell with the values reported in the literature. (F) Cycling performance of SP-COF/Zn||I2 pouch cell. To match the possibly practical application scenarios, a pouch cell (6 cm by 6 cm, three-layer stacked design) based on SP-COF/Zn||I2 was constructed (Fig. 6C). The components of the pouch cell are shown in fig. S36 (A to D). The SP-COF/Zn||I2 pouch cell was assembled with three I2 cathodes, four SP-COF/Zn anodes, and six separators, achieving an N/P ratio of 1.85, an I2 loading of 6 g, and a total capacity of 1.2 Ah (Fig. 6D and table S2). SP-COF/Zn||I2 pouch cell maintains 90% of its capacity after 220 cycles, and the minor thickness change of reacted iodine cathode is detected (Fig. 6F and figs. S36, E and F, and S37). Under this condition, the pouch cell achieves an energy density (calculated on the basis of the total mass of the Zn anode and iodine active material) of 156 Wh kg−1(cathode+anode), surpassing the performances of previously reported systems as far as we know (Fig. 6E and table S3) (35–44). Compared with recently reported COF-based Zn interfaces (45, 46), the isotropic spiral-potato–like SP-COF exhibits distinct material design, differing from conventional nanostructures. Its scalable spray-coating method ensures high Zn2+ conductivity and uniform coverage over large Zn electrodes, unlike thick coatings prepared via drop-casting or polyvinylidene difluoride–assisted dispersions. Mechanistically, this unique nanostructure enables nearly unidirectional Zn deposition. Considering the high safety of AZIBs, thus, fabricated pouch cell with high energy density enabled by SP-COF layer holds much promise in competing with the traditional lithium battery systems with safety risks. DISCUSSION In summary, we have successfully developed a SP-COF as an effective interfacial modifier to address the critical challenges of high zinc utilization and uniform Zn nucleation/growth in Ah-level AZIBs. Owing to its unique spiral-potato–like nanostructure and even zincophilic sites, the SP-COF enables isotropic interfacial regulation to achieve nearly unidirectional Zn deposition following the theoretical 3D nucleation curve. Notably, symmetric cell based on SP-COF/Zn electrode displays ultrastable plating/stripping over 2800 hours at 10 mA cm−2. The successfully assembled Ah-level triple stacked layer pouch cell (SP-COF/Zn||I2, 1.2 Ah) demonstrates stable cycling over 200 cycles with an ultrahigh zinc utilization rate of 40%. The pouch cell achieves an unprecedented energy density of 156 Wh kg−1. This work highlights the significant potential of rationally engineered COF-based interfacial architectures in enabling high energy density, large-scale AZIBs, offering a promising strategy for their practical application. MATERIALS AND METHODS All solvents and reagents were sourced commercially and used as received without further purification. Zn-TAPP, TAPP, TDA-CHO, and BTTDA-CHO were obtained from Shanghai Kylpharm Co. Ltd. Zinc sulfate heptahydrate (ZnSO4·7H2O), EtOH, o-DCB, n-BuOH, MeCN, MES, tetrahydrofuran (THF), 1,4-dioxane, trifluoroacetic acid, and acetic acid (≥99.8%) were obtained from Sigma-Aldrich. Iodine (I2) and polytetrafluoroethylene were purchased from Aladdin Industrial Corporation. Celgard 2400 and Super P were acquired from commercial suppliers. Zinc and Ti foils were obtained from Saibo Electrochemical Materials. Synthesis of SP-COF To synthesize SP-COF, BTTDA-CHO (10.6 mg, 0.02 mmol), Zn-TAPP (11 mg, 0.02 mmol), and a mixed solvent of o-DCB (0.5 ml) and n-BuOH (0.5 ml) were introduced into a Pyrex tube. The mixture was sonicated for 15 min, followed by three freeze-pump-thaw cycles in a N2 liquid bath at 77 K. The reaction was carried out at 120°C for 3 days. After cooling to room temperature, the product was filtered; washed with THF, DMA, and acetone; and then dried under vacuum to yield the final product. Synthesis of CC-COF To synthesize CC-COF, BTTDA-CHO (10.6 mg, 0.02 mmol), Zn-TAPP (11 mg, 0.02 mmol), and a mixed solvent of MES (0.25 ml) and MeOH (0.75 ml) were introduced into a Pyrex tube. The following experimental steps were identical to that of SP-COF. Synthesis of NP-COF To synthesize NP-COF, BTTDA-CHO (10.6 mg, 0.02 mmol), Zn-TAPP (11 mg, 0.02 mmol), and n-BuOH (1 ml) were introduced into a Pyrex tube. The following experimental steps were identical to that of SP-COF. Synthesis of NC-COF To synthesize NC-COF, BTTDA-CHO (10.6 mg, 0.02 mmol), Zn-TAPP (11 mg, 0.02 mmol), and a mixed solvent of MES (0.1 ml) and n-BuOH (0.9 ml) were introduced into a Pyrex tube. The following experimental steps were identical to that of SP-COF. Synthesis of TAPP-COF To synthesize TAPP-COF, BTTDA-CHO (10.6 mg, 0.02 mmol), TAPP (11 mg, 0.02 mmol), and a mixed solvent of o-DCB (0.5 ml) and n-BuOH (0.5 ml) were introduced into a Pyrex tube. The following experimental steps were identical to that of SP-COF. Synthesis of TDA-COF To synthesize TDA-COF, TDA-CHO (10.2 mg, 0.02 mmol), Zn-TAPP (11 mg, 0.02 mmol), and a mixed solvent of o-DCB (0.5 ml) and n-BuOH (0.5 ml) were introduced into a Pyrex tube. The following experimental steps were identical to that of SP-COF.