Towards practical aqueous zinc metal pouch cell batteries via weakly bound water-mediated shear-thickening electrolyte

Abstract

Disordered dendrite growth and corrosion reactions of zinc negative electrodes remain critical challenges in aqueous zinc metal batteries. Regulating water states in shear-thickening non-Newtonian fluid electrolytes has emerged as a promising strategy to simultaneously suppress dendrite growth and corrosion for zinc negative electrodes. Herein, we design a multifunctional shear-thickening non-Newtonian fluid electrolyte, based on carboxymethyl cellulose and sulfonate silicon oxide nanoparticles, which addresses both issues through hierarchical regulation of water molecule states. Carboxymethyl cellulose converts free water into weakly bound water, thereby suppressing water-induced parasitic reactions. Concurrently, sulfonated SiO₂ nanoparticles form an integrated shear-thickening network with amylopectin and carboxymethyl cellulose while introducing abundant surface –SO₃⁻ groups that disrupt the strongly bound water layer at the zinc interface. This architecture enables localized mechanical stiffening at dendrite tips without compromising ionic conductivity. As a result, Zn | |Zn symmetric cells exhibit stable Zn plating/stripping for 900 h at 50 mA cm⁻² and 25 mAh cm⁻², and Zn | |I₂ pouch cells with a capacity of 1.5 Ah maintain stability over 200 cycles at 20 mA cm⁻². These findings offer a alternative pathway toward corrosion-resistant, mechanically adaptive aqueous and practical zinc pouch cells systems.

Zinc dendrite growth and corrosion limit zinc metal electrode performance. Here, authors design a weakly bound water-mediated shear-thickening electrolyte that integrates chemical regulation with mechanical reinforcement to suppress dendrites and corrosion, enabling stable zinc metal electrodes.

Introduction

Aqueous zinc metal batteries (ZMBs) have attracted increasing attention as promising candidates for large-scale energy storage, due to their inherent safety, low cost, and environmental friendliness^1,2. Despite these advantages, the practical deployment of ZMBs is hindered by persistent challenges associated with the zinc negative electrodes. Chief among these are the uncontrolled growth of zinc dendrites and parasitic side reactions, including hydrogen evolution reaction (HER) and zinc corrosion^3–5. These issues not only degrade cycling stability but also pose safety risks, such as gas accumulation and battery swelling^6,7. Notably, dendrite formation and side reactions often reinforce one another, highlighting the need for integrated strategies to stabilize the zinc interface and extend battery lifespan^8–10.

The relationship between water binding states and electrochemical stability is well established in theory and supported by extensive prior work. Free water molecules possess high mobility and weak coordination constraints, enabling them to readily access the electrode surface and participate in parasitic reactions. In ZMBs, HER proceeds through water reduction at the negative electrodes, and the resulting OH⁻ species subsequently react with Zn²⁺ and electrolyte anions to form corrosion byproducts. The activity of water molecules, therefore, governs both the thermodynamics and kinetics of these processes. Free water molecules exhibit the highest electrochemical activity because they are minimally constrained by hydrogen bonding. Consequently, reducing free water content and regulating water binding states have emerged as effective strategies for suppressing both HER and zinc corrosion^11–13. To address dendrite growth, shear-thickening non-Newtonian fluid electrolytes (ST-NNFEs) have emerged as a promising solution^14–17. Under static or low-stress conditions, ST-NNFEs exhibit low viscosity, which promotes uniform zinc-ion flux and favorable interfacial compatibility. When localized stress arises at emerging dendrite tips, the electrolyte rapidly stiffens, physically hindering further propagation¹⁸. However, achieving simultaneous suppression of water-induced side reactions while preserving shear-thickening functionality remains challenging. This difficulty originates from the multifaceted role of water molecules within ST-NNFEs¹⁹. On one hand, the water content and its binding state are essential in determining the shear-thickening characteristics. Insufficient water limits particle dispersion, leading to ineffective thickening, while excess water increases interparticle spacing and lubrication, thereby reducing mechanical responsiveness²⁰. On the other hand, free water molecules are known to promote HER and zinc corrosion, further compromising zinc electrode stability^21,22. Precise regulation of water binding states is therefore essential for balancing mechanical protection and electrochemical stability in ST-NNFEs.

In this work, we develop a weakly bound water-induced shear-thickening electrolyte (Amy-CMC/SiO₂-SO₃Zn) that integrates water-state regulation with stress-responsive rheological behavior to simultaneously suppress dendrite growth and mitigate water-induced side reactions. By optimizing the composition of amylopectin (Amy), carboxymethyl cellulose (CMC), and sulfonated silica nanoparticles (SiO₂-SO₃H), the electrolyte undergoes rapid mechanical stiffening under transient strain, effectively inhibiting dendrite propagation. Meanwhile, the abundant polar groups in CMC, including carboxyl and hydroxyl groups, undergo rapid mechanical stiffening under transient strain, effectively inhibiting dendrite propagation. The incorporated SiO₂-SO₃H nanoparticles not only reinforce the shear-thickening framework through physical interaction with the polymer matrix but also provide surface sulfonate groups that disrupt the hydration environment at the zinc interface, further suppressing corrosion. Through this coordinated chemical and mechanical regulation, the electrolyte maintains high ionic conductivity while enhancing interfacial stability. As a result, the Amy-CMC/SiO₂-SO₃Zn electrolyte enables stable, dendrite-free zinc deposition. Zn | |Zn symmetric cells exhibit a prolonged plating/stripping process for over 900 h at 50 mA cm⁻² and 25 mAh cm⁻², while Zn | |I₂ pouch cells (1.5 Ah) achieve 200 stable cycles at 20 mA cm⁻². This work offers an alternative perspective on resolving the corrosion-mechanics trade-off in ST-NNFEs and establishes a robust design strategy for practical zinc pouch cells.

Results

Targeted regulation of weakly bound water

Figure 1a schematically illustrates the detrimental roles of water molecules in conventional zinc negative electrodes, which can be classified into three types: strongly bound water, weakly bound water, and free water^23,24. Strongly bound water molecules form an interconnected network covering the zinc negative electrodes' surface. Free water molecules continuously replenish this network, sustaining parasitic side reactions. Additionally, free water molecules coordinate with zinc ions to form robust hydration shells, which impede ion diffusion and hinder zinc migration on the electrode surface. In contrast, weakly bound water represents an optimal intermediate state that balances reactivity suppression with ion transport: these water molecules are sufficiently constrained to reduce their electrochemical activity, yet remain mobile enough to facilitate efficient Zn²⁺ transport. To address these intertwined issues, we designed a composite ST-NNFE, denoted as Amy-CMC/SiO₂-SO₃Zn, as shown in Fig. 1b. In this system, CMC rich in polar functional groups (e.g., –COOH, –OH) acts as a structural scaffold, effectively converting free water into weakly bound water, thereby mitigating water-induced side reactions. Meanwhile, SiO₂-SO₃H nanoparticles act as structural fillers to enhance the shear-thickening response of the electrolyte, while their surface –SO₃⁻ groups disrupt the hydrogen-bond network associated with strongly bound water at the zinc interface. Supplementary Fig. 1 displays the hierarchical morphology of the SiO₂ nanoparticles, while Supplementary Fig. 2 reveals their porous architecture, which provides abundant anchoring sites for –SO₃⁻ groups. Furthermore, the porous framework accommodates a significant number of zinc ions, which is expected to promote efficient zinc-ion transport.

Fig. 1. Design concept and physical characterization of Amy-CMC/SiO₂-SO₃Zn.

a Schematic diagram for the different states and behavior of water molecules on traditional zinc negative electrodes. b Regulating mechanism of water molecule state and behavior in Amy-CMC/SiO₂-SO₃Zn electrolyte. c Rheological analysis of Amy-CMC/SiO₂-SO₃Zn electrolyte at 25 °C. d Raman spectra of different electrolytes, red, blue, and yellow peaks denote strong, medium, and weak hydrogen bonding interactions, respectively. e T₂ inversion spectra of LF-NMR and the percentage of strongly bound water, weakly bound water, and free water for different electrolytes, the calculation results are shown in Supplementary Table 1.

To investigate the shear-responsive behavior of the electrolytes, rheological measurements were performed. As shown in Fig. 1c, the Amy-CMC/SiO₂-SO₃Zn electrolyte exhibits frequency-dependent viscoelasticity at 25 °C. Both storage modulus (G’) and loss modulus (G”) increase with frequency beyond 1 Hz, indicating enhanced elastic and viscous resistance under mechanical stimulation. Notably, the loss factor (G”/G’) decreases from 1.1 to 0.1 (Supplementary Fig. 3), reflecting a transition from viscous-dominated to elastic-dominated behavior. This transition ensures fluid-like ion transport under resting conditions (low-frequency regime, <1 Hz) and rapid mechanical stiffening under localized dendritic stress (high-frequency regime, >10 Hz). Rheological measurements were further conducted at 0 °C and 50 °C to evaluate the practical applicability of the electrolyte (Supplementary Fig. 4). The Amy-CMC/SiO₂-SO₃Zn electrolyte preserves its characteristic shear-thickening behavior across this temperature range, with the transition from a viscous-dominated to an elastic-dominated response consistently maintained at all tested temperatures. This thermal robustness ensures reliable mechanical reinforcement against dendrite growth under realistic operating conditions. In contrast, the pure Amy electrolyte exhibits a decrease in G’ and G” with increasing frequency, accompanied by a sharp increase in the loss factor (>1) above 10 Hz (Supplementary Fig. 5). This response suggests a predominantly viscous character due to the high mobility of free water molecules, which impedes chain reorganization under dynamic stress. The addition of CMC alters the viscoelastic properties of the electrolyte. As shown in Supplementary Fig. 6a, increasing CMC content strengthens the elastic and viscous responses while maintaining elasticity-dominated behavior at high frequencies. However, when CMC content reaches 10 wt%, the resulting increase in crosslinking density reduces adaptability to frequency-dependent stress (Supplementary Fig. 6b). Therefore, an optimized formulation with 1 wt% CMC and SiO₂-SO₃H was selected to balance shear responsiveness and water regulation.

Fourier-transform infrared (FTIR) spectroscopy was employed to probe hydrogen-bond interactions in the various electrolyte formulations (Supplementary Fig. 7). The O–H stretching vibration shifts from 3230 cm⁻¹ in Amy to 3260 cm⁻¹ in Amy-CMC, and further to 3270 cm⁻¹ in Amy-CMC/SiO₂-SO₃Zn. Similarly, the O–H bending vibration shifts from 1630 cm⁻¹ in Amy to 1640 cm⁻¹ and 1650 cm⁻¹ in the CMC-containing and SiO₂-SO₃Zn electrolytes, respectively. These shifts are attributed to the introduction of polar groups (–OH and –COOH) by CMC, which strengthen hydrogen bonding and reduce vibrational frequencies^25,26. The presence of SiO₂-SO₃H further enhances hydrogen bonding, leading to additional spectral shifts due to interactions with surface –SO₃⁻ groups²⁷. Raman spectroscopy provides further evidence of water state modulation (Fig. 1d). The broad O–H stretching band (3000–3700 cm⁻¹) was deconvoluted into three peaks, corresponding to strong (3223 cm⁻¹), medium (3407 cm⁻¹), and weak (3560 cm⁻¹) hydrogen bonds^28,29. Strong hydrogen bonds are primarily associated with strongly bound water, while medium and weak hydrogen bonds represent weakly bound water and free water, respectively^30–32. The fraction of strong hydrogen bonds decreases from 34% in Amy and 33% in Amy-CMC to 31% in Amy-CMC/SiO₂-SO₃Zn. Similarly, the proportion of weak hydrogen bonds drops from 10% (Amy) and 4% (Amy-CMC) to 2% in the Amy-CMC/SiO₂-SO₃Zn electrolyte. In contrast, medium hydrogen bonds increase from 56% and 63% to 67%, suggesting a favorable transition toward a stable weakly bound water matrix. These results demonstrate that CMC reduces both strongly bound and free water content, while the inclusion of SiO₂-SO₃H further enhances water confinement.

Low-field nuclear magnetic resonance (LF-NMR) spectroscopy was used to quantify the distribution of water states. As shown in Fig. 1e, the T₂ relaxation time reflects proton mobility and water binding strength^33,34. Strongly bound water corresponds to T₂ values between 0.1 and 10 ms, weakly bound water lies between 10 and 100 ms, and free water appears beyond 100 ms^35,36. The area under each peak indicates the relative content of each water state. In the Amy-CMC/SiO₂-SO₃Zn electrolyte, two distinct peaks are observed at T₂₁ (0.01–5.1 ms) and T₂₂ (7.3–95.4 ms), representing strongly bound and weakly bound water, with relative proportions of 23% and 77%. In Amy-CMC, the same peaks are observed with similar positions and proportions (25% and 75%). However, in Amy, the spectrum reveals three components for 27% strongly bound water, 61% weakly bound water, and 12% free water. These findings confirm that CMC promotes the conversion of free water into weakly bound water, while SiO₂-SO₃H further enhances water regulation by reducing strongly bound water content and eliminating free water. The interactions between electrolyte components and water molecules were further analyzed through adsorption energy calculations (Supplementary Fig. 8, Supplementary Data 1). The strong interaction between CMC and free water molecules underscores the crucial role of CMC in transforming free water into a less reactive, weakly bound state.

Regulation of corrosion inhibition and zinc deposition process

To evaluate zinc metal corrosion under static conditions, Zn | |Zn symmetric cells were assembled and their interfacial impedance was monitored at 30 min intervals. As shown in Supplementary Fig. 9, throughout the monitoring period, the resistance of the Amy electrolyte continues to increase rapidly, suggesting progressive electrode corrosion. In contrast, both the Amy-CMC and Amy-CMC/SiO₂-SO₃Zn electrolytes show markedly slower resistance growth, with the Amy-CMC/SiO₂-SO₃Zn electrolyte ultimately stabilizing after an initial period, indicating effective suppression of corrosion. To further elucidate interfacial processes occurring at the electrode surface, distribution of relaxation time (DRT) analysis was conducted^37,38. As shown in Fig. 2a–c, both Amy-CMC/SiO₂-SO₃Zn and Amy-CMC electrolytes exhibit a single weak peak at τ = 0.01 s, while the Amy electrolyte displays two distinct peaks at τ = 0.01 and 1 s. The additional relaxation peak in the latter case suggests the formation of a substantial quantity of surface byproducts caused by parasitic corrosion reactions^39,40.

Fig. 2. Verification of the corrosion inhibition effect and regulation of zinc ion deposition.

DRT analysis of Zn | |Zn symmetric cells in a Amy-CMC/SiO₂-SO₃Zn, b Amy-CMC, and c Amy electrolytes by operando EIS evolution during standing at 30 min intervals. d XRD patterns of zinc electrodes using different electrolytes after EIS testing. In situ Raman spectra of H₂O signals at the electrode/electrolyte interface in e Amy and f Amy-CMC/SiO₂-SO₃Zn electrolytes, when Zn plating/stripping at 10 mA cm⁻². g In situ DEMS measurements of Zn | |Zn symmetric cells using different electrolytes, when Zn plating/stripping at 10 mA cm⁻². h Desolvation activation energies, and i CA curves of different electrolytes under an applied potential of −150 mV. All the tests were conducted under 26 °C ± 1 °C.

This observation was corroborated by X-ray diffraction (XRD) analysis (Fig. 2d), which reveals no detectable byproducts for the Amy-CMC/SiO₂-SO₃Zn and Amy-CMC electrolytes. In contrast, the characteristic peak corresponding to Zn₄SO₄(OH)₆·xH₂O (PDF no. 39-0688) appears in the Amy electrolyte, confirming corrosion-induced byproduct formation⁴¹. Tafel and linear sweep voltammetry (LSV) measurements (Supplementary Figs. 10 and 11) show that both Amy-CMC/SiO₂-SO₃Zn and Amy-CMC electrolytes yield lower corrosion current densities and higher corrosion potential compared to the Amy counterpart. Moreover, in situ Raman spectroscopy was employed to probe the dynamic evolution of water molecules at the electrode/electrolyte interface during Zn plating/stripping at 10 cm⁻² (Fig. 2e, f, Supplementary Fig. 12). The broad band in the range of 3000–3700 cm⁻¹ corresponds to the ν_O–H stretching modes of water. Compared with the Amy electrolyte, the interfacial ν_O–H signal in the Amy-CMC/SiO₂-SO₃Zn electrolyte exhibits a pronounced intensity decrease during continuous Zn plating/stripping, indicating a reduced population of electrochemically active water molecules at the electrode surface. This observation demonstrates that the Amy-CMC/SiO₂-SO₃Zn electrolyte effectively regulates the interfacial reaction environment by limiting water accessibility. Collectively, these results highlight the ability of adsorbed CMC chains and –SO₃⁻ functional groups to restructure the interfacial hydrogen-bond network and spatially confine water molecules at the electrode/electrolyte interface. Furthermore, the in situ differential electrochemical mass spectrometry (DEMS) measurements quantitatively confirm the suppressed hydrogen gas evolution during operation (Fig. 2g). For cells with Amy electrolyte, a strong H₂ response can be detected when Zn plating/stripping at 10 mA cm⁻², while no obvious H₂ signal can be seen during the entire cycle of the cells with Amy-CMC/SiO₂-SO₃Zn electrolyte, indicating that the HER is inhibited. These results highlight the role of CMC in immobilizing free water as weakly bound water, thereby minimizing its corrosive effects. The inclusion of SiO₂-SO₃Zn further enhances this suppression while also improving zinc ion transport. Electrochemical impedance spectroscopy (EIS) reveals ionic conductivities of 22.6, 17.5, and 17.2 mS cm⁻¹ for the Amy-CMC/SiO₂-SO₃Zn, Amy-CMC, and Amy electrolytes, respectively (Supplementary Fig. 13, Eq. (1)). Chronoamperometry (CA) and EIS-based calculations of zinc ion transference numbers reveal enhanced migrating ability in the Amy-CMC/SiO₂-SO₃Zn electrolyte (Supplementary Fig. 14, Eq. (2)). Temperature-dependent EIS measurements were used to determine the activation energy for ion transport (Fig. 2h, Supplementary Fig. 15, Eq. (3)). The activation energies for the Amy-CMC/SiO₂-SO₃Zn, Amy-CMC, and Amy electrolytes were 15.8, 28.9, and 43.2 kJ mol⁻¹, respectively. The lower activation energy in Amy-CMC/SiO₂-SO₃Zn indicates reduced ion diffusion barriers, likely resulting from the disruption of strong hydrogen bonding at the zinc/electrolyte interface by CMC and SiO₂-SO₃H, thereby facilitating faster interfacial transport^42,43. Figure 2i compares the zinc nucleation behavior of different electrolytes under an applied potential of −150 mV. The Amy-CMC/SiO₂-SO₃Zn electrolyte exhibits a characteristic indicative of progressive three-dimensional nucleation, alongside a significantly lower steady-state current density compared to the Amy-CMC and Amy electrolytes. This reduced current density and stabilized nucleation morphology are attributed to suppressed parasitic reactions (e.g., HER) at the electrode surface, which enables more uniform and stable Zn²⁺ nucleation. The above results demonstrate that the Amy-CMC/SiO₂-SO₃Zn electrolyte not only suppresses HER and zinc corrosion but also significantly enhances interfacial zinc ion transport.

Performance testing and electrode characterization of zinc deposition/stripping

To evaluate its influence on overall electrochemical performance, Zn | |Zn symmetric cells were cycled under various current densities. At a current density of 10 mA cm⁻² with a capacity of 5 mAh cm⁻² (denoted as 10@5), cells using the Amy electrolyte fail after 400 h (Supplementary Fig. 16). The Amy-CMC electrolyte extends Zn plating/stripping life to 1000 h, whereas the Amy-CMC/SiO₂-SO₃Zn electrolyte enables stable Zn plating/stripping for over 4000 h. Under more demanding conditions of 50@25, the Amy electrolyte fails to sustain normal operation. The Amy-CMC electrolyte shows limited stability, lasting approximately 200 h, while the Amy-CMC/SiO₂-SO₃Zn system maintains stable Zn plating/stripping for more than 900 h with minimal voltage polarization (Fig. 3a and Supplementary Fig. 17). The practicality of the Amy-CMC/SiO₂-SO₃Zn electrolyte under thin-electrode conditions was further evaluated using Zn | |Zn symmetric cells assembled with zinc foils of 10, 20, and 50 μm thickness at 10@5. These foil thicknesses correspond to markedly different depths of discharge (DOD) of approximately 85.4%, 42.7%, and 17.1%, respectively. As shown in Supplementary Fig. 18, all cells exhibit stable voltage profiles throughout Zn plating/stripping, demonstrating the robust adaptability and practical applicability of the Amy-CMC/SiO₂-SO₃Zn electrolyte across a broad range of zinc foil thicknesses. The rate performance of the electrolytes was further examined under current densities ranging from 1 to 50 mA cm⁻² (Fig. 3b). The Amy-CMC/SiO₂-SO₃Zn-based cells consistently exhibit low voltage hysteresis across the tested range, highlighting the enhancement in zinc ion transport dynamics.

Fig. 3. The Zn deposition/stripping behavior.

a The Zn plating/stripping performance of the Zn | |Zn symmetrical cells with different electrolytes at 50@25. b Rate performance of the Zn | |Zn symmetrical cells under varying current densities. c Electrochemical performance comparison of the Zn | |Zn symmetric cells with other reported electrolytes, the corresponding numerical data can be found in Supplementary Table 2^64–73. Quasi in situ AFM testing of zinc electrodes deposited under a deposition current density of 10 mA cm⁻² at 26 ± 1 °C, in (d–f) Amy electrolyte for (d) 100, (e) 500, and (f) 1000 s; g–i Amy-CMC electrolyte for (g) 100, (h) 500, and (i) 1000 s; and j–l Amy-CMC/SiO₂-SiO₃Zn electrolyte for (j) 100, (k) 500, and (l) 1000 s.

Zn | |Cu asymmetric cells were also assembled and tested under 10@5 to evaluate the reversibility of zinc plating and stripping. As shown in Supplementary Fig. 19, cells containing the Amy and Amy-CMC electrolytes exhibit lower cycle lifetimes and average Coulombic efficiencies (CEs) of 94.1% and 98.2%, respectively. These relatively low and unstable CE values are attributed to continuous parasitic side reactions, including HER and corrosion. In contrast, the Zn | |Cu cells with the Amy-CMC/SiO₂-SO₃Zn electrolyte demonstrate a significantly prolonged cycle life and a high average CE of 99.6%. This improvement is likely due to the combined suppression of side reactions and enhanced zinc ion transport kinetics. To comprehensively assess the cycling performance of Zn | |Cu asymmetric cells employing the Amy-CMC/SiO₂-SO₃Zn electrolyte, additional tests were conducted at reduced current densities of 1@0.5 and 2@1. As shown in Supplementary Figs. 20 and 21, the cells with Amy-CMC/SiO₂-SO₃Zn electrolyte exhibit good cycling stability and maintain a high CE of 99.7% under these conditions. These results further demonstrate the strong practical applicability of the Amy-CMC/SiO₂-SO₃Zn electrolyte. The nucleation overpotential of zinc was further investigated by varying the applied current density. As shown in Supplementary Fig. 22, across all tested current densities (1, 5, and 10 mA cm⁻²), the Zn | |Cu asymmetric cells containing the Amy-CMC/SiO₂-SO₃Zn electrolyte consistently exhibit lower nucleation overpotentials. This observation indicates improved zinc ion diffusion at the electrode-electrolyte interface and a more favorable nucleation process in the Amy-CMC/SiO₂-SO₃Zn electrolyte. Compared with previously reported gel electrolytes and those employing shear-thinning or shear-thickening properties, both symmetric and asymmetric cells using the Amy-CMC/SiO₂-SO₃Zn electrolyte exhibit competitive performance in terms of achievable current density, areal capacity, and cycle life (Fig. 3c, and Supplementary Table 2 and 3). This performance advantage originates from the synergistic integration of water-state regulation and shear-thickening behavior. While most reported non-Newtonian electrolytes primarily rely on mechanical suppression of dendrite growth, our electrolyte design concurrently addresses the critical challenge of water-induced side reactions. By converting free water into weakly bound states through tailored molecular interactions, the Amy-CMC/SiO₂-SO₃Zn electrolyte effectively suppresses HER and zinc corrosion, which are particularly severe at high current densities. Meanwhile, the shear-thickening response provides dynamic mechanical resistance against dendrite penetration. This dual-function mechanism enables highly efficient zinc deposition/stripping under demanding conditions, thereby delivering stable cycling performance.

The quasi in situ atomic force microscopy (AFM) results provide direct and compelling evidence of the distinct shear-thickening function of the Amy-CMC/SiO₂-SO₃Zn electrolyte. As summarized in Fig. 3d–f and Supplementary Table 4, under a deposition current density of 10 mA cm⁻², the zinc electrode in the baseline Amy electrolyte exhibits progressively unstable growth behavior. A mossy morphology emerges within 100 s (Ra = 121 nm), which evolves into a rough, bulky structure with continuously increasing surface roughness (Ra = 206 nm at 500 s; Ra = 306 nm at 1000 s). A similar trend of surface roughening and non-uniform deposition is observed in the Amy-CMC electrolyte (Fig. 3g-i), where blocky deposits form and the roughness increases to Ra = 150 nm at 500 s and 243 nm at 1000 s. In contrast, zinc deposition in the Amy-CMC/SiO₂-SO₃Zn electrolyte follows a fundamentally different evolution pathway (Fig. 3j–l). Although initial nucleation produces block-like features within 100 s (Ra = 76 nm), the subsequent growth stage exhibits a pronounced self-smoothing behavior. After 500 s, zinc deposition becomes delocalized from the initial protrusions, leading to a reduced surface roughness (Ra = 67 nm). By 1000 s, a highly uniform and compact zinc layer is formed, with the roughness further decreasing to Ra = 55 nm. This real-time, quasi in situ observation provides visual and quantitative evidence that the shear-thickening property of Amy-CMC/SiO₂-SO₃Zn electrolyte actively promotes uniform zinc deposition⁴⁴.

The morphologies of zinc electrodes after deposition/stripping at different length scales were examined using scanning electron microscopy (SEM) and confocal laser scanning microscopy (CLSM). The electrode cycled in the Amy-CMC/SiO₂-SO₃Zn electrolyte exhibits a dense and uniform zinc deposition morphology (Supplementary Fig. 23a). In contrast, the electrode from the Amy-CMC system displays a more porous and irregular surface, accompanied by localized corrosion features and flake-like byproducts (Supplementary Fig. 23b). For the Amy electrolyte, severe dendritic growth and extensive parasitic side reactions are clearly observed (Supplementary Fig. 23c). These morphological differences are further corroborated by CLSM analysis. The electrode cycled in the Amy-CMC/SiO₂-SO₃Zn electrolyte maintains a relatively smooth surface with an average roughness of approximately 0.18 μm (Supplementary Fig. 24a), whereas electrodes cycled in the Amy-CMC and Amy electrolytes exhibit pronounced surface undulations, with average roughness values of approximately 1.41 and 2.77 μm, respectively (Supplementary Fig. 24b and 24c). These results indicate that the Amy-CMC/SiO₂-SO₃Zn electrolyte effectively suppresses the development of surface irregularities during cycling, consistent with the presence of shear-thickening-induced mechanical regulation. To further visualize the dynamic deposition behavior, in situ optical microscopy was performed to monitor real-time zinc plating morphology (Supplementary Fig. 25). At 10 mA cm⁻², markedly different behaviors are observed. In the Amy electrolyte, zinc deposition becomes increasingly heterogeneous with time, and after ~40 min, localized protrusions emerge on the deposited layer and continue to grow, representing the early stages of dendrite formation. By contrast, the electrode operated in the Amy-CMC/SiO₂-SO₃Zn electrolyte exhibits uniformly distributed deposition throughout the entire process, without observable protrusions or surface instabilities. This sustained morphological uniformity is consistent with effective mechanical suppression of dendrite growth enabled by the shear-thickening electrolyte.

Verification of water molecule regulation and mechanical function

Molecular dynamics (MD) simulations were carried out to better understand the molecular-level behavior of zinc ions and interfacial interactions (Supplementary Data 1). Simulation snapshots illustrating the structural organization of the three electrolytes are shown in Fig. 4a, and Supplementary Fig. 26a and 26b. The radial distribution function and coordination number of zinc ions in the Amy electrolyte are shown in Supplementary Fig. 26c. A coordination number of 5.3 between zinc ions and water molecules is observed, indicating that zinc ions are tightly surrounded by water molecules, forming a stable solvation shell. Upon introducing CMC, the coordination number decreases slightly to approximately 5.0 (Supplementary Fig. 26d), suggesting competitive coordination between CMC and water molecules that partially displaces the water from the zinc ion solvation shell. In the Amy-CMC/SiO₂-SO₃Zn electrolyte, the coordination number further decreases to about 4.8 (Fig. 4b), reflecting the entry of –SO₃⁻ groups into the solvation shell. These groups effectively compete with water molecules for zinc coordination, thereby modifying the solvation environment. The spatial distribution of zinc ions, water molecules, and other electrolyte components near the electrode interface is shown in Fig. 4c, and Supplementary Fig. 26e, f. In the Amy system, zinc ions and water molecules show strong coordination near the electrode surface, with intensity gradually decreasing with distance. In the Amy-CMC system, this coordination at the electrode surface is weakened, as CMC molecules displace water and form a new coordination network with zinc ions, while the coordination in the bulk electrolyte becomes more prominent. This indicates disruption of the interfacial water network and enhanced weakly bound water interactions in the electrolyte. In the Amy-CMC/SiO₂-SO₃Zn system, the coordination of zinc and water shifts farther from the metal surface, where the surface of the zinc electrode is predominantly occupied by –SO₃⁻ groups.

Fig. 4. Theoretical calculation of water molecule regulation and verification of dendrite suppression by shear thickening properties.

a The MD simulation snapshot of Amy-CMC/SiO₂-SO₃Zn electrolyte. b The coordination number of zinc ions and the radial distribution function in Amy-CMC/SiO₂-SO₃Zn electrolyte. c Proportions of different coordination pair between Zn surface and the Amy-CMC/SiO₂-SO₃Zn electrolyte. d Depth profile curve of ToF-SIMS in negative mode of OH⁻, –SO₃⁻, and O⁻ fragments for the electrode with Amy-CMC/SiO₂-SO₃Zn electrolyte after Zn plating/stripping 10 cycles at 10@5. EQCM analysis for Amy-CMC/SiO₂-SO₃Zn electrolyte of the e frequency and f dissipation change versus time under 26 °C ± 1 °C. Force field simulations of Zn deposition during different diffusion time in (g–i) ST-NNFEs and j–l electrolyte without shear-thickening behavior.

To verify that the –SO₃⁻ groups disrupt the interfacial hydrogen-bond network of water molecules, the structure and chemical states of cycled Zn electrodes were examined using X-ray photoelectron spectroscopy (XPS) and time-of-flight secondary ion mass spectrometry (ToF-SIMS). In the Zn 2p spectrum prior to cycling, dominant peaks at 1021.8 and 1044.7 eV are observed, corresponding to metallic Zn (Supplementary Fig. 27a). After cycling, additional peaks emerge at 1022.4 and 1045.7 eV, which can be assigned to Zn²⁺ species in coordinated chemical environments. The appearance of these features indicates the formation of coordination interactions between Zn²⁺ and functional groups derived from the electrolyte. Analysis of the S 2p spectrum provides further insight (Supplementary Fig. 27b). Before cycling, a characteristic doublet at 169.5/170.3 eV is detected, originating from SO₄²⁻ species in the ZnSO₄ electrolyte. After cycling, two new spectral features appear, a peak centered at ~167.5 eV, assigned to –SO₃⁻ groups, and a doublet at 161.6/162.6 eV corresponding to Zn–S bonds. The presence of –SO₃⁻ signals on the cycled electrode confirms that the sulfonated SiO₂ nanoparticles participate actively at the electrode-electrolyte interface. More importantly, the emergence of Zn–S bonds provides direct chemical evidence for coordination between –SO₃⁻ groups and Zn²⁺ ions, supporting the proposed mechanism in which sulfonate groups modify the Zn²⁺ solvation environment^6,39. The O 1 s spectrum before cycling can be deconvoluted into three components (Supplementary Fig. 27c). The main peak at 531.4 eV is attributed to C = O bonds from the carboxyl groups of CMC, the peak at 532.1 eV originates from S–O bonds associated with SO₄²⁻ species, and a weaker component at 530.8 eV corresponds to Zn–O species⁴⁵. After cycling, all three components remain present, while the enhanced Zn–O contribution, together with the newly observed Zn–S bonds in the S 2p spectrum, indicates that sulfur- and oxygen-containing functional groups jointly participate in tailoring the interfacial coordination environment of Zn²⁺. ToF-SIMS analysis was further employed to probe the interfacial distribution and functionality of –SO₃⁻ groups. As shown in Fig. 4d and Supplementary Fig. 28 and 29, the ion signals of –SO₃⁻ and O⁻ exhibit pronounced spatial overlap across the analyzed region, indicating the persistent presence of sulfonate groups at the electrode-electrolyte interface during Zn plating/stripping. Notably, the OH⁻ signal remains weak throughout the entire depth profile, suggesting effective suppression of water-related side reactions, particularly during the early stages of cycling. Collectively, these complementary surface-sensitive characterizations demonstrate that –SO₃⁻ groups are actively involved at the zinc interface, where they modulate the local water structure and Zn²⁺ coordination environment. This interfacial regulation reduces water reactivity while maintaining efficient Zn²⁺ transport, thereby contributing to the enhanced electrochemical stability of the zinc negative electrodes.

These changes collectively indicate that the Amy-CMC/SiO₂-SO₃Zn electrolyte effectively regulates the solvation structure and interfacial ion adsorption, thereby improving zinc ion mobility and electrochemical performance. Accordingly, the mechanism for zinc ion diffusion and migration in the Amy-CMC/SiO₂-SO₃Zn electrolyte is proposed. As illustrated in Supplementary Fig. 30a, the interfacial region between the zinc electrodes and electrolyte is divided into two functional zones, a diffusion layer and a compact layer. In the diffusion layer, the presence of hydration shells increases the effective volume of zinc ions, reducing their diffusion rate. In the compact layer, a networked water structure forms an ordered shield that restricts zinc ion migration. In the Amy-CMC/SiO₂-SO₃Zn system, the competition between CMC and water molecules weakens zinc ion hydration and destabilizes the structured water layer. At the same time, the conversion of free water into weakly bound water minimizes the replenishment of hydration layers, resulting in faster diffusion of zinc ions. In the compact layer, the –SO₃⁻ groups on the surface of SiO₂-SO₃H disrupt the ordered networked water structure and introduce additional ion-conductive pathways. These effects synergistically improve zinc ion migration along the electrode surface (Supplementary Fig. 30b), contributing to the enhanced electrochemical kinetics and interfacial stability observed in this system.

To elucidate the role of the shear-thickening behavior of the Amy-CMC/SiO₂-SO₃Zn electrolyte in dendrite suppression, electrochemical quartz crystal microbalance with dissipation monitoring (EQCM) measurements were combined with COMSOL simulations. As shown in Fig. 4e and f and Supplementary Figs. 31 and 32, the multi-harmonic EQCM responses reveal fundamentally different interfacial mechanical characteristics. For the Amy-CMC/SiO₂-SO₃Zn electrolyte, the normalized frequency shift (|Δf_n/n | ) increases with increasing harmonic order (n), while the dissipation factor (ΔD) decreases correspondingly. This behavior indicates that the interfacial layer becomes stiffer and less dissipative at higher acoustic shear rates, which is characteristic of shear-thickening behavior. In contrast, the Amy electrolyte exhibits a decreasing |Δf_n/n| with increasing n, a hallmark of a viscous or weakly structured interface that softens under shear^46–48. Additionally, we performed EQCM measurements on pure ZnSO₄ electrolyte. In this system, |Δf_n/n| and |ΔD_n| show no dependence on harmonic order, consistent with the formation of a rigid passivation film on the electrode surface. The contrasting harmonic dependencies among these three systems provide clear differentiation. The harmonic-dependent stiffening observed in Amy-CMC/SiO₂-SO₃Zn originates from its intrinsic shear-thickening rheology rather than passive interphase formation (Supplementary Note 1). Complementary COMSOL simulations further support this mechanical suppression mechanism by modeling non-Newtonian fluid dynamics at dendrite tips. As shown in Fig. 4g-i, the mechanical resistance of the ST-NNFEs progressively intensifies and becomes highly localized at electrode protrusion tips as deposition proceeds. This localized stiffening effectively suppresses the further evolution of surface protrusions into dendrites. In stark contrast, simulations of an electrolyte lacking shear-thickening behavior show unchecked amplification of surface irregularities that rapidly develop into pronounced dendrites (Fig. 4j-l). In addition, analysis of the current density distribution reveals a coupled electrochemical effect (Supplementary Fig. 33 and 34). In the ST-NNFEs system, mechanical inhibition of tip growth concurrently alleviates current density inhomogeneity across the electrode surface, thereby promoting more uniform zinc ion deposition. Conversely, in conventional electrolytes, current density becomes increasingly concentrated at advancing dendrite tips, creating a positive feedback loop that accelerates non-uniform growth.

To comprehensively elucidate the synergistic roles of chemical regulation and mechanical reinforcement in suppressing side reactions and dendrite growth, a schematic is provided to delineate the specific function of each electrolyte component in breaking the corrosion-dendrite feedback loop (Fig. 5). In this framework, CMC and –SO₃⁻ groups suppress water-induced corrosion and promote a more homogeneous Zn²⁺ distribution by regulating water binding states and disrupting the Zn²⁺ hydration shell, thereby stabilizing the initial zinc nucleation process. Once local deposition inhomogeneities emerge, the shear-thickening response is activated, leading to a rapid, localized increase in viscosity that physically impedes further dendrite elongation. Importantly, this mechanical hardening is dynamic and self-adaptive, occurring preferentially at regions of concentrated mechanical stress.

Fig. 5. Schematic diagram of dendrite suppression in Amy-CMC/SiO₂-SO₃Zn electrolyte.

a In conventional electrolytes, the vicious cycle driving the disordered growth of dendrites, involving water-induced corrosion, imbalanced charge and ion transfer, and heterogeneous Zn nucleation and growth. b The Amy-CMC/SiO₂-SO₃Zn electrolyte interrupts this cycle by regulating water molecule, modulating interfacial ion transport, and coupling mechanical properties, thereby enabling uniform Zn deposition.

To directly evaluate the contribution of shear-thickening behavior, we designed a critical control experiment in which the chemical composition was maintained while the shear-thickening property was deliberately eliminated by adjusting the CMC content. FTIR spectra confirm that the control sample closely matches the Amy-CMC/SiO₂-SO₃Zn electrolyte in functional group composition (Supplementary Fig. 35). Both systems exhibit characteristic absorption bands associated with –SO₃⁻ groups (~1180 cm⁻¹) and –OH groups (~3300 cm⁻¹), with similar peak positions and intensities, indicating comparable surface functional group densities. LF-NMR measurements further demonstrate that the control sample possesses a nearly identical water-state distribution to the Amy-CMC/SiO₂-SO₃Zn electrolyte (Supplementary Fig. 36), with complete elimination of free water in both systems. This confirms that chemical regulation of water molecules remains effective in the absence of shear-thickening. Electrochemical characterization shows that the ionic conductivity and Zn²⁺ transference number of the control sample are only slightly lower than those of the Amy-CMC/SiO₂-SO₃Zn electrolyte, yet substantially higher than those of the Amy and Amy-CMC electrolytes (Supplementary Fig. 37 and 38, and Supplementary Table 5). Despite these similarities, rheological measurements reveal a decisive difference. The control electrolyte does not exhibit shear-thickening behavior, its storage modulus consistently exceeds the loss modulus across the entire frequency range, and the loss factor remains below 0.5, indicative of a non-shear-responsive material (Supplementary Fig. 39). Correspondingly, its electrochemical performance is markedly inferior. Under 10@5 conditions, the Zn plating/stripping stability of the control sample closely resembles that of the Amy-CMC electrolyte and is substantially shorter than that achieved with the Amy-CMC/SiO₂-SO₃Zn electrolyte (Supplementary Fig. 40). This performance disparity becomes even more pronounced at higher current densities (50@25), where localized stress concentrations intensify, and mechanical adaptability becomes essential. These results demonstrate that chemical and mechanical regulation play complementary and indispensable roles in achieving long-term zinc electrode stability. While water-state regulation suppresses corrosion and facilitates uniform nucleation, the dynamic mechanical hardening afforded by shear-thickening provides an additional, stress-responsive barrier against dendrite propagation. This chemo-mechanical synergy underpins the robust cycling stability of the Amy-CMC/SiO₂-SO₃Zn electrolyte, particularly under demanding operating conditions.

Performance testing of practical zinc pouch cells

To evaluate the contribution of the Amy-CMC/SiO₂-SO₃Zn electrolyte to the overall performance of full cells, Zn | |I₂ batteries were assembled and systematically tested. Cyclic voltammetry (CV) measurements were carried out at scan rates ranging from 0.2 to 1 mV s⁻¹ (Fig. 6a, b, and Supplementary Fig. 41). Both the Amy-CMC/SiO₂-SO₃Zn and Amy electrolyte-based systems exhibit stable redox peaks across varying scan rates, indicating good reversibility. However, the Zn | |I₂ battery with the Amy-CMC/SiO₂-SO₃Zn electrolyte shows a noticeably narrower reduction peak width, suggesting faster redox kinetics and enhanced electrode-electrolyte interactions^49,50. The rate capability of the Zn | |I₂ cells was further examined (Fig. 6c). At a specific current of 0.2 A g⁻¹, the Zn | |I₂ battery with the Amy-CMC/SiO₂-SO₃Zn electrolyte delivers an area capacity of approximately 3.2 mAh cm⁻², slightly higher than the 3.1 mAh cm⁻² achieved with the Amy electrolyte. At an elevated specific current of 5 A g⁻¹, the battery using the Amy-CMC/SiO₂-SO₃Zn electrolyte retains a capacity of 1.9 mAh cm⁻², whereas the capacity of the battery using the Amy electrolyte drops sharply to 0.8 mAh cm⁻². This performance degradation can be attributed to insufficient zinc ion transport at high current densities, which leads to severe polarization and capacity fading in the absence of effective interfacial regulation. The galvanostatic charge-discharge profiles at 5 A g⁻¹ are shown in Fig. 6d. The Zn | |I₂ battery with the Amy-CMC/SiO₂-SO₃Zn electrolyte exhibits a much lower voltage polarization compared to its counterpart, further confirming improved ion diffusion and reduced resistance across the interface. To assess self-discharge behavior, static tests of the Zn | |I₂ batteries were conducted. As shown in Supplementary Fig. 42, the voltage-time profile of the battery with Amy-CMC/SiO₂-SO₃Zn exhibit a slow voltage decay compared to that of Amy. The battery with Amy-CMC/SiO₂-SO₃Zn electrolyte also maintains a high CE of 93.2% after 48 h of rest (Fig. 6e). In contrast, the CE of the battery with the Amy electrolyte decreases to 75.8%, reflecting substantial self-discharge caused by parasitic reactions during idle storage.

Fig. 6. The electrochemical performance of Ampere-hour pouch cells.

Contour maps of CV curves for Zn | |I₂ batteries with a Amy-CMC/SiO₂-SO₃Zn and b Amy electrolytes. c Rate performance, d the charge and discharge curves at 5 A g^–1, and e self-discharge curves of Zn | |I₂ batteries with different electrolytes. f Cycling stability of the 1.5 Ah Zn | |I₂ pouch cell at 20 mA cm⁻². The insets are the photograph of the pouch cell consisting of two positive electrodes and three negative electrodes, and the voltage vs time curves for the 190 ~ 200 h. g Performance comparison of the Zn | |I₂ pouch cell with other reported ones.

For practical evaluation, a calendar-aging constant-current cycling protocol was employed, in which Zn | |I₂ full cells using different electrolytes were aged for 2 h after every 5 cycles. Under this testing condition, cells with the Amy-CMC/SiO₂-SO₃Zn electrolyte maintain a high average CE of 99.1% with minimal capacity decay throughout the test. In contrast, cells using the Amy electrolyte exhibit a significantly lower average CE of 89.1% and pronounced capacity fading after 60 cycles (Supplementary Fig. 43). Furthermore, high-capacity Zn | |I₂ pouch cells exceeding 1.5 Ah were fabricated (inset of 6 f). As shown in Fig. 6f, the pouch cell containing the Amy-CMC/SiO₂-SO₃Zn electrolyte demonstrates stable cycling over 200 cycles at 20 mA cm⁻², with a maintained discharge capacity of 1.5 Ah and a minimal capacity decay rate of 0.004% per cycle. In contrast, cells using the Amy and conventional ZnSO₄ electrolytes exhibit only moderate cycling stability, with pronounced capacity decay within 50 cycles (Supplementary Fig. 44), further highlighting the strong potential of the Amy-CMC/SiO₂-SO₃Zn electrolyte for practical applications. In addition, the pouch cell delivers an initial volumetric energy density of 70 Wh L⁻¹ (based on the total volume of all internal components excluding the packaging films, as detailed in Supplementary Note 2 and Supplementary Table 6). The corresponding initial specific energy is 33.6 Wh Kg⁻¹ at the whole-cell level and 119.8 Wh Kg⁻¹ when calculated based on the mass of the active materials. This performance demonstrates competitive advantages relative to previously reported high-performance Zn | |I₂ pouch cells, particularly regarding key metrics, such as active material loading and areal capacity (Fig. 6g). To further examine mechanical robustness and safety, puncture tests were carried out at different impact rates (Supplementary Fig. 45)^51,52. At higher puncture velocities, cells with the Amy-CMC/SiO₂-SO₃Zn electrolyte exhibit a rapid increase in mechanical stress, which is consistent with the expected shear thickening behavior of the non-Newtonian fluid. Voltage monitoring during puncture events reveals a sharp voltage drop to zero at a puncture rate of 1 mm min⁻¹ (Supplementary Fig. 46a), indicative of immediate short-circuit failure. Interestingly, at 100 mm min⁻¹, the cell shows an initial voltage fluctuation followed by a recovery phase, implying partial impact resistance (Supplementary Fig. 46b). This behavior suggests that the shear-thickening response of the electrolyte contributes to mechanical damping, offering additional safety advantages under abusive mechanical conditions. Overall, these findings demonstrate that the Amy-CMC/SiO₂-SO₃Zn electrolyte not only enhances electrochemical performance in Zn | |I₂ batteries but also improves safety through its unique rheological properties.

Discussion

In summary, we present a rationally designed weakly bound water-induced ST-NNFE that mitigates corrosion by regulating the binding state of water molecules. Simultaneously, the electrolyte’s inherent shear-thickening properties further inhibit dendrite growth, thereby improving electrode stability. In the Amy-CMC/SiO₂-SO₃Zn electrolyte, the strong interaction between CMC and water effectively transforms free water into weakly bound water, thereby inhibiting hydrogen evolution and zinc surface passivation. The embedded SiO₂-SO₃Zn serves two synergistic functions: 1) it disrupts the interfacial hydrogen-bond network within the electrical double layer, thereby enhancing Zn²⁺ transport kinetics; 2) it forms a mechanically responsive network with the polymer matrix that stiffens under localized stress, enabling effective suppression of dendrite propagation while preserving high ionic conductivity. As a result, Zn | |Zn symmetric cells using this electrolyte achieve dendrite-free cycling for 4000 h at 10@5, and for over 900 h under more demanding conditions of 50@25. In full Zn | |I₂ batteries, the CMC/SiO₂-SO₃Zn electrolyte enables favorable rate capability and long-term cycling stability. High-capacity pouch cells with approximately 1.5 Ah capacity maintain stability over 200 cycles at 20 mA cm⁻², with a minimal capacity decay rate of just 0.004% per cycle. Moreover, puncture tests reveal that the electrolyte exhibits impact-responsive mechanical reinforcement when subjected to high puncture rates (greater than 100 mm min⁻¹), providing additional protection against mechanical failure. This work provides a smart strategy for tuning the water structure in non-Newtonian fluid electrolytes and offers an alternative pathway toward the development of stable and safe zinc pouch cells for practical applications.

Methods

Preparation of SiO₂-SO₃Zn nanoparticles

Porous SiO₂ nanoparticles were first obtained via hydrolysis-polycondensation of tetraethyl orthosilicate using a light-assisted method⁵³. Hexadecyltrimethylammonium bromide (0.7 mmol, ≥99%, Aladdin Chemical Co. Ltd) and urea (2.5 mmol, ≥99%, Aladdin Chemical Co. Ltd) were dissolved in deionized water (5 mL). Separately, tetraethyl orthosilicate (0.5 mL, ≥99%, Aladdin Chemical Co. Ltd) was mixed with anisole (5.0 mL, ≥99.5%, Aladdin Chemical Co. Ltd) and 1-hexanol (0.25 mL, ≥99%, Aladdin Chemical Co. Ltd). These solutions were combined in a sealed flask and stirred vigorously. The mixture was irradiated with visible light at 90 °C for 20 h under continuous stirring. The resulting precipitate was collected by vacuum filtration, washed repeatedly with deionized water, and air-dried. Finally, the product was calcined at 575 °C for 5 h in air to remove the surfactant template, yielding flower-like porous SiO₂ nanoparticles. These SiO₂ nanoparticles were then immersed in excess H₂SO₄ for 24 h and separated by centrifugation (rotation speed ~6150 g, 20 min). Afterward, the particles were washed three times with deionized water and vacuum-dried at 70°C to yield SiO₂-SO₃H. To obtain SiO₂-SO₃Zn nanoparticles, the SiO₂-SO₃H were dispersed into the pre-prepared electrolytes.

Preparation of electrolytes

Amy (≥95%), CMC (M. W. 250000), and ZnSO₄·7H₂O ( ≥ 99%) were purchased from Aladdin Chemical Co. Ltd., a 2 M ZnSO₄ solution was prepared using deionized water. For Amy electrolyte, 3 g of Amy was mixed with 3 mL of 2 M ZnSO₄ solution, sonicated for 30 min, and left undisturbed for 4 h. For Amy-CMC electrolytes, CMC was added to 4 mL of 2 M ZnSO₄ at varying concentrations (1, 5, and 10 wt%), followed by stirring and sonication for 15 min. Then, 3 g of Amy was introduced, and the mixture was further sonicated for 30 min before resting for 4 h. For Amy-CMC/SiO₂-SO₃Zn electrolyte, 1 wt% CMC and 1 wt% SiO₂-SO₃H were added to 4 mL of 2 M ZnSO₄, stirred, and ultrasonicated for 15 min. Subsequently, 3 g of Amy was added and sonicated for another 30 min. The final mixture was allowed to stand for 4 h to ensure structural uniformity.

Preparation of I₂ electrodes

For high loading I₂ electrodes, an electrodeposition strategy was applied^54,55. A dry-blended mixture of YP50F (Canrd Technology Co. Ltd.), acetylene black (Canrd Technology Co. Ltd.), and PVDF (Canrd Technology Co. Ltd.) was roll-pressed onto Ti mesh substrates (100 mesh, Canrd Technology Co. Ltd.) by roller press (Shenzhen Kejing Star Technology Co. Ltd.). The resulting large-area electrodes were subjected to constant-current electrodeposition at 5 mA cm⁻² from an electrolyte containing 0.5 M KI and 1 M ZnSO₄. The deposition process was carried out in an apparatus connected to the Neware battery testing system, and the experiment was conducted at room temperature. The I₂ mass loading was adjusted by deposition time. After deposition, the electrodes were rinsed with deionized water and dried at 40°C for 12 h. I₂ electrodes were prepared with different mass loadings to suit specific electrochemical characterizations in this work. For CV analysis, a low loading (~1 mg cm⁻²) was utilized to better resolve the redox mechanisms. To evaluate rate performance and self-discharge, the areal loading was increased to approximately 17.5 mg cm⁻². The above I₂ electrodes are only loaded on one side. Calendar life was tested using a configuration identical to the pouch cells, featuring double-sided deposition with a single-sided loading of 37.5 mg cm⁻².

Materials characterization

The morphological and compositional characteristics of the samples were investigated using SEM (FEI Verios G4) equipped with EDS at 5 kV. CLSM was carried out on a DCM8 system. Quasi in situ AFM measurements were performed using a Dimension Icon microscope equipped with PeakForce QNM technology and a Scanasyst-Fluid probe. HRTEM and associated EDS mapping were performed using a double Cs-corrector TEM (FEI Themis Z) operated at 300 kV. Crystalline phase identification was performed by XRD (Bruker D8 Advance) with Cu Kα radiation (λ = 0.154056 nm). Scans were performed over a 2θ range of 5-80° with a step size of 0.013° (5712 data points) and a counting time of 0.1 s per step. FTIR spectra were acquired using a Bruker Tensor 27 spectrometer. EQCM tests were carried out using a Biolin QSense system. The measurements were performed using a three-electrode configuration consisting of a gold-coated quartz crystal (working electrode), a platinum plate (counter electrode), and an Ag/AgCl reference electrode. The gold sensor was cleaned by UV/ozone treatment for 30 min, followed by spin-coating the electrolyte at 3000 rpm for 50 s. EQCM measurements were coupled with CV, as shown in Supplementary Fig. 31. ToF-SIMS analysis was conducted on a Tescan AMBER GMH system using a Ga⁺ ion beam (30 kV, 1 nA). XPS analysis was conducted on a Thermo Scientific K-Alpha spectrometer with Al Kα radiation (hν = 1486.6 eV). Binding energies were referenced to C 1 s at 284.8 eV. Samples were transferred from an argon-filled glovebox using an air-tight vessel to minimize air exposure. DEMS measurements were performed using a QSA 100Li system (Hiden Analytical). High-purity argon was used as the carrier gas, and the Zn deposition/stripping process was conducted at a current density of 10 mA cm⁻². Raman spectra were conducted on a WITec Alpha300R system employing a 532 nm laser. In situ Raman spectroscopy was acquired using a HORIBA HR evolution Raman microscope with a 532 nm excitation. Viscoelastic properties were measured using an Anton Paar MCR 302e rheometer. LF-NMR T₂ relaxation measurements were performed using a MesoMR23-060H-1 spectrometer. The acquisition parameters were as follows: 90° pulse width (P1) = 6.8 μs, 180° pulse width (P2) = 10.8 μs, echo time (TE) = 0.2 ms, number of echoes (NECH) = 7000, recycle delay (TW) = 4500 ms, number of scans (NS) = 4, and receiver bandwidth (SW) = 125 kHz.

Assembly of cells

Symmetric and asymmetric cells were assembled in 2032-type coin cell configurations. Each cell consisted of a stainless steel spacer (1.0 mm thickness), a stainless steel spring, the working electrodes, and a glass fiber separator (Whatman GF/D, thickness ~0.4 mm, Canrd Technology Co. Ltd.). Zinc foil electrodes were mechanically polished with 10000-grit sandpaper and cleaned with ethanol. Symmetric cells used two 100 μm thick Zn foils (ϕ = 12 mm, Canrd Technology Co. Ltd.). Asymmetric Zn | |Cu cells used 100μm Zn and 10μm Cu foils (ϕ = 16 mm, Canrd Technology Co. Ltd.). The mass of the electrolyte was fixed to 0.1 g. For Zn | |I₂ coin cells assembly, polished Zn electrodes were coupled with I₂-loaded electrodes (ϕ = 12 mm). For the 1.5 Ah pouch cells, three positive electrodes and two negative electrodes were used using glass fiber as the separators (Whatman GF/D). Al packaging bag was used to accommodate the cell. The parameter of the components is listed in Supplementary Table 6. The sealing procedure was conducted by the automated vacuum sealing machine (Shenzhen Kejing Star Technology Co. Ltd.). All the assembling procedures were completed in atmosphere at room temperature.

Electrochemical measurements

All electrochemical measurements were performed using a Neware battery testing system and a Metrohm Autolab PGSTAT302N workstation. All battery cycling experiments were conducted in a thermostatic chamber at 26 ± 1 °C. Electrochemical tests of full cells were performed after resting for 4 h. EIS was conducted over a frequency range of 100 kHz to 0.01 Hz (12 data points per decade of frequency) with a 10 mV amplitude. The cells firstly rested for 5 min until the open circuit voltage became stable. Zn | |I₂ cells were cycled between 0.6-1.6 V (vs. Zn²⁺/Zn).

Ionic conductivity (σ) was calculated using a general equation (Eq. (1)):

$$\mathrm{\sigma }=\frac{\mathrm{L}}{\mathrm{RS}}$$

where R is resistance, L is electrolyte thickness, and S is the electrode-electrolyte contact area.

Zn²⁺ transference numbers (t_Zn²⁺) were calculated using equation (Eq. (2)):

$${\text{t}}_{{\text{Zn}}^{2+}=\frac{{\text{I}}_{\text{s}}\left(△\text{V}-{\text{I}}_0{\text{R}}_0\right)}{{\text{I}}_{\text{o}}\left(△\text{V}-{\text{I}}_{\text{s}}{\text{R}}_{\text{s}}\right)}}$$

where ΔV = 25 mV, I₀ and R₀ are initial current and resistance, and I_S and R_S are steady-state values. Activation energies (E_a) were obtained using the Arrhenius equation (Eq. (3)):

$$\frac{1}{{\text{R}}_{\text{ct}}}=\text{A}{\text{e}}^{\frac{{-\text{E}}_{\text{a}}}{\text{RT}}}$$

where R_ct is the charge transfer resistance, R = 8.314 J mol⁻¹ K⁻¹, T is temperature, and A represents the pre-exponential factor.

Computational detail

MD simulations were performed using the GROMACS with a 1 fs timestep⁵⁶. A cutoff of 1.0 nm was used for van der Waals and electrostatics. Electrostatic interactions were computed using the particle mesh Ewald method. The nonbonded interactions included van der Waals and electrostatic interactions. The cubic box dimensions were 6.4 × 5.6 × 10 nm³, with periodic boundary conditions. System 1 consisted of one Zn electrode, 364 Zn²⁺ ions, 364 SO₄²⁻ ions, 10 Amy molecules, and 10,000 H₂O molecules. System 2 contained one Zn electrode, 364 Zn²⁺ ions, 364 SO₄²⁻ ions, 10 Amy molecules, 2 CMC molecules, and 10,000 H₂O molecules. System 3 comprised one Zn electrode, 369 Zn²⁺ ions, 364 SO₄²⁻ ions, 10 Amy molecules, 2 CMC molecules, 10 SiO₂-SO₃²⁻ units, and 10,000 H₂O molecules. The OPC3 water model was used, and the interaction parameters for various ions were derived from the GAFF and Merze force fields⁵⁷. During the NVT ensemble simulations, temperature control was achieved using the V-rescale thermostat with a time constant of 2 ns. The system was evolved for 20 ns to reach thermodynamic stability. Structural and dynamical properties were subsequently evaluated based on the last 5 ns of the trajectory, after confirming equilibration⁵¹.

DFT calculations were performed to investigate the adsorption energy between H₂O molecules and various ions using the CASTEP method. The calculations employed the Generalized Gradient Approximation with the Perdew-Burke-Ernzerhof functional, which provides an accurate representation of exchange-correlation energy^58–61. A plane-wave basis set with a cutoff energy of 600 eV was used, along with ultrasoft pseudopotentials. Geometry optimizations were performed using the Broyden-Fletcher-Goldfarb-Shannon algorithm⁶². A 3 × 3 × 1 k-point mesh was adopted for all energy minimization calculations. Geometry optimization was deemed converged when the Hellmann-Feynman force acting on each atom was lower than 0.03 eV Å⁻¹. Meanwhile, the convergence thresholds for the total energy and ionic displacement were set to 1 × 10⁻⁵ eV per atom and 0.001 Å per atom, respectively⁶³.

Finite element simulations were carried out using COMSOL Multiphysics to investigate dendrite growth behavior in ZMBs. The model integrates three coupled physics modules: (1) a tertiary current distribution interface combined with the Nernst-Planck equations to describe ion transport and electrochemical reactions; (2) a deformed geometry interface to track the morphological evolution of the electrode surface during deposition; and (3) a solid mechanics interface to capture the mechanical response of the electrolyte. To represent the viscoelastic behavior of ST-NNFEs, the electrolyte was approximated as a hyperelastic material at the microscopic scale. The simulation domain consists of a 2D geometry representing a vertical cross-section of the cell, with the bottom boundary defined as the zinc electrode surface and the top boundary corresponding to the bulk electrolyte. The diffusion coefficient of Zn²⁺ ions was set to 1 × 10⁻⁹ m² s⁻¹. To account for the intrinsic inhomogeneity of electrodeposition, the local current density along the electrode surface was prescribed using a Gaussian distribution (Eq. (4)):

$$i={-i}_0\exp \left(-\frac{x_2}{{2\sigma }^2}\right)$$

where i₀ = 1000 A m⁻², σ represents the standard deviation controlling the distribution width, and x denotes the lateral coordinate along the electrode surface. A key feature of the ST-NNFEs is stress-mediated suppression of dendrite growth. Accumulated mechanical stress inhibits vertical dendrite propagation and promotes denser zinc deposition. This effect was modeled by coupling the local zinc deposition density (ρ_Zn) to the compressive stress (σ_c) according to the following relationship (Eq. (5)):

$${\mathrm{\rho }}_{\mathrm{Zn}}={\mathrm{\rho }}_{\mathrm{Zn0}}\left({\xi \sigma }_{\mathrm{c}}+1\right)$$

where ρ_Zn0 is the stress-free reference density of deposited zinc, and ξ is a density-stress coupling coefficient. Furthermore, uniform zinc deposition reduces current-density non-uniformity. To capture this coupled electrochemical-mechanical feedback, a dynamic correction to the current-density distribution was introduced, in which the standard deviation σ evolves as a function of dendrite peak height. This feedback mechanism reflects the self-regulating effect of surface smoothing on current redistribution during zinc deposition.

Author contributions

W.A. and Z.D. conceived and supervised the project; G.G. designed the experiments; G.G. performed the experiments with the help of X.H., Y.G., F.L., J.B., Z.Z. and X.Y.; B.L. carried out the DFT calculations; All authors discussed the results and contributed to writing the manuscript; All authors have approved the final version of the manuscript.

Peer review

Peer review information

Nature Communications thanks Olivier fontaine, Jiaojiao Deng and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. A peer review file is available.

Funding

This work was financially supported by the National Natural Science Foundation of China (22279104, 62288102 and 61935017) and the National Key Research and Development Program of China (2020YFA0709900).

Data availability

All data are available in the main text or the supplementary materials. The source data for Figs and Supplementary Figs generated in this study are provided in the Source Data file. Source data are provided with this paper.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s41467-026-74014-y.