Graphene acid-enhanced interfacial layers with high Zn²⁺ ion selectivity and desolvation capability for corrosion-resistant Zn-metal anodes†

Kailai Xia^a, Liuyan Li^a, Yanbin Qiu^a, Jianqiang Weng^a, Shengtao Shen^a, Meixin Chen^a, Yuhang Zhuang^a, Yeye Wen*^b, Chengkai Yang*^a, Zheyuan Liu^a, Mingmao Wu*^ac and Zhigang Zou^d
aKey Laboratory of Advanced Materials Technologies, International (HongKong Macao and Taiwan) Joint Laboratory on Advanced Materials Technologies, College of Materials Science and Engineering, Fuzhou University, Fuzhou, 350108, P. R. China. E-mail: wumm20@fzu.edu.cn; chengkai_yang@fzu.edu.cn
^bAdvanced Research Institute of Multidisciplinary Science, Beijing Institute of Technology, Beijing, 100081, P. R. China. E-mail: wenyy@bit.edu.cn
^cFujian Science & Technology Innovation Laboratory for Optoelectronic Information of China, Fuzhou, Fujian 350108, P. R. China
^dDepartment of Physics, National Laboratory of Solid State Microstructures, Nanjing University, Nanjing, 210093, P. R. China

First published on 3rd August 2024

---

---

Introduction

Aqueous Zn-metal batteries (AZBs), emerging as viable alternatives to lithium-ion batteries, are increasingly considered promising candidates for large-scale grid storage owing to their high safety, low cost, and fair energy density.^1–3 However, several challenges, such as non-uniformity plating/stripping of Zn²⁺ ions and corrosion reactions triggered by aqueous electrolytes, significantly impede the advancement of AZBs.^4–11 The uncontrolled growth of Zn dendrites not only diminishes anode utilization but also poses risks such as separator failure and even short circuiting. What's worse, the corrosion reactions, including the hydrogen evolution reaction (HER) and other parasitic reactions involving electrolyte ions, exacerbate the unevenness of Zn anodes and lead to continuous electrolyte depletion.¹² To address these issues, it is crucial to identify and understand the Zn deposition chemistry and interfacial processes.

To date, various strategies have been investigated and conducted, such as the introduction of artificial interfacial layers, adjustment of electrolyte additives, and modification of Zn anode structures.^13–22 Among them, establishing a favorable interfacial layer is considered a highly reproducible and convenient technique for enhancing the stability of Zn anodes. And numerous interfacial protection layers have been developed, including surface coatings, artificial solid electrolyte interphase, and polymer membranes. Most of these layers demonstrate the capability to suppress Zn dendrite formation by providing a homogenized electric field distribution across the electrode surface and mitigating direct electrode–electrolyte contact.^23–25 However, it should be noted that the observed dendrite suppression often results from the inhibition of ion migration and charge transfer due to the passivation effect of these interfacial layers, leading to slower Zn²⁺ ion deposition. Essentially, although the delayed Zn deposition process can achieve partial inhibition of Zn dendrites at the early stage of the plating/stripping process, the issues of parasitic reactions caused by non-Zn components remain unresolved, which will inevitably lead to severe corrosion and dendrite growth after long-term cycling. Additionally, the introduction of interfacial layers also presents other challenges, such as interfacial adhesion, structural stability during long-term cycling, and side reaction inhibition capability. Thus, the design of the interfacial layer emerges as a systematic engineering challenge. This entails the interfacial layer to have the following functions: (i) selective and rapid transport/transfer of Zn²⁺ ions; (ii) good deformation resistance for long-term cycling; (iii) strong corrosion resistance.

In this context, cellulose nanofibers (CNFs), as a natural and cost-effective polymer, possess a combined profile of rich zincophilic functional groups, stable adhesion to Zn substrates, and an adequate porous structure, making them widely used in AZB manufacturing. Unfortunately, a pure CNF interfacial layer exhibits weak anion repulsive capability, resulting in an electrochemically inactive cation flux and a low Zn²⁺ transference number (t_Zn²⁺) of 0.2–0.4.²⁶ Previous studies have shown that this deficient t_Zn²⁺ tends to trigger the quick growth of Zn dendrites.²⁷ The CNF cannot also refuse water molecules to penetrate the inner Helmholtz plane (IHP), exacerbating the surface corrosion of Zn anodes. Additionally, the brittle attribute of CNF films presents challenges in withstanding repeated volumetric deformation of Zn-metal, thereby causing unstable cycling performances that may culminate in potential short-circuiting. Recently, some studies have focused on the modification of CNFs to improve the Zn²⁺ ion selectivity by grafting specific functional groups, including ester groups, amino groups, and sulfonic acid groups.^28–30 However, the brittleness of CNF membranes remains unresolved. Furthermore, while these methods have shown improvement in t_Zn²⁺, the values were limited to around 0.7, which may be due to inherent constraints of the grafting method, hindering the density of anion repulsive functional groups.

Here, by using carboxyl-enriched graphene acid (GA) as a filter and accelerator for Zn²⁺ ion transport, we develop a toughened, porous, and stable CNF/GA composite coating for Zn anode protection (Fig. 1a). In this CNF/GA coating, the CNF serves as the matrix, providing good contact with the Zn anode and abundant porous structures. GA, as a secondary oxidized graphene oxide (GO), has a large number of carboxyl groups. It can effectively cross-link CNFs and increase the strength and toughness of the coating layer, thus enhancing its structural stability during repeated Zn plating/stripping. Meanwhile, carboxyl-enriched GA facilitates the transport/transfer of Zn²⁺ ions. Density functional theory (DFT) calculations reveal a binding energy of −9.15 eV between Zn²⁺ and –COOH, notably lower than that of Zn²⁺ with H₂O (−4.59 eV) and Zn²⁺ with –OH (−5.74 eV). This promotes the desolvation process of Zn²⁺ ions, thereby inhibiting parasitic reactions involving water (Fig. 1b). Moreover, abundant carboxyl groups serve as an electrostatic shield, preventing anions from entering the IHP. This significantly enhances the selectivity of Zn²⁺ ions and boosts t_Zn²⁺ up to 0.81, reducing the generation of by-products (Fig. 1b). In addition, with CNF/GA coatings, Zn²⁺ ion flux is further optimized, which is conducive to the tight deposition of Zn²⁺ ions on the surface of the Zn anode along the (002)_Zn crystal planes, effectively avoiding growth of Zn dendrites (Fig. 1c). As a result, the Zn‖Zn symmetric cell with CNF/GA coatings shows an ultra-long lifetime of over 2900 h at a current density of 1 mA cm⁻², three times as high as the Zn‖Zn symmetric cell using CNF coatings and exceeding most symmetric cells using graphene-based and cellulose-based interfacial layers. Moreover, when the CNF/GA-protected Zn anodes are used in full cells, a stable capacity of 145 mA h g⁻¹ is obtained after 600 cycles. Additionally, the CNF/GA coatings exhibit a great self-discharge inhibition capability. Even after a 24 h period of standing, the full cell still demonstrates a capacity retention of 99%.

Experimental

Materials

1200 mesh graphite powder was purchased from Qingdao Huatai Lubrication and Sealing Technology Co., Ltd. Nanocellulose was purchased from Zhejiang Jinjiahao Green Nanomaterial Co., Ltd. KMnO₄, H₂O₂ (30%), H₂SO₄ (98%) and HCl (36–38%) were purchased from Sinopharm Chemical Reagent Co., Ltd. ZnSO₄·7H₂O (99.5%) and vanadium oxide (V₂O₅, 99.7%) were purchased from Aladdin. PVDF (≥99.5%) was purchased from Hefei Kejing Materials Technology Co., Ltd. Zn foil (99%), Cu foil (99%), Ti foil (99%), and stainless steel foil (99%) were purchased from Shengshida Materials Technology Co., Ltd. The glass microfiber filter membrane and non-woven fabric diaphragm were purchased from Whatman. Ketjen Black, CR2032 coin cells, Celgard diaphragm, aluminium tabs and aluminium plastic film cell bags were purchased from Guangdong Canrd New Energy Technology Co., Ltd.

Preparation of GA

Preparation of different anodes

Commercial Zn foil was polished to eliminate the passivation layer and subsequently cleaned with deionized water and ethanol. 1 g of 0.5 wt% aqueous solution of GA was added to 8 g of 2 wt% dispersion of CNF. The mixture was thoroughly blended by continuous stirring at 30 °C for 5 hours. Next, the prepared mixture was uniformly coated onto the surface of the Zn foil by the blade coating method. The modified Zn foil denoted as CNF/GA@Zn was then dried in a vacuum oven at 30 °C for 8 hours to remove the solvent. For comparison, Zn foil coated with CNFs was prepared and named CNF@Zn.

Preparation of the NaV₃O₈·1.5H₂O (NVO) cathode

1g of commercial V₂O₅ powder was added into 15 mL of NaCl aqueous solution (2 M). After stirring for 96 h at 30 °C, the suspension was centrifuged and washed with deionized water several times. Finally, the black-red product (NaV₃O₈·1.5H₂O) was obtained by freeze-drying for 3 days. To prepare the cathode, the NaV₃O₈·1.5H₂O was ground with acetylene black and PVDF adhesive in NMP in a weight ratio of 7:2:1 until the mixture formed a uniform slurry without any visible particles. Subsequently, the slurry was evenly cast onto stainless steel foil using the blade coating method. After vacuum drying at 60 °C for 12 h, the NaV₃O₈·1.5H₂O cathode was obtained. The mass loading of NaV₃O₈·1.5H₂O electrodes was 1.4 ± 0.2 mg cm⁻².

Battery assembly

A symmetric cell was fabricated by pairing two Zn electrodes of 12 mm diameter and 80 μm thickness in a CR2032 coin cell. Glass fiber and 2 M ZnSO₄ were used as the separator and electrolyte, respectively. Zn–Cu and Zn–Ti half cells are assembled using Zn anodes and copper or titanium cathodes. The full cell was assembled using the prepared NaV₃O₈·1.5H₂O cathode and Zn anode. For comparison, the Zn electrodes were replaced by CNF@Zn and CNF/GA@Zn, and the symmetric, half and full batteries were assembled in the same way. Additionally, symmetric cells with a non-woven fabric separator were assembled using the same way except that the glass fiber separator was replaced by the non-woven fabric separator. The pouch cell was assembled using the NaV₃O₈·1.5H₂O cathode, CNF/GA@Zn anode, Celgard separator, 2 M ZnSO₄ electrolyte, aluminium tabs, and aluminium plastic film cell bags.

Materials characterization

X-ray diffraction (XRD) patterns of the samples were recorded on a benchtop X-ray diffractometer (RIGAKU-Miniflex II) with Cu Kα radiation (λ = 1.5406 Å) in the range of 2θ = 5–90°. The morphology and structure of the samples were characterized using a field emission scanning electron microscope (Carl Zeiss Supra55) operated at 10 kV. The morphology and height distribution of GA were characterized by atomic force microscopy (Agilent 5500). The morphology and structure of GA and GO were characterized using a transmission electron microscope (Talos F200i). The three-dimensional morphology of the Zn anode surface was obtained using a confocal laser scanning microscope (KEYENCE VK-X150). X-ray photoelectron spectroscopy (XPS) measurements were performed on a Thermo ESCALAB 250XI spectrometer using non-monochromatic Al Kα X-rays as the excitation source and selecting C 1s (284.8 eV) as the reference line. The contact angle was measured on a Dataphysics OCA20 contact angle meter. Fourier transform infrared spectroscopy (FTIR) mapping was performed with a NEXUS 670 FT-IR instrument. Circular samples with a diameter of 12 mm and a thickness of 30 μm were fixed onto the puncture strength test device and punctured with a steel needle at a speed of 50 mm min⁻¹ for puncture tests. The zeta potential of CNF and CNF/GA solutions was measured using a zeta potential and nanoparticle sizer (NanoPlus3).

Electrochemical tests

Zn‖Zn, Zn‖Cu, Zn‖ Ti and Zn‖ NVO configured button cells (type CR2032) were assembled in an air atmosphere and electrochemical measurements were performed on a Neware (CT-4800) battery cycler. CE was evaluated on asymmetric Zn‖Cu button cells. Galvanostatic charge/discharge testing of full cells was performed at a current density of 1 A g⁻¹ over a voltage range of 0.3–1.5 V. CV measurements were performed on a CHI660E electrochemical workstation. EIS tests were performed from 100 kHz to 0.01 Hz using a CHI660E electrochemical workstation. In the charge transfer activation test, the Arrhenius curves was obtained by fitting the R_ct at different temperatures from 35 to 85 °C by the Arrhenius formula and E_a was obtained by calculating the slope of the curve. In the AC voltammetry test, the frequency was 6 Hz, the amplitude was 5 mV, and the potential range was extended from 0.9 to 0.1 V compared to Zn²⁺/Zn. Constant potential current–time transient curves were measured at a fixed overpotential of −150 mV. The ionic conductivity (σ) of the electrolyte was calculated from the following equation:

	(1)

where R_b is the ohmic resistance according to Nyquist plots by measuring the first intersection on the X-axis, L is the thickness of the sample and S is the contact area.

The transference number of Zn²⁺ ions (t_Zn²⁺) was calculated based on the following equation:

	(2)

In the formula, ΔV represents the applied voltage (25 mV); I₀ and R₀ represent the initial current and resistance, respectively; I_s and R_s represent steady-state current and resistance, respectively.

The activation energy (E_a) was calculated based on the following equation:

k = Ae−Ea/RT	(3)

k represents the rate constant, R represents the molar gas constant (8.314 J mol⁻¹ k⁻¹), T represents the thermodynamic temperature, E_a represents the apparent activation energy and A represents the frequency factor, respectively.

Computational methods

The density functional theory (DFT) calculations were performed using the Gaussian and CP2K programs. Structure optimizations were carried out at the DZVP level. All self-consistent loops were iterated until the total energy difference of the systems between the adjacent iterating steps was less than 1 × 10⁻⁶ eV. The 1 × 1 × 1 Monkhorst–Pack k-point grid for the Brillouin zone was used for k-point sampling. We have employed the density functional theory calculations within the generalized gradient approximation (GGA) using the PBE formulation. We have chosen the projected augmented wave (PAW) potentials to describe the ionic cores and take valence electrons into account using a plane wave basis set with a kinetic energy cutoff of 500 eV. Energy barriers were examined by linear and quadratic synchronous transit methods in combination with the conjugated gradient (CG) refinement.

The energy (ΔE) of binding energy and adsorption energy is defined, in which E_AB represents the energy of Zn²⁺ when adsorbed with H₂O/–OH/–COOH, E_A represents the energy of Zn²⁺, and E_B represents the energy of H₂O/–OH/–COOH.

ΔE = EAB − EA − EB	(4)

Results and discussion

Preparation and characterization of GA and CNF/GA

The molecular structure evolution during the preparation of GA and the schematic of CNF/GA coating preparation are depicted in Fig. 2a. Specifically, GA was synthesized via a continuous second oxidation method.³¹ The first oxidation is the conventional GO preparation according to the classical Hummers' method. During this process, graphite (1200 mesh, Fig. 2b) was oxidized and exfoliated into single-layered GO with an average size of 1 μm (Fig. 2c). Subsequently, a second oxidation step was conducted under identical conditions to further reduce the size of GO to 100 nm, yielding GA (Fig. 2d). Relative to the intermediate GO, the carbon skeleton of GA is greatly disrupted, resulting in increased edge defects, in which more oxygen-containing functional groups are generated (Fig. 2a). Transmission electron microscope (TEM) images reveal microstructural differences between GO and GA. As shown in Fig. 2e and S1.†, GO exhibits the typical sheet-like structure as previously reported, whereas GA exhibits a three-dimensional (3D) porous aggregation composed of numerous curved sheets. This supports the extremely disrupted structure of GA, wherein numerous small sheets form tight interconnections through hydrogen bonds generated from oxidation groups on the exposed edges of GA.³¹ Atomic force microscope (AFM) measurements further confirm this observation, revealing that the microstructure of GA in the thickness direction forms aggregates with surface porosity (Fig. S2†). Benefiting from the hydrophilicity conferred by the abundant edges and oxygen-containing functional groups of GA, it readily mixed with CNFs. Following blade coating, a translucent CNF/GA composite coating with a thickness of 30 μm was prepared (Fig. S3†). Scanning electron microscope (SEM) images demonstrate that the CNF/GA coatings retain the porous network of CNF morphology (Fig. 2a). In comparison with the pure CNF coatings, the incorporation of GA strengthens the composite coatings with a 59% increase in the toughness, providing enhanced puncture resistance (Fig. 2f).

X-ray diffraction (XRD), Fourier transform infrared (FTIR), and X-ray photoelectron spectroscopy (XPS) characterization experiments were carried out to elucidate the structure and chemistry evolution of the obtained samples (Fig. 2g–i). In the XRD patterns, graphite exhibits a sharp (002) diffraction peak at 2θ ≈ 26.5°, indicative of a d-spacing of 0.335 nm.³² Upon the first oxidation of graphite, lots of oxygen-containing functional groups are introduced between adjacent GO layers, causing the (002) reflection to shift to a lower value of 11.3° (d-spacing = 0.803 nm) (Fig. 2g). Further oxidation leads to the disruption of this layered structure, resulting in a broad diffraction peak centered at 29° for GA, consistent with the aggregated structure observed via TEM and AFM. For pure CNF coatings, the prominent diffraction peaks are located at 16.3° and 22.5° corresponding to (110) and (200) planes.³³ After mixing with GA, a slight leftward shift of the (200) diffraction peak is observed, attributed to the widened interlayer spacing of the CNF/GA coating owing to the spacing effect of GA. Fig. 2h shows FTIR spectra of GO, GA, CNF, and CNF/GA. The typical GO exhibits characteristic absorptions including C–O–C (∼1000 cm⁻¹), C–O (1230 cm⁻¹), bending modes of water molecules (∼1620 cm⁻¹), and CO (1740–1720 cm⁻¹) bonds.³⁴ Comparatively, GA displays significant enhancement at these oxygen-containing peaks, especially at the peak of 1720 cm⁻¹, suggesting a dramatic increase in the proportion of –COOH and –CO groups after second oxidation. This phenomenon is further corroborated by XPS analysis; the split-peak areas of O–CO bonds at 287.8 and 289.0 eV increase from 5.5% in GO to 12.6% in GA, while the split-peak areas of C–C, CC decrease from 53.7% in GO to 39.1% in GA (Fig. 2i and S4†). Additionally, the higher oxidation degree of GA is evident from its lighter coloration, indicating a greater degree of fragmentation of conjugated graphene domains and an increased coverage of oxidized functional groups (Fig. S5†). Upon the formation of the CNF/GA composite coating, the resulting samples exhibit characteristic absorption peaks corresponding to both CNFs and GA in the FTIR spectra, proving their firm bonding (Fig. 2h).

Corrosion resistance of CNF/GA coatings

To assess the protective efficacy of CNF/GA coatings on Zn anodes, the corrosion resistance of the coatings in electrolytes was initially examined. Fig. 3a–c and S6† show the SEM and digital images of the surface morphologies of bare Zn foil, Zn foil with CNF coatings (denoted as CNF@Zn), and Zn foil with CNF/GA coatings (denoted as CNF/GA@Zn) after soaking in 2 M ZnSO₄ electrolyte for 7 days and subsequent removal of surface coatings. Upon soaking, the bare Zn surface exhibits the presence of a host of flaky by-products, suggesting a galvanic cell reaction involving H₂O/O₂ at the Zn foil interface (Fig. 3a). In contrast, both CNF and CNF/GA coatings demonstrate an inhibitory effect on air and solution penetration, as evidenced by the relatively smooth morphologies of Zn surfaces (Fig. 3b–c). In particular on the surface of Zn in CNF/GA@Zn, there are almost no visible by-products. This is attributed to the incorporation of GA. It can increase the density of anions on the coating surface (Fig. S7†), thereby further avoiding corrosion caused by SO₄²⁻ ions and the consequent generation of by-products such as Zn₄SO₄(OH)₆·5H₂O. Furthermore, XRD tests verify the protective effect of the coatings (Fig. 3d). The diffraction peaks of Zn₄SO₄(OH)₆·5H₂O are almost not observed in the XRD patterns of CNF/GA@Zn, while they are distinct in the XRD patterns of bare Zn and CNF@Zn.

The HER stands as another primary corrosion factor causing cell failure in AZBs.³⁵ To evaluate this, the HER tendency of bare Zn, CNF@Zn and CNF/GA@Zn was studied using linear sweep voltammetry (LSV) at a scan rate of 10 mV s⁻¹, employing a two-electrode system (Fig. 3e).³⁶ LSV curves reveal that CNF/GA@Zn shows a more negative potential than bare Zn and CNF@Zn at a specific current density, representing lower H₂ generation on CNF/GA@Zn.³⁷ Also, Tafel fitting confirms the corrosion resistance of the CNF/GA coatings via linear polarization tests in a three-electrode system (Fig. 3f).³⁸ The corrosion current density of CNF/GA@Zn is only 0.25 mA cm⁻², significantly lower than that of bare Zn (1.74 mA cm⁻²) and CNF@Zn (0.34 mA cm⁻²), indicating the effective corrosion inhibition provided by the introduction of GA.

Furthermore, the impact of the coatings on nucleation behavior was investigated by chronoamperometry (CA) at a constant voltage of −150 mV. Fig. 3g shows that the current density of bare Zn continuously increases over 26 s, signifying rapid and uncontrollable 2D diffusion during nucleation.³⁹ In this process, Zn²⁺ ions diffuse to favorable nucleation sites and grow into dendrites, resulting in an expanding area exposed to aqueous electrolytes.⁴⁰ Conversely, CNF@Zn and CNF/GA@Zn exhibit a brief period of 2D diffusion followed by a stable 3D diffusion, indicating the stable diffusion nucleation of Zn²⁺ ions. Further comparison between CNF@Zn and CNF/GA@Zn shows that the addition of GA results in a higher steady-state current, implying improved transport kinetics of Zn²⁺ ions.⁴¹ This current trend is also evident in the ionic conductivity (Fig. S8†) and cyclic voltammetry (CV) profiles (Fig. 3h), where CNF/GA@Zn exhibits a larger peak current density compared to CNF@Zn, supporting enhanced Zn²⁺ transport in CNF/GA@Zn. However the highest current in bare Zn is attributed to the increased localized current caused by the inhomogeneous deposition of Zn. Moreover, electrochemical impedance spectroscopy (EIS) analysis illustrates the lowest charge transfer resistance (R_ct) of CNF/GA@Zn (Fig. 3i), further validating the low Zn²⁺ transfer barrier facilitated by GA. All these results demonstrate that while CNFs can stabilize the Zn anode to a certain extent, the introduction of GA further optimizes Zn²⁺ ion flow and the uniformity of Zn deposition, promising to serve as a great Zn anode protective layer.

Effect of GA on the desolvation process of Zn²⁺ ions

In order to clarify the source of the excellent corrosion resistance of CNF/GA coatings, we constructed a simplified molecular structure model of GA based on the results of material characterization and relevant literature (Fig. 4a).³¹ The intrinsic origin of the ability of CNF/GA to inhibit the HER and parasitic reactions was explored by simulating the interactions between electrodes and Zn²⁺ solvated molecules using density functional theory (DFT) (Fig. 4b, S9 and 10†). Typically, each Zn²⁺ can form a solvation shell with 6 molecules, including water molecules and anions. During Zn²⁺ deposition, some coordinated molecules, such as H₂O, would be released. Previous reports have shown that the high desolvation energy of Zn²⁺ ions through the outer Helmholtz plane (OHP) would induce local polarization and discontinuous Zn²⁺ transfer, thereby increasing the risk of notorious corrosion and uneven deposition.²⁷ Fig. 4c illustrates the calculated desolvation energies of Zn²⁺ ions across three different anodes. It is evident that CNFs reduce the desolvation energy of Zn²⁺ ions compared to bare Zn anodes. Notably, the addition of GA further reduces the desolvation energies, primarily because of the strong binding energy between Zn²⁺ and the rich carboxyl groups of GA (Fig. 4d), which facilitates the fast Zn²⁺ transfer (Fig. S11†).

Not only do theoretical simulations provide this opinion, but experimental results also support the promoted Zn²⁺ transfer assisted by CNF/GA coatings. In the bare Zn symmetric cell, a routine t_Zn²⁺ of 0.49 is measured, consistent with standard aqueous solutions, owing to the large desolvation barrier of solvated Zn²⁺ ions. In contrast, CNF coatings provide a higher t_Zn²⁺ of 0.63, attributed to their zincophilic functional groups. Furthermore, CNF/GA coatings exhibit a t_Zn²⁺ of 0.81, significantly surpassing both bare Zn and CNF coatings (Fig. 4e and S12†). More importantly, such a high t_Zn²⁺ approaches the single-ion conducting behavior, demonstrating the significant preponderance of GA for unique Zn²⁺ selectivity and promotion. Additionally, it is commonly believed that the desolvation process of Zn²⁺ is the rate-limiting step for charge transfer. To investigate the desolvation processes of these three anodes, EIS measurements were performed at increasing temperatures from 35 to 85 °C. The activation energy (E_a) representing the desolvation barrier can be calculated from the variation in R_ct using the Arrhenius equation. The results show that the E_a of CNF/GA@Zn is 15.71 kJ mol⁻¹, far lower than that of bare Zn (33.57 kJ mol⁻¹) and CNF@Zn (20.02 kJ mol⁻¹) (Fig. 4f and S13†), implying that CNF/GA coatings contribute to the removal of solvated molecules from the solvation shell of Zn²⁺ and enhance deposition kinetics, which is consistent with the theoretical simulations.⁴² Moreover, having composite coatings consisting of two hydrophilic materials, CNF/GA@Zn exhibits strong hydrophilicity, with a small contact angle of 46.7°, lower than that of bare Zn (90°) and CNF@Zn (53.6°) (Fig. 4g). A previous study has demonstrated that better wettability means the Zn anode is in fuller contact with the electrolyte, leading to the uniform deposition of Zn.⁴³

CNF/GA@Zn half-cell cycle stability

To assess the effect of CNF/GA@Zn coatings on the rate performance and cycling stability in the cells, long-term galvanostatic plating/stripping experiments using symmetric cells were performed across a wide range of current densities. Initially, at a lab-level current density of 1 mA cm⁻² (Fig. 5a and b), the Zn‖Zn cell exhibits a polarization of 73.3 mV in the early cycle, followed by a gradual increase in polarization and eventual failure after only 109 h. The CNF@Zn‖CNF@Zn cell with the CNF coating layer shows improved cyclability. However, its cycling life remains below 900 h, indicating that the efficacy of CNF coatings in protecting the Zn anode and suppressing dendritic growth remains limited. In contrast, the CNF/GA@Zn‖CNF/GA@Zn cell demonstrates an exceptionally extended stable cycling life exceeding 2900 h, with a low voltage hysteresis of 46.6 mV, verifying the uniform Zn²⁺ flux and quite fast transport kinetics facilitated by CNF/GA coatings. Further increasing the current density and specific capacity to practical operating conditions at 5 mA cm⁻² and 5 mA h cm⁻² (Fig. 5c), or even under extremely high conditions of 10 mA cm⁻² and 10 mA h cm⁻² (Fig. S14†), the CNF/GA@Zn symmetric cell still delivers ultra-long lifetimes of 850 h or 380 h, respectively. This remarkable Zn anode protective capability not only surpasses the performances of Zn‖Zn and CNF@Zn‖CNF@Zn cells but also outperforms some recently reported excellent Zn anode protection schemes including graphene-based and cellulose-based interfacial layers (Fig. 5d and Table S1†).^{28–30,44–49} Additionally, the good stability of the CNF/GA@Zn‖CNF/GA@Zn cell was confirmed using thinner non-woven fabric separators with a thickness of only 100 μm compared to the common glass fiber separators (650 μm), which requires coatings with superior dendrite suppression capabilities. Fig. S15.† illustrates the comparison of cycling performances between symmetric cells with non-woven fabric separators, wherein the CNF/GA@Zn‖CNF/GA@Zn cell exhibits the most stable cycling life over 1800 h, 100 times longer than that of the Zn‖Zn cell.

Furthermore, the rate performance of symmetric cells was evaluated at different current densities ranging from 1 to 10 mA cm⁻² with a capacity of 1 mA h cm⁻². As shown in Fig. 5e, when the current density increased to 5 mA cm⁻², the Zn‖Zn cell short-circuits quickly. Simultaneously, the polarization of CNF@Zn‖CNF@Zn cells deteriorates, eventually leading to cell failure. In contrast, the CNF/GA@Zn‖CNF/GA@Zn cell exhibits stable voltage profiles at all the current densities without abnormal fluctuations, indicating its great Zn²⁺ flux tolerance. Additionally, asymmetrical cells were prepared to study the coulombic efficiency (CE) variations under the protection of coatings.⁵⁰ As shown in Fig. 5f, the plating/stripping curves of half-cells were recorded at a current density of 1 mA cm⁻² and a specific capacity of 1 mA h cm⁻². The CNF/GA@Zn‖Cu asymmetrical cell exhibits a continuously stable lifetime over 1500 h, and the average CE remains as high as 99.11% (Fig. 5f). By contrast, the Zn‖Cu and CNF@Zn‖Cu cells suffer from fluctuated CE and low lifespan of only 100 h, which is in coincidence with the results of symmetric cells, indicating that the CNF/GA coating can effectively inhibit severe side reactions and dendrites.

Exploration of the Zn nucleation mechanism on the CNF/GA@Zn surface

For a deeper understanding of the advantages of Zn deposition under CNF/GA protection, the surface morphologies of bare Zn, CNF@Zn and CNF/GA@Zn were directly evaluated after 120 minutes of plating/stripping at a current density of 1 mA cm⁻² (1 cycle). To avoid observation bias caused by localized areas, a confocal laser scanning microscope (CLSM) was employed, covering statistical areas exceeding 1 × 1 mm² (Fig. 6a). By converting the 2D images into 3D contour maps, where color variations represent undulating morphologies (Fig. 6b), it is evident that the surface of bare Zn exhibits dendritic protrusions. These dendrites display a maximum altitude difference of approximately 34 μm, indicative of a large surface roughness of 2.27. Inversely, the application of CNF coating homogenizes the local deposition of Zn²⁺ ions with a surface roughness of 1.64. The further introduction of GA accelerates and stabilizes the Zn²⁺ ion flow, resulting in the smallest height difference (around 12 μm) and the lowest surface roughness (1.17) of CNF/GA@Zn. This conclusion is supported by high-magnification SEM images (Fig. S16†). Even after the 300 h cycling stability test, the SEM images of CNF/GA@Zn still show an intact surface with almost no corroded holes or glass fiber adhesion (Fig. S17†).

Analyzing the deposition kinetic behaviors provides insights into the underlying factors influencing the formation of diverse surface morphologies. Nucleation overpotentials (Δη) were evaluated at three different current densities and specific capacities (1 mA cm⁻², 1 mA h cm⁻²; 5 mA cm⁻², 5 mA h cm⁻²; and 10 mA cm⁻², 10 mA h cm⁻²) (Fig. 6d and Fig. S18†). For CNF/GA@Zn, the Δη increases with rising current densities, measuring 28.4 mV, 32.9 mV, and 36.0 mV. Bare Zn exhibits Δη values of 33.9 mV, 37.7 mV, and 55.1 mV, while CNF@Zn demonstrates Δη values of 47.5 mV, 48.1 mV, and 74.9 mV, respectively. An intriguing observation emerges: upon introducing the CNF coating, Δη increases, yet the resultant deposition morphology indicates that the CNF coating fosters a uniform deposition of Zn, yielding a smoother surface compared to bare Zn. More interestingly, following the introduction of GA into CNF, the Δη is notably reduced, even lower than that of bare Zn once again.

Indeed, Δη is defined as the disparity in driving force between the initial growth of crystal nuclei and progressive nucleation.⁵¹ In many studies, the introduction of an interfacial layer often increases the diffusion barrier and decreases the migration rate of Zn²⁺ ions. Consequently, this delays the time required for Zn²⁺ ions to attain the critical concentration for nucleation and deposition. However, deposition uniformity is not solely dependent on Δη. This is because any interfacial layer can homogenize the electric field on the Zn foil surface on account of the obstruction of Zn²⁺ ion migration (Fig. 6d), restricting the rapid growth of large crystal nuclei, as evidenced by the smaller current observed in the CA test (Fig. 3g). At this juncture, any coating can confer a certain degree of protection to the Zn anode. However, it is essential to note that Zn anode protection should not indiscriminately impede the flow of Zn²⁺ ions (Fig. 6d). Rather, it should facilitate the stable deposition and transport/transfer of Zn²⁺ ions while ensuring their uniform flow. The rational introduction of GA in CNF coating layers significantly alleviates the diffusion barrier of Zn²⁺ ions. Moreover, the fast desolvation process and excellent ion selectivity accelerate the concentration of Zn²⁺ ions at the interface to the critical concentration, thereby effectively reducing the nucleation overpotential. This, in turn, promotes the deposition of Zn in fine-grained form, preventing the growth of dendrites (Fig. 6d). In situ optical microscopy tests proved this; at a high current density of 10 mA cm⁻², CNF/GA@Zn and bare Zn showed very different growth morphologies. Compared to the granular aggregated growth on the bare Zn surface, the Zn on the surface of CNF/GA@Zn tended to grow in layers stacked parallel to the surface, and after 1 h CNF/GA@Zn showed a flatter surface, while the bubbles due to the hydrogen evolution reaction could not be observed during the whole process (Fig. S19†).

Furthermore, the quantitative Zn deposition results were assessed by XRD. Fig. 6e maps the presence of by-product diffraction peaks on the surfaces of bare Zn and CNF@Zn, while none is observed on CNF/GA@Zn. Fig. 6f depicts the relative intensities between (101)_Zn and (002)_Zn crystal planes (I₁₀₁/I₀₀₂) after the 1st, 25th, and 50th cycles. To ensure accuracy, the I₁₀₁/I₀₀₂ is obtained from three different positions (Fig. S20†).⁵¹ The results show that I₁₀₁/I₀₀₂ values of bare Zn and CNF@Zn continuously increase with the number of cycles. Conversely, for CNF/GA@Zn, the values of I₁₀₁/I₀₀₂ decrease from 1.25 to 0.85, indicating that the uniform deposition process would induce (002)_Zn crystal plane growth in CNF/GA@Zn, contributing to corrosion resistance.^52,53 Meanwhile, the digital images of the three different anodes after the 50th cycle are displayed in Fig. 6g. The bare Zn anode has lost its original shape due to uneven nucleation and stripping. The CNF@Zn anode exhibits a bumpy and reflective surface, and only CNF/GA@Zn shows a flat and non-adherent surface after removing the coating.

CNF/GA@Zn full cell performance

The excellent protection capability of CNF/GA coatings encourages evaluation of the practical performances of full cells (Zn‖NVO, CNF@Zn‖NVO, and CNF/GA@Zn‖NVO) with NaV₃O₈·1.5H₂O (NVO) serving as the cathode material. XRD results confirm the high purity of the NVO crystal structure (Fig. S21†). In CV curves (Fig. 7a) at a scan rate of 0.1 mV s⁻¹, the presence of similar redox peaks across the three different anodes indicates the minimal impact of the coatings on the energy storage mechanism of NVO. Notably, CNF/GA@Zn‖NVO exhibits the highest current density at the oxidation peak of 1.22 V, suggesting the improved transport kinetics of Zn²⁺ ions facilitated by CNF/GA coatings. Moreover, the effect of CNF/GA coatings on Zn²⁺ ion transfer was further validated using the galvanostatic charge/discharge (GCD) curves (Fig. 7b) and Nyquist diagrams (Fig. 7c), wherein the CNF/GA@Zn//NVO cell shows the lowest voltage polarization and the lowest R_ct of 142 Ω. Furthermore, Fig. 7d presents a comparison of the rate performance of the three full cells. The CNF/GA@Zn‖NVO cell shows specific capacities of 279, 242, 198, 163, and 120 mA h g⁻¹ at 0.1, 0.2, 0.5, 1, and 2 A g⁻¹, respectively, with over 96% capacity retention upon restoration of current density to 0.1 A g⁻¹. In contrast, the Zn‖NVO and CNF@Zn‖NVO cells exhibit inferior rate performance with only 74% and 84% capacity retention, respectively. Additionally, the long-term cycling performance tested at 1 A g⁻¹ is shown in Fig. 7e. The CNF/GA@Zn‖NVO cell maintains a specific capacity of 145 mA h g⁻¹ after 600 cycles, while rapid capacity degradation is observed in the Zn‖NVO cell, with only 67 mA h g⁻¹ remaining after 600 cycles. And the CNF@Zn‖NVO cell even experiences short-circuiting after 385 cycles.

Furthermore, the effectiveness of the CNF/GA coatings was further investigated by self-discharge tests. As depicted in Fig. 7f, after a 24 h period of standing, the CNF/GA@Zn‖NVO cell still demonstrates a capacity retention of 99%, significantly surpassing the 71% and 83% retention observed for Zn‖NVO and CNF@Zn‖NVO, respectively, as well as the capacity retention reported in most previous studies (Fig. 7g and Table S2†).^28,54–60 This excellent self-discharge inhibition capability of the CNF/GA coatings indicates that GA not only effectively promotes the transfer and deposition of Zn²⁺ ions during dynamic charge/discharge processes, but also inhibits the dissolution of Zn²⁺ ions and parasitic reactions in the static state. Finally, a pouch cell was assembled using CNF/GA@Zn as the anode and NVO as the cathode, successfully powering an electronic watch (Fig. 7h). These outstanding results demonstrate the broad advantages of CNF/GA@Zn anodes across various scenarios, highlighting their significant potential for the development of high-performance AZBs.

Conclusions

A novel Zn-metal anode interfacial layer, highly resistant to corrosion, was developed using GA and CNFs. In this CNF/GA coating, the CNF provides strong interactions with Zn foil and a porous structure conducive to Zn²⁺ ion transport. The GA component can effectively regulate the Zn²⁺ flux through the zincophilic carboxyl groups. As a result, the CNF/GA coatings not only enhance the Zn²⁺ ion selectivity and transport/transfer kinetics but also homogenize the Zn deposition on the Zn anode. Notably, this differs significantly from conventional interfacial layers or coatings, which merely rely on the inhibitory effect of the interfacial layer to delay Zn²⁺ ion transport and transfer. Therefore, they fail to prevent the corrosion of Zn anodes and achieve efficient selective Zn²⁺ ion transport. In sharp contrast, the CNF/GA coatings demonstrate a high t_Zn²⁺ of 0.81 and excellent corrosion resistance capability. Accordingly, the symmetric cell with CNF/GA coatings achieves a high cycling life of 2920 h at a current density of 1 mA cm⁻². Even under extreme conditions of 5 mA cm⁻², 5 mA h cm⁻² and 10 mA cm⁻², 10 mA h cm⁻², it remains stable for 850 h and 380 h, outperforming most Zn anodes with graphene-based and cellulose-based interfacial layers. Furthermore, the full cell with CNF/GA@Zn anodes demonstrates a high specific capacity of 145 mA h g⁻¹ even after 600 cycles at 1 A g⁻¹ and exhibits an ultra-low self-discharge behavior with 99% capacity retention after standing for 24 h. These exceptional results distinguish this CNF/GA coating from ordinary interfacial layers and coatings, offering a valuable insight into the fundamental protection of Zn anodes for the widespread application of AZBs.

Data availability

The data supporting this article have been included as part of the ESI.†

Author contributions

Kailai Xia: investigation, conceptualization, writing – original draft. Liuyan li: data curation, validation. Yanbin Qiu, Jianqiang Weng: methodology. Shengtao shen: visualization, data curation. Meixin Chen: data curation. Yuhang Zhuang: software, resources. Zheyuan liu, Zhigang Zou: resources, formal analysis. Mingmao Wu, Chengkai Yang, Yeye Wen: data curation, resources, writing – review & editing.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was financially supported by the National Natural Science Foundation of China (No. 22105040 and 22305017), the Fujian Science & Technology Innovation Laboratory for Optoelectronic Information of China (2021ZZ127), the Natural Science Foundation of Fujian Province of China (2021J01588), and the National Key Research and Development Program of China (2020YFA0710303).

Notes and references

1. D. Chen, P. Qing, F. Tang, H. Yu, P. He, H. Huang, Z. Wu, F. Sun, W. Wei, X. Ji and L. Chen, Mater. Today Energy, 2023, 33, 101272 CrossRef CAS .
2. M. Zhang, H. Hua, P. Dai, Z. He, L. Han, P. Tang, J. Yang, P. Lin, Y. Zhang, D. Zhan, J. Chen, Y. Qiao, C. C. Li, J. Zhao and Y. Yang, Adv. Mater., 2023, 35, 2208630 CrossRef CAS PubMed .
3. J. Hao, L. Yuan, Y. Zhu, M. Jaroniec and S.-Z. Qiao, Adv. Mater., 2022, 34, 2206963 CrossRef CAS PubMed .
4. J. Fan, J. Bai, H. Ma, X. Tian and B. Xu, Energy Storage Mater., 2023, 55, 12–20 CrossRef .
5. Z. Cai, J. Wang and Y. Sun, eScience, 2023, 3, 100093 CrossRef .
6. J. Hao, X. Li, X. Zeng, D. Li, J. Mao and Z. Guo, Energy Environ. Sci., 2020, 13, 3917–3949 RSC .
7. Q. Yang, Q. Li, Z. Liu, D. Wang, Y. Guo, X. Li, Y. Tang, H. Li, B. Dong and C. Zhi, Adv. Mater., 2020, 32, 2001854 CrossRef CAS PubMed .
8. L. Wang, M. Peng, J. Chen, T. Hu, K. Yuan and Y. Chen, Adv. Mater., 2022, 34, 2203744 CrossRef CAS PubMed .
9. Y. Tian, Y. An, C. Liu, S. Xiong, J. Feng and Y. Qian, Energy Storage Mater., 2021, 41, 343–353 CrossRef .
10. Y. Yang, C. Liu, Z. Lv, H. Yang, Y. Zhang, M. Ye, L. Chen, J. Zhao and C. C. Li, Adv. Mater., 2021, 33, 2007388 CrossRef CAS PubMed .
11. L. Zhou, F. Yang, S. Zeng, X. Gao, X. Liu, X. Cao, P. Yu and X. Lu, Adv. Funct. Mater., 2022, 32, 2110829 CrossRef CAS .
12. G. Liang, F. Mo, X. Ji and C. Zhi, Nat. Rev. Mater., 2021, 6, 109–123 CrossRef CAS .
13. X. Zhang, J. Li, H. Ao, D. Liu, L. Shi, C. Wang, Y. Zhu and Y. Qian, Energy Storage Mater., 2020, 30, 337–345 CrossRef .
14. Z. Yi, J. Liu, S. Tan, Z. Sang, J. Mao, L. Yin, X. Liu, L. Wang, F. Hou, S. X. Dou, H.-M. Cheng and J. Liang, Adv. Mater., 2022, 34, 2203835 CrossRef CAS PubMed .
15. S. Tan, Z. Sang, Z. Yi, J. Guo, X. Zhang, P. Li, W. Si, J. Liang and F. Hou, EcoMat, 2023, 5, e12326 CrossRef CAS .
16. J. Yu, Z. Yi, X. Yan, R. Chen, S. Tan, P. Li, T. Zhang, H. Zhang, J. Liang and F. Hou, Small, 2024, 2402055 CrossRef PubMed .
17. M. Liu, W. Yuan, G. Ma, K. Qiu, X. Nie, Y. Liu, S. Shen and N. Zhang, Angew. Chem., Int. Ed., 2023, 62, e202304444 CrossRef CAS PubMed .
18. Z. Yi, G. Chen, F. Hou, L. Wang and J. Liang, Adv. Energy Mater., 2021, 11, 2003065 CrossRef CAS .
19. Q. Li, Y. Zhao, F. Mo, D. Wang, Q. Yang, Z. Huang, G. Liang, A. Chen and C. Zhi, EcoMat, 2020, 2, e12035 CrossRef CAS .
20. H. Yu, Y. Chen, H. Wang, X. Ni, W. Wei, X. Ji and L. Chen, Nano Energy, 2022, 99, 107426 CrossRef CAS .
21. Y. Qiu, X. Zheng, R. Zhang, Q. Lin, M. Li, J. Luo, S. Yang, Z. Liu, Q. Wang, Y. Yu and C. Yang, Adv. Funct. Mater., 2024, 34, 2310825 CrossRef CAS .
22. J. Weng, W. Zhu, K. Yu, J. Luo, M. Chen, L. Li, Y. Zhuang, K. Xia, Z. Lu, Y. Hu, C. Yang, M. Wu and Z. Zou, Adv. Funct. Mater., 2024, 34, 2314347 CrossRef CAS .
23. H. Yu, Q. Li, W. Liu, H. Wang, X. Ni, Q. Zhao, W. Wei, X. Ji, Y. Chen and L. Chen, J. Energy Chem., 2022, 73, 565–574 CrossRef CAS .
24. A. Chen, C. Zhao, Z. Guo, X. Lu, N. Liu, Y. Zhang, L. Fan and N. Zhang, Energy Storage Mater., 2022, 44, 353–359 CrossRef .
25. Y. An, Y. Tian, K. Zhang, Y. Liu, C. Liu, S. Xiong, J. Feng and Y. Qian, Adv. Funct. Mater., 2021, 31, 2101886 CrossRef CAS .
26. S. Yang, Y. Zhang, Y. Zhang, J. Deng, N. Chen, S. Xie, Y. Ma and Z. Wang, Adv. Funct. Mater., 2023, 33, 2304280 CrossRef CAS .
27. L. Jiang, D. Li, X. Xie, D. Ji, L. Li, L. Li, Z. He, B. Lu, S. Liang and J. Zhou, Energy Storage Mater., 2023, 62, 102932 CrossRef .
28. L. Kang, J. Zheng, K. Yue, H. Yuan, J. Luo, Y. Wang, Y. Liu, J. Nai and X. Tao, Small, 2023, 19, 2304094 CrossRef CAS PubMed .
29. X. Liu, Q. Han, Q. Ma, Y. Wang and C. Liu, Small, 2022, 18, 2203327 CrossRef CAS PubMed .
30. X. Ge, W. Zhang, F. Song, B. Xie, J. Li, J. Wang, X. Wang, J. Zhao and G. Cui, Adv. Funct. Mater., 2022, 32, 2200429 CrossRef CAS .
31. O. Jankovský, M. Nováček, J. Luxa, D. Sedmidubský, V. Fila, M. Pumera and Z. Sofer, Chem.–Eur. J., 2016, 22, 17416–17424 CrossRef PubMed .
32. B. Gurzęda, P. Florczak, M. Kempiński, B. Peplińska, P. Krawczyk and S. Jurga, Carbon, 2016, 100, 540–545 CrossRef .
33. W. Chawalitsakunchai, P. Dittanet, S. Loykulnunt, S. Tanpichai and P. Prapainainar, IOP Conf. Ser.: Mater. Sci. Eng., 2019, 526, 012019 CAS .
34. J. Chen, F. Chi, L. Huang, M. Zhang, B. Yao, Y. Li, C. Li and G. Shi, Carbon, 2016, 110, 34–40 CrossRef CAS .
35. P. Sun, L. Ma, W. Zhou, M. Qiu, Z. Wang, D. Chao and W. Mai, Angew. Chem., Int. Ed., 2021, 60, 18247–18255 CrossRef CAS PubMed .
36. W. Yuan, G. Ma, X. Nie, Y. Wang, S. Di, L. Wang, J. Wang, S. Shen and N. Zhang, Chem. Eng. J., 2022, 431, 134076 CrossRef CAS .
37. W. Nie, H. Cheng, Q. Sun, S. Liang, X. Lu, B. Lu and J. Zhou, Small Methods, 2023, 2201572,  DOI:10.1002/smtd.202201572 .
38. M. Qiu, P. Sun, A. Qin, G. Cui and W. Mai, Energy Storage Mater., 2022, 49, 463–470 CrossRef .
39. Z. Cao, X. Zhu, D. Xu, P. Dong, M. O. L. Chee, X. Li, K. Zhu, M. Ye and J. Shen, Energy Storage Mater., 2021, 36, 132–138 CrossRef .
40. Z. Zhao, J. Zhao, Z. Hu, J. Li, J. Li, Y. Zhang, C. Wang and G. Cui, Energy Environ. Sci., 2019, 12, 1938–1949 RSC .
41. X. Chen, X. Shi, P. Ruan, Y. Tang, Y. Sun, W.-Y. Wong, B. Lu and J. Zhou, Small Sci., 2023, 3, 2300007 CrossRef CAS .
42. X. Xie, S. Liang, J. Gao, S. Guo, J. Guo, C. Wang, G. Xu, X. Wu, G. Chen and J. Zhou, Energy Environ. Sci., 2020, 13, 503–510 RSC .
43. Q. Yang, Q. Li, Z. Liu, D. Wang, Y. Guo, X. Li, Y. Tang, H. Li, B. Dong and C. Zhi, Adv. Mater., 2020, 32, 2001854 CrossRef CAS PubMed .
44. H. Dong, X. Hu, R. Liu, M. Ouyang, H. He, T. Wang, X. Gao, Y. Dai, W. Zhang, Y. Liu, Y. Zhou, D. J. L. Brett, I. P. Parkin, P. R. Shearing and G. He, Angew. Chem., Int. Ed., 2023, 62, e202311268 CrossRef CAS PubMed .
45. W. Xu, X. Liao, W. Xu, K. Zhao, G. Yao and Q. Wu, Adv. Energy Mater., 2023, 13, 2300283 CrossRef CAS .
46. J. Cao, D. Zhang, C. Gu, X. Zhang, M. Okhawilai, S. Wang, J. Han, J. Qin and Y. Huang, Nano Energy, 2021, 89, 106322 CrossRef CAS .
47. Y. Wang, X. Lin, L. Wang, Y. Yang, Y. Zhang and A. Pan, Adv. Funct. Mater., 2023, 33, 2211088 CrossRef CAS .
48. J.-H. Wang, L.-F. Chen, W.-X. Dong, K. Zhang, Y.-F. Qu, J.-W. Qian and S.-H. Yu, ACS Nano, 2023, 17, 19087–19097 CrossRef CAS PubMed .
49. J. Zhou, M. Xie, F. Wu, Y. Mei, Y. Hao, R. Huang, G. Wei, A. Liu, L. Li and R. Chen, Adv. Mater., 2021, 33, 2101649 CrossRef CAS PubMed .
50. C. Liu, X. Xie, B. Lu, J. Zhou and S. Liang, ACS Energy Lett., 2021, 6, 1015–1033 CrossRef CAS .
51. X. Yu, Z. Li, X. Wu, H. Zhang, Q. Zhao, H. Liang, H. Wang, D. Chao, F. Wang, Y. Qiao, H. Zhou and S.-G. Sun, Joule, 2023, 7, 1145–1175 CrossRef CAS .
52. C. Han, W. Li, H. K. Liu, S. Dou and J. Wang, Nano Energy, 2020, 74, 104880 CrossRef CAS .
53. K. E. K. Sun, T. K. A. Hoang, T. N. L. Doan, Y. Yu, X. Zhu, Y. Tian and P. Chen, ACS Appl. Mater. Interfaces, 2017, 9, 9681–9687 CrossRef CAS PubMed .
54. X. Cai, X. Wang, Z. Bie, Z. Jiao, Y. Li, W. Yan, H. J. Fan and W. Song, Adv. Mater., 2024, 36, 2306734 CrossRef CAS PubMed .
55. H. Qin, W. Kuang, N. Hu, X. Zhong, D. Huang, F. Shen, Z. Wei, Y. Huang, J. Xu and H. He, Adv. Funct. Mater., 2022, 32, 2206695 CrossRef CAS .
56. Q. Guan, J. Li, L. Li, P. Chai, Y. Li, S. Zhang, X. Yu, L. Bao, J. Peng and X. Li, Chem. Eng. J., 2023, 476, 146534 CrossRef CAS .
57. T. C. Li, C. Lin, M. Luo, P. Wang, D.-S. Li, S. Li, J. Zhou and H. Y. Yang, ACS Energy Lett., 2023, 8, 3258–3268 CrossRef CAS .
58. Z. Jiao, X. Cai, X. Wang, Y. Li, Z. Bie and W. Song, Adv. Energy Mater., 2023, 13, 2302676 CrossRef CAS .
59. Y. Yu, P. Zhang, W. Wang and J. Liu, Small Methods, 2023, 7, 2300546 CrossRef CAS PubMed .
60. C. Meng, W. He, L. Jiang, Y. Huang, J. Zhang, H. Liu and J.-J. Wang, Adv. Funct. Mater., 2022, 32, 2207732 CrossRef CAS .

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4ta03599b

This journal is © The Royal Society of Chemistry 2024