An ultrathin nanocellulosic ion redistributor for long-life zinc anode

and efficient strategy to homogenize the ion flux by building an ultrathin cellulose nanofiber (U-CNF, thickness of less than 1 μm) ion redistributor from waste biomass, which shows remarkable talent in addressing the dendrite formation issue without compromising the low-cost and environmentally friendly nature of AZBs. With this interfacial design, exceptional long-cycle-life (over 2500 h at 1 mA cm −2 ), excellent rate capability (low voltage hysteresis of 238 mV at 40 mA cm −2 ), and high reversibility (average Coulombic efficiency of 99.14% over 850 cycles) of Zn plating/strip-ping performance are synchronously realized. Furthermore, the long-term cyclability and reliability of practical Zn||MnO 2 batteries are also demonstrated with the incorporation of the U-CNF ion redistributor. Considering the simplicity and effectiveness, this proof-of-concept study may imply a promising physical route toward dendrite-free Zn anode for AZBs


INTRODUCTION
The fast-growing energy consumption and extensive environmental concerns have stimulated the increasing demand for renewable energy sources and low-carbon emissions. 1,2Thus, low-cost, highly safe, high-efficiency, and environmentally friendly energy storage systems are urgently pursued. 3,47][8] However, the practical application of AZBs is hindered by the Zn anode troubles, including uncontrolled Zn dendrites growth, Zn corrosion, and hydrogen evolution reaction associated with the use of aqueous electrolyte, among which dendrite growth is generally considered the leading cause of considerably shortened cycling lifespan. 9,10uring the past decade, extensive efforts have been devoted to inhibiting dendrite growth for Zn anode protection and the major strategies focus on the following aspects: (i) developing new electrolyte systems, [11][12][13] (ii) engineering structuralized current collector, [14][15][16] and (iii) constructing artificial interfacial layers. 17For instance, a recently reported multi-layer hierarchical Zn-alginate polymer electrolyte is demonstrated to enable a stable dendritefree Zn plating/stripping lifetime of >500 h. 13 Compared to new electrolyte systems and structuralized current collectors that are usually at the cost of the intrinsic merits of AZBs (i.e., high ionic conductivity, cost-effectiveness, and intrinsic safety), constructing artificial interfacial layers on the Zn anode or separator is considered to have higher practicability in terms of cost, scalability, and feasibility.As a representative example, Xiong et al. 18 recently developed a facile and controllable galvanic replacement method to construct various metallic interlayers on the Zn anode, which was demonstrated to effectively inhibit the Zn corrosion and dendrite formation via regulating the Zn deposition behavior.
As a vital part of a battery, the separator plays the primary role of preventing physical contact between two electrodes and meanwhile allowing the transport of ions.Commercial glass fiber possessing good hydrophilicity, high porosity, and the capability of delivering high ionic conductivity naturally has arisen as the obvious choice since the booming of AZBs research.For this reason, research on separators for AZBs has long been overlooked until some striking progress on separators emerged.For example, an electrospun polyacrylonitrile/graphene oxide separator containing abundant zincophilicity cyano ligands was found to be capable of regulating the migration and distribution of Zn 2+ in the separator, extending the cycle life of Zn symmetrical cells to an unprecedented value of 13000 h (at 1 mA cm −2 ). 19Song et al. 20 revealed that glass fiber functionalized with a Zr-based MOF could favor the preferential growth of the Zn(002) crystal texture, allowing the stable operation of Zn symmetrical cells at 2 mA cm −2 over 1650 h.These advances highlight the potential and significance of upgrading the separator to improve the cyclability of AZBs.
Unlike the previous efforts on functional separators in which laborious, costly, and environmentally unfriendly chemical reactions were usually involved, herein, a facile strategy of constructing an ultrathin cellulose nanofiber (U-CNF) ion redistributor was proposed to homogenize Zn deposition and thus extend the cycling lifetime of Zn anode. 213][24][25][26] As a proof-of-concept demonstration, CNF used in this work was obtained from the waste biomass, i.e., spent bacterial cellulose masks and residues of wood and bamboo.It was coated onto the glass fiber separator via a facile and scalable solution casting method, forming a ~1-μm interfacial U-CNF layer (Figure 1A).In addition to a systematic investigation on the beneficial effects of U-CNF on the basic properties of the baseline glass fiber separator (denoted as w/o for short in all the figures hereafter), we experimentally and computationally demonstrated the unique ability of U-CNF in inhibiting Zn dendrite growth and prolonging Zn cycling lifespan.Further considering the inexpensiveness and environmental benignity of CNF as well as the simplicity of the solution casting technique (Figures 1B-D), the practicability of U-CNF ion redistributor in practical pouch-type AZBs was demonstrated as a proof of concept.

Characterization of the morphology and property improvements with the introduction of U-CNF
As depicted in Figure 1, the CNF layer is in-situ constructed by a facile and scalable solution casting method.The optical image of the partially coated separator in Figure 2A visually exhibits the appearance evolution after the introduction of the coating layer.CNF derived from spent bacterial cellulose masks (Figure S1) forms a transparent film on the surface of white-colored glass fiber as shown in the optical images (Figure 2A).The scanning electron microscopy (SEM) image with a clear dividing line in Figure 2B further shows the much smoother and denser surface of the CNF layer than that of glass fiber.Figure S2A shows the glass fiber decorated with the U-CNF layer can partially reflect the sunlight, indicating the much smaller roughness of the U-CNF layer, which is in good agreement with the SEM and optical microscopical observations (Figures S2B-D).Additionally, the layer covered on the separator can greatly reduce the surface roughness as revealed by the threedimensional (3D) height images and relative curves (Figure S3).In detail, the three parameters of the layer (surface arithmetic mean deviation (S a ), the surface height of irregularities (S z ), and surface root mean square deviation (S q )) are several times smaller than glass fiber, indicating the much smoother surface of the layer.C and Si element distribution from scanning electron microscopy/energy-dispersive X-ray spectrometry (SEM-EDS) mapping in Figure 2C and the characteristic peaks of C−H at 2920 cm −1 and Si−O−Si at 1060 cm −1 (Figure S4) further confirm the successful construction of CNF layer. 27,28Compared to glass fiber with the non-uniform distributed large pores (Figure 2B), the CNF layer possesses more uniform distributed nanoscale pores, due to the closely stacked nanofibers (Figure 2D).In addition, CNF derived from other waste biomass such as wood and bamboo residues can also constitute the layer with uniform nano-scale pores (Figure S5).The smooth layer with uniform nano-scale pores distribution is conducive to the uniform redistribution of ions. 29Additionally, as shown in Figure 2E, the thickness of the CNF layer is less than 1 μm, which will not crowd the battery's internal space.Despite the ultralow thickness, U-CNF can protect the glass fiber from breakdown under severe deformations (Figure 2F and Figure S6).
As electrolyte wettability greatly affects the electrochemical performance of AZBs, 29,30 the electrolyte contact angle, electrolyte absorption, and electrolyte immersion-height tests were conducted to investigate the wettability of the U-CNF.The electrolyte (3 M ZnSO 4 ) contact angle measurements in Figures 2G and H reflect the hydrophilicity of glass fiber (0° after 30 s) and U-CNF modified separator (28° after 30 s).Notably, the electrolyte immersion-height tests in Figure 2G and H exhibit that the U-CNF layer can be wellwetted, as the wet height reaches 59 mm after 10 min, while that of glass fiber is only 53 mm.In addition, the electrolyte absorption ability of the separator decorated with U-CNF is up to 1728 ± 252%, similar to glass fiber (1726 ± 5%) (Figure S7).According to the above results, the U-CNF layer does not sacrifice the excellent wettability of the separator.The ionic conductivity of the separator, closely related to its electrolyte absorption ability, is also a crucial parameter for evaluating the separator performance.As seen from Figure S8, the ionic conductivity of 3 M ZnSO 4 absorbed in glass fiber decorated with U-CNF layer is 3.1 mS cm -1 , which is close to that of bare glass fiber (3.3 mS cm -1 ), indicating the negative effect of the U-CNF layer on ionic conductivity is almost negligible.
Favorable mechanical performance for the separator, especially the side facing the Zn anode, is crucial to prevent Zn dendrite growth and avoid short circuits. 30,31The Young's modulus of U-CNF is estimated to be as high as 10 GPa according to the tensile strain-stress curve shown in the inset of Figure 2I, despite being lower than that of metallic Zn (108 GPa), 32 it is believed that the Zn dendrite growth would be substantially suppressed with the presence of U-CNF layer.Benefiting from the high mechanical strength of 284.0 MPa of the U-CNF layer, a 1-fold increase in tensile strength from 0.6 to 1.

ARTICLE
The Innovation Materials 1(2): 100029, September 20, 2023 3 (Figure S9) are realized with the introduction of the U-CNF layer.The digital photographs of separator samples before and after tensile tests are shown in Figure S10.As seen, numerous broken fibers are pulled out from the fracture edge of glass fibers along the tensile direction, while for glass fibers with the U-CNF layer, brittle fracture as indicated by a flat fracture edge of the U-CNF layer is revealed, which is believed to be an important basis for the enhanced mechanical strength.The thermogravimetric (TG) curves (Figure S11) of glass fiber with and without U-CNF show that, excluding the mass loss of water, no sharp decrease occurs before 250 °C, suggesting the good thermal stability of the U-CNF layer.Associated with the advantageous properties of uniform pore distribution, favorable electrolyte wettability, and excellent mechanical properties, our proposed U-CNF shows great potential to serve as a function layer to improve the electrochemical performance of AZBs.

Electrochemical performance of the cell with U-CNF ion redistributor
To evaluate the electrochemical performance of the U-CNF layer as an ion redistributor in AZBs, the Zn||Zn symmetrical cells and Zn||Cu half cells with/without U-CNF were assembled with 3 M ZnSO 4 electrolyte.The longterm interfacial stability was investigated via the Zn symmetrical cells with the U-CNF ion redistributor.During the cycling process, Zn 2+ migrates reversely between the two Zn metal anodes at a high current density of 1 mA cm −2 with a capacity of 1 mAh cm −2 in each half cycle.The voltage versus time plot depicted in Figure 3A shows that the Zn||Zn cell without U-CNF can only survive for 296 h with a high overpotential of 31 mV, suggesting that the failure caused by short circuit may result from the accumulated Zn dendrites.In sharp contrast, the Zn||Zn symmetrical cell assembled with the U-CNF ion redistributor renders ultralong and stable cycle life over 2500 h with a lower overpotential of 17 mV.Electrochemical impedance spectroscopy (EIS) was recorded to investigate the influence of U-CNF on the interfacial resistance during Zn cycling.As shown in Figure S12, compared to the large fluctuation in the interfacial resistance (R int ) for Zn symmetrical cell with glass fiber separator, the presence of U-CNF allows the Zn symmetrical cell to deliver smaller and stable R int upon cycling, indicating the ability of CNF in favoring fast and stable Zn plating/stripping kinetics.Even increasing the current density to 10 mA cm −2 within the high capacity of 5 mAh cm −2 , compared with the control cells, the symmetrical cells with U-CNF ion redistributor offer a lower initial overpotential of 51 mV (versus 68 mV for glass fiber) coupled with over 150 h cycle life (versus 57 h for glass fiber) in Figure 3B.The overpotential and lifespan of the symmetrical cells demonstrate the good ability of the U-CNF ion redistributor to inhibit dendrite growth.
To explore the rate performance of U-CNF ion redistributor, the Zn deposition and exfoliation reversibility were evaluated under more aggressive conditions through cycling the Zn symmetrical cells at current densities ranging from 5 to 40 mA cm −2 with a fixed areal capacity of 5 mAh cm −2 as exhibited in Figure 3C.Even though the cells with different separators exhibit similar low overpotential at the current densities of 5 mA cm −2 , the voltage hysteresis of the cell with glass fiber increases sharply once the current density increases to 10 mA cm −2 .By sharp contrast, the cell with the U-CNF shows a more stable and superior reversible Zn deposition/exfoliation performance with 238 mV voltage hysteresis at 40 mA cm −2 and no signs of dendriteinduced short-circuiting, indicating that the U-CNF ion redistributor facilitates the excellent rate performance for practical AZBs.Coulombic efficiency (CE) reflecting the amount of irreversible Zn metal consumption during cycling has been recognized as a vital parameter, directly affecting the life span of the AZBs. 33,34Zn||Cu cells using glass fiber with and without U-CNF were assembled to evaluate the Zn reversibility.The Cu electrode was first plated at a current density of 1 mA cm −2 for 1 h followed by stripped to a 0.8 V cutoff voltage.Less initial capacity loss before the stabilization of CE is of great importance to reduce the negative/positive electrode capacity (N/P) ratio and thus improve the energy density of practical batteries.As compared in the inset of Figure 3D and Figure S13, in the presence of U-CNF, we see a massive increase in the initial Coulombic efficiency (ICE) from 70.4% to 85.5%, 89.6%, and 92.5% for U-CNF prepared from bacterial cellulose (BC), wood, and bamboo, respectively.For the separator without U-CNF, Zn plating/stripping CE experiences a slow increase and remains < 99.0% after 100 cycles, fast short-circuiting failure and a low aver-age CE of 98.24% within 197 cycles are observed.In contrast, impressively, the CE can reach up to > 99.0% within 15 cycles and stabilize at 99.3% to 99.8% in the presence of U-CNF (Figures 3D and E).The cycling lifespan is significantly improved regardless of the starting materials of CNF.The BCbased U-CNF allows the achievement of an average CE of 99.14% within a cycling lifespan of 850 cycles.Despite the relatively shortened cycling lifespan for Zn||Cu cells with U-CNF prepared from bamboo and wood, impressively high average CEs of 99.58% and 99.62% are achieved, respectively.More importantly, out of our expectations, much smaller overpotentials are achieved in the presence of U-CNF, demonstrating the uniform ion flux enabled by U-CNF contributes to faster Zn plating/stripping kinetic (Figures S14 and S15).
To evaluate the CE at larger current densities, the rate performance of the Zn||Cu half cells at current densities ranging from 1 to 10 mA cm -2 with a fixed areal capacity of 1 mAh cm -2 as exhibited in Figure S16.The measured average CE at first five cycles of the Zn||Cu cell without U-CNF is 93.2% (1 mA cm -2 ), 97.9% (2 mA cm -2 ), 99.1% (5 mA cm -2 ), and 99.5% (10 mA cm -2 ).By contrast, the cell with U-CNF shows higher CE values of 94.5%, 98.6%, 99.4%, and 99.7% at the corresponding current densities.The higher CE values and lower overpotential (Figure S16) demonstrate that the Zn plating/stripping reversibility and energy efficiency are substantially improved by the U-CNF ion redistributor, regardless of the operating current density.Taking a

ARTICLE
The Innovation Materials 1(2): 100029, September 20, 2023 5 comprehensive comparison of the Zn reversibility in terms of CE within initial cycles, stable CE value, and cycling lifespan, we conclude the U-CNF ion redistributor does have the capability of enabling exceptional long-life and high-reversibility performance for the Zn anode.As the stable Zn plating/stripping CE and total cumulative capacity are generally considered as two critical indicators for Zn battery performance, they are selected to compare our achievement and previously reported results in Figure 3F. 5,7,11,35- 46It is found that our results of ~99.6% average CE with 800 mAh cm −2 cumulative capacity and 99.14% average CE with 1700 mAh cm −2 cumulative capacity outperform most previous efforts on improving Zn anode cyclability and reversibility, indicating the great importance and feasibility of ion redistributor in promising practical AZBs.

Experimental and computational investigations of the Zn deposition behavior
In situ optical microscopy observation was first conducted to characterize the dendrite suppression efficacy of the U-CNF ion redistributor.As shown in Figure 4A, with plating at a constant current density of 1 mA cm −2 , scattered Zn protuberances appear and experience continuous uncontrolled growth, and the increase in Zn thickness after 90 minutes far exceeds the theoretical thickness of the deposited capacity of Zn.For Zn decorated with U-CNF, in contrast, no visible Zn dendrites are observed in Figure 4B within the whole observation time, intuitively proving the U-CNF ion redistributor's exceptional ability in enabling uniform Zn deposition.To get microscopic details of the Zn deposit morphology in the presence of U-CNF, post-mortem SEM analysis was conducted on the Zn anode after 20 plating/stripping cycles at 1 mA cm −2 and a half-cycle duration time of 1 h.As shown in Figure 4C, massive randomly distributed Zn deposits with diameters of several tens of micrometers are observed in the absence of U-CNF.Some of the large-sized deposits have grown inside the glass fibers, which are prone to penetrate the separator and result in a fast short circuit.In the case that a U-CNF ion redistributor is incorporated, the Zn deposition layer remains flat and compact upon cycling (Figure 4D, Figure S17), which agrees well with the in situ optical observation result, indicating a typical 3D diffusion mechanism for Zn deposition enabled by the U-CNF layer, as further supported by the chronoamperometry (CA) test result that a short period (20 s) of plane diffusion process followed by a dominant 3D diffusion process was observed (Figure S18).Furthermore, the surface roughness (mainly indicated by S a ) of the cycled Zn electrodes is quantitatively evaluated using a 3D surface imaging technique.
As obtained from the 3D height images of cycled Zn electrodes (Figure S19), a massive decrease in S a from 8.4 to 2.8 μm is realized by the U-CNF ion redistributor.If evaluating the roughness based on the surface height of irregularity (S z ), the effect of U-CNF on enabling smooth Zn deposition behavior is more pronounced (S z decreases from 68.7 to 21.0 μm).X-ray diffraction (XRD) was further employed to compare the crystal structure of Zn deposits with and without the assistance of the U-CNF ion redistributor.The (002)/(100) peak intensity ratio of the Zn electrode with glass fibers is 1.33, and that of the Zn electrode coupling with the U-CNF ion redistributor increases to 1.42 (Figure 4E), revealing the ability of U-CNF to favor dendritefree Zn (002) electrodeposition.The variation of preferential crystal orientation is also explored by the relative texture coefficients, which were calculated by the peak intensities of XRD from (hkl) crystallographic planes of the Zn deposits according to the intensities of the standard Zn sample taken from a powder diffraction file (PDF) card (PDF# 65-5973).The results in Figure 4F show that the texture coefficient of the (002) plane increase from 24.3 to 28.1 under the function of the U-CNF ion redistributor, indicating the exposure of more preferential (002) planes.
Density function theory (DFT) calculation was conducted to get theoretical insights into the effects of U-CNF ion redistributor on regulating Zn plating behavior.As shown in Figure 4G, the calculation model consists of a vertical cellulose molecule adsorbed on the Zn(002) surface and a moving pathway of Zn 2+ with a distance varying from 14 to 2 angstrom from the Zn surface.As shown in Figure 4H, an energy barrier of ~0.67 eV appeared when a Zn 2+ transfer through the surrounding area of the cellulose molecule, proving a hindering effect of the cellulose on the diffusion behavior of Zn 2+ .Such an energy barrier would contribute to a sluggish Zn deposition dynamic in the surrounding area of cellulose molecules, favoring the uniform deposition of Zn 2+ .In contrast, the adsorption process of Zn 2+ on the surface of bare Zn presents a smooth energy profile without an energy barrier, indicating an unlimited Zn 2+ adsorption process on the surface of bare Zn, which would lead to uncontrolled dendrite growth (Figure 4H).
At a large scale, the regulation ability of the U-CNF ion redistributor on Zn electrodeposition behavior was theoretically investigated by the finite element method (FEM) simulations.Generally, Zn nucleation and deposition are closely related to the electric field distributions in the vicinity of the Zn electrode/electrolyte interface.In the modeling where a separator featuring a large pore size is adopted (simulating the glass fiber), severely uneven electric field distribution is observed along the Zn surface, and the highest elec- tric intensity appears at the edge of Zn deposits due to the well-known "tip effect" (Figure 4I).In the presence of the interfacial layer with uniform nanosized mass transportation channels (simulating the U-CNF ion redistributor), significantly reduced electric field inhomogeneity, especially along the electrode/electrolyte interface (Figure 4J), is realized, showing a much less pronounced tip effect.Following a similar changing regulation with the electric field, the current density distribution is further homogenized despite the inevitable influence of the uneven Zn deposits (Figure S20).The Zn 2+ concentration at the Zn electrode/electrolyte interface is slightly improved in the presence of the ion redistributor (Figure S21), which favors better Zn deposition kinetics and thus leads to a smaller overpotential, agreeing with the experimental investigation results of Zn cycling.Complementary to experimental investigations, the above simulation results provide solid theoretical proof to strengthen the validity of our concept that the U-CNF layer can function as an efficient ion redistributor, enabling uniform Zn 2+ deposition to realize exceptionally long Zn electrode cycling lifespan and high efficiency.

Practical prospect of the U-CNF ion redistributor
As a proof-of-concept demonstration, the practical application of U-CNF ion redistributor was studied in both coin-and pouch-type Zn||MnO 2 cells using a baseline aqueous electrolyte of 3 M ZnSO 4 .As revealed by the cycling results of Zn||MnO 2 cells shown in Figure 5A, with the introduction of the U-CNF ion redistributor, it can be observed that reasonable initial discharge specific capacities of 275.5 and 181.9 mAh g -1 at 0.2 and 1 A g -1 are delivered, respectively, both are slightly higher than those delivered by the control cells (using glass fiber separator), suggesting that the presence of U-CNF would not hamper the fast ion transport across the separator.After 100 charge/discharge cycles at 0.2 A g −1 , the Zn||MnO 2 cell with U-CNF retains 95.4% of its capacity (highest value of 308.5 mAh g −1 achieved at 44th cycle); at 1 A g −1 , the discharge capacity stabilizes at ~150 mAh g −1 over 400 cycles after a sharp decrease within the first dozen of cycles, corresponding to a high capacity retention ratio of 81.7%.Impressively, almost overlapped charge/discharge curves are observed (Figure 5B), jointly pointing to the highly reversible Zn plating/stripping chemistry in the presence of U-CNF.In contrast, serious capacity decay, especially at the lower current density of 0.2 A g −1 , was observed (Figure S22).Since no optimization was conducted to inhibit the Mn dissolution, compared to a capacity decay rate of 0.69% at 0.2 A g −1 for the control cell, the significantly lowered capacity decay rate of 0.046% enabled by the U-CNF ion redistributor is mainly ascribed to its talent in favoring uniform Zn deposition as previous text proposed.This is further evidenced by the fact that the XRD patterns of carbon nanotubes (CNT)/MnO 2 electrodes cycled with glass fiber and glass fiber with U-CNF are almost the same (Figure S23).
To further evaluate the application of U-CNF in practical AZBs, Zn||MnO 2 cells with a theoretical N/P ratio of 4.0 (mass loading of MnO 2 : 5 mg cm −2 ; thickness of Zn: 10 μm) were assembled and tested.As shown in Figures 5C  and D, the initial specific discharge areal capacity of the Zn||MnO 2 is around 1.4 mAh cm −2 at 0.2 A g −1 , corresponding to a high depth of discharge of 24% for Zn anode and a high specific capacity of ~280 mAh g −1 for MnO 2 cathode and, encouragingly, a reasonable high cycling lifetime of ~60 cycles (80% capacity retention) was realized.
Benefiting from the significantly improved mechanical strength, pouchtype Zn||MnO 2 full cells remain operational even being subjected to severe bending deformation (Figure 5E).In a daily life scenario simulation, three Zn||MnO 2 pouch cells were connected in series to charge a smartwatch with a rated voltage of 5 V (Figure 5F).To summarize, the beneficial effects of the introduction of the U-CNF ion redistributor are qualitatively expressed through a radar graph, it is seen that compared to the glass fiber, an all-round improvement in terms of both basic properties (mechanical strength and wettability) and electrochemical performance (Zn reversibility and cycling lifespan) was achieved (Figure 5G).

DISCUSSION
In summary, we have proposed an ion redistributor strategy to homogenize Zn metal electrodeposition.As a proof-of-concept investigation, an ultrathin CNF layer is in-situ constructed on the baseline glass fiber separator, and it is demonstrated that the introduction of the U-CNF layer warrants the exceptionally high rate capability (~ 238 mV voltage hysteresis at 40 mA cm −2 ), long cycling lifespan (2500 h at 1 mA cm −2 ), and high plating/stripping reversibility (an average CE of 99.14% over 850 cycles) for Zn cycling.We experimentally and computationally demonstrate the unique talent of U-CNF in homogenizing the current across the Zn electrode surface as an ion redistributor, thus favoring a dendrite-free Zn cycling behavior.The practicability of the ion redistributor strategy is demonstrated by the good cycling stability (capacity retention of ~82% after 400 cycles) and mechanical flexibility of Zn||MnO 2 cells.The direct conversion of waste biomass into high valueadded products is also well in line with the sustainable and carbon neutrality goal, providing another benefit for our strategy.Further considering the simplicity, inexpensiveness, and environmental benignity of constructing a nanocellulosic ion redistributor, this work opens future exploration directions toward high-performance sustainable metal-anode battery chemistries.

2 Figure 1 .
Figure 1.Construction process and the function of a U-CNF ion redistributor in AZBs (A) Schematic illustration of the preparation and the proposed Zn deposition behavior in the presence of a U-CNF ion redistributor.(B) Digital photograph shows the massive production of CNF dispersion obtained from biomass waste.(C and D) Demonstration of the scalable preparation process of U-CNF via a solution cast technique.

Figure 2 .
Figure 2. Characterization of the morphology and property changes of glass fiber with the introduction of U-CNF (A) Digital photograph, (B) SEM image, and (C) corresponding EDS mapping of the glass fiber in half area decorated with CNF.Top-view SEM image (B), and EDS mapping (C) of the separator partially decorated with CNF (right).(D) Topview and (E and F) cross-section view SEM images of glass fiber decorated with U-CNF in naturally placed (E) and bending state (F).The inset of (E) shows the high-magnification SEM image of the U-CNF layer.A comprehensive study of the wettability of glass fiber without (G) and with (H) the U-CNF layer.(I) Tensile stress-strain curves of glass fiber with and without the U-CNF layer.The inset of (I) presents the tensile stress-strain curve of U-CNF.

Figure 3 .
Figure 3. Cyclability of Zn anode with and without a U-CNF ion redistributor Cycling performances of Zn symmetrical cells at (A) 1 mA cm −2 with a capacity of 1 mAh cm −2 and (B) 10 mA cm −2 with a capacity of 5 mAh cm −2 .(C) Rate capability of Zn||Zn symmetrical cells at current densities of 5, 10, 20, and 40 mA cm −2 with a fixed capacity of 5 mAh cm −2 .(D and E) CE of Zn||Cu cells at 1 mA cm −2 with a capacity of 1 mAh cm −2 .(E) is an enlarged view of d in the range of 350th to 390th cycle.(F) Comparison of the Zn cycling performance in terms of CE and cumulative capacity with previous reports.

Figure 4 .
Figure 4. Experimental and computational investigations of the influence of U-CNF ion redistributor on the Zn deposition chemistry (A-D) In-situ optical observation of the Zn deposition process and the SEM images of the Zn surface after 20 plating/striping cycles in Zn symmetrical cells.(A and C) for glass fiber and (B and D) for glass fiber decorated with U-CNF.(E) XRD pattern of the Zn electrode cycles with and without U-CNF.(F) Calculated texture coefficients of Zn(002).(G) Model of DFT calculations: a vertical cellulose molecule adsorbed on the Zn(002) surface and a moving pathway of Zn 2+ with distance vary from 14 to 2 angstrom from the Zn surface and (H) results of potential energy change at a different distance calculated by DFT.Finite element modeling results of (I) the electrical field.(J) Electric field along the in-plane direction.

Figure 5 .
Figure 5. Evaluation of practical application prospect of the U-CNF ion redistributor (A) Comparison of cyclability of Zn||MnO 2 cells using glass fiber separator with/without the U-CNF layer.(B) Representative charge/discharge voltage profiles as a function of cycle number.(C and D) Cyclability and charge/discharge voltage profiles of Zn||MnO 2 full cell with a theoretical N/P ratio of 4.0.(E) Digital photograph records the open-circuit potential of the Zn||MnO 2 pouch cell under 180°.(F) Digital photograph shows the smartwatch is charged by three pouch cells connected in series.(G) Comparison in basic properties and electrochemical performance indices of glass fiber with and without U-CNF ion redistributor in terms of mechanical property, wettability, Zn reversibility, cycling lifespan, and cycling stability.