Lamella-nanostructured eutectic zinc–aluminum alloys as reversible and dendrite-free anodes for aqueous rechargeable batteries

Metallic zinc is an attractive anode material for aqueous rechargeable batteries because of its high theoretical capacity and low cost. However, state-of-the-art zinc anodes suffer from low coulombic efficiency and severe dendrite growth during stripping/plating processes, hampering their practical applications. Here we show that eutectic-composition alloying of zinc and aluminum as an effective strategy substantially tackles these irreversibility issues by making use of their lamellar structure, composed of alternating zinc and aluminum nanolamellas. The lamellar nanostructure not only promotes zinc stripping from precursor eutectic Zn88Al12 (at%) alloys, but produces core/shell aluminum/aluminum sesquioxide interlamellar nanopatterns in situ to in turn guide subsequent growth of zinc, enabling dendrite-free zinc stripping/plating for more than 2000 h in oxygen-absent aqueous electrolyte. These outstanding electrochemical properties enlist zinc-ion batteries constructed with Zn88Al12 alloy anode and KxMnO2 cathode to deliver high-density energy at high levels of electrical power and retain 100% capacity after 200 hours.

Here are some of my comments: 1. In figure 3, EIS was used to reveal the oxidation resistances of Zn metal and Zn/Al alloys. To be honest, these results are confusing. First of all, the author did not demonstrate how they test the EIS. Are they symmetric cells or not? Did the immersion and oxidation process happen inside a sealed cell or outside in the electrolytes? The author should give the model of the equivalent circuit and calculate the resistance from different components separately. Simply comparing the value does not make sense. And also, charge transfer resistance does not mean the oxidation resistance. Actually, the author should also explain what the oxidation resistance is. 2. In the discussion part of Figure 3 and Figure 4, the author should give more explanation about the reason for the different phenomenon, but not just present the results. 3. When discussing the effect of O2, especially during the electrochemical process in a sealed cell, the author should consider more than just the O2 in the electrolyte. The author could calculate the amount of O2 that dissolved in the electrolyte and compare it with the anode to see whether there is enough O2 to oxidize the alloy. Besides that, the reaction between the alloy and the electrolyte and H2O also need to be taken consideration. 4. The author claimed the high columbic efficiency of the alloy anode; however, no data was demonstrated for the columbic efficiency of the anode. 5. As for the full cell performance, Figure 5b is a little confusing. How did the author get the specific capacity versus the Scan rate? 6. Generally, when the loading of the cathode is only 1.0 mg/cm2, which is much less the anode, the electrochemical performance usually relies on the cathode part but anode part. As in many previous reports, Zn foil can also support a good electrochemical performance. So, in this paper, the author should give more strong and supportive data to demonstrate the superiority of the alloy anode, for example, limiting the anode/cathode mass ratio.

Reply:
We thank the reviewer for finding interest of our work. We also appreciate the reviewer for his/her constructive comments and suggestions. Following these valuable and insightful comments/suggestions, we have supplemented some experiments, including XRD, Raman and XPS characterizations on the Zn 88 Al 12 and Zn electrodes before and after the cycling test of Zn stripping/plating, SEM characterization of Zn 88 Al 12 electrode during Zn stripping and then plating, electrochemical characterization on the Zn stripping/plating behavior of Zn 88 Al 12 and Zn electrodes in the mixed electrolyte of 2 M ZnSO 4 and 0.2 M MnSO 4 , self-discharge performance of Zn 88 Al 12 /K x MnO 2 full batteries, as well as SEM and XRD characterization on the K x MnO 2 cathodes and ICP-OES measurement of electrolyte in Zn 88 Al 12 /K x MnO 2 and Zn/K x MnO 2 full batteries after cycling measurements. Based on these new results, we have completely revised the manuscript. The detailed corrections are listed below.
(1) Why the alloy can show this lamellar structure and how it can influence the reversible Zn plating/stripping. Solid evidence and discussions should be provided.

Reply:
We appreciate the reviewer for the constructive suggestion. As a model system, Zn-Al system is a typical bimetallic alloy with poor miscibility. At the eutectic Zn/Al composition of 88/12 (at %), there takes place a eutectic reaction at the eutectic temperature, in which both Zn and Al constituents crystallize simultaneously from their molten liquid solution. The Zn 88 Al 12 alloy often forms a parallel array of the two coexisting phases of Zn and Al lamellas that grow side by side as a result of the balance between the lateral diffusion of excess Zn and Al in the liquid just ahead of the solid/liquid interface and the creation of Zn/Al interfacial area during the solidification process. The eutectic structure composed of alternating Zn and Al lamellas enlists the Zn 88 Al 12 alloy electrode to effectively circumvent the irreversibility issues of monometallic Zn anode for potential applications in rechargeable Zn-based batteries. Therein, the Zn lamellas supply the Zn 2+ charge carriers with the Zn stripping and the residual Al lamellas with self-grown Al 2 O 3 shells form lamellar nanopatterns. In view that the insulating Al 2 O 3 shells substantially block the electron transfer from Al to the Zn 2+ cations, there forms a positively electrostatic shield around the Al/Al 2 O 3 lamellas to guide the uniform Zn deposition at their interlamellar spacing along the Zn precursor sites. To further demonstrate these processes, we have carried out supplementary SEM characterization on the eutectic Zn 88 Al 12 electrode after the Zn stripping and then plating. As shown in supplementary Figure 9a, the Zn-stripped Zn 88 Al 12 electrode displays evidently lamellar nanopatterns of the residual Al/Al 2 O 3 lamellas. While the Zn-stripped Zn 88 Al 12 recovers to the initial morphology with a smooth surface as a result of the Al/Al 2 O 3 lamellas guiding the subsequent Zn electrodeposition in the plating process (supplementary Figure 9b).
(2) The polarization of CD curves of the symmetric batteries using ZnAl alloy was decreased compared to Zn/Zn batteries, though the enhancement was not obvious, and the reason of positive effect from the alloy should be included.

Reply:
We thank the reviewer for the comment. Following the comment, we have explained the positive effect of alloy in the text. The less polarization of Zn 88 Al 12 is probably due to the unique eutectic structure of alternating Zn and Al lamellas. Therein, the constituent Al lamellas not only protect against the passivation of the electroactive Zn but reduce the local current density of Zn stripping/plating via the formation of core/shell Al/Al 2 O 3 lamellar nanopatterns (supplementary   (4) The calculation of energy density should also be based on the overall mass of anode and cathode, not just the mass of cathode.
Reply: Following this comment, we have calculated the energy density on the basis of the total mass of anode and cathode. The overall energy density of the full Zn 88 Al 12 /K x MnO 2 battery can reach ~142 Wh kg −1 by lowering the anode-to-cathode mass ratio to 3:1 (supplementary Figure 19).
(5) What about the self-discharge behavior of the Zn/KxMnO 2 battery, since both tests in the absence and presence of O 2 had been presented. The self-discharge mechanism should be discussed.

Reply:
We appreciate the reviewer for the suggestion, according to which we have performed self-discharge measurement of the Zn 88 Al 12 /K x MnO 2 battery in the 2 M ZnSO 4 and 0.2 M MnSO 4 electrolytes with/without the presence of O 2 . Supplementary Figure 20 shows the self-discharge performance of the Zn 88 Al 12 /KxMnO 2 batteries. In the electrolyte with the absence of O 2 , the voltage dramatically drops to ~1.481 V from 1.800 V in ~13 h. This is slower than the Zn 88 Al 12 /K x MnO 2 battery with the O 2 -present electrolyte, of which the voltage decreases to ~1.472 V in ~6 h. The initial voltage drop is probably due to the pseudocapacitive discharge behavior of K x MnO 2 , which is boosted by the presence of O 2 . Nevertheless, the Zn 88 Al 12 /K x MnO 2 batteries in the electrolyte with/without the presence of O 2 exhibit a voltage plateau with very low self-discharge, with ~0.1 mV h −1 , in the subsequent 600 h, because of ultralow insertion kinetics of Zn 2+ , which is almost independent of O 2 in the electrolyte.
(6) In page 9, the authors mentioned that "This observation is probably due to the absence of passivation film on the electroactive Zn surface, which result in the less polarization." What is the passivation film? XPS data before and after cycling test should be included.

Reply:
We thank the reviewer for the constructive comment and suggestion, according to which we have carried out supplementary XRD, Raman and XPS characterizations on the Zn 88 Al 12 and Zn electrodes before and after cycling test. The detailed results are shown in supplementary Figure 10 and Figure 11. After cycling test, the monometallic Zn electrode displays neoformative peaks in both XRD patterns and Raman spectrum compared with the one before cycling test (supplementary Figure 10b,d). These neoformative diffraction peaks are in agreement with the line patterns of Zn 4 SO 4 (OH) 6 ⋅H 2 O (JCPDS 39-0690), demonstrating the formation of Zn 4 SO 4 (OH) 6 ⋅H 2 O on the Zn metal (supplementary Figure 10b). Raman spectrum of the tested Zn electrode reveals that in addition to the Zn 4 SO 4 (OH) 6 ⋅H 2 O, there is ZnO to form on the surface during the cycling test (supplementary Figure 10d). In contrast, the Zn 88 Al 12 alloy electrode after cycling test exhibits almost the same XRD patterns as the one before cycling test (supplementary Figure 10a). Furthermore, the tested Zn 88 Al 12 electrode does not display additional characteristic Raman bands in addition to the weak ones that is due to the Al 2 O 3 (supplementary Figure 10c). These facts confirm the absence of passivation film on the Zn 88 Al 12 electrode, which results in the less polarization. The XPS spectra of the monometallic Zn and the eutectic Zn 88 Al 12 alloy before and after cycling test are shown in supplementary Figure 11. The Zn 2p XPS demonstrates that the surface Zn on the pristine monometallic Zn electrode is mainly in metallic state in addition to a little oxidized state due to the general surface oxidation (supplementary Figure 11b). While for the Zn electrode after cycling test, the surface Zn is in completely oxidized state because of the formation of  Reply: We appreciate the reviewer for approving our eutectic strategy to be potentially a universal way for Zn anodes. In view that both Zn-Li and Zn-Mg alloy systems have similar eutectic structures, they are expected to hold a promise to circumvent the dendrite problem of Zn anode.
(8) There was no capacity fading at 0.5 A g -1 in the ZnAl/KxMnO 2 battery, but the Zn/K x MnO 2 battery degraded extremely fast. What are the reason besides the effect from the anode? The cathode and electrolyte should be analyzed and discussed for both batteries.

Reply:
We appreciate the reviewer for the valuable comment and suggestion. Following this suggestion, we have carried out SEM and XRD characterizations on the K x MnO 2 cathodes in the Zn 88 Al 12 /K x MnO 2 and Zn/K x MnO 2 full batteries after cycling test. Supplementary Figure 22a,c show the typical SEM images of the K x MnO 2 cathodes in the Zn 88 Al 12 /K x MnO 2 and Zn/K x MnO 2 full batteries, displaying the same morphology. Furthermore, these tested K x MnO 2 cathodes exhibit the same XRD patterns as the as-prepared one (supplementary Figure 22b,d). In addition, ICP-OES analysis demonstrates that the tested electrolyte keeps almost the same concentration of Zn 2+ and Mn 2+ ions as the initial one. These facts indicate that the fast capacity degradation of Zn/K x MnO 2 battery at 0.5 A g −1 results from the monometallic Zn anode, which undergoes severe irreversibility issues associated with the dendrite formation and growth and side reaction of Zn 4 SO 4 (OH) 6 ⋅H 2 O and ZnO (supplementary Figure 10b,d and Figure 11a).

(9) The XRD and XPS of the symmetric cells long-term Zn stripping/plating cycling measurements should be provided to see whether there are side reactions on the anodes.
Reply: We thank the reviewer for the constructive suggestion, following which we have carried out supplementary XRD and XPS characterizations on the symmetric cells of Zn 88 Al 12 and Zn electrodes after the cycling measurement of Zn stripping/plating. At the same time, we also carried out Raman spectroscopy measurements. The detailed results are shown in supplementary Figure 10 and Figure  11. The Zn 88 Al 12 electrode displays almost the same XRD patterns before and after the cycling test of Zn stripping/plating, demonstrating that there does not occur evident side reactions on the Zn 88 Al 12 alloy, in addition to the formation of Al 2 O 3 on the Al lamellar nanopatterns (supplementary Figure 10c). While for the monometallic Zn electrode, there occurs evident side reactions to form Zn 4 SO 4 (OH) 6 ⋅H 2 O and ZnO. As shown in supplementary Figure 10b,d for the XRD patterns and the Raman spectrum of the cycled Zn electrode, there appear neoformative diffraction peaks and characteristic Raman bands that correspond to Zn 4 SO 4 (OH) 6 ⋅H 2 O in addition to the ZnO. As a consequence, X-ray photoelectron spectroscopy (XPS) analysis demonstrates the complete Zn 2+ state of surface Zn in the Zn electrode (supplementary Figure 11a). While the Zn 2p XPS of the Zn 88 Al 12 reveals that the surface Zn maintains almost the same chemical states, i.e., in both metallic Zn and Zn 2+ , as the one before cycling test (supplementary Figure 11c,e). the Al primarily becomes Al 3+ as a result of the formation of Al 2 O 3 shell (supplementary Figure  11d,f).

Reviewer #2 (Remarks to the Author):
In this paper, the authors show that eutectic-composition alloying of zinc and aluminum could be an effective strategy to address the irreversibility of Zn anode. The lamellar structure composed of alternating Zn and Al is interesting. The dendrite formation seems to be suppressed in this alloy anode. The discussion of the effect of O 2 in the electrolyte is also instructive. However, some important comments need to be addressed before this paper can achieve the high standard of this reputable journal.
Here are some of my comments: Reply: We appreciate the reviewer for his/her insightful and constructive comments and finding our work of interest. According to these comments/suggestions, we have carried out supplementary experiments, including the determination of the amount of O 2 in electrolyte, the electrochemical measurements to demonstrate the coulumbic efficiency of Zn 88 Al 12 alloy electrode, as well as the overall energy density of the full Zn 88 Al 12 /K x MnO 2 batteries with various anode-to-cathode mass ratios. Based on these new results and the reviewer′s valuable suggestions, we have completely revised the manuscript. The detailed corrections are listed below.
(1) In figure 3, EIS was used to reveal the oxidation resistances of Zn metal and Zn/Al alloys. To be honest, these results are confusing. First of all, the author did not demonstrate how they test the EIS. Are they symmetric cells or not? Did the immersion and oxidation process happen inside a sealed cell or outside in the electrolytes? The author should give the model of the equivalent circuit and calculate the resistance from different components separately. Simply comparing the value does not make sense. And also, charge transfer resistance does not mean the oxidation resistance. Actually, the author should also explain what the oxidation resistance is.

Reply:
We appreciate the reviewer for the insightful comments. Following the comments, we have presented a detailed method to perform the EIS measurement in Result and Method sections. We also have corrected the presentation to clearly explain the oxidation-resistance capability of the eutectic Zn 88 Al 12 alloy electrode. Our EIS measurements were performed in sealed cells with O 2 -or N 2 -saturated aqueous 2 M ZnSO 4 electrolytes at room temperature, respectively, on the basis of a classic three-electrode configuration with Pt foil as the counter electrode and an Ag/AgCl electrode as the reference electrode. The frequency ranges from 100 kHz to 10 mHz and the amplitude is 10 mV. Because of the high chemical activity of Zn and Al, there always exists an oxide layer on the electrode surface, as demonstrated by XPS analysis in supplementary Figure 11b,e,f. Nevertheless, the Zn 88 Al 12 alloys exhibit superior oxidation-resistance capability in aqueous electrolyte compared with the monometallic Zn electrode. This is demonstrated by EIS analysis in Figure 3a and supplementary Figure 7b, which compare the EIS spectra of Zn 88 Al 12 alloys with those of hypoeutectic Zn 50 Al 50 alloy and monometallic Zn in the O 2 -present electrolytes for 1 and 10 h, respectively. Based on the equivalent circuit with the general descriptors: the intrinsic resistance of solution and electrodes (R I ), the charge-transfer resistance (R CT ), the double-layer capacitance (C F ) and the Warburg impedance (Z W ) (supplementary Figure 7a), the EIS spectra are analyzed using the complex nonlinear least-squares fitting method. Supplementary Figure 7c compares the R I values of all Zn-based electrodes immersed in the O 2 -present electrolyte for 1 h, wherein the Zn 88 Al 12 with λ = ~450 nm has the lowest R I value because of the outstanding oxidation-resistance property. Even extending the immersion time to 10 h, the Zn 88 Al 12 with λ = ~450 nm still maintains ~11 Ω whereas the Zn electrode has the R I value to increase to ~22 Ω. The large change of R I value indicates the inferior oxidation-resistance capability of monometallic Zn. Owing to their different oxidation-resistance properties, there form distinct oxide layers to depress the Zn stripping/plating kinetics, indicated by the increase of R CT value. When immersed in the O 2 -present electrolyte for 1 and 10 h, the Zn 88 Al 12 with interlamellar spacing of ~450 nm exhibit exceptional stability with the R CT value changing from ~32 Ω to ~36 Ω, in sharp contrast with the monometallic Zn electrode with a remarkable change of R CT from ~96 Ω to ~177 Ω (Figure 3b). These observations demonstrate that the eutectic Zn 88 Al 12 alloy electrode exhibit superior oxidation-resistance capability relative to the monometallic Zn electrode.
(2) In the discussion part of Figure 3 and Figure 4, the author should give more explanation about the reason for the different phenomenon, but not just present the results.
Reply: We appreciate the reviewer for the constructive suggestion. According to this suggestion, we have presented more discussion on the EIS results and electrochemical performance. The details are included in the text.
(3) When discussing the effect of O 2 , especially during the electrochemical process in a sealed cell, the author should consider more than just the O 2 in the electrolyte. The author could calculate the amount of O 2 that dissolved in the electrolyte and compare it with the anode to see whether there is enough O 2 to oxidize the alloy. Besides that, the reaction between the alloy and the electrolyte and H 2 O also need to be taken consideration.

Reply:
We appreciate the reviewer for the constructive suggestion. Following this suggestion, we have measured the concentration of O 2 in the O 2 -present ZnSO 4 electrolyte (~16.59 mg L −1 ) and calculated the amount of O 2 to be ~8.295 × 10 −6 g, or ~2.59 × 10 −7 mol, in the sealed cell with electrolyte of 0.5 mL. It means that there is enough O 2 to oxidize the surface atoms of the Zn 88 Al 12 alloy with the dimension of 0.5 cm × 0.5 cm × 40 μm, wherein the number of surface atoms is estimated to be ~1.36 × 10 −9 mol. In view that the standard equilibrium potential of Al 3+ /Al (−1.66 V versus SHE) is much lower than that of Zn 2+ /Zn (−0.76 V versus SHE), the Al surface atoms on the Al lamellas prefer to firstly react with O 2 to form Al 2 O 3 during the Zn stripping via the reaction Zn → Zn 2+ + 2e − . Meanwhile, some Zn 2+ ions take part in the irreversible reaction, Zn 2+ + OH − → Zn(OH) 2 + 2e − → ZnO + H 2 O, to form ZnO. Therefore, there may take place reactions: Zn 88 Al 12 + O 2 + H 2 O → Al 2 O 3 + Zn(OH) 2 + Zn 2+ + e − and Zn(OH) 2 + 2e − → ZnO + H 2 O.
(4) The author claimed the high coulombic efficiency of the alloy anode; however, no data was demonstrated for the columbic efficiency of the anode.

Reply:
We thank the reviewer for the suggestion. The coulombic efficiency of the eutectic Zn 88 Al 12 alloy anode has been shown in supplementary Figure 12. It demonstrates that the Zn stripping/plating on the eutectic Zn 88 Al 12 alloy is highly reversible, with the CE approaching 100 %, during the cycling test for more than 100 cycles.
(5) As for the full cell performance, Figure 5b is a little confusing. How did the author get the specific capacity versus the Scan rate?
Reply: We thank the reviewer for the comment. The specific capacity shown in Figure 5b is calculated according to the CV curves at various scan rates (supplementary Figure 17a,b).
(6) Generally, when the loading of the cathode is only 1.0 mg/cm2, which is much less the anode, the electrochemical performance usually relies on the cathode part but anode part. As in many previous reports, Zn foil can also support a good electrochemical performance. So, in this paper, the author should give more strong and supportive data to demonstrate the superiority of the alloy anode, for example, limiting the anode/cathode mass ratio.

Reply:
We appreciate the reviewer for the constructive suggestion, following which we have adjusted the anode-to-cathode mass ratio from 40:1 to 3:1. Based on the overall mass of anode and cathode, we have calculated the overall energy densities.
The detailed results are shown in supplementary Figure 19. The overall energy density of the full Zn 88 Al 12 /K x MnO 2 battery can reach ~142 Wh kg −1 by lowering the anode-to-cathode mass ratio to 3:1.
I am ok with the revised manuscript.