Implications of Testing a Zinc-Oxygen Battery with Zinc Foil Anode Revealed by Operando Gas Analysis.

Zinc-oxygen batteries are seen as promising energy storage devices for future mobile and stationary applications. Introducing them as secondary battery is hindered by issues at both the anode and cathode. Research efforts were intensified during the past two decades, mainly focusing on catalyst materials for the cathode. Thereby, zinc foil was almost exclusively used as the anode in electrochemical testing in the lab-scale as it is easy to apply and shall yield reproducible results. However, it is well known that zinc metal reacts with water within the electrolyte to form hydrogen. It is not yet clear how the evolution of hydrogen is affecting the performance results obtained thereof. Herein, we extend the studies and the understanding about the evolution of hydrogen at zinc by analyzing the zinc-oxygen battery during operation. By means of electrochemical measurements, operando gas analysis, and anode surface analysis, we elucidate that the rate of the evolution of hydrogen scales with the current density applied, and that the roughness of the anode surface, that is, the pristine state of the zinc foil surface, affects the rate as well. In the end, we propose a link between the evolution of hydrogen and the unwanted impact on the actual electrochemical performance that might go unnoticed during testing. Thereof, we elucidate the consequences that arise for the working principle and the testing of materials for this battery type.

Introduction

Metal–oxygen batteries are seen as promising energy storage devices and possess theoretical energy densities that can be higher than that of lithium-ion batteries.[1,2] However, their application as electrically rechargeable battery technology—both for aqueous metal–oxygen batteries and nonaqueous metal–oxygen batteries—is hindered so far because fundamental and technical challenges remain.[3−6] Major drawbacks are energy efficiency, the need for purified oxygen, as well as the poor cycle life due to electrolyte decomposition.[7,8] Especially, in the case of organic Li–O2 batteries, it is well known that high overpotentials during charge decompose the organic electrolyte.[9−11] Thereby, using operando techniques allows better understanding the processes involved.[12−17]

Among the aqueous metal–oxygen batteries, the alkaline zinc–oxygen (Zn–O2) battery has been seen as highly promising[8,18−20] but exhibits a similar, equally grave side reaction: the evolution of hydrogen (HER) on the active material Zn occurs because Zn is reacting with water from the alkaline aqueous electrolyte to form hydrogen (H2) and zinc hydroxide species (Zn(OH)2)[21−23] as follows

Thereby, not only the solvent in the electrolyte is consumed and gas bubbles are formed but the precious active material Zn at the anode is converted to an inactive, insulating material in the process. This loss of active material ultimately leads to a severe decrease in usable capacity. A detailed description of the working principle of the Zn–O2 battery is provided in the Supporting Information (see reactions S1–S3).

The HER is considered since decades to be responsible for the self-discharge of the Zn–O2 battery.[24−27] Efficient strategies were developed to suppress the HER and thus to diminish the degradation of the anode.[28] The strategies aim to reduce the activity of H2O and hence the rate of the HER, for example, by adding inorganic and organic inhibitors to the Zn metal powder or paste or by using metal alloys as the Zn anode.[24,25,29−32] It is well known that the “HER also consumes electrons during the charging process” (Chen et al.[33]) and that the electrons could rather be invested into the HER if all ZnO is converted back to Zn.

Interestingly, the major research focus for Zn–O2 batteries is on cathode materials and not on the anode with its deleterious HER and possible implications. Often new catalysts and cathode concepts are proposed and investigated to propel the much needed research on the sluggish oxygen reduction reaction (ORR) and oxygen evolution reaction (OER).[34−41] A recent benchmarking study on anode concepts for Zn–O2 batteries revealed that 65% of the published research articles during the past 20 years used simple Zn foil anodes in lab-scale research (21 out of 32 articles; see also Table S1 in the Supporting Information), although it is not the state-of-the-art anode and facilitates the HER as the side reaction.[42]

Thus, the purpose of our study is to shed light on the gas that is evolved at the anode during battery testing with the commonly used zinc foil. We aim to gain understanding about the governing factors for the electrochemical formation of hydrogen in practically relevant operating conditions. By means of electrochemical analysis, pressure change measurements, and operando differential electrochemical mass spectrometry (DEMS), we elucidate the implications for the underlying working principle of the battery. In the end, our study might help transfer the gained knowledge to apply improved cycling protocols and setups for testing novel materials for the anode and cathode of zinc–oxygen batteries.

Results and Discussion

Battery Cycling

Figure a shows cycling results obtained at 1.0 mA cm–2 for a fixed time limit (1 h discharge followed by 1 h charge). This cycling procedure, known as shallow cycling, is commonly applied for testing the performance of Zn–O2 batteries (as well as other newly developed batteries in lab-scale; compare Table S1) but does not resemble a practical use case. Stable cycling with flat discharge and charge plateaus for 60 cycles over the course of 120 h (the inset in Figure a shows the same cycling within 10 h) can be observed. The corresponding capacity withdrawn is 0.79 mA h for each discharge and charge.

Figure 1

Cycling data of Zn–O2 batteries with Zn foil as the anode: (a) at 1.0 mA cm–2 (discharge/charge step 0.79 mA h, corresponding to a DoD of 0.09%), the inset on the right shows the same cycling within 10 h; (b) first at 1.0 mA cm–2 and then at 5.0 mA cm–2 (discharge step limited to 0.9 V). The charge step is not limited by a cut-off potential so that—unintentionally—more capacity than previously discharged might be retained during charge. (c) Uncommon cycling protocol starting to charge a pristine, fully charged battery at 1.0 mA cm–2, whereas a second potential plateau above 2.00 V can be observed.

Figure b reveals the cycling performance if discharge is not limited by time but rather by potential (cut-off potential at 0.90 V). First, a Zn–O2 battery is discharged at 1.0 mA cm–2 until the cut-off potential is reached at about 2.5 h or 2 mA h. The subsequent charge is yielding the same amount of capacity at a stable potential plateau of about 1.95 V, whereas a second potential plateau emerges from there at 2 mAh (red-colored area).

Next, the cell is discharged again. This time, the charging current is changed to 5.0 mA cm–2. We can observe a first potential plateau exhibiting the same capacity as the preceding discharge with a potential value marginally higher than for 1.0 mA cm–2 and a second potential plateau which scales with the current density applied. It is to be noted that we observe only one potential plateau when charging a Zn–O2 battery with large excess of the electrolyte (compare Figure S2; charging at 10 mA cm–2 and 1 mL of electrolyte instead of a few μL). By implication, a large excess of the electrolyte might facilitate other, generally unwanted electrochemical processes, shifting the potential to another plateau, that is, the second plateau.

Figure c shows results obtained for a very different cycling protocol that is intentionally chosen to evoke side reactions in the battery: A pristine—and thus fully charged—Zn–O2 battery is charged further at 1.0 mA cm–2. Thereby, the cell potential increases to values around 2.20 V and is stable for approximately 1 h, corresponding to 0.79 mA h. Afterward, the battery is discharged, achieving approximately 4.38 mA h of discharge capacity. Subsequently, the same amount of capacity is charged at a cell potential of around 1.95 V. After 11.5 h, when the charge capacity is equal to the previously obtained discharge capacity—the cell potential shifts to another plateau above 2.00 V and charge continues steadily, that is, the battery is overcharged. However, the electrons invested during charge must stem from other reactions than the reduction of ZnO so that this mode of operation can not be of interest because it possibly leads to the degradation of the battery itself.

The second potential plateau that appears during the second charging step can also be observed if MnO2 is used as the catalyst at the cathode instead of the perovskite-type catalyst (compare Figure S3; same cycling profile for a Zn–O2 cell with commercially available cathode). Furthermore, the second potential plateau does not appear when discharge and charge are limited to a certain threshold value for the cell potential (compare Figure S4; cycling between 0.90 and 2.10 V). It remains to be an open question whether studies employing novel catalyst materials for the cathode of Zn–O2 batteries do reflect the true impact of the material on the electrochemistry if tested in a two electrode setup.

Gas Analysis

To further understand the underlying processes during charge, we monitored the pressure inside the oxygen reservoir of a closed Zn–O2 battery and applied operando DEMS analysis for an open cell setup. Figure a shows the obtained results for the operando pressure measurement. The constant slope of pressure during discharge is in line with the expected flat plateaus for the cell potential and implies that O2 is consumed in the oxygen reservoir as intended during ORR. During charge, the pressure increases almost linearly because of the release of O2 ascribed to the OER. However, at 3.4 mA h of charge (corresponding to ∼9 h of operation; starting point of the charge at the second potential plateau), the slope of the pressure β increases noticeably by a factor of 2.7 from β1 to β2, indicating that more gas is formed during charge at the second potential plateau in comparison to the charge at the first plateau.

Figure 2

Extended gas analytic during cycling: (a) pressure change in a closed cell system (during cycling at 1.0 mA cm–2). (b) Ion currents for m/z = 32 (assigned to oxygen) and m/z = 2 (assigned to hydrogen) by DEMS (during cycling at 2.5 mA cm–2; raw data in Figures S6, and S7 a). (c) Ion currents for the same m/z values at the OCV mode between 0 and 0.5 h, at 0% DoD, then charging further at 2.5 mA cm–2 between 0.5 and 1.5 h, ending in the OCV mode again.

To enable better comparison and quantification of the slope of the pressure change for a different anode geometry, we divide the slope β by the geometric surface area of the respective anode (Zn foil and Zn sponge as well as Zn particles; see also Figure S5), referred to as β*.

The value for β3* = β2* – β1* (with β1* = 23.56 mbar h–1 cm–2 and β2* = 64.17 mbar h–1 cm–2, respectively) obtained during charge at the second potential plateau is approximately 40.61 mbar h–1 cm–2 and corresponds to approximately 12.94 μmol h–1 cm–2 of evolved gas—even after ZnO as the designated source of oxygen in the overall cell reaction has been converted completely.

It can be presumed that the additional increase of pressure stems from the formation of hydrogen by means of the well-known electrochemical HER. This assumption is supported by comparing β3* to the pressure slope that is obtained for the chemical reaction of H2O from the electrolyte with the Zn foil anode (only 8.79 mbar h–1 cm–2 as shown in Figure S5a). The chemical formation of H2 is much slower compared to the electrochemical formation of H2 during overcharging the Zn–O2 battery at 1.0 mA cm–2 as shown in Figure a. The rate of the electrochemical formation of H2 increases almost linearly with the current density applied (see Figure S5b; 200.57 mbar h–1 cm–2 at 5.0 mA cm–2). The finding that the rate of the electrochemical formation of hydrogen is faster than the chemical one is in contrast to the results from Einerhand et al.,[27] whereas they draw their conclusions not from a battery during operation. Besides, a different electrolyte, namely, 8 M KOH saturated with 1 M zincate, is used, for which the amount of zincate in the electrolyte should reduce the activity of H2O. This relation should diminish the apparent rate of the electrochemical hydrogen evolution.[25]

It is worth mentioning at this point that H2 evolved in a closed battery system could be used to electrochemically recombine with O2 at the cathode to form H2O again: Müller et al. proposed that this intriguing recombination reaction can be used in a large-scale Zn–O2 battery stack to compensate for large losses of H2O from the electrolyte because of the HER and Zn corrosion.[52] Deiss et al. included the necessity for overcharging with the benefit of recombination to H2O into their model-based analysis of Zn–O2 battery cycling.[53] However, we do not observe that H2O can be regained in our lab-scale battery setup, indicated by the constant increase in gas pressure during cycling as shown in Figure a with the closed battery setup.

The results of the DEMS analysis, shown in Figure b, give additional information on the type of gas inside the battery during OCV (open circuit conditions without current flow), discharge, and charge: during OCV, a small amount of H2 (ascribed to the ion current at m/z = 2) is detected as can be expected from zinc corrosion. The amount of O2 (ascribed to the ion current at m/z = 32) decreases as expected with an almost constant slope during discharge in each cycle, whereas only a small and constant amount of H2 is detected. However, a drastic increase of evolved H2 can be observed in each cycle during charge, while O2 is evolved and can be detected as expected as well. Even by using a Zn sponge as the state-of-the-art anode, which is doped with In and Bi as inhibitors to suppress HER, we observe the same trend for the evolution of H2 (compare Figures S7b and S8).

Charging a battery further that is already at 100% state-of-charge after assembly does not show a significant increase of the ion current for O2 (compare Figure c for using a pristine Zn anode). Instead, the ion current ascribed to H2 is increasing while the battery is charged. This observation implies that the formation of the H2 species is caused by an electrochemical reaction, which we will further elucidate in the following.

To suppress the formation of H2, we assembled a Zn–O2 battery with 800 ppm of polyethylene glycol in the electrolyte (see Figure S10) and recorded the respective pressure change in the O2 reservoir. Adding a surfactant to the electrolyte decreases the amount of additionally evolved gas in general, which also holds good for our results. Using Zn particles that are surface-treated with indium (In) and bismuth (Bi) or using poly acrylic acid (PAA) as the electrolyte additive does not have a significant impact on the rate of gas evolution during the OCV mode (see Figure S10c). Because further optimization of the electrolyte is out of the scope of this study, we refer the reader to further study about electrolyte additives.[55−59]

Impact of Anode Roughness on Charge

To further shed light on the evolution of H2 during charge, we apply discharge and charge cycling for a Zn–O2 battery with a reference electrode. Besides, we use ex situ SEM analysis of the Zn anode to link morphological changes on the electrode surface to the electrochemical results obtained. The results are depicted in Figure a,b.

Figure 3

(a) Cycling a Zn–O2 battery at 1.0 mA cm–2 with a Zn foil as the anode and a Ni(OH)2/NiOOH reference electrode soaked with 4 mol dm–3 KOH. (b) SEM analysis of the anode in the pristine state and after various steps of cycling (corresponding electrochemical data shown in Figure S9).

The performed EIS measurements (Figure S11) reveal that the increase of the potential during charge at the second potential plateau is not due to an increase of the internal resistance of the battery. This result and the aforementioned cycling profiles imply that there is a relation between the charge mechanism at the second potential plateau and an increase in overpotential because of the presence of the electrochemically induced HER. For better comparison, the overpotential for the hydrogen evolution on Zn metal is given in a previous study by Lee.[54]

Figure a reveals the contribution of each electrode to the full cell potential: the cathode potential remains almost constant during each step of discharge and charge and does not fluctuate. Besides, the overvoltage at the cathode is much higher than for the anode during operation, which is well known for Zn–O2 batteries.[53] The potential of the anode is not constant during charge at the second potential plateau and fluctuates significantly, which can be correlated to evolving gas bubbles. This random gas formation can decrease the electrochemical active surface area and thus increase the potential at the anode.

The SEM analysis, shown in Figure b (corresponding electrochemical data in Figure S9; corresponding elemental analysis of the agglomerates formed is shown in Figure S12), reveals that the surface of the anode has changed during charge at the second potential plateau: a pristine anode possesses a smooth surface with a defined surface area, which leads to cycling with stable cell potentials. After one full discharge/charge cycle, the electrode surface is covered with unevenly distributed agglomerates of Zn and ZnO that have formed randomly during stripping and plating. The structural properties of the agglomerates are investigated by X-ray diffraction analysis (XRD) (results of the analysis for similar cycling data are shown in Figure S13). We observe that indeed ZnO is formed on the anode surface after discharge and that ZnO is formed to a lesser extent for overcharging a battery with a Zn foil anode.

Figure shows that the value of the second potential plateau can be shifted if the surface of the pristine Zn anode is roughened (here intentionally etched with HCl before use; increase in surface roughness between the pristine and the etched state shown in Figure b,c). Cycling a battery with the roughened anode surface leads to a decrease in the overpotential during charge at the second potential plateau by 60 mV. The potential at the anode scales with the active surface area: an increased surface area will provide more reaction sites, whereas only the same number of electrons participates in the electrochemical reaction (constant current operation). Thus, more of the reactants in the liquid electrolyte can reach the higher amount of reaction sites, lowering presumably the concentration overpotential and thus yielding a reduced overpotential in total.

Figure 4

Cycling a Zn–O2 battery at 1.0 mA cm–2 with a Zn foil as the anode starting with charging, whereas the DoD is 0% after assembly: (a) pristine state. (b) Etched Zn foil as the anode (the etching solution was 0.5 M HCl). (c,d) SEM images of the surface of both the pristine and the as-prepared/etched Zn foil anode.

H2 is formed electrochemically from the solvent H2O during the overcharge of a Zn–O2 battery, which is in line with previous studies, and is known for decades already.[54,60,61] It is to be noted that the chemical reaction of Zn metal with H2O is often assumed to be the predominant process.[27,62−64]

In Figure S14, we summarize the reactions known to take place at the anode of the Zn–O2 battery. It can be presumed that the impact of the electrochemically induced HER[54,60,61] has severe consequences for the cycling of the battery because the reaction consumes solvent water in the aqueous electrolyte. In this view, it might be challenging to estimate the practical cycling stability and lifetime of Zn–O2 batteries if Zn foil is used as the anode because it enables the consumption of the solvent H2O in the electrolyte and gas evolution at the same time (compare also Figure S2; charging a cell with a large excess of the liquid electrolyte progresses without a change in the charging plateau). Using defined amounts of the electrolyte instead of an excess electrolyte can help notice whether unwanted overcharge takes place or not—complete consumption of the electrolyte would then yield early battery failure and reasonable, practical cycle numbers.

By implication, testing a novel catalyst material at the cathode by means of a full cell test with the Zn foil anode might not reveal the full impact of the cathode because slight changes on the anode surface (different distributor; variation in surface treating before use) will affect the overall cell potential.

Conclusions

We aim to extend the already existing knowledge about the evolution of hydrogen during the operation of zinc–oxygen batteries. By applying a comprehensive set of electrochemical measurements, morphology analysis, and gas analytic, we show a link between the evolution of hydrogen and the processes at the anode. We observe by means of DEMS that the ion current signal ascribed to hydrogen is higher during charge than during discharge and conclude that hydrogen mainly stems from a deleterious electrochemical reaction at the anode. This process and its polarization behavior both depend on the current as well as on the surface area of the anode. By implication, some of the electrons used to recharge aqueous zinc–oxygen batteries can be easily dissipated into deleterious side reactions that, in the end, can consume water until the electrolyte is depleted.

The results gained help to put the use of zinc foil as the anode for performance testing of zinc–oxygen batteries into perspective: once all zinc oxide is converted back to zinc, an increase in overpotential due to the presence of the electrochemically induced evolution of hydrogen sets in. The starting point for this increase is not detectable in the electrochemical data if a large excess of the electrolyte is used. The findings can be used to carefully select the testing protocols and cell setups for zinc–oxygen batteries to reveal their true performance and the impact of novel materials introduced.

Methods

Electrode Preparation

The Zn foil anodes (2 mm thickness, >99.99%, ChemPur, Germany) were polished before use with SiC paper (5 μm, Buehler); the preparation of Zn sponge anodes (used for better comparison to state-of-the-art cycling performance; based on the work by Parker and colleagues[43,44]) and gas diffusion electrodes (GDEs) with the Sr2CoO3Cl catalyst was carried out as described previously.[45,46] HCl (0.5 M) (reactant grade) was used to chemically etch off the surface of pristine Zn anodes.

The Zn foil anode is not the state-of-the-art electrode regarding cycling stability and performance but was chosen for the purpose of the herein conducted fundamental study to shed light on the otherwise less-distinct HER during the operation. The reasons for using Zn foil as the anode in Zn–O2 battery research are as follows: easy handling and cutting to almost any size; no need for additional host material or current collector while providing mechanical stability. On the downside, no full discharge to 100% of the discharge capacity is possible, as well as fast passivation[47,48] with ZnO can occur (compare diminished passivation with 3D-structured and Zn particle anodes).[43,44,49]

Assembly of Zn–O2 Cells

Zn–O2 cells were prepared as CR2032 coin-type cells (TOB New Energy Limited) or as in-house-modified Swagelok-type cells (i.e., known as “Giessen cell”)[50] using the following assembly protocol: The freshly polished Zn foil (12 mm diameter) anode was placed on top of a tin (Sn) disk current collector (125 μm thickness, 99.99+%, ChemPur, 16 mm diameter for CR2032, 13 mm diameter for Swagelok-type) and covered with 150 μL of 4 mol dm–3 KOH aqueous electrolyte. Four laminated (16 mm diameter for CR2032, 13 mm diameter for Swagelok-type) nonwoven separators (Celgard 5550) were soaked in 4 mol dm–3 KOH solution and used as the separator between the Zn anode, and the GDE. 800 ppm polyethylene glycol (PEG 600, Sigma-Aldrich) was used as the additive in 4 mol dm–3 KOH solution to suppress HER where mentioned. Additional 25 μL of KOH solution (4 mol dm–3) were used to wet the exposed separator membrane surface again before placing the GDE (10 mm diameter) and closing the cell. A titanium mesh (10 mm diameter, 0.076 mm wire thickness, Alfa Aesar) was used as the current collector at the cathode side. As particular modification in the coin-type cell, a stainless steel spacer (15.5 mm diameter, 0.2 mm thick, MTI Corporation) was placed underneath the Sn current collector. The void space between the spacer, Sn current collector, and Zn anode was kept dry, and the infiltration of the electrolyte was prevented by using an insulating tape made of polytetrafluoroethylene. A stainless steel spring (15.4 mm diameter, 1.1 mm thick, MTI Corporation) was on top of the titanium mesh current collector. Finally, the coin cell was automatically closed with a crimping machine at a pressure equivalent to 1 ton per coin cell area (MTI Corporation). A Ni(OH)2/NiOOH reference electrode (Panasonic) was used as the quasi-reference electrode in the three-electrode measurements with a Swagelok-type cell.

To evaluate and report data on Zn−O2 batteries, guidelines introduced in a very recent perspective article by Parker et al.,[51] were followed.

Pressure Measurement

Time-dependent pressure change inside the gas volume of the Swagelok-type cells was recorded using an PAA-33X absolute pressure sensor, a K104B USB computer adapter (both Omega Engineering), and the software ControlCenterSeries30 (Keller AG) during cycling (compare Figure S1a for the measurement setup).

Electrochemical Characterization

Electrochemical characterization was performed using a SP-300 potentiostat/galvanostat (Biologic) and a BCS-805 battery cycler (Biologic) at room temperature. The Zn–O2 coin cells were cycled inside a sealed gas container [total volume of approximately 0.5 L, filled with humidified O2 (purity 5.0, Praxair)]. The Swagelok-type cells—with a gas reservoir of approximately 8 mL—were likewise filled with pure humidified O2 at 1 bar. Zn–O2 cells were galvanostatically cycled at ±1.0, 5.0, and 10.0 mA cm–2 with the cut-off potential during discharge set to 0.90 V.

Where needed, capacity values referred to in this work are based on the mass of active material Zn used. Overcharging occurs if the mA h/gZn gained during discharge is gained back in the subsequent charge step, added a surplus amount of mA h/gZn in capacity.

Differential Electrochemical Mass Spectrometry

Lysgaard et al. used additives on the Zn anode surface and determined the amount of H2 that was formed by using DEMS by accumulating gas within 1 h of charge for each data point.[21] We apply a continuous flow DEMS setup, that is, an open system to investigate the gas composition in electrically rechargeable Zn–O2 cells with resolution in time in the range of several minutes (see also Figure S1b for details on the measurement setup). After preparation of the electrodes (Zn foil or Zn sponge, GDE and Sn disk current collector; all 10 mm diameter) and separators (three anion-exchange membranes 16 mm diameter, A-201 (Tokuyama) or nonwoven separators (Celgard 5550) soaked with the electrolyte as previously reported[46]), the cell elements were transferred to a glovebox with a nitrogen atmosphere and assembled as the Zn–O2 cell as described above. Then, the Zn–O2 cell was connected to the DEMS setup, which consists of a BaSyTec cycling station for electrochemical analysis, a mass flow controller (control unit PR4000, MKS), and the mass spectroscopy unit (Prisma QMA 200, Pfeiffer Vacuum GmbH). The cell was initially fed with 3.5 mL min–1 of synthetic air (80% N2, 20% O2) to ensure optimal reactant distribution; the flow was reduced to 0.5 mL min–1 after 15 min, allowing an optimal determination of the gas composition. To prevent drying out of the cell, the gas flow was directed through a humidifier (tank with deionized H2O) and then directed into the cell. For better data visualization, the current signal for m/z = 32 (assigned to oxygen) was drift-corrected because a linear increase of the total ion current I0—originating from the measurement setup and not from the electrochemical cell—was observed (compare raw data in the Supporting Information).

Scanning Electron Microscopy

Microstructure images of Zn foil anodes (before and after cycling) were obtained on a Merlin high-resolution scanning electron microscope (SEM; Carl Zeiss AG, Germany).

Energy-Dispersive X-Ray Spectroscopy

Extended surface analysis by means of energy dispersive X-ray spectroscopy (EDS) mapping was performed with a 50 mm2 Silicon Drift Detector X-Max (Oxford Instruments) during SEM analysis at the very same samples.

X-Ray Diffraction

An Empyrean X-ray powder diffractometer (Cu Kα, 40 kV, 40 mA; PANalytical) was used to provide structural data on the surface of Zn foil anodes.