Novel Insights into Energy Storage Mechanism of Aqueous Rechargeable Zn/MnO2 Batteries with Participation of Mn2+

Highlights Pourbaix diagram of Mn–Zn–H2O system was used to analyze the charge–discharge processes of Zn/MnO2 batteries. Electrochemical reactions with the participation of various ions inside Zn/MnO2 batteries were revealed. A detailed explanation of phase evolution inside Zn/MnO2 batteries was provided. Electronic supplementary material The online version of this article (10.1007/s40820-019-0278-9) contains supplementary material, which is available to authorized users.

Many efforts have been made to reveal the energy storage mechanisms of Zn/MnO 2 ZIBs. Up to now, three types of energy storage mechanisms were proposed, including (i) Zn 2+ insertion/extraction into/from MnO 2 [8,[33][34][35][36], (ii) conversion between MnO 2 and MnOOH with the participation of H + [37], and (iii) co-insertion of H + and Zn 2+ [38]. Mechanisms (i) and (ii) explain the formation of ZnMn 2 O 4 and MnOOH as discharging products on MnO 2 cathode in ZIBs, respectively, while cannot explain that there are two redox processes during one charge/discharge cycle of ZIBs. Mechanism (iii) seems to be capable of explaining the coexistence of ZnMn 2 O 4 and MnOOH as discharging products on MnO 2 cathode, but deeper analysis will find that it is not accurate: The mechanism deems that potential of Zn 2+ insertion is lower than that of H + insertion (this means that MnOOH forms before ZnMn 2 O 4 once the battery discharge process begins), being conflicted to the experimental result that MnOOH appears latter than ZnMn 2 O 4 . In short, the current mechanisms are unsatisfactory to explain genuine charge/discharge process in ZIBs, mainly because they were proposed based on a simplistic view that the insertion of Zn 2+ and H + and the phase change from MnO 2 to ZnMn 2 O 4 or MnOOH are highly reversible. Furthermore, to achieve satisfactory cyclic stability and rate performance of the Zn/ MnO 2 ZIBs, Mn 2+ ions are always introduced in the electrolyte [37]. However, electrochemical reactions inside the ZIBs become more complicated in such cases, thus corresponding energy storage mechanism has not been clearly revealed. Therefore, it is necessary to re-examine the thermodynamic and kinetic characteristics of Zn/MnO 2 ZIBs to propose a reasonable Zn 2+ storage mechanism.
In fact, for the active materials in aqueous ZIBs and some other rechargeable aqueous batteries, their structure and phase generally undergo complex changes during charge/discharge processes (e.g., the active materials can interact with not only metal ions, but also H + , OH − , and water molecules) [39]. This is an important reason why the energy storage mechanism of MnO 2 cathode in ZIBs is still inconclusive [40][41][42][43]. Besides general experimental techniques such as cyclic voltammetry (CV) and galvanostatic charge-discharge (GCD) tests, Pourbaix diagram (E-pH diagram) has been widely used to study electrochemical reactions in aqueous solution [44][45][46][47]. The electrochemical reductive products of active materials can be predicted according to the thermodynamics, which is beneficial for us to understand the charge/ discharge process. Therefore, we combined experimental methods with the E-pH diagram of the Mn-Zn-H 2 O system together to comprehensively analyze the charge/discharge processes of MnO 2 cathode in ZIBs and tried to reveal the authentic energy storage mechanism.
Herein, based on comprehensive analysis methods including electrochemical analysis and E-pH diagram, etc., we provide novel insights into the energy storage mechanism of Zn/MnO 2 batteries with the co-participation of Zn 2+ , H + , Mn 2+ , SO 4 2− , and OH − . During the first discharge process,

Electrochemical Characterizations
The cathode was prepared by coating a mixture paste of 70 wt% of MnO 2 powder, 20 wt% of acetylene black, and 10 wt% of LA133 binder on a stainless steel foil and dried overnight under vacuum conditions at 80 °C. In the prepared cathode, the mass loading of MnO 2 is around 1 mg cm −2 . Zn/MnO 2 ZIBs were assembled based on MnO 2 cathode, metallic Zn foil anode, air-laid paper separator, and zinc salt solution electrolyte (2 M ZnSO 4 or 2 M ZnSO 4 +0.5 M MnSO 4 solution). The assembled ZIBs were kept more than 4 h before electrochemical measurements. The CV and GCD tests were performed on a Bio-logic VMP3 multichannel electrochemical station and a Land CT2001 battery tester, respectively. CV tests of the prepared MnO 2 cathode were also carried out in a three-electrode system, in which a platinum plate as the counter electrode, a saturated calomel electrode (SCE) as the reference electrode, and 50 mL electrolyte was applied.

Material Characterizations
Microstructure and composition were characterized by XRD (Rigaku 2500) with Cu-Kα radiation operating at 40 kV and 100 mA within an angle range of 10° to 70° at a scan speed of 5° min −1 . Micro-morphology was observed by field emission scanning electron microscopy (SEM, Zeiss Supra 55) and TEM (Tecnai G2 F30). Element content in electrodes and electrolytes was analyzed by inductively coupled plasma atomic emission spectrometry (ICP-AES).

Characterization of MnO 2
The MnO 2 material used in this work was synthesized through a chemical co-precipitation method. XRD pattern and micro-morphology observations in Fig. 1a-c show that the synthesized MnO 2 powder is crystalline α-MnO 2 nanorod with a diameter of 10-60 nm and length of several hundred nanometers. From the XRD pattern in Fig. 1a, it seems that the strongest peak is the one at ~ 12.8°, but if the background is taken into account, the strongest peak is still the one at ~ 37.5°, which matches well with the α-MnO 2 (PDF# 44-0141). In addition, since the as-prepared sample is nanobelts, (110) plane (corresponding to the diffraction peak at ~ 12.8°) is considered as preferred orientation, thus leading to high diffraction intensity. A similar phenomenon was also observed for some other MnO 2 nanomaterials [38]. From the high-resolution TEM (HRTEM) image in Fig. 1d, the crystal planes (121) and (330) of the α-MnO 2 with a corresponding interplanar spacing of 0.238 nm and 0.233 nm respectively are observed, and high crystallinity of the as-synthesized α-MnO 2 sample is also confirmed.

Electrochemical Analysis
We first studied the electrochemical behaviors of MnO 2 cathode in two different electrolytes, including 50 mL 2 M ZnSO 4 (Fig. 2a) and 50 mL 2 M ZnSO 4 + 0.5 M MnSO 4 mixture solution ( Fig. 2b-d). The capacity and rate performance of the MnO 2 cathode in ZnSO 4 + MnSO 4 electrolyte are exhibited in Fig. S1. Note that MnO 2 cathode would dissolve in ZnSO 4 electrolyte during charge/discharge processes, as detected by ICP-AES tests in Table S1. (This has also been pointed out in previous researches.) [37] With the addition of Mn 2+ in the electrolyte, the redox peaks in CV curves (except for the 1st CV cycle) in Fig. 2a, b become more obvious, and meanwhile, the gap between oxidation peak and reduction peak becomes smaller, which indicates that the reversibility of electrochemical process gets better. In Fig. 2a, the reduction peak is much stronger than the oxidation peak in the first cycle, which means that discharge products cannot be electrochemically oxidized completely. By contrast, the intensity of the reduction peak is close to that of the oxidation peak in Fig. 2b. Above phenomenon suggests that Mn 2+ concentration in electrolyte plays a crucial role in the first discharge/charge process [48]. There are two pairs of redox peaks when the electrode discharges/ charges in the 2 M ZnSO 4 + 0.5 M MnSO 4 electrolyte (Fig. 2b). The reduction peaks at low and high potentials are denoted as R 1 and R 2 , respectively, and the oxidation peaks at low and high potentials are denoted as O 1 and O 2 , respectively. Considering that electro-deposition of Mn 2+ will occur only when the cathode potential reaches about There is a dip and a platform in the initial GCD curve (Fig. 2c) and the reaction type of R 1 (at about 1.2 V) and R 2 (at about 1.4 V) are further studied by the constant voltage discharge test (Fig. 2d). The current changes greatly when the battery is discharged at 1.2 V at which R 1 will happen, and it keeps almost flat when discharged at 1.4 V at which R 2 will occur. This indicates that a heterogeneous reaction occurs during R 1 and a homogeneous reaction occurs during R 2 . Such a heterogeneous reaction between solid phases accompanying with nucleation process and 20 2 Theta (degree) electro-crystallization process will cause the formation of the dip and steep curve in the discharge curve in Fig. 2c [49]. With the increasing CV cycles (Fig. 2b), the peak current intensity of R 1 and O 1 is getting weaker, while the peak current intensity of R 2 and O 2 becomes stronger. Therefore, the redox reactions R 1 /O 1 and R 2 /O 2 are more likely to be independent of each other. That the initial process differs from the subsequent one can also be seen from Fig. 2c (the red circle). R 2 is weaker in the initial discharge process than that in the second one, which indicates that a new phase may generate as active materials.

Phase Evolution of Cathode in the Initial Discharge Process
ex situ XRD tests of the cathodes at different charge/discharge states support were performed. As shown in Fig.  S3, when the cathode is initially discharged from 1.9 to 1.4 V, no new phase produces, while the cathode is further discharged to 1.0 V, several new diffraction peaks occur, indicating the appearance of new phases. In the charging process, some diffraction peaks cannot be detected, which means the disappearance of some phases. After 100 charges/discharge cycles, the XRD pattern is not in conformity with the XRD patterns of the cathode at the original state and fully charged state in the 1st charge process. These demonstrate that the cathode undergoes a complicated phase evolution. In the following, phase evolution of MnO 2 cathodes during the 1st discharge process, the 1st charge process and subsequent discharge/charge processes were investigated in detail.
To investigate phases evolution of cathode in the initial discharge process in 2 M ZnSO 4 + 0.5 M MnSO 4 electrolyte, R 1 and R 2 reactions (as defined in Fig. 2, the same hereinafter) are separately studied. Only R 2 occurs when the voltage of the battery is above 1.4 V. No new phase produces are seen from the XRD pattern (Fig. 3a). Nevertheless, characteristic peaks of the MnO 2 active material such as the peaks at 28° and 38° shift (inset of Fig. 3a), which is attributed to the change of the layer spacing of MnO 2 . This means that the active material undergoes structure change. The nanorods become shorter (Fig. 3b) which is greatly different from that of the as-prepared material (Fig. 1b). Energy dispersive spectrometer (EDS)  Fig. S4 suggests that the molar ratio of Zn, Mn, and S is approximately 1:16:0. To exclude the influence of electrolyte's absorption, we immersed the electrodes in 2 M ZnSO 4 for 48 h and washed several times with deionized water. Neither Zn nor S element is found in the SEM-EDS result (Fig. S5). Thus, the existence of Zn in the cathode when discharging to 1.4 V is caused by the insertion of zinc ions in MnO 2 , instead of zinc ion adsorption on the cathode surface. In other words, zinc ion insertion happens when the voltage of the Zn/MnO 2 battery is above 1.4 V. The process of zinc-ion insertion in MnO 2 can be written as: R 1 reaction is then studied. When the battery is discharged to 1.2 V and then to the 1 V at constant current several new phases (XRD patterns in Fig. 3c) appear, which are confirmed as BZSP (PDF#39-0688), α-MnOOH (PDF#24-0713), and α-Mn 2 O 3 (PDF#44-1442). There are many large hexagonal nanosheets in the SEM image (Fig. 3d). The Zn, O, and S element distribute evenly over the whole hexagonal nanosheets (Fig. S6b, c, e, f). And the atomic proportion of Zn to S is about 4:1 from the SEM-EDS results (Fig.  S6d). Combined with the XRD result, we conclude that the (1) hexagonal nanosheets are BZSP. The structure evolution of cathode in the first discharge is shown in Fig. 3e. The existence of α-MnOOH and α-Mn 2 O 3 is further demonstrated by HRTEM (Fig. 4). The α-Mn 2 O 3 is semi-coherent with the α-MnOOH phase (Fig. 4c). These substances and their reactions can be written as [21,28,37]: Phases in regions e, f, g, and h in Fig. 4c, d marked by yellow dash can be identified as α-MnOOH, α-Mn 2 O 3 , α-Mn 2 O 3 , and α-Zn x MnO 2 , respectively, through fast Fourier transform (FFT) in Fig. 4e-h (detailed calculation procedures are given in Table S2-S4). From the TEM-EDS result, the Zn element can be found in both regions A and B and there is no S element in these regions (Fig. 4b), further confirming that zinc-ion inserts into the nanorods. Besides, the generation of BZSP and α-MnOOH indicates that H + and Zn 2+ participate in the reaction. From the above discussions, zinc ion insertion in α-MnO 2 occurs around 1.4 V versus Zn 2+ / Zn to generate α-Zn x MnO 2 , and proton conversion reaction

Phase Evolution of Cathode in the First Charge Process
To detect the phase evolution of MnO 2 cathode during the charge process, the battery is discharged to 1.0 V and then charged to 1.9 V in 2 M ZnSO 4 and 2 M ZnSO 4 + 0.5 M MnSO 4 electrolyte, respectively. The discharging products of Mn 2 O 3 , BZSP, and MnO 2 still exist on the charged cathode, and besides, new phase ZnMn 3 O 7 ·3H 2 O generates (Fig. 5a). It is worth noting that when adding Mn 2+ in the electrolyte, the diffraction peaks of ZnMn 3 O 7 ·3H 2 O become strong while the diffraction peaks of BZSP become weak (Fig. 5a). The morphology of cathode changes greatly (from hexagonal nanosheets to ball-like nanoflowers) during the initial charging process as shown in Figs. S7a-f and S8a-d. Figure S7a- Fig. 5c) as expected. ZnMn 3 O 7 ·3H 2 O is also found in HRTEM in Fig. 5d. Combining with the result of XRD, the nanoflower is ZnMn 3 O 7 ·3H 2 O. The dissolution of BZSP (nanosheets in Fig. S8a-d) and the occurrence of the new phase (nanoflowers in Fig. S8a- (Fig. S10). The structure evolution of cathode in the first charge is shown in Fig. 5e. In other words, besides H + and Zn 2+ , Mn 2+ also participates in the reactions that occur during the discharge/charge process as the conversion reactions are supposed as follows: The generation of ZnMn 3 O 7 ·3H 2 O during the first charge process can explain the phenomenon that the second discharge curve of the battery is different from the first discharge curve. In a word, during the first charging process, Zn x MnO 2 and MnOOH reversibly become α-MnO 2 with the extraction of Zn 2+ and H + , while ZnMn 3 O 7 ·3H 2 O acts as the host for the insertion of Zn 2+ forms through the reaction between Mn 2+ and BZSP.

Phase Evolution of Cathode in Subsequent Discharge/Charge Processes
The phases of the battery system after 100 charges/discharge cycles are also studied. XRD patterns of the discharged cathode after 100 cycles in 2 M ZnSO 4 + 0.5 M MnSO 4 solution in Fig. 6a imply that the main phases  (Fig. S11a, b). ZnMn 2 O 4 is on the surface of the α-MnO 2 in HRTEM (Fig. 6b), so it may be  (Fig. 6c, d). Thus, Zn 2 Mn 3 O 8 reacts with Mn 2+ to form ZnMn 2 O 4 . (This will be further analyzed in the Mn-Zn-O diagram in the following part.) There are two kinds of nanoparticles after 100 cycles (Fig. S12), which may be converted from BZSP and Mn 2+ as nanosheets and nanoparticles surround each other. Combining the XRD result (Fig. 6a), the phase is Zn 2 Mn 3 O 8 . Thus, Zn 2 Mn 3 O 8 is generated from the reactions between BZSP and Mn 2+ . The structure evolution of cathode after 100 cycles is shown in Fig. 6e. Reactions between them are as follows: To sum up, within the continuous charge/discharge process, ZnMn 2 O 4 and Zn 2 Mn 3 O 8 as host for insertion of Zn 2+ further generate on the surface of MnO 2 , which implies that the phase change of MnO 2 cathode is irreversible.

Thermodynamic Analysis
When dynamic conditions are met, the phases can be predicted from Zn-Mn-O diagram since the system of the Zn/ ZnSO 4 + MnSO 4 /MnO 2 battery reaches an equilibrium state  (Fig. 7a). Detailed density functional theory calculation and theoretical analysis of MnO 2 as a cathode of ZIBs are given in Discussion S1 and S2 in Supporting Information). There are two paths to the reduction of  (Fig. 7b).
Overall, we combined electrochemical analysis, phase identification with E-pH diagram of the Mn-Zn-H 2 O system together to analyze charge/discharge processes of aqueous rechargeable Zn//MnO 2 batteries and revealed complicated phase evolution of the cathode (i.e., what new phases will form and how can they form in different charge/discharge stages). We obtained some different conclusions from previous literature. For example, Sun et al. thought that the conversion of H + occurs before Zn 2+ insertion [38]. But we find that Zn 2+ insertion occurs before the conversion of H + in the first discharge process, and this is confirmed by thermodynamic analysis. Besides, previous literature deemed that the disappearance of BZSP is always caused by the change in electrolyte pH [34], but we find that BZSP can react with Mn 2+ in the electrolyte to form a new phase of ZnMn 3 O 7 .

Conclusions
Based on experimental results and theoretical analysis of Zn/ MnO 2 ZIBs with the mixture electrolyte of ZnSO 4 + MnSO 4 aqueous solution, we found that the mechanism in ZIBs is dynamic and the phase transformation at MnO 2 cathode is irreversible during charge/discharge processes. Not only H + and Zn 2+ but also Mn 2+ in the electrolyte take part in the reactions. In the first discharge process, Zn x MnO 2 , MnOOH, Mn 2 O 3 , and by-product BZSP generate, and then in the first charge process, α-MnO 2 and ZnMn 3 O 7 ·3H 2 O appear. In the following charge/discharge processes, ZnMn 2 O 4 and ZnMn 3 O 8 are further generated on the surface of MnO 2 and serve as the hosts for Zn 2+ insertion. The mechanism becomes dynamic and complex because of the co-participation of the insertion process, conversion reaction, and oxidation reactions. The aforementioned phase changes inside ZIBs are well explained by the Mn-Zn-O phase diagram and the E-pH diagram. This work can provide guidance for continual research from the following aspects. (i) The research method combining electrochemical analysis and phase identification with E-pH diagram together can be used to analyze charge/discharge processes of other electrochemical energy storage systems, such as aqueous rechargeable Zn//V 2 O 5 batteries. (ii) According to the proposed energy storage systems in this work, at least two approaches can be applied to enhance cycling performance of ZIBs: One is adding Mn 2+ to promote the disappearance of BZSP, and the other one is adding pH buffer into the electrolytes or preparing solid electrolytes to prohibit the generation of OH-and BZSP.