Promoting the reversibility of electrolytic MnO₂-Zn battery with high areal capacity by VOSO₄ mediator

INTRODUCTION

Lithium-ion batteries have emerged as the leading power source due to their exceptional energy density^[1,2]. However, the flammability and toxicity of organic electrolytes have raised significant safety concerns, restricting their widespread use in energy storage applications^[3,4]. In light of these challenges, aqueous zinc (Zn)-ion batteries have gained considerable attention. Zn, with its low reduction potential (-0.763 V vs. SHE) [Equation (1)], abundant reserves, and affordable price, stands out as an attractive option. Furthermore, the use of aqueous electrolytes ensures inherent safety and environmental friendliness^[5-9].

In the realm of Zn-ion batteries, MnO₂ has traditionally served as a cathode material. It exhibits a theoretical capacity of 308 mAh g^-1 by allowing the intercalation of Zn²⁺ and/or H⁺ ions while reducing Mn⁴⁺ to Mn^3+[10,11]. Recently, a novel electrolytic mechanism of MnO₂ [Equation (2)] has been reported. The two-electron transfer reaction involves the deposition and dissolution of MnO₂, resulting in an impressive theoretical capacity of 616 mAh g^-1[12-18]. Notably, the MnO₂-Zn battery benefits from the high reduction potential of Mn²⁺/MnO₂ (1.228 V), enabling it to achieve a high voltage output of 1.991 V^[19-28].

However, electrolytic MnO₂-Zn batteries exhibit a limited cycling lifespan due to the incomplete dissolution of deposited MnO₂. This issue is exacerbated by the accumulation of MnO₂ particles, which can detach from the current collector and block ion transportation channels in the separator. Consequently, substantial polarization and rapid capacity degradation occur in the battery^[29,30].

To address this prominent issue, various ions have been introduced into the electrolyte of electrolytic MnO₂-Zn batteries to enhance their electrochemical performance^[31-36]. For instance, the incorporation of Ni²⁺ ions serves as a catalyst for the Mn²⁺/MnO₂ reaction kinetics, enabling charge and discharge at a rate of 50 C while maintaining a lifespan of over 450 cycles at an areal capacity of 1 mAh cm^-2[31]. Additionally, Al³⁺ ions play a vital role in promoting MnO₂ dissolution by creating oxygen vacancies during deposition, leading to improved cycling stability over 2,000 cycles at an areal capacity of 2 mAh cm^-2[36]. Another approach involves the utilization of redox mediators such as I^- and Br^- ions^[30,33], which interact with MnO₂ to prevent its accumulation. This mechanism has shown promising results, extending the lifespan of electrolytic MnO₂-Zn batteries to over 100 cycles at an areal capacity of 5 mAh cm^-2. Nevertheless, achieving enhanced reversibility and cyclic stability, particularly at high areal capacities, remains a significant challenge.

In this study, we employ VOSO₄ as a redox mediator to enhance the dissolution of MnO₂ in the electrolytic MnO₂-Zn battery. VOSO₄ is commonly utilized as an active component in the cathode of vanadium-based flow batteries due to its favorable reversibility and stability^[37,38], Given that the reduction potential of the VO²⁺/VO₂⁺ reaction is slightly lower than that of the Mn²⁺/MnO₂ reaction (1.01 V vs. 1.228 V), VOSO₄ is a suitable choice as a redox mediator for the Mn²⁺/MnO₂ reaction. Through electrochemical measurements, ex-situ UV-visible (UV-vis) spectroscopy, X-ray diffraction (XRD), and scanning electron microscopes (SEM), we validate the interaction between VO²⁺ and MnO₂, which effectively mitigates the accumulation of MnO₂. By employing this redox mediator, the MnO₂/VOSO₄-Zn battery exhibits exceptional redox reversibility and excellent cycling stability, even at high areal capacities. Specifically, it achieves 900 cycles at 5 mAh cm^-2 and 500 cycles at 10 mAh cm^-2. Notably, even in the flow battery device, the battery demonstrates stable cycling performance for over 300 cycles at 20 mAh cm^-2.

EXPERIMENTAL

RESULTS AND DISCUSSION

The electrochemical performance of the Mn²⁺/MnO₂ was first studied using a three-electrode system. The electrolyte consisted of 0.5 M MnSO₄, 1 M Na₂SO₄, and 0.5 M H₂SO₄ (referred to as E-Mn). During the galvanostatic discharge/charge measurements [Figure 1A], E-Mn exhibits a cyclic life of less than 200 cycles when charged to 5 mAh and discharged to 0.2 V vs. SCE at an applied current density of 10 mA cm^-2. The limited cyclic life can be attributed to incomplete MnO₂ dissolution [Figure 2A]. As revealed in the SEM image [Supplementary Figure 1], a thick layer of MnO₂ remains on the carbon fiber after 30 cycles. Therefore, enhancing MnO₂ dissolution is imperative to extend the cyclic life of Mn²⁺/MnO₂.

Figure 1. (A) Charge/discharge profiles, (B) CV curves of E-Mn, E-VO, and E-MnVO at 0.5 mV s^-1 and (C) cyclic performance of E-Mn and E-MnVO at 5 mAh cm^-2 and 10 mA cm^-2. (D) Optical images, (E) UV-vis spectra, and (F) discharge profiles of VOSO₄ solution, MnO₂ suspension, and a mixture of VOSO₄ and MnO₂.

Figure 2. Schematic illustration of the charge/discharge process of the (A) MnO₂-Zn and (B) MnO₂/VOSO₄-Zn batteries.

The electrolyte, denoted as E-VO, consists of 0.05 M VOSO₄, 1 M Na₂SO₄, and 0.5 M H₂SO₄. E-VO displays a pair of peaks at 0.75/0.71 V vs. SCE on the CV curve, corresponding to the redox reaction of VO²⁺/VO₂⁺ [Figure 1B]. When charged/discharged in the range of 0.2-1.1 V at a current density of 10 mA cm^-2, E-VO exhibits a stable discharge capacity of ~0.42 mAh cm^-2 over 4,000 cycles [Supplementary Figure 2]. Comparatively, the VO²⁺/VO₂⁺ redox couple demonstrates a lower redox potential and improved cyclic stability than Mn²⁺/MnO₂, indicating that VOSO₄ is a suitable redox mediator for the conversion reaction of Mn²⁺/MnO₂.

The cyclic performance of E-Mn at different VOSO₄ concentrations was evaluated [Supplementary Figure 3]. The E-Mn system with a VOSO₄ concentration of 0.02 M displays a relatively low coulombic efficiency and a short lifespan. While an addition of 0.1 M VOSO₄ improves the coulombic efficiency and cyclic stability, it results in a relatively low energy density. Consequently, a concentration of 0.05 M VOSO₄ was determined to be the optimal choice for the system, termed E-MnVO. The initial charge curve of E-MnVO exhibits a slope followed by a plateau at 1.13 V [Figure 1A], corresponding to the oxidation of VO²⁺ prior to that of Mn²⁺. The oxidation of VO²⁺ contributes a capacity of ~0.5 mAh cm^-2, consistent with the charge capacity delivered by E-VO [Supplementary Figure 2]. Initially, E-MnVO displays low discharge capacities. After a few cycles of activation, the discharge capacity of E-MnVO gradually increases and stabilizes. This phenomenon may be attributed to the initial cycles where an excessive amount of VO²⁺ ions cannot be fully oxidized into VO₂⁺ ions during the charging process. These excess VO²⁺ ions react with MnO₂, consequently depleting MnO₂, but without contributing to the discharge capacity. As a result, the discharge capacity is relatively lower during the initial cycles. However, after some cycles, the oxidation of VO²⁺ ions contribute more capacity in the charge process as indicated by the lowered charge plateau, which may be because residual MnO₂ trap Mn²⁺ ions in the electrolyte. The remaining VO²⁺ ions are reduced in the electrolyte after the charge process, therefore leading to an increase of the discharge capacity. The discharge curve of the 100th cycle [Figure 1A] presents two plateaus, indicating the successive electrochemical reductions of MnO₂ (1.06 V) and VO₂⁺ (0.69 V). This cycle delivers a capacity of 4.75 mAh cm^-2 with a coulombic efficiency of 95%. Moreover, the charge/discharge platforms of E-MnVO maintain stability throughout cycling, with a cyclic life exceeding 1,000 cycles [Figure 1A]. In contrast, the charge/discharge platforms of the E-Mn gradually increase over cycling, suggesting a decrease in pH value due to the accumulation of MnO₂. The capacity of E-Mn remains stable for only 200 cycles [Figure 1C]. SEM images [Supplementary Figure 4] demonstrate the accumulation of thick layers of MnO₂ on or exfoliating from the carbon fibers in E-Mn after 200 cycles. However, with the addition of VOSO₄, the accumulation of MnO₂ on the carbon felt significantly decreases after 200 cycles.

To elucidate the underlying mechanism of the VOSO₄ mediator, we conducted an experiment by adding MnO₂ powder into the blue VOSO₄ solution. As a result, the precipitate dissolved [Figure 1D], leading to a change in solution color to light yellow. Accordingly, the distinct peak on the UV-vis spectrum of VOSO₄, corresponding to VO²⁺ (~770 nm)^[37], disappeared upon the addition of MnO₂, suggesting a reaction between VOSO₄ and MnO₂ [Figure 1E]. Furthermore, we examined the discharge curves of the VOSO₄ solution, MnO₂ suspension, and a mixture of VOSO₄ and MnO₂. Individually, VOSO₄ and MnO₂ exhibited negligible capacities [Figure 1F], attributed to the reductive state of VO²⁺ and the inadequate contact between MnO₂ suspended in the electrolyte and the carbon felt. However, when the mixture was formed, it became transparent and displayed a plateau at 0.70 V, indicating MnO₂ oxidizes VO²⁺ to VO₂⁺, which can subsequently undergo electrochemical reduction.

Therefore, the working principle of the VOSO₄ mediator to improve the dissolution of MnO₂ is depicted in Figure 2B. During the charging progress, the oxidation of VO²⁺ precedes the deposition of MnO₂ on the carbon felt, generating VO₂⁺ [Equation (3)]. Conversely, during discharge, MnO₂ dissolves into Mn²⁺ prior to the reduction of VO₂⁺. The reduction product VO²⁺ then reacts with the residual MnO₂, leading to the formation of VO₂⁺ and Mn²⁺ [Equation (4)]. Moreover, the resulting VO₂⁺ is subjected to another round of electrochemical reduction [Equation (3)], steadily providing a continuous supply of VO²⁺ to react with MnO₂ until all the remaining MnO₂ is consumed. As a result, this process promotes the reversibility of the MnO₂ deposition and dissolution.

To investigate the impact of the VOSO₄ mediator in the MnO₂-Zn full cell, we introduced VOSO₄ into the catholyte of a two-electrode configuration [Supplementary Figure 5]. First, we assessed the capacity contribution and stability of 0.05 M VOSO₄ in the MnO₂-Zn full cell. The VOSO₄-Zn batteries exhibit excellent cyclic stability, retaining a reversible capacity of 0.3 mAh cm^-2 even after 3,000 cycles [Supplementary Figure 6]. The MnO₂/VOSO₄-Zn battery, charged to 1 mAh cm^-2 at 10 mA cm^-2, exhibits stable charge/discharge performance for over 3000 cycles, surpassing the cyclic life of the MnO₂-Zn battery [Figure 3A and Supplementary Figure 7]. When the charging capacity increases to 5 mAh cm^-2, the MnO₂-Zn battery exhibits a high discharge plateau at 1.95 V and coulombic efficiency of 95% at 10 mA cm^-2 in the first cycle. However, after 150 cycles, the discharge plateau declines to 1.69 V, and the gap between the charge and discharge plateaus widens to 0.76 V [Figure 3B]. This decline in performance can be attributed to the accumulation of undissolved MnO₂, leading to serious polarization and high charge transfer resistance [Supplementary Figure 8], which results in a substantial loss in capacity. In comparison, the use of the VOSO₄ mediator reduces the discharge voltage. However, the MnO₂/VOSO₄-Zn battery demonstrates remarkable stability in charge/discharge cycles, exceeding 900 cycles and maintaining a high coulombic efficiency of approximately 98% [Figure 3C and D]. On the other hand, the energy efficiency of the MnO₂-Zn battery gradually declines over cycling, averaging around 60%. In contrast, the MnO₂/VOSO₄-Zn battery displays a stable energy efficiency of approximately 75% throughout cycling [Supplementary Figure 9].

Figure 3. Electrochemical performance of the MnO₂-Zn and MnO₂/VOSO₄-Zn batteries: (A) Cycling performance at 1 mAh cm^-2, (B and C) charge and discharge curves, and (D) cyclic performance at 5 mAh cm^-2 and 10 mA cm^-2, (E and F) discharge curves at 5 mAh cm^-2 and various current densities, and (G) cyclic performance at 10 mAh cm^-2 and 10 mA cm^-2.

Figure 3E and F compares the rate capability of the MnO₂-Zn battery and MnO₂/VOSO₄-Zn battery, ranging from 5 to 50 mA cm^-2. As a current density of 50 mA cm^-2, the MnO₂/VOSO₄-Zn battery exhibits a discharge plateau at 1.77 V, maintaining a reversible capacity of 3.50 mAh cm^-2 and an energy density of 5.42 mWh cm^-2. In contrast, the MnO₂-Zn cell displays a lower discharge voltage of 1.67 V, with a reversible capacity of 3.09 mAh and an energy density of 4.63 mWh cm^-2. Additionally, even when charged to 10 mAh cm^-2 at 1 C [Figure 3G], the MnO₂/VOSO₄-Zn battery exhibits a high coulombic efficiency of 97 % and an energy efficiency of 79% after 500 cycles [Supplementary Figure 10]. On the other hand, the MnO₂-Zn battery fails to deliver capacity after 100 cycles. These comparisons of the charge/discharge performance clearly highlight the effectiveness of VOSO₄ as a mediator in enhancing the cyclic stability of the MnO₂-Zn battery.

To investigate the influence of VOSO₄ on the deposition and dissolution behavior of Mn²⁺, extensive characterization was performed on the cathodes of the MnO₂/VOSO₄-Zn and MnO₂-Zn batteries using XRD, SEM, and XPS after charge and discharge cycles. Following the 20th charging at 5 mAh cm^-2 and 10 mA cm^-2, diffraction peaks corresponding to ε-MnO₂ (JCPDS #30-0820) can be observed in the XRD pattern of the cathode from the MnO₂/VOSO₄-Zn battery [Figure 4A]. The broad peaks with low intensity indicate low crystallinity of the deposited ε-MnO₂. SEM images [Figure 4B and C] reveal a dense layer of MnO₂ layer composed of nanoparticles. XPS analysis of the charged cathode shows peaks at 84.16 and 89.26 eV in the Mn 3s region, corresponding to Mn⁴⁺ [Figure 4D]. Peaks at 529.9 and 531.3 eV in the O1s region correspond to the Mn-O-Mn and Mn-OH bonds, respectively [Figure 4E]^[12]. Notably, the energy splitting of Mn⁴⁺ peaks in the charged cathode of the MnO₂/VOSO₄-Zn battery is higher than that in the MnO₂-Zn battery (5.10 eV vs. 4.99 eV) [Supplementary Figure 11], suggesting the formation of lower-valence Mn during the deposition^[33,39]. The VOSO₄ mediator introduces more defects in MnO₂, thereby promoting its dissolution. After subsequent discharge to 1.0 V, the peaks corresponding to ε-MnO₂ disappear in the XRD pattern [Figure 4A], and the surface of the carbon fiber appears bare and smooth in the SEM image [Figure 4F, Supplementary Figure 12]. Additionally, the peaks associated with Mn⁴⁺, Mn-O-Mn, and Mn-OH disappear in the XPS spectrum, indicating good reversibility of MnO₂ deposition and dissolution. After 100 cycles, the ε-phase MnO₂ peaks are nearly absent in the XRD pattern of the cathode [Figure 5A], and only a few residual MnO₂ particles are observed in the SEM image [Figure 5B and C]. These SEM and XRD measurements conclusively confirmed that MnO₂ can be completely dissolved by the VOSO₄ mediator during cycling.

Figure 4. Characterizations of the cathode in the MnO₂/VOSO₄-Zn battery in the 20th cycle: (A) XRD patterns at different states, (B) SEM and (C) EDS elemental mapping images after charge, (F) SEM image after discharge, XPS spectra in the region of (D) Mn 3s and (E) O 1s.

Figure 5. Comparisons of the cathodes in the MnO₂/VOSO₄-Zn and MnO₂-Zn batteries after cycling: (A) XRD patterns, SEM images after (B) 10 and (C) 100 cycles of the cathode in the MnO₂/VOSO₄-Zn battery. (D) XRD patterns, SEM images after (E) 10 and (F) 100 cycles of the cathodes in the MnO₂-Zn battery.

Regarding the cathode of the MnO₂-Zn battery, the XRD pattern consistently exhibits increasingly intense diffraction peaks corresponding to MnO₂ after 20 cycles [Supplementary Figure 13] and 100 cycles [Figure 5D]. SEM images reveal a remaining MnO₂ layer on the carbon fiber with a thickness of approximately 6 μm after 20 cycles [Supplementary Figures 14 and 15], which further grows to 14 μm after 100 cycles [Figure 5E and F]. Furthermore, some MnO₂ particles even detach from the carbon felt, obstructing ion transport channels on the membrane [Figure 6A]. The accumulation and detachment of MnO₂ ultimately lead to the failure of the MnO₂-Zn battery.

Figure 6. (A) Comparisons of Nafion membranes obtained from the cycled MnO₂/VOSO₄-Zn and MnO₂-Zn batteries before and after immersion in E-MnVO. (B) Discharge curves and (C) cyclic performance of the failed MnO₂-Zn battery after addition of VOSO₄ at 5 mAh cm^-2 and 10 mA cm^-2. (D) Charge/discharge curves and (E) cyclic performance of MnO₂/VOSO₄-Zn flow battery at 20 mAh cm^-2 and 10 mA cm^-2. (F) Comparison of the electrochemical performance of reported electrolytic MnO₂-based batteries.

Figure 6A provides a comparison of the Nafion membranes from the MnO₂/VOSO₄-Zn and MnO₂-Zn batteries after 100 cycles. The membrane in the MnO₂/VOSO₄-Zn battery remains transparent, while the membrane in the MnO₂-Zn battery turns black due to the migration of MnO₂. However, the blackened membrane regains transparency after immersion in E-MnVO. The MnO₂-Zn battery experienced failure after 218 cycles. When VOSO₄ is added to the catholyte of the failed MnO₂-Zn battery, the battery exhibits a discharge capacity of 31 mAh cm^-2. Because a significant amount of MnO₂ accumulated on the carbon felt or suspended within the catholyte reacts with VO²⁺ ions, generating VO₂⁺ ions, which contribute discharge capacity through the electrochemical reduction. After VOSO₄ completely transforms the accumulated MnO₂ into Mn²⁺, the battery is successfully revived and demonstrates stable charge/discharge behavior in subsequent cycles [Figure 6B and C]. Conversely, replacing the zinc foil in the failed MnO₂-Zn battery does not resolve the abnormal charge/discharge performance [Supplementary Figure 16]. This comparison indicates that the failure of the MnO₂-Zn battery is primarily attributed to the accumulation of undissolved MnO₂ rather than the degradation of the zinc anode. Hence, the use of the VOSO₄ mediator proves effective in extending the lifespan of the MnO₂-Zn battery.

Additionally, we introduced the VOSO₄ mediator into the catholyte of the MnO₂-Zn flow battery [Supplementary Figure 17], resulting in the achievement of even higher area capacity. As depicted in Figure 6D and E, when the charge capacity is increased to 20 mAh cm^-2, the flow battery exhibits stable cyclic performance with a coulombic efficiency of approximately 97% and an energy density of 30.32 mWh cm^-2 after 300 cycles. Remarkably, even at 30 mAh cm^-2, the flow battery achieves 90 stable cycles with excellent reversibility [Supplementary Figure 18]. Furthermore, the summarized results in Figure 6F and Supplementary Table 1 illustrate that our assembled MnO₂/VOSO₄-Zn battery displays outstanding performance in terms of areal capacity, reversibility, and cycle life compared with the reported electrolytic MnO₂-based batteries. To upscale the battery, a carbon felt measuring 64 cm² (8 cm × 8 cm) was utilized as the current collector for the cathode, resulting in the assembly of a larger-scale MnO₂/VOSO₄-Zn battery with a charge capacity of 600 mAh [Supplementary Figure 19]. This battery demonstrates good cyclic stability, maintaining approximately 500 mAh and 751 mWh over 20 cycles, highlighting the potential of the MnO₂/VOSO₄-Zn battery for practical energy storage applications.

CONCLUSIONS

In conclusion, our study demonstrates significant improvements in the reversibility and cycling stability of MnO₂-Zn batteries through the utilization of redox mediator VOSO₄. The oxidation reaction involving VO²⁺ effectively interacts with residual MnO₂, facilitating its dissolution. As a result, when this mediator strategy is adopted in the MnO₂-Zn battery, it exhibits an impressive long cycle life of 500 cycles, even at a high area capacity of 10 mAh cm^-2, while maintaining a coulombic efficiency of approximately 97%. Furthermore, the MnO₂/VOSO₄-Zn flow battery demonstrates excellent cycle stability, maintaining performance over 300 cycles at 20 mAh cm^-2. These findings provide valuable insights into the potential large-scale application of electrolytic MnO₂-Zn batteries.