2022 Volume 90 Issue 6 Pages 067002
For practical room-temperature magnesium rechargeable battery operation, utilizing the nano-sized MgMn2O4 spinel as cathode material is effective way to have high rate-capabilities; however, electrolyte decomposition at the cathode/electrolyte interface is inevitable. In this work, a Mg-Fe binary oxide is coated onto nano-sized MgMn2O4 to form stable cathode electrolyte interface (CEI). This developed cathode material suppresses the electrolyte decomposition and improves the cyclability at room temperature.
With the implementation of the decarbonization policies worldwide, the role and demand for energy storage batteries is growing and changing significantly. Therefore, further improvements in battery technology are required. Lithium-ion batteries are well-known and widely used in several industries.1 Despite lithium-ion batteries have been explored for their larger storage capacity, longer battery life, and faster recharging speed, there are still some issues that cannot be underestimated: the high price of Li metal, the need for safety measurements, and the lack of energy density.
Magnesium rechargeable batteries (MRBs), which use Mg metal as the anode, are attracting attention as post-Li-ion batteries. MRBs have been widely explored because of their high volumetric energy density (3,833 vs. 2,046 mAh cm−3 for Li metal), low cost of Mg metal (one-tenth of that of Li metal), and high safety (higher melting point for Mg metal than that of Li metal).2–4 In 2000, Aurbach et al. reported the first prototype MRB using Mo6S8 as the cathode, which achieved reversible charge-discharge cycles at room temperature and was the most successful cathode regarding cyclability.5 However, its nominal terminal voltage and theoretical energy density were of 1.0–1.3 V and 135 Wh kg−1, which is about half compared with those of current Li-ion batteries. More recently, researchers have been actively pursuing the development of high-voltage and high-energy density cathode materials for MRBs. Among the many reports on cathode materials, such as spinel-type,6–10 layered-type,11,12 and polyanion-type,13,14 spinel-type MgM2O4 (M = Cr, Mn, Fe, or Co) has attracted considerable attention owing its high-voltage operation. In particular, MgMn2O4 (MMO) is expected to be a high-energy cathode material as it can be used in both Mn2+/Mn3+ and Mn3+/Mn4+ redox reactions within the potential window of a standard electrolyte (<3.5 V vs. Mg/Mg2+).15–17
Nonetheless, the abovementioned oxide-type cathode materials show poor rate capabilities for practical use at room temperature, due to the sluggish diffusion rate of Mg-ion in the solid derived from the large charge density of Mg2+.18 One effective solution is to utilize nanosized cathode materials to reduce the diffusion distance in the solid.19–21 In our previous work, a cubic nano-spinel MMO cathode with a size of approximately 5 nm was synthesized via alcohol reduction and exhibited a superior rate capability at room temperature when operated using a half-cell with an acetonitrile-based electrolyte.22 The acetonitrile-based electrolyte is suitable for MRB half-cell tests owing its large potential window for oxidation, whereas this electrolyte cannot be applied for full-cell tests, since Mg2+ deposition hardly occurs on the anode.23 Instead, ether-based electrolytes, such as Mg[B(HFIP)4]2/triglyme (HFIP: hexafluoroisopropyl), are suitable for full-cell tests as their electrolytes are relatively easy for magnesium deposition/dissolution.24 However, for full-cell test using ether-based electrolytes, the cathode hardly cycles because of the catalytic oxidative decomposition of the electrolyte during charge. Therefore, a stable cathode/electrolyte interface (CEI) must be established; the CEI requires high ion conductivity, large potential window, low reactivity with the electrolyte or cathode material, and uniform coating on nano-sized particles. Theoretical studies25,26 suggest Mg(PO3)2 and MgAl2O4 as candidates for high-voltage cathode materials from Mg2+ conductivity and electrochemical stability windows. Since the core material is spinel-type oxide MMO, oxide-type MgAl2O4 can be suitable to form close interface, considering its affinity. However, our preliminary experiment showed that MgAl2O4 coating to MMO (MAO@MMO) did not show cyclability improvement, though first discharge capacity was increased (Fig. S1). Therefore, another oxide-type coating materials should be applied.
The present study was focused on the Mg-Fe binary oxide (MFO). In the typical case of cathode materials, the valence band maximum of transition metal is sufficiently high to decompose the electrolyte instead of Mg extraction from the cathode.27 In contrast to MMO, spinel-type MgFe2O4 does not easily decompose the electrolyte, having the largest overpotential for oxidative electrolyte decomposition, suggesting that its composition is suitable for stabilizing MRBs. Han et al. reported the cathodic cyclability was enhanced by replacing Fe into the Mn site of MMO, where the contact area between Mn and the electrolyte was reduced.28 Therefore, we consider that a core-shell structure with a nano-sized core material (core: nano-sized MMO, shell: MFO) is the most effective way to achieve a high capacity with cyclability, due to the effects of reducing Mg-ion-diffusion distance (nano-sized core material) and suppressing the electrolyte decomposition (core-shell structure).29 Herein, we develop a nano-spinel MMO cathode material coated with MFO. Due to the high reactivity between the MFO/MMO interface, a subsequent coating process is performed using alkoxides30 (Fig. 1): i) Mg alkoxide reacts with MMO surface by dispersing MMO into Mg alkoxide solution, followed by hydrolysis to form Mg hydroxide-coated MMO; ii) Fe alkoxide reacts with the Mg-rich surface in a similar manner to form Mg-Fe bilayer coated MMO; iii) Low temperature calcination promoting the dehydration to obtain Mg-Fe binary oxide-coated MMO (MFO@MMO). We report the suppress of electrolyte decomposition by the MFO coating layers and the improvement of cyclability performed in a full-cell test.
Schematic illustration of Mg-Fe binary oxide coating process onto MgMn2O4.
As our previous reports stated,22 MMO synthesized via alcohol reduction is attributed to a single phase of metastable cubic spinel with the Fd–3m space group (Fig. S2). Note that the 220 diffraction at 31° is too weak and broad to observe. This suggests that Jahn-Teller distortion of Mn3+ is suppressed and a metastable cubic structure is obtained.31 After Mg hydroxide-coating onto MMO, no change in the MMO crystalline phase was observed (Fig. S3), and peaks attributable to the Mg-layered product were observed, which were presumed to be Mg-Mn-layered double hydroxide generated from the reaction between partially dissolved Mn3+ (formed during the reflux process) and Mg ethoxide.32 After the Mg-Fe bilayer coating and the following calcination, the peaks attributable to the layered product disappeared (Fig. S3 and Fig. 2a), indicating that the Fe and Mg reacted to form the Mg-Fe binary oxide phase. Additionally, Rietveld refinements showed that the metastable cubic phase did not change after the coating process (Fig. 2a). Indeed, the selected area electron diffraction (SAED) analysis of MFO@MMO particles (Fig. 2b) revealed three diffraction rings that corresponded to the obtained XRD pattern. The specific surface areas of MMO and MFO@MMO, as well as the primary particle size, were determined using Brunauer-Emmett-Teller (BET) method using N2 gas adsorption and Scherrer’s equation based on the XRD patterns, respectively (Table S1). The larger crystallite size and smaller specific surface area of MFO@MMO compared with those of MMO were probably due to the calcination process, where crystal growth and aggregation occurred. Transmission electron microscopy (TEM) and cross-sectional images (Fig. S5 and Fig. 2c) showed that hundreds of micron-sized secondary particles were formed by the aggregation of 5 nm-sized primary particles. Their lattice fringe of 0.25 nm was attributable to the d value with the 311 plane of the cubic MMO spinel (36.5° in the XRD patterns in Fig. 2a). Furthermore, cross-sectional TEM images showed 2.6 nm-amorphous-like layers deposited on the MMO surface (Fig. 2c). Even though the layer of Mg@MMO was relatively thick as shown in the XRD pattern (Fig. S3), the layer of MFO@MMO was thinner. This was probably because that the coating layer became thinner from Mg@MMO to MFO@MMO by dissolving the Mg-layered product at the reflux treatment. To confirm the amorphous-like layers, the elemental distributions of the MFO@MMO surface and its interior were investigated. Figure S5 shows the results of TEM–energy dispersive spectroscopy (EDS) analysis. All the elements were uniformly distributed on the surface, even with nanometre-order observations. On the other hand, the depth direction analysis using X-ray photoelectron spectroscopy analysis with Ar+ sputtering shows that compared with Mn, Mg and Fe were more abundant near the surface of MFO@MMO than inside the particles (Fig. 2d). According to the above analyses, amorphous-like Mg-Fe binary oxide is successfully coated onto the MMO surface.
(a) XRD pattern of MFO@MMO with fitting curve by Rietveld refinement. (b) SAED pattern of MFO@MMO. (c) Cross-sectional TEM image of MFO@MMO. (d) Elemental XPS depth profile of MFO@MMO.
The MRB cathode performance was evaluated using a full-cell type charge-discharge test, starting from discharge, operated at room temperature. Figure 3a shows the voltage curves of the MMO and MFO@MMO electrodes at a current density of 10 mA g−1. In these curves, no plateaus were observed during the discharge process, probably due to the sluggish Mg-ion kinetics, even though the primary particle size was 5 nm. The discharge capacity of MMO was 100 mAh g−1 in the first cycle, which decreased to approximately half in the second cycle, although the first recharge capacity was 270 mAh g−1. Note that the charging capacity is regulated to the theoretical capacity of MMO (270 mAh g−1) because charging curve shows the long plateau due to the side reaction of electrolyte decomposition. This indicates that the oxidative decomposition of the electrolyte occurred in the 3.5 V-plateau region, which was accelerated by the active material in contact with the electrolyte. In contrast, for MFO@MMO, the first discharge capacity improved to approximately 150 mAh g−1, probably because of the formation of a stable cathode electrolyte interface. Figure 3b and Fig. S6 show the electrochemical impedance spectroscopy (EIS) results of the MMO and MFO@MMO cathodes. MFO@MMO had a lower interface resistance, and Mg-ion could be easily inserted into MFO@MMO to improve the discharge capacity. Notably, the second discharge capacity of MFO@MMO was much higher than that of MMO, demonstrating that the MFO coating layer suppressed the electrolyte decomposition, improving the cyclability (Fig. S7). Furthermore, when compared with single-oxide coatings at second discharge behaviors, that is, Mg-oxide@MMO and Fe-oxide@MMO particles, the rate of decrease in discharge capacity with MFO@MMO is the smallest of the three. Then, MFO coating was the most effective to improve the cyclability among the three coatings (Fig. 3a and Fig. S8). This implies that the key factors of coating are not only reducing the contact between Mn and the electrolyte, but also having Mg-ion conduction pass by using Mg-containing binary oxide.
(a) Voltage curves of MMO and MFO@MMO particles. (b) EIS of MMO and MFO@MMO at pristine state. (c, d) Mn K-edge XANES spectra of (c) MMO and (d) MFO@MMO. (e, f) B 1s XPS spectra of (e) MMO and (f) MFO@MMO.
Figures 3c and 3d show the Mn K-edge X-ray absorption near edge structure (XANES) spectra of MMO and MFO@MMO during the first discharge-charge cycle. In both cathode materials, the Mn K-edge spectra shifted to lower energy positions after discharge, indicating the reduction of Mn. Regarding MMO, however, the XANES spectra did not change after re-charge, indicating that Mn reoxidation reaction was not achieved; Mg-ions were hardly extracted from the rock-salt structure under room-temperature conditions and that the charge current was mainly used for the decomposition of the electrolyte. However, the second discharge capacity of MMO was 50 mAh g−1. The second discharge capacity is derived from non-reduced MMO phase remained after the first discharge. In contrast, MFO@MMO was relatively re-oxidized after charge because the MFO coating prevented the side reaction of electrolyte decomposition. Furthermore, the Fe K-edge spectra did not shift (Fig. S9); thus, Fe is not a redox species. Indeed, these particles, with an interesting redox mechanism, are different from the previously reported Mg(FexMn1−x)2O4.28 Thus, the MFO layer works as just suppressing the electrolyte decomposition.
To investigate the electrolyte decomposition products, X-ray photoelectron spectroscopy measurements were conducted on the cathode electrode. Figures 3e and 3f shows the B 1s spectra of MMO and MFO@MMO, which were expected to follow the decomposition products of Mg[B(HFIP)4]2. For MMO, a peak at 192 eV was observed after discharge and charge, suggesting the decomposed products of the electrolyte formed during both reductive and oxidative conditions. In contrast, the spectra of MFO@MMO hardly showed any peaks, even after re-charging, suggesting that decomposition of the electrolyte was successfully prevented; thus, the MFO coating was effective against not only oxidative decomposition but also reductive decomposition.
In conclusion, we successfully synthesized Mg-Fe binary oxide-coated nanosized MgMn2O4 spinel and demonstrated the cyclability improvement of an MRB full-cell operated at room temperature. Moreover, the coated cathode material was shown to be repeatably charged and discharged at room temperature. These findings provide new foundations for designing cathode materials for MRBs to operate at room temperature.
We thank Shun Ito for his technical support with the TEM measurements.
The data that support the findings of this study are openly available under the terms of the designated Creative Commons License in J-STAGE Data at https://doi.org/10.50892/data.electrochemistry.19743361.
Reona Iimura: Investigation (Lead), Methodology (Equal), Writing – original draft (Lead)
Hiroaki Kobayashi: Conceptualization (Lead), Funding acquisition (Lead), Methodology (Equal), Project administration (Lead), Writing – review & editing (Lead)
Itaru Honma: Supervision (Lead), Writing – review & editing (Supporting)
There are no conflicts to declare.
Advanced Low Carbon Technology Research and Development Program: JPMJAL1301
R. Iimura: ECSJ Student Member
H. Kobayashi and I. Honma: ECSJ Active Members
