Manganese-Based Oxide Cathode Materials for Aqueous Magnesium-Ion Batteries

Abstract

Aqueous magnesium-ion batteries (AMIBs) are promising for next-generation energy storage technologies due to their high safety, low cost, high theoretical energy density, and environmental friendliness. In particular, manganese-based oxides have attracted much attention due to the abundant resources, high theoretical capacity, and environmental friendliness. This paper provides a comprehensive overview of manganese-based oxide cathode materials for AMIBs, including the crystal structure, electrochemical performance, optimization strategies, and electrode reaction mechanisms. Meanwhile, recent research progress of AMIB full cells based on Mn-based oxide cathode materials is summarized. Finally, the challenges and future perspectives of Mn-based oxide cathode materials for AMIBs are discussed. This review will provide a valuable reference and source of inspiration for future research of manganese-based oxide cathode materials for AMIBs.

1. Introduction

The sustainable energy transition is an important path to the development of modern society, and advanced energy storage technology is the key to efficiently utilizing green, renewable energy [1]. Aqueous magnesium-ion batteries (AMIBs) are promising in next-generation energy storage due to multiple advantages: (1) The non-flammable nature of aqueous electrolytes makes AMIBs inherently safe [2,3,4,5]; (2) The high ionic conductivity of aqueous electrolytes bodes well for the rates [6,7]; (3) The abundant reserves of magnesium, low cost of aqueous electrolytes, and simple assembly conditions give AMIBs a significant cost advantage [8,9,10,11]; (4) AMIBs can potentially achieve a high energy density, if a magnesium metal anode with high theoretical capacity (3833 mAh cm⁻³ and 2205 mAh g⁻¹) and low electrode potential (−2.37 vs. SHE) can be used [12,13,14,15]; (5) Both magnesium and aqueous electrolytes are environmentally benign. However, the development of AMIBs faces multiple challenges, and cathode materials constitute a key limiting factor. The electrochemical properties of cathode materials for AMIBs are typically consist of the slow diffusion kinetics of divalent Mg ions, the dissolution of transition metals in aqueous solutions, and the interfacial side reactions [16,17,18]. Therefore, significant research effort is being devoted to developing high-performance cathode materials for AMIBs.

Research on cathode materials for AMIBs has hitherto mainly focused on manganese-based oxides [17,18,19,20], vanadium-based oxides [4,19,20], and Prussian blue analogs [21,22,23]. For example, manganese-based oxides have attracted significant attention due to the abundant resources, low cost, and high theoretical capacity. Boasting rich oxidation states of manganese, such as +2, +3, and +4, there are many types of manganese-based oxides, such as MnO₂ (oxidation state of +4), MgMn₂O₄ (oxidation state of +3), Mn₂O₃ (oxidation state of +3), Mn₃O₄ (oxidation state of +3 and +2), and MnO (oxidation state of +2) [24]. Meanwhile, MnO₂ possesses various polymorphs [6]. The wide variety of species and crystal structures facilitates the regulation of electrochemical properties. However, pristine manganese-based oxides typically exhibit poor cycling stability and rate capability as a result of manganese dissolution, poor electronic conductivity, and sluggish Mg-ion diffusion kinetics [18,25,26]. Therefore, many optimization strategies, such as interlayer regulation [27,28], crystal defect engineering [26,29], heteroatom doping [30,31], and nanocomposite construction [25,32,33], have been proposed to improve the electrochemical properties, and significant progress has been made.

To develop high-performance manganese-based oxide cathode materials in the future, it is imperative to understand their crystal structure, electrode reactions and mechanisms, electrochemical properties, and optimization strategies. In this review, we aim to summarize the progress of Mn-based oxides as cathode materials for AMIBs. The crystal structures and classifications of Mn-based oxide cathode materials are discussed. Recent research on various Mn-based oxides is summarized, including their electrochemical properties, optimization, reaction mechanisms, and application to full cells. Finally, future perspectives of Mn-based oxide cathode materials for AMIBs are discussed.

2. Classification and Crystal Structure of Mn-Based Oxide Cathode Materials

Mn-based oxide cathode materials for AMIBs mainly include MnO₂, MgMn₂O₄, Mn₃O₄, Mn₂O_3, and MnO (Figure 1). Crystalline MnO₂ is formed by connecting [MnO₆] units based on edge-sharing or corner-sharing. Furthermore, different connecting modes produce various polymorphs, such as α, β, γ, δ, τ, and λ phases [6,24]. α-MnO₂ has a [2 × 2] tunnel structure with a size of 4.6 × 4.6 Å, which facilitates the intercalation/deintercalation of Mg ions [31]. For α-MnO₂, K ions are typically required as pillars in the channels to stabilize the tunnel structure. Although β-MnO₂ is the most stable phase thermodynamically, it exhibits the narrowest [1 × 1] (2.3 × 2.3 Å) tunnel structure formed by corner sharing of [MnO₆] units, being unfavorable to the intercalation/deintercalation of Mg ions [34]. γ-MnO₂, the most common polymorph of manganese dioxide in the battery industry, has a mixed tunnel structure with [1 × 1] (2.3 × 2.3 Å) and [1 × 2] (2.3 × 4.6 Å) unit cells [35]. The crystal structure of ε-MnO₂ is very similar to that of γ-MnO₂, both of which are mixed tunnel structures, but ε-MnO₂ has more structural defects [36,37]. Todorokite-MnO₂ (τ-MnO₂) has a [3 × 3] tunnel structure with the largest channel size (7 × 7 Å) [38]. However, as with α-MnO₂, τ-MnO₂ tunnels often contain a large number of impurity cations and water molecules to maintain structural stability. Unlike the aforementioned polymorphs with a tunnel structure, δ-MnO₂ has a layered structure with an interlayer spacing of about 7 Å [27]. The large interlayer spacing and interlayer water of δ-MnO₂ facilitate the diffusion and storage of Mg ions, thus δ-MnO₂ has attracted a lot of attention in the field of AMIBs. λ-MnO₂ is usually obtained from the spinel LiMn₂O₄ after removing lithium ions and inheriting the cubic spinel structure [39]. MgMn₂O₄ and Mn₃O₄ are manganese-based spinel oxides, often having distorted structures, on account of the Jahn–Teller effect of Mn³⁺ ions [40,41]. Mn₂O₃ primarily displays a crystal structure of an orthorhombic bixbyite at room temperature [42], and MnO exhibits a typical face-centered cubic structure [43]. In both structures, the Mg migration paths are blocked, rendering the materials unfavorable for Mg-ion insertion/extraction. In summary, based on the crystal structure, Mn-based oxide cathode materials can be classified into the following types (Figure 1): (1) Tunnel-type, such as α-MnO₂ ([2 × 2] tunnel), β-MnO₂ ([1 × 1] tunnel), γ-MnO₂ ([1 × 2] tunnel), and τ-MnO₂ ([3 × 3] tunnel); (2) Layer-type, such as δ-MnO₂; (3) Spinel-type, such as λ-MnO₂, MgMn₂O₄, and Mn₃O₄; (4) Other-type, such as Mn₂O₃ and MnO.

Figure 1.

Classification and crystal structure of Mn-based oxide cathode materials.

3. Recent Development of Mn-Based Oxide Cathode Materials for AMIBs

3.1. Tunnel-Type Mn-Based Oxides

3.1.1. α-MnO₂

As early as 2009, Xu et al. have investigated the electrochemical characteristics of α-MnO₂ in an electrolyte containing Mg-ion (0.1 M Mg(NO₃)₂), observing ideal capacitive behavior with good cycling stability (5000 cycles) [44]. Sun et al. have reported that the capacitance of α-MnO₂ in the 0.5 M MgSO₄ electrolyte is lower than that in the 0.5 M Mg(NO₃)₂ electrolyte, but higher than that in the 0.25 M Na₂SO₄ or Al₂(SO₄)₃ electrolyte [45]. Meanwhile, α-MnO₂ with lower crystallinity exhibits better electrochemical properties. Zhang et al. have studied Mg-OMS-2 (Mg-α-MnO₂) in different aqueous electrolytes (MgCl₂, Mg(NO₃)₂, and MgSO₄) and found better electrochemical properties in the Mg(NO₃)₂ electrolyte [46]. These results indicate that the electrolyte composition affects the electrochemical properties of α-MnO₂.

To improve rate capability and cycling stability, some α-MnO₂-based nanocomposites are used as cathode materials in AMIBs. For example, Zhang et al. have fabricated the Mg-OMS-2/graphene composite with better rate capability and cycling than Mg-OMS-2 [46]. In a 0.5 M Mg(NO₃)₂ electrolyte, the Mg-OMS-2/graphene composite exhibits a high capacity of 232.4 mAh g⁻¹ at 20 mA g⁻¹ and a capacity retention of 93.0% after 300 cycles at 100 mA g⁻¹. Moreover, the transformation from the tetragonal phase to the cubic and hexagonal phases after Mg-ion intercalation is observed. Jia et al. have prepared vertically aligned α-MnO₂ nanosheets on carbon nanotubes (α-MnO₂/CNT) as cathode materials for AMIBs [47]. In the 1 M MgSO₄ electrolyte, α-MnO₂/CNT exhibits better capacity (144.6 mAh g⁻¹) than α-MnO₂ (87.5 mAh g⁻¹) at 500 mA g⁻¹, and even at 10,000 mA g⁻¹, α-MnO₂/CNT shows a capacity of about 60 mAh g⁻¹ and a capacity retention of 85% after 1000 cycles.

Doping has been utilized to ameliorate the electrochemical properties of α-MnO₂. Ding et al. have reported that partial substitution of Al for Mn in α-MnO₂ (Figure 2a) creates more oxygen vacancies (Figure 2b) and regulates the diffusion of Mg ions (Figure 2c) [30]. Al shrinks the lattice due to the shorter length of the Al-O bond compared to the Mn-O bond, thereby improving the stability of α-MnO₂ during ion insertion/extraction. As a result, Al-doped α-MnO₂ exhibits a high capacity of 197.02 mAh g⁻¹ at 100 mA g⁻¹ (Figure 2d), good rate capability with a capacity of 90.5 mAh g⁻¹ at 1000 mA g⁻¹, and stable cycling with 82% retention after 2500 cycles. Wang et al. have found that Fe doping (15 mol%) enlarges the lattice of α-MnO₂ and alleviates Jahn–Teller distortions, consequently improving Mg-ion diffusion kinetics and structural stability [31]. After Fe doping (15 mol%), the capacity of α-MnO₂ at 2000 mA g⁻¹ improves from 47.2 mAh g⁻¹ to 99.6 mAh g⁻¹, and the capacity retention after 1000 cycles increases from 44.3% to 85.2%. The reaction of α-MnO₂ stems from the co-intercalation of Mg²⁺/H⁺. In addition, V-doped and Nb-doped α-MnO₂ deliver improved rate performance in AMIBs, although their cycling stability needs further optimization [48]. Besides doping of Mn sites in α-MnO₂, doping in tunnel sites (K sites) has been investigated. Zhao et al. have studied Mg-intercalation α-MnO₂ with Mg ions in tunnel sites as cathode materials for AMIBs [49]. Compared to pristine α-MnO₂, Mg-intercalation α-MnO₂ exhibits an enhanced capacity of 419.8 mAh g⁻¹ at 100 mA g⁻¹ and retention of 111.6 mAh g⁻¹ even at 4000 mA g⁻¹. The in situ Raman spectra indicate the reversible phase transformation between α-MnO₂ and δ-MnO₂ during the charging/discharging process. However, a more detailed study of the electrode reaction mechanism of α-MnO₂ is still needed. Recently, Gu et al. used F doping to enhance the magnesium storage capacity of α-MnO₂ and prepared fluorine-regulated MnO_x/KMnF₃ heterostructures (Figure 2e,f) [25]. The high-strength Mn-F bonds formed in the MnO₂ lattice locally enhance the structural stability by strengthening the tunnel structure and suppressing structural collapse during cycling. As a result, the fluorine-regulated MnO_x/KMnF₃ heterostructure achieves good cycling stability with a capacity retention of 89.6% after 1800 cycles at 1000 mA g⁻¹.

Figure 2.

(a) Crystal structure of Al-doped α-MnO₂, (b) electron paramagnetic resonance spectra, (c) calculated Mg-ion diffusion barriers, and (d) charging/discharging curves at 100 mA g⁻¹ of α-MnO₂ and Al-doped α-MnO₂. Reproduced from [30], with permission under CC BY 4.0, (e) structural diagram and (f) TEM image of fluorine-regulated MnO_x/KMnF₃ heterostructure, reproduced from [25], with permission under CC BY 4.0.

3.1.2. β-MnO₂

Zhang et al. have evaluated the electrochemical performance of β-MnO₂ (Mg-OMS-7) and investigated the size effect [34]. Compared to micro-sized β-MnO₂, nano-sized β-MnO₂ (1.6-Mg-OMS-7) has better electrochemical properties. Furthermore, 1.6-Mg-OMS-7 displays a high capacity of over 280 mAh g⁻¹ at 10 mA g⁻¹ in a 0.2 M Mg(NO₃)₂ electrolyte and a capacity retention of 94.1% after 200 cycles at 100 mA g⁻¹ in the 1 M Mg(NO₃)₂ electrolyte. The electrochemical properties of 1.6-Mg-OMS-7 in Mg(NO₃)₂ and MgCl₂ electrolytes are better than those in the MgSO₄ electrolyte. Li et al. have improved the electrochemical properties of β-MnO₂ in the MgSO₄ electrolyte by regulating the nanostructure and Mn valence [50]. The nanostructure with a larger specific surface area exhibits a higher capacitive capacity, and a suitable Mn³⁺ concentration promotes the Faradic reaction. Therefore, β-MnO₂ with a larger specific surface area and Mn³⁺ concentration of 16% exhibits a high capacity of 260 mAh g⁻¹ at 200 mA g⁻¹ in the 1 M MgSO₄ electrolytes, and a capacity of 68 mAh g⁻¹ at 10,000 mA g⁻¹. However, cycling stability worsens since the capacity retention after 1000 cycles at 2000 mA g⁻¹ is only 65%.

3.1.3. τ-MnO₂

Zhang et al. have studied τ-MnO₂ (Mg-OMS-1) nanowires and nanosheets as cathode materials for AMIBs in different electrolytes [38,51]. In the 0.2 M MgCl₂ electrolyte, τ-MnO₂ nanosheets show a reversible capacity of 300 mAh g⁻¹ at 10 mA g⁻¹. However, the capacity of τ-MnO₂ nanosheets decreases to 115 mAh g⁻¹ when the current density increases to 100 mA g⁻¹. To improve the properties of τ-MnO₂, τ-MnO₂/graphene composites [49] and Ni-doped τ-MnO₂ [50] have been fabricated, but the improvements are limited.

3.1.4. ε-MnO₂

Liu et al. have synthesized binder-free ε-MnO₂ nanoflakes on carbon cloth by potentiostatic electrodeposition [52]. Compared to commercial electrolytic MnO₂, binder-free ε-MnO₂ shows better capacity, rate performance, and cycling stability. In the 1 M MgCl₂ electrolyte, a capacity of 259.3 mA h g⁻¹ is achieved at 500 mA g⁻¹. Moreover, reversible transformation between ε-MnO₂ and MgMn₂O₄ occurs during discharging/charging. Furthermore, nanostructure regulation [53] and Fe doping [37] have been demonstrated to improve the electrochemical properties of ε-MnO₂. However, compared to α-MnO₂, research on other tunnel-type MnO₂ for AMIBs is relatively limited.

3.2. Layer-Type Mn-Based Oxides

Nam et al. have prepared Mg-δ-MnO₂, also called Mg-birnessite MnO₂ (Mg-B), from Mn₃O₄ using aqueous electrochemical transformation and evaluated its magnesium storage capacity [54]. In a 0.5 M Mg(ClO₄)₂ aqueous electrolyte, Mg-B displays a high capacity of 231.1 mAh g⁻¹ at 100 mA g⁻¹ and high rate with a capacity of 88.6 mAh g⁻¹ at 2000 mA g⁻¹. After 10,000 cycles at 2000 mA g⁻¹, the capacity retention of Mg-B is 62.5%. Water in the electrolyte plays an important role in improving magnesium storage in Mg-B. In the Mg(ClO₄)₂/acetonitrile (ACN) electrolyte without water, Mg-B exhibits a low capacity of 56.8 mAh g⁻¹ at 100 mA g⁻¹. Sun et al. have demonstrated that magnesium storage in Mg-δ-MnO₂ is based on the intercalation reaction in aqueous electrolyte and conversion reaction in a non-aqueous electrolyte (Figure 3a) [55]. In a 0.5 M Mg(ClO₄)₂ aqueous electrolyte, the interlayer spacing of Mg-δ-MnO₂/CC contracts from 7 Å to 4.86 Å after Mg ion intercalation and expands to 10 Å after Mg ion extraction. Meanwhile, a glide of the MnO₂ planes occurs during the intercalation/extraction of Mg ions. However, structural assessment of the discharged product (Mg-δ-MnO₂ after Mg-ion intercalation) is inadequate. The interlayer spacing change in δ-MnO₂ during intercalation/de-intercalation of Mg ions reported in many subsequent studies is smaller than that reported by Sun et al. [29,56,57,58,59]. The intercalation species in δ-MnO₂ as cathode materials in AMIBs is also an unanswered question worth exploring. Based on the electrochemical quartz crystal microbalance (EQCM), Wang et al. have proposed that some water molecules are intercalated into δ-MnO₂ along with magnesium ions (Figure 3b) [27]. Liu et al. have verified the consequent H⁺ and Mg²⁺ insertion reaction in δ-MnO₂ by comparing the electrochemical properties of δ-MnO₂ in different aqueous electrolytes and Mg concentrations at different charge/discharge states (Figure 3c) [60]. During the discharge process, the H⁺ insertion occurs first at a slightly higher potential, followed by magnesium ion insertion. The intercalation species tend to be hydrated H⁺ and hydrated Mg²⁺, but more evidence is needed. Moreover, the effect of electrolyte composition on the Mg²⁺/H⁺ co-insertion is also unclear.

Figure 3.

(a) Schematic of the magnesium storage reaction mechanism of δ-MnO₂ in aqueous and non-aqueous electrolytes. Reproduced from [55] with permission. (b) Schematics of Mg²⁺/H₂O co-insertion mechanism of δ-MnO₂ in aqueous electrolyte. Adapted from [27], with substantial modifications. (c) Schematic of Mg²⁺/H⁺ co-insertion mechanism of δ-MnO₂ in aqueous electrolyte. Adapted from [60], with substantial modifications.

Interlayer regulation is an efficient strategy to improve the electrochemical performance of layered materials, and this is no exception for δ-MnO₂ in AMIBs. δ-MnO₂ with different interlayer cations (Li⁺, Na⁺, K⁺, and Mg²⁺) has been investigated as cathode materials for AMIBs [28,54,61,62]. Huang et al. have compared the electrochemical properties of δ-MnO₂ with different interlayer cations (Li⁺, Na⁺, and K⁺) synthesized by the same electrodeposition method [28]. K⁺-intercalated δ-MnO₂ exhibits a higher capacity and rate performance. Xiong et al. have fabricated tetramethyl ammonium (TMA⁺) ions-intercalated δ-MnO₂ (TMA-MnO₂) with an expanded interlayer spacing of 0.93 nm (Figure 4a) and investigated the metal-ion (Li⁺, Na⁺, K⁺, and Mg²⁺) storage behavior [63]. Compared to pristine δ-MnO₂ (K-MnO₂) with an interlayer spacing of 0.69 nm and protonated δ-MnO₂ (H-MnO₂) with an interlayer spacing of 0.72 nm, TMA-MnO₂ has better electrochemical characteristics. Compared with Na⁺ and K⁺ storage, TMA-MnO₂ shows a higher capacity for Mg²⁺ storage at a low current density, albeit with poorer rate performance. The magnesium storage capacity retention of TMA-MnO₂ is as high as 80% after 5000 cycles at 5000 mA g⁻¹. Moreover, the water content interlayer plays an important role in the magnesium storage capacity of δ-MnO₂. Sun et al. have studied two δ-MnO₂ materials with different interlayer water contents [56]. Compared to Mg_0.61MnO₂·0.42H₂O (MMO-2), Mg_0.58MnO₂·0.56H₂O (MMO-1), with a higher interlayer water content, has a larger interlayer space (∼9.70 Å), which is favorable to the diffusion of Mg ions. Consequently, MMO-1 shows a better rate performance. In a 0.5 M MgCl₂ electrolyte, MMO-1 exhibits a capacity of 98.3 mAh g⁻¹ at 1000 mA g⁻¹, which is higher than that of MMO-2 (74.4 mAh g⁻¹).

Figure 4.

(a) Schematic of the synthetic process and crystal structure of TMA-MnO₂. Reproduced from [63] with permission. (b) Mn 3d-band center values of δ-MnO₂, W-doped δ-MnO₂ (W-MnO₂), and O_d-WMO, (c) shift diagram of the Mn 3d-band center, (d) Mg-ion adsorption energy, (e) Mg-ion diffusion barriers and (f) energy barrier for Mn vacancy formation in δ-MnO₂ and O_d-WMO. Reproduced from [29], with permission under CC BY 4.0.

Nanostructure regulation is an effective approach to improve the properties of materials [64], which has also been used for δ-MnO₂ in AMIBs. Shi et al. have fabricated hierarchical Mg-birnessite nanowall arrays with enriched (010) planes (ECMB) in situ by electrochemical conversion from Mn₃O₄ nanowall arrays [58]. Compared to Mg-birnessite nanosheets prepared by a traditional hydrothermal synthesis, ECMB has a smaller nanosheet size and enriched active (010) planes, facilitating fast Mg²⁺ intercalation/deintercalation. ECMB shows significant improvements in capacity and rate performance. In the 0.5 M Mg(ClO₄)₂ aqueous electrolyte, ECMB exhibits a high capacity of 255.1 mAh g⁻¹ at 200 mA g⁻¹, which remains at 55.4 mAh g⁻¹ at 8000 mA g⁻¹. Moreover, EMCB exhibits significantly improved cycling stability, with a capacity retention of 73.6% after 3000 cycles at 2000 mA g⁻¹. Some δ-MnO₂-based nanocomposites have been synthesized as cathode materials for AMIBs, for example, δ-MnO₂@carbon molecular sieves (CMS) composite [65], δ-MnO₂/multiwalled carbon nanotubes (MWCNTs) composite [57,66], δ-MnO₂/reduced graphene oxide (rGO) composite [67], and K-δ-MnO₂/mulberry-like carbon (K-MnO₂/HMC) composite [68]. Compared to the corresponding pristine δ-MnO₂, these composites have better electrochemical properties. For example, δ-MnO₂@CMS shows a high capacity of ~270 mAh g⁻¹ at 50 mA g⁻¹ in a 0.5 M Mg(NO₃)₂ electrolyte, which remains at 140 mAh g⁻¹ at 1000 mA g⁻¹ [65]. However, the cycling stability needs improvement.

Recently, crystal defect engineering has been adopted to optimize the electrochemical properties of δ-MnO₂ in AMIBs. Ren et al. have synthesized oxygen-deficient W-doped δ-MnO₂ (O_d-WMO) as cathode materials [29]. The combination of W doping and oxygen defects leads to the upward shift in the d-band center of Mn (Figure 4b,c), the narrowing of the bandgap, and the strengthening of Mn-O bonds (Figure 4f), thereby improving the adsorption capacity for Mg²⁺ (Figure 4d), charge transfer, ion transport kinetics (Figure 4e), and structural durability (Figure 4f). O_d-WMO shows a high capacity (185.2 mAh g⁻¹ at 100 mA g⁻¹), enhanced rate performance (101.2 mAh g⁻¹ at 2000 mA g⁻¹), and good cycling stability (capacity retention of 92% after 1500 cycles). Suitable Fe doping improves the capacity and rate capability of δ-MnO₂ [31]. δ-MnO₂ with 5 mol% Fe exhibits a high capacity of 263.5 mAh g⁻¹ at 100 mA g⁻¹, which retains 145.8 mAh g⁻¹ at 2000 mA g⁻¹. However, a larger Fe concentration leads to a phase transformation from the δ- to α-phase.

Anionic doping is another effective technique. Zhang et al. have synthesized F-doped δ-MnO₂ (MnO_2−xF_y) from K_1.33Mn₈O₁₆ by thermal treatment with NH₄F using it as a cathode material for AMIBs [69]. During the thermal process, many oxygen vacancies are introduced into MnO_2−xF_y, and the morphology changes from nanorods to a three-dimensional hierarchical structure. Compared to K_1.33Mn₈O₁₆, MnO_2−xF_y exhibits improved specific capacitance and rate performance. The combination of crystal defect engineering and nanocomposite construction further improves the electrochemical properties of δ-MnO₂. Liu et al. have fabricated V-doped and O vacancy-rich δ-MnO₂/CNTs/rGO (V-O_vac-MnO₂/CG) composites [59]. V doping promotes the generation of O vacancies to lower the electrostatic barrier for ion transport and enhance the electronic conductivity. Meanwhile, CNTs and rGO improve the electrical conductivity. Therefore, the V-O_vac-MnO₂/CG composite achieves a high discharge capacity of 398 mAh g⁻¹ at 100 mA g⁻¹ and good cycling stability with a capacity retention of 81% after 500 cycles at 200 mA g⁻¹. Even at 2000 mA g⁻¹, the V-O_vac-MnO₂/CG composite exhibits a reversible capacity of 143 mAh g⁻¹. Layered A_xMnO₂ (A = Na or K) without interlayer water has been investigated, but the electrochemical properties are poorer than those of δ-MnO₂ [70,71], demonstrating that the water interlayer plays an important role in the superior electrochemical performance of δ-MnO₂.

3.3. Spinel-Type Mn-Based Oxides

3.3.1. λ-MnO₂

Yuan et al. have synthesized λ-MnO₂ nanoparticles from spinel LiMn₂O₄ by acid leaching to remove Li and determined the electrochemical properties in aqueous electrolytes containing different Mg salts (MgSO₄, MgCl₂, and Mg(NO₃)₂) [39]. λ-MnO₂ has better properties in the 1 M MgCl₂ electrolyte, as manifested by a high discharge capacity of ~480 mAh g⁻¹ at 13.6 mA g⁻¹. When cycling at 136 mA g⁻¹, the capacity of λ-MnO₂ remains at 155.6 mAh g⁻¹ after 300 cycles. Tekin et al. have reported that λ-MnO₂ delivers better cycling performance in the 1 M MgSO₄ electrolyte than 1 M MgCl₂ and 1 M Mg(NO₃)₂ electrolytes [72]. λ-MnO₂ is detached from the current collectors in MgCl₂ and Mg(NO₃)₂ electrolytes but not in MgSO₄. In addition, λ-MnO₂ can be prepared from MgMn₂O₄ using an acid treatment, but MgMn₂O₄ can be directly used in the cathode of AMIBs [40]. Therefore, most λ-MnO₂ used for AMIBs is still prepared from LiMn₂O₄. Chen et al. have investigated the effect of particle size on magnesium storage in λ-MnO₂ [73]. Compared to micron-sized λ-MnO₂, nano-sized λ-MnO₂ has a higher reversible capacity, especially the capacity contributed from Faradaic process. The phase transition from the cubic to the tetragonal spinel structure during Mg-ion intercalation is studied by X-ray diffraction (XRD) and atomic-resolution scanning transmission electron microscopy (STEM) imaging. Zhang et al. have synthesized λ-MnO₂/MWCNTs [74]. In a 0.5 M MgSO₄ electrolyte, the λ-MnO₂/MWCNTs composite displays an initial discharge capacity of 124.1 mAh g⁻¹ at 1000 mA g⁻¹ and capacity retention of 86.2% after 1000 cycles.

3.3.2. MgMn₂O₄

Sinha et al. have synthesized MgMn₂O₄ from LiMn₂O₄ by electrochemical conversion and determined the electrochemical properties in the 1 M Mg(NO₃)₂ aqueous electrolyte [75]. However, the obtained MgMn₂O₄ exhibits a low discharge capacity of only 42 mAh g⁻¹. MgMn₂O₄ can be synthesized by the sol–gel method, hydrothermal method, and other techniques. MgMn₂O₄ nanoparticles directly synthesized by Cabello et al. by the sol–gel method show a discharge capacity of 150 mAh g⁻¹ in a 3 M Mg(NO₃)₂ electrolyte [40]. Unfortunately, the capacity decays below 100 mAh g⁻¹ after 20 cycles, indicating poor cycling stability.

To improve the electrochemical properties of MgMn₂O₄ in AMIBs, several MgMn₂O₄-based nanocomposites have been prepared. For example, Liu et al. have fabricated MgMn₂O₄ nanoparticles/rGO (MgMn₂O₄/rGO) composite [33]. Compared to MgMn₂O₄ nanoparticles, MgMn₂O₄/rGO exhibits a higher capacity and rate. In a 0.5 M MgCl₂ electrolyte, the specific capacity of MgMn₂O₄/rGO is 211.8 mAh g⁻¹ at 50 mA g⁻¹ and remains at 140.1 mAh g⁻¹ at 1000 mA g⁻¹. Moreover, a capacity of 119.5 mAh g⁻¹ is observed after 1000 cycles at 1000 mA g⁻¹. Javadian et al. have synthesized MgMn₂O₄ hollow spheres/rGO composite and evaluated its electrochemical properties in a 1 M MgSO₄ mixed electrolyte of water/acetonitrile (1:10 in volume) [76]. The composite exhibits a high capacity of 305 mAh g⁻¹ at 300 mA g⁻¹ with a capacity retention of ~95% after 100 cycles. Zhang et al. have synthesized the MgMn₂O₄/MWCNTs composites from LiMn₂O₄/MWCNTs by electrochemical conversion [77,78]. The MgMn₂O₄/MWCNTs composites obtained from commercial LiMn₂O₄ exhibit a high capacity of over 400 mAh g⁻¹ at 50 mA g⁻¹, good cycling stability with a retention rate of 73.3% after 1000 cycles at 1000 mA g⁻¹ [77]. However, the Coulombic efficiency of MgMn₂O₄/MWCNTs composites is low, especially at low current density, indicating that irreversible reaction or side reaction may occur during the discharge process. Recently, MgMn₂O₄/polyaniline (PANI) has been prepared as a cathode material for AMIBs [79]. The optimized MgMn₂O₄/PANI composite delivers a high capacity of 287 mAh g⁻¹ at 250 mA g⁻¹ and 73% capacity retention after 250 cycles at 500 mA g⁻¹.

Cationic substitution is an effective route to enhance the structural stability and electrochemical kinetics of LiMn₂O₄ and it is also excepted to improve the electrochemical properties of MgMn₂O₄ in AMIBs. In this case, Zhang et al. have synthesized MgFe_xMn_2−xO₄ (x = 0.67, 1, 1.33, 1.6) nanoparticles as cathode materials for AMIBs [80]. MgFe_1.33Mn_0.67O₄ (MFM-2) has better electrochemical characteristics, including higher capacity, enhanced rate capability, and cycling stability. MFM-2 exhibits a high capacity of 88.3 mAh g⁻¹ after 1000 cycles at 1000 mA g⁻¹ without obvious attenuation. Compared to MgMn₂O₄, MgMn_1.5Ni_0.5O₄ has better cycling stability [81]. After 500 cycles at 1000 mA g⁻¹, the capacity retention of MgMn_1.5Ni_0.5O₄ is 92.9%, but that of MgMn₂O₄ is only about 30%. The properties of MgMn_1.5Ni_0.5O₄ can be further improved using a polypyrrole (PPy) coating [82], which improves not only the electrical conductivity but also the structural stability by alleviating expansion/contraction and Mn dissolution. Consequently, the optimized MgMn_1.5Ni_0.5O₄@PPy composite exhibits a high capacity of 133.0 mAh g⁻¹ with 90.8% capacity retention after 2000 cycles at 1000 mA g⁻¹. Zhao et al. have incorporated oxygen vacancies into MgMn₂O₄ (Figure 5a) [26]. DFT calculations indicate that oxygen vacancies reduce the Mg²⁺ diffusion barrier (Figure 5b), narrow the bandgap, and generate favorable adsorption sites for Mg²⁺ by inducing charge redistribution (Figure 5c). Therefore, the optimized oxygen-deficient MgMn₂O₄ exhibits an improved capacity of 230.8 mAh g⁻¹ at 100 mA g⁻¹ and retains a capacity of 84 mAh g⁻¹ at 2000 mA g⁻¹.

Figure 5.

(a) Synthesis of oxygen-deficient MgMn₂O₄ (defect-MMO), (b) Mg-ion diffusion energy barriers, and (c) structure and charge density difference distribution diagrams of MgMn₂O₄ (MMO) and defect-MMO. Reproduced from [26], with permission under CC BY 4.0. (d) Schematic of the structural evolution from spinel MgMn₂O₄ to a layered structure during the electrochemical process. Reproduced from [83], with permission under CC BY 4.0.

The electrochemical reaction mechanism of MgMn₂O₄ in AMIBs remains controversial, with three mechanisms proposed. (1) Tetragonal spinel MgMn₂O₄ transforms to cubic spinel λ-MnO₂ after Mg-ion extraction during charging, and the transformation is reversible [40,77,78]. (2) Tetragonal spinel MgMn₂O₄ transforms irreversibly to the birnessite-type layered phase during early cycling (Figure 5d) [83,84]. (3) Tetragonal spinel MgMn₂O₄ irreversibly transforms into MnO₂ (α- or γ-phase) after Mg-ion extraction during charging, and MnO₂ undergoes a combined reaction involving intercalation from MnO₂ to MgMnO₂ and conversion from MnO₂ to MgO and MnO during discharging [76,79]. The evidence supporting the third mechanism is limited, consisting only of poor-quality ex situ XRD data (Table 1), therefore it may be a flawed conclusion. The complete spinel to layer transformation (second mechanism) typically requires dozens of cycles [84], but most mechanistic studies only focus on the first or first two cycles, thus failing to observe the spinel to layer transformation. Before transforming into a layered structure, MgMn₂O₄ may undergo reversible magnesium ion extraction (first mechanism). These may be the reasons why different studies report different results. To determine the more precise electrode reaction mechanism of MgMn₂O₄ further research is needed.

Compared to the tetragonal spinel MgMn₂O₄, the cubic inverse-spinel Mg₂MnO₄ containing a higher concentration of Mn⁴⁺ and is expected to expedite ion diffusion. Sun et al. have studied Mg₂MnO₄ [85]. In the 1 M MgSO₄ + 0.1 M MnSO₄ electrolyte, the Mg₂MnO₄ exhibits a capacity of 116.1 mAh g⁻¹ at 1C (160 mA g⁻¹), which remains 60.8 mAh g⁻¹ even at 20 °C (3200 mA g⁻¹), demonstrating good rate performance. However, the theoretical capacity of Mg₂MnO₄ is only 160 mAh g⁻¹ (based on the Mn⁴⁺/Mn³⁺ redox couple), which is much lower than that of other manganese oxides and is unfavorable for achieving higher energy density.

Table 1.

The proposed reaction mechanisms for different Mn-based oxides.

Mn-Based Oxides	Proposed Reaction Mechanisms	Evidences	Ref.
α-MnO2	Mg2+ insertion	Ex situ XRD, XPS, HRTEM	[25,47]
α-MnO2	Mg2+/H+ co-insertion	Ex situ XRD, XPS, HRTEM, DFT calculations	[31]
α-MnO2	Mg2+ insertion with partial transformation to new cubic phase and hexagonal phase	Ex situ XRD	[46]
α-MnO2	Mg2+ insertion with reversible transformation to layered birnessite	In situ Raman	[49]
β-MnO2	Mg2+ insertion	Ex situ XRD, XPS	[34,50]
τ-MnO2	Mg2+ insertion with partial transformation to layered birnessite	Ex situ XPS, XRD	[51]
τ-MnO2	Mg2+ insertion	Ex situ XPS, XRD	[38,86]
ε-MnO2	Mg2+ insertion with reversible transformation to MgMn2O4	Ex situ XPS, XRD and EDS	[52]
δ-MnO2	Mg2+ insertion	In situ Raman, Ex situ XRD, XPS, HRTEM, ICP, Raman, STEM, EDS, EELS	[29,56,57,58,59,63,65,66,87,88]
δ-MnO2	Mg2+/H2O co-insertion	EQCM	[27]
δ-MnO2	Mg2+/H+ co-insertion	GITT, ex situ XPS, EDS	[60]
δ-MnO2	Mg2+ insertion with reversible transformation to unknown new phase	Ex situ XRD	[55]
λ-MnO2	Mg2+ insertion with cubic to tetragonal phase transformation	Ex situ XRD, STEM, XPS	[40,73,74]
MgMn2O4	Mg2+ extraction	Ex situ XRD, XPS	[26,80]
MgMn2O4	Mg2+ extraction with tetragonal to cubic phase transformation	Ex situ XRD	[77,78]
MgMn2O4	Irreversible transformation to layered birnessite during the initial cycling	In situ Raman, ex situ XRD, HRTEM, TG-DTA, FT-IR, SEM, XPS	[80,84]
MgMn2O4	Irreversible transformation to MnO2 (α- or γ-phase) during Mg-ion extraction and a combined reaction involving insertion to MgMnO2 and conversion to MgO and MnO during discharging.	Ex situ XRD	[76,79]
Mn3O4	Irreversible transformation to layered birnessite during the initial cycling	Ex situ XRD, HRTEM	[41,86]
Mn3O4	Mg2+ insertion	Ex situ XRD, XPS	[88]
Mn3O4	Mg2+ insertion with Mn(II) dissolution induced tetrahedral vacancies as insertion sites	Ex situ XRD, HRTEM, XANES, XPS, Raman	[18]
Mn2O3	Mg2+ insertion	Ex situ XRD, XPS	[32,89]
MnO	Mg2+ insertion	Ex situ XRD, XPS	[90]

3.3.3. Mn₃O₄

Besides Mg-Mn spinel oxides, spinel Mn₃O₄ has been investigated as the cathode material in AMIBs. In an aqueous electrolyte containing Mg²⁺, Mn₃O₄ transforms to birnessite-MnO₂ (δ-MnO₂), accompanied by Mn²⁺ dissolution during the electrochemical process [89]. Therefore, some δ-MnO₂ materials are obtained from Mn₃O₄ by the electrochemical conversion method discussed in Section 3.2 [27,54,58]. Pan et al. have discovered that Mn₃O₄ nanoparticles electrochemically transform into the amorphous-crystalline Mn₃O₄ composite (Mn₃O₄-A) when cycling in the NaSO₄ aqueous electrolyte [91]. Moreover, as a cathode material in AMIBs with the MgSO₄ aqueous electrolyte, Mn₃O₄-A (crystalline part) retains its tetragonal spinel structure during cycling rather than transforming into a layered structure. In addition, although Mn₃O₄ transforms to δ-MnO₂ during cycling, its initial structural characteristics still have a significant impact on the electrochemical performance. Ding et al. have fabricated Mn₃O₄ with cation-anion dual defects (CADDs-Mn₃O₄) and better properties than commercial Mn₃O₄ [41]. CADDs-Mn₃O₄ shows a high capacity of 320.93 mAh g⁻¹ at 100 mA g⁻¹ in the 0.5 M MgSO₄ electrolyte. Moreover, CADDs-Mn₃O₄ exhibits a capacity of ~200 mAh g⁻¹ even after 2000 cycles at 1000 mA g⁻¹. Li et al. have found that introducing a small amount of Cs in the hydrothermal process can induce the phase transition from Mn₂O₃ to Mn₃O₄, thereby obtaining Cs-doped Mn₃O₄ [90]. In a 0.5 M MgSO₄ aqueous electrolyte, Cs-doped Mn₃O₄ has better properties than Mn₂O₃. However, it is uncertain whether the electrochemical improvement stems from Cs doping or a phase transition.

Pan et al. have investigated Mn dissolution and structure transformation in Mn₃O₄ during cycling [18]. Dissolution of Mn(II) during the electrochemical process causes the morphological transformation (Figure 6b–d) and the formation of tetrahedral vacancies (Figure 6e). The vacancies not only expedite Mg-ion diffusion but also provide storage sites. During initial cycling, the capacity of Mn₃O₄ increases with cycles (Figure 6a). Notably, the spinel framework remains the same, rather than transforming into a layered structure during 4000 CV cycles at 25 mV s⁻¹. Furthermore, the collapse of the spinel framework may be responsible for the fading capacity during subsequent cycling. Adding MnSO₄ to the electrolyte can stabilize the spinel framework by inhibiting the dissolution of Mn(III) via the common-ion effect (Figure 6f). Therefore, the capacity retention of Mn₃O₄ after 2000 cycles at 1000 mA g⁻¹ increases from 42% to 94.9% after adding 0.1 M MnSO₄ into the electrolyte.

Figure 6.

(a) Cycling performance of Mn₃O₄, (b–d) TEM images and low-pass filtered HR-TEM images of Mn₃O₄ after different CV cycles, (e) schematic diagram of the crystal structure evolution of Mn₃O₄, (f) cycling performance of Mn₃O₄ in electrolytes with and without the MnSO₄ additive. Reproduced from [18], with permission under CC BY 4.0.

3.4. Other Mn-Based Oxides

Other Mn-based oxides, such as Mn₂O₃ and MnO, have been employed as cathode materials for AMIBs. In order to promote Mg-ion transport kinetics, Li et al. have constructed Mn₂O₃@TiO₂@MXene composites with dual heterogeneous interfaces, thereby creating built-in electric fields and reducing the diffusion energy barrier [92]. The Mn₂O₃@TiO₂@MXene composites achieve an improved capacity of 241.5 mAh g⁻¹ in a 0.5 M MgSO₄ aqueous electrolyte compared to Mn₂O₃. After 1350 cycles at 1000 mA g⁻¹, the Mn₂O₃@TiO₂@MXene composites still exhibit a capacity retention of 85.7%. An organic-inorganic coupling strategy has been proposed to improve magnesium-ion storage in Mn₂O₃ by fabricating organic-inorganic ethylenediamine-Mn₂O₃ (EDA-Mn₂O₃) composites [32]. EDA adsorbed on the surface of Mn₂O₃ can capture Mg ions by the coordination effect of the terminal biprotonic amine, promoting the electrochemical reaction kinetics. Consequently, EDA-Mn₂O₃ delivers enhanced capacity, rate performance, and cycling stability. Meanwhile, the reversible structural evolution of Mn₂O₃ during Mg-ion insertion/extraction is confirmed. Liu et al. reported the MnS/MnO heterostructure with dual-ion defects (DID-MnS/MnO) for aqueous magnesium-ion storage [93]. Compared to MnS, DID-MnS/MnO exhibits a highly enhanced capacity of 237.9 mAh g⁻¹ at 100 mA g⁻¹ and maintains a capacity retention of 94.7% after 1200 cycles at 1000 mA g⁻¹.

3.5. Electrolyte Effects

The electrolyte, as the ion transport medium between cathode and anode, has a decisive influence on the ion migration kinetics, electrochemical stability window, electrode/electrolyte interface characteristics, and reaction mechanism of the battery system. Therefore, the design and optimization of the electrolyte system is crucial for the practical application of manganese-based oxide cathode materials for AMIBs.

The electrolyte composition has a significant impact on the solvation structure of magnesium ions, thereby affecting the desolvation energy barrier. For example, the SO₄²⁻ anion exhibits a significantly positive Jones-Dole B coefficient, classifying it as a strongly hydrated anion, which makes it difficult to competitively replace water molecules in the first solvation shell of magnesium ions, resulting in a stable, dense Mg[(H₂O)₆]²⁺ configuration [46,94]. The abundant coordinated water and strong Mg–O bonds significantly raise the desolvation energy barrier. In contrast, Cl⁻ and NO₃⁻ anions have lower Jones-Dole B coefficients (mostly negative), indicating weaker self-hydration, thereby they can partially replace water molecules in the first solvation shell of magnesium ions, effectively reducing the desolvation energy barrier. Therefore, Mn-based oxide cathode materials usually displays lower capacity and poor rate performance in aqueous MgSO₄ electrolytes than in aqueous MgCl₂ and Mg(NO₃)₂ electrolytes with same concentration.

In addition, the electrolyte composition design can regulate the proton activity of the system, which significantly affect the proton insertion or co-insertion behavior [95,96]. Usually, the capacity contribution from proton insertion decreases with the increase in pH (decrease in proton activity). Meanwhile, the proton activity will affect the dissolution behavior of the cathode or anode materials [97,98]. However, the specific correlation between the proton activity of the electrolyte and the electrochemical performance of the manganese-based oxide cathode materials for AMIBs still lacks systematic research.

Mn dissolution is a key bottleneck limiting the electrochemical performance of Mn-based oxide cathode materials for AMIBs. Optimizing electrolyte composition is an effective strategy to inhibit this problem. Adding an appropriate amount of MnSO₄ into an electrolyte can effectively suppress manganese dissolution by the common ion effect [18,85]. In addition, by adding appropriate electrolyte additives, a protective electrode-electrolyte interphase can be formed in situ, which can also effectively inhibit the manganese dissolution [99].

Although conventional dilute aqueous electrolytes exhibit high ionic conductivities, their electrochemical stable window (ESW) is restricted by the theoretical decomposition voltage of H₂O (1.23 V). Broadening the ESW of aqueous electrolytes is key to improving the operating voltage and energy density of AMIBs. Recently, some electrolytes with expanded ESW have been designed for AMIBs, including super-concentration (water-in-salt) electrolyte [100], eutectic electrolyte [101], organic solvent-in-water electrolyte [102], and quasi-solid-state electrolyte [103]. However, the electrochemical properties of Mn-based oxide cathode materials in these electrolytes still need further systematic study.

4. Full Cells Based on Mn-Based Oxide Cathode Materials

The research on full-cell batteries is an important step towards the practical application of manganese-based oxide cathode materials for AMIBs. Currently, anode materials used to assemble full cells with Mn-based oxide cathode materials are mainly organic compounds, such as 3,4,9,10-perylenetetracarboxylic dianhydride (PTCDA), 3,4,9,10-perylenetetracarboxylic diimide (PTCDI), and polyimide (PI), as well as vanadium-based oxides like VO₂, FeVO₄, and VO_x. Some studies employ activated carbon anodes paired with manganese oxide cathodes, but the resulting devices are magnesium-ion hybrid capacitors rather than AMIBs. Therefore, these devices will not be discussed in this section. To achieve higher energy density, some researchers have paired Mg metal anodes with Mn-based oxide cathode materials for full-cell assembly, but this places greater demands on the electrolyte. In this section, the recent progress of AMIB full cells is discussed.

Vanadium-based oxide anode materials have been widely used in AMIBs due to their suitable electrode potential and high theoretical capacity. Zhang et al. have assembled the FeVO₄·0.9H₂O/graphene||Mg-OMS-1 (τ-MnO₂) full battery with the 1 M MgSO₄ electrolyte, exhibiting a discharge capacity of 65.1 mAh g⁻¹ and energy density of 58.5 Wh kg⁻¹ at 50 mA g⁻¹ [104]. In this section, capacity and energy density are based on the total active mass of the cathode and anode, unless otherwise specified. The anode materials are changed from FeVO₄·0.9H₂O/graphene to FeVO₄/C during the assembly of the FeVO₄/C||Mg-OMS-1 full battery, which shows a discharge capacity of 78.2 mAh g⁻¹ and energy density of 70.4 Wh kg⁻¹ [105]. The results demonstrate that anode materials also play an important role in AMIB full cells. Compared to FeVO₄, VO₂ is used more extensively in AMIBs because of its higher capacity, which is favorable for achieving higher energy density. For example, the VO₂||δ-MnO₂@CMS full battery exhibits a higher capacity of 112.2 mAh g⁻¹ [65]. Subsequently, various AMIB full cells based on the VO₂ anode and Mn-based oxide cathode have been assembled, for instance, VO₂||δ-MnO₂/MWCNT [57], VO₂||Ni-doped τ-MnO₂ [106], VO₂||Li_0.21MnO₂·H₂O (Li-δ-MnO₂) [61], VO₂||MgMn₂O₄/MWCNTs [78], VO₂||MgMn_1.5Ni_0.5O₄ [81], and so on. In order to improve the properties of the VO₂||δ-MnO₂ full battery, Li et al. have proposed an over-discharge pretreatment strategy [87]. After the over-discharge treatment, both Mg-δ-MnO₂ and Ca/Mg-δ-MnO₂ cathode materials are controllably cleaved into ultrafine nanoparticles. The difference is that Mg-δ-MnO₂ becomes amorphous, while the layered structure of Ca/Mg-δ-MnO₂ is retained. This is attributed to the pillar effect of pre-intercalated Ca-ion in Ca/Mg-δ-MnO₂. The VO₂||Ca/Mg-δ-MnO₂ full battery after the pretreatment delivers an enhanced capacity of 113.2 mAh g⁻¹ and an improved rate (45.8 mAh g⁻¹ even at 1000 mA g⁻¹). Moreover, a high capacity retention of 72.5% is achieved after 1000 cycles at 50 mA g⁻¹. Meanwhile, the over-discharge treatment of VO₂||δ-MnO₂ full battery using an electrolyte containing Cl⁻ (2 M Mg(NO₃)₂ + 25 mM MgCl₂) produces different results [107]. After the over-discharge pretreatment with Cl⁻ in the electrolyte, the δ-MnO₂ turns into amorphous ultrafine nanoparticles, but a crystalline V₆O₁₃ layer is generated on the surface. Meanwhile, the VO₂ anode is transformed into polycrystalline vanadium oxide containing mixed-valence states (+4 and +5). The generation of the V₆O₁₃ layer is attributed to the dissolution of vanadium species from the vanadium-based anode into the electrolyte mediated by chloride ions. Therefore, the full-cell capacity is enhanced to 120.62 mAh g⁻¹. However, the Coulombic efficiency (below 90%) and cycling stability (37.52% retention after 700 cycles) are unsatisfactory. The low-temperature electrochemical properties of the AMIB full battery composed of Mn-based oxide cathode materials are evaluated. Yang et al. have reported that the VO₂||δ-MnO₂ full battery with the 4 M MgCl₂ aqueous electrolyte can work in a wide temperature range from +25 to −50 °C [108]. At −20 °C, the full battery capacity is 97.9 mAh g⁻¹ at 100 mA g⁻¹ and 25.5 mAh g⁻¹ at 3000 mA g⁻¹, and after 1000 cycles at 2000 mA g⁻¹, 90% capacity retention is achieved. A flexible quasi-solid-state AMIB based on the polyacrylamide/MgCl₂ hydrogel electrolyte exhibits exciting properties, including low-temperature performance, mechanical flexibility, and safety.

Organic compounds constitute another important class of anode materials for AMIBs. Liu et al. have assembled the PTCDA||ε-MnO₂ full battery with 1 M MgCl₂ electrolyte, which exhibits good rates (246.6 mAh g⁻¹ at 500 mA g⁻¹ and 63.3 mAh g⁻¹ at 6000 mA g⁻¹, based on the mass of the cathode material) and cycling stability (capacity retention of 72.6% after 800 cycles) [52]. The energy density of the PTCDA||ε-MnO₂ full battery is 98.6 Wh kg⁻¹. Compared to PTCDA, PTCDI typically exhibits better cycling stability in aqueous batteries, which is beneficial for the full battery. Therefore, more researchers have paired the PTCDI anode with the Mn-based oxide cathode in the AMIB full battery. Various AMIB full cells based on the PTCDI anode and Mn-based oxide cathode have been reported, such as PTCDI||MgMn₂O₄ [84], PTCDI||MgMn₂O₄/rGO [67], PTCDI||Mg-intercalation α-MnO₂ [49], PTCDI||MnO_2−xF_y [69], PTCDI||O_d-WMO [29], and PTCDI||Mn₃O₄[18]. Among these, the PTCDI||MgMn₂O₄ full battery using a 0.5 M Mg(NO₃)₂ electrolyte delivers an energy density of 80.9 Wh kg⁻¹ at 648 W kg⁻¹ (based on the mass of cathode material) and a capacity retention of 80.9% even after 5000 cycles at 1000 mA g⁻¹ [84]. The PTCDI||MnO_2−xF_y full battery with the 1 M Mg (NO₃)₂ electrolyte achieves a high energy density of 106.1 Wh kg⁻¹ at 0.5 kW kg⁻¹ and a capacity retention of 88.1% even after 10,000 cycles at 5000 mA g⁻¹ [69]. Adding manganese salts to the electrolyte can inhibit manganese dissolution, thereby improving the cycle performance of manganese-based oxide cathode materials and AMIB full cells [18,85]. The PI||Mg₂MnO₄ full battery using 1 M MgSO₄ + 0.1 MnSO₄ electrolyte has good rate performances and a high capacity retention of 89% after 10,000 cycles at 100 C (18,000 mA g⁻¹) [85]. In contrast, the PI||Mg₂MnO₄ full battery with 1 M MgSO₄ electrolyte, without MnSO₄, exhibits poor rate capability and rapid capacity decay.

Although some AMIBs with full cells based on Mn-based oxide cathodes and V-based or organic anodes exhibit good rate performance and cycling stability, their energy density is limited by the low average discharge voltage (typically not exceeding 1 V). Mg metal is the ultimate negative electrode sought after for aqueous magnesium-ion batteries because it can achieve high operating voltage and high energy density, but it undergoes serious side reactions and passivation in traditional aqueous electrolytes. Xu et al. have proposed an organic solvent-in-water electrolyte (saturated MgCl₂ + 1 m MnCl₂ in H₂O:poly(ethylene glycol) 400 = 1:1, denoted as SIW-2) to avoid the passivation of Mg metal anode, and a high-voltage and stable aqueous Mg||MnO₂ battery (Figure 7a) is obtained [102]. The Mg||MnO₂ battery using SIW-2 electrolyte displays a discharge plateau at ~2.5 V and a discharge capacity of ~500 mAh g⁻¹ (based on the mass of MnO₂ on the cathode) (Figure 7b) with a capacity retention of around 99% after over 1000 cycles. The work demonstrates the feasibility of constructing high-performance AMIB full cells using Mg metal anodes and Mn-based oxide cathodes.

Figure 7.

(a) Schematic illustration of the structure and working mechanism of Mg||MnO₂ battery using the SIW electrolyte and (b) Discharge curves of Mg||MnO₂ batteries with different electrolytes. Reproduced from [102], with permission.

Energy density is one of the most important indicators for batteries in practical application. Most reported energy density of AMIB full cells are calculated merely based on the active mass of cathode or the total active mass of cathode and anode, whereas practical energy density should take the total weight of the cathode, anode, electrolyte, separator, current collectors and packaging into consideration. However, most reported AMIB full cells have low active materials loading, excessive electrolyte dosage and limited capacity (mAh-level), resulting in extremely low practical energy density. Therefore, improving the practical energy density requires not only enhancing the specific capacity and potential gap between cathode and anode materials, but also elevating active material loading, optimizing electrolyte dosage and developing high-capacity (Ah-level) pouch-cell configurations. For future practical application, the initial goal of AMIBs should be to compete with lead-acid batteries. When the practical energy density of AMIBs exceeds 40 Wh/kg, it will demonstrate significant comprehensive competitiveness due to its advantages such as environmental friendliness and abundant reserves. Currently, there is still a long way to go.

5. Conclusions and Outlook

Manganese-based oxides are an important class of cathode materials for AMIBs due to the abundant resources, low cost, high theoretical capacity, and environmental friendliness. Manganese-based oxide cathode materials for AMIBs are reviewed, including the tunnel type (α-MnO₂, β-MnO₂, γ-MnO₂, and τ-MnO₂), layer type (δ-MnO₂), spinel-type (λ-MnO₂, MgMn₂O₄, and Mn₃O₄), and other types (Mn₂O₃ and MnO) of manganese-based oxides. The crystal structure, electrochemical properties, optimization strategies, and electrode reaction mechanisms of various manganese-based oxide cathode materials are discussed. By using various efficient optimization strategies, including interlayer regulation, crystal defect engineering, heteroatom doping, and nanocomposite construction, many optimized manganese-based oxides achieve significant improvements in electrochemical performance (Table 2). Meanwhile, many AMIB full cells based on manganese-based oxide cathode materials have been reported (Table 3), and some of them show good rate capability and cycling stability. However, despite recent advances, the further development of Mn-based oxide cathode materials for AMIBs remains challenging.

Although the electrochemical properties of Mn-based oxide cathode materials for AMIBs have been improved by various optimization techniques, they still cannot meet the demands of practical applications. Mn dissolution, poor electronic conductivity, and slow Mg-ion diffusion kinetics are still the major issues. The Mn dissolution causes fast capacity fading, while poor electronic conductivity and slow Mg-ion diffusion kinetics limit the performance rate. For further improvements, research on other aqueous batteries, especially aqueous zinc-ion batteries, can provide valuable insights. In addition, artificial intelligence (AI)-assisted design based on machine learning and prediction may be helpful in developing high-performance Mn-based oxide cathode materials for AMIBs. However, the challenge lies in obtaining reliable data sources and building suitable learning models. In addition, the practical application potential of these optimization strategies needs further consideration. The nanostructure or nanocomposite construction cannot solve the slow magnesium-ion diffusion kinetics and will lead to a series of issues including reduced tap density, increased specific surface area causing more side reactions, and easy agglomeration. Interlayer regulation, crystal defect engineering, and heteroatom doping are all optimization strategies at the crystal level that can effectively improve the magnesium-ion diffusion kinetics in the lattices of Mn-based oxides. However, interlayer regulation is only applicable to layered materials, and increasing the interlayer spacing leads to a decrease in tap density. In contrast, doping and crystal defect engineering seem more suitable for future practical applications, and some doping will also induce crystal defects.

Understanding the electrode reaction mechanism of Mn-based oxide cathode materials is crucial to optimizing their electrochemical properties. In this respect, the electrode reaction mechanism of some manganese-based oxide cathode materials is still unclear or controversial (Table 1). For example, whether the intercalated species are Mg ions, protons, or both. Meanwhile, the different electrolytes and operating conditions may lead to different electrode reaction mechanisms. Therefore, more in-depth investigations are required.

To develop high-performance AMIBs based on Mn oxide cathodes, research on anode materials and electrolytes is also crucial. Anode materials with lower electrode potentials and high capacity are key to achieving high energy density. Mg metal is the ultimate negative electrode sought after, but it undergoes serious side reactions and passivation in traditional aqueous electrolytes. Therefore, electrolyte limitation is the biggest challenge for developing AMIB full cells using Mn-based oxide cathodes and Mg metal anodes. Excitingly, recent work demonstrates that electrolyte engineering and the design of artificial solid electrolyte interphases are effective approaches to alleviate these issues. However, more research is needed to achieve better improvement results.

Moreover, there is a gap between laboratory research and practical application. The mass loading of Mn-based oxide cathode materials in most previous literature is less than 5 mg cm⁻² (Table 2), which is too low for practical applications. Even if the specific capacity of manganese-based oxide cathode materials reaches 300 mAh g⁻¹, the areal capacity of commercial lithium-ion batteries (over 3.0 mAh cm⁻²) would still require a mass loading exceeding 10 mg cm⁻². However, the electrochemical properties of manganese-based oxides for AMIBs at high mass loadings (>10 mg cm⁻²) remain poorly understood. It is worth noting that some progress has been made in high mass loading manganese oxide cathodes for zinc-ion batteries, which can provide a reference for manganese-based oxide cathode materials for AMIBs [109]. Meanwhile, a pouch-cell demonstration of AMIBs full cell with manganese-based oxide cathode materials has been reported, but the capacity of the reported pouch cells is relatively low. The demonstration of high-capacity pouch cells (e.g., Ah-level) is very important for assessing the practical application potential of manganese-based oxide cathode materials. To achieve higher energy density, parameters such as active material loading, electrolyte-to-capacity ratio, and N/P ratio need to be considered. The long-term cycling stability under practical conditions is also important. In addition, prior to commercial application, exploring the large-scale synthesis of the high-performance manganese-based oxide cathode materials is necessary.

In summary, although manganese-based oxide cathode materials have great potential for AMIBs, much more work is needed to deepen our understanding of them, achieve further breakthroughs, and ultimately bring them to commercial fruition. Meanwhile, this review serves as a reference and a source of inspiration for future research on manganese-based oxide cathode materials for AMIBs.

Table 2.

Electrochemical properties of various Mn-based oxide cathode materials for AMIBs.

Materials	Electrolyte	Active-Material Mass Loading (mg cm−2)	Reference Electrode	Potential Range (V, vs. RE)	Capacity (mAh g−1)/Current Density (mA g−1)	Cycle Number/Remained Capacity (mAh g−1)/Retention Rate/Current Density (mA g−1)	Ref.
α-MnO2/graphene	0.5 M Mg(NO3)2	~5.0	SCE	−0.65–0.75	232.4/20, 115/100	300/-/93.0%/100	[46]
α-MnO2/CNT	1.0 M MgSO4	-	SCE	−0.6–1.4	144.6/500, 59.7/10,000	1000/48.3/85%/10,000	[47]
Mg-intercalation α-MnO2	1.0 M MgCl2	-	Graphite	−0.6–1.0	~419.8/100, 111.6/4000	1000/-/95.4%/1000	[49]
Al-doped α-MnO2	0.5 M MgSO4	-	Hg/HgO	−0.8–1.1	197.02/100, 90.50/1000	2500/-/82%/1000	[30]
10% Nb-doped α-MnO2	1.0 M Mg(NO3)2	~3.5	SCE	−0.6–0.9	252.6/20, 168.7/100	200/65.1/38.6%/100	[48]
10% V-doped α-MnO2	1.0 M Mg(NO3)2	~3.5	SCE	−0.6–0.9	265.9/20, 178.7/100	200/79.4/36.8%/100	[48]
15 mol% Fe-doped α-MnO2	1 M MgCl2	~2.0	Ag/AgCl	−0.65–1.0	205.9/100, 171.8/200, 118.6/500, 101/1000, 99.6/2000	1000/84.9/85.2%/2000	[31]
Fluorine-regulated MnOx/KMnF3	0.5 M MgSO4	~1.3	Hg/HgO	−0.8–1.1	142/100, 106/200, 85.2/300, 73.5/500, 60.9/800, 54.8/1000	1800/-/89.6%/1000	[25]
β-MnO2 (Mg-OMS-7) nanorods	0.2 M Mg(NO3)2	~7.0	SCE	−0.65–0.75	283.1/10, 102/100	200/97.21/95.3%/100	[34]
β-MnO2 porous nanoflowers	1 M MgSO4	~2.0	Ag/AgCl	−0.2–1.0	260/200, 151/1000, 68/10,000	1000/169/65%/2000	[50]
τ-MnO2 (Mg-OMS-1) nanobelts	0.5 M Mg(NO3)2	~7.0	SCE	−0.65–0.8	243/10, 149/20, 93.5/100	200/-/90.4%/100	[38]
τ-MnO2 (Mg-OMS-1) nanosheets	0.2 M MgCl2	~7.0	SCE	−0.65–0.8	300/10, 115/100	300/-/86%/100	[51]
τ-MnO2 (Mg-OMS-1)/graphene	0.5 M Mg(NO3)2	-	SCE	−0.65–0.8	194.1/20, 100.1/100	200/92.5/92.5%/100	[86]
Ni-doped τ-MnO2	0.5 M Mg(NO3)2	6.0–7.0	SCE	−0.65–0.75	318/10	300/-/89.7%/100	[106]
ε-MnO2 on carbon cloth	1.0 M MgCl2	2.0	Ag/AgCl	−0.65–1.0	259.3/500, 114.4/2000	400/-/94.3%/2000	[52]
ε-MnO2 nanoflowers	1.0 M MgSO4	1.0	SCE	−0.5–1.3	230.9/1000	-	[53]
Li-δ-MnO2 nanowires	0.5 M Mg(NO3)2	-	SCE	−0.65–0.75	226.7/20, 165.8/100	300/93.4/56.3%/100	[61]
K-δ-MnO2 nanoparticles	1 M MgSO4	~1.5	Ag/AgCl	−0.2–1.1	163/100, 185/1000, 78/10,000	1000/160/86.7%/1000	[62]
δ-MnO2 nanoflowers	1.0 M MgCl2	2.0	Ag/AgCl	−0.65–1.0	252.1/50, 198.2/100, 150.8/200, 125.4/500, 109.7/1000, 95.2/2000	800/-/54.4%/1000	[60]
δ-MnO2 from γ-MnS	1.0 M Mg(NO3)2	~1.5	Ag/AgCl	−0.2–1.0	~360/616	50/~200/-/308	[87]
Mg-δ-MnO2	0.5 M Mg(ClO4)2	-	Ag/AgCl	−0.3–1.1	231.1/100, 216.8/200, 172.4/500, 138.4/1000, 112.3/2000	500/-/78.3%/500, 10,000/-/62.5%/2000	[54]
Mg-δ-MnO2 with larger interlayer space (∼9.70 Å)	0.5 M MgCl2	~6.0	SCE	−0.8–1.0	169.3/50, 153.2/100, 98.3/1000	100/~150/-/100	[56]
Mg-δ-MnO2/carbon cloth	0.5 M Mg(ClO4)2	~1.5	Ag/AgCl	−0.3–1.3	~150/500	160/~45/-/500	[55]
TMA-MnO2	1.0 M MgSO4	-	Ag/AgCl	−0.6–0.9	110.8/232.69, 50.3/2326.9	500/63.3/-/465.38	[27]
δ-MnO2 nanosheet arrays on Ti foi	0.5 M Mg(ClO4)2	-	Ag/AgCl	−0.3–1.1	250/100, 210/200, 150/500, 135/1000, 90/2000	1500/63/84%/3000	[110]
ECMB	0.5 M Mg(ClO4)2	1.0–2.0	Ag/AgCl	−0.2–1.3	255.1/200, 203.5/500, 166.4/1000, 122.7/2000, 85.6/4000, 55.4/8000	3000/-/73.6%/2000	[58]
δ-MnO2@CMS	0.5 M Mg(NO3)2	~3.0–5.0	SCE	−0.6–0.8	270.3/50	300/~136/60.5%/50	[65]
δ-MnO2/MWCNT	0.5 M MgSO4	~3.9	SCE	−0.7–1.0	313.2/50, 189.9/100, 56.8/1000	500/-/~100%/1000	[57]
δ-MnO2@MWCNTs/CC	0.5 M MgSO4	4.4	SCE	−0.7–1.0	246.7/50, 75.2/1000	2000/60.5/80%/1000	[66]
δ-MnO2/rGO composite	2 M MgSO4 + 2 M MgOAc	-	Ag/AgCl	0–1.0	88.5/1000, 80.2/2000, 74.9/3000, 70.8/4000, 67.3/5000	-	[67]
K-MnO2/HMC	1.0 M MgSO4	~1.0	Hg/HgO	−0.5–1.3	168/500, 132/1000, 89/3000, 69/5000, 52/8000, 45.3/10,000	-	[68]
K+-intercalated δ-MnO2 on graphitic nanofibers	1.0 M MgSO4	1.5	Ag/AgCl	0–0.8	91.18/50, 51.88/4000	-	[28]
MnO2−xFy (F-doped δ-MnO2)	1.0 M Mg(NO3)2	~2.0	SCE	0–1.0	491.5 F g−1/500, 123.5 F g−1/20,000	10,000/-/83.9%/5000	[69]
Od-WMO	1.0 M MgSO4	-	Ag/AgCl	−0.3–1.1	185.2/100, 153.8/200, 133.5/500, 116.9/1000, 101.2/2000	1500/-/92%/500	[29]
V-Ovac-MnO2/CG	0.5 M MgSO4	~3.1	SCE	−0.82–0.85	398/100, 295/200	500/239/81%/200	[59]
Micron-sized Na0.7MnO2.05	1 M MgSO4 + 0.008 M sodium lauryl sulfate	6.0–8.0	SCE	0–0.8	40/20, 35/40, 22/100, 13/200	1000/10.2/59.8%/200	[70]
λ-MnO2 nanoparticles	1.0 M MgCl2	~7.0	SCE	−0.8–1.0	478.4/13.6, 288.0/136, 253.9/408	300/155.6/-/136	[39]
λ-MnO2 nanoparticles	1.0 M MgSO4	~2.5–3.0	SCE	−0.8–0.4	220/0.025C, 180/0.05C, 161/0.1C, 88/0.2C	100/-/52.5%/0.1C	[72]
λ-MnO2 nanoparticles	1.0 M Mg(NO3)2	~3.0	Ag/AgCl	−0.2–1.0	352/27, 293/90, 217/270, 110/1350	100/110/59%/270	[73]
λ-MnO2/MWCNTs	0.5 M MgSO4	3.5–5.0	SCE	−0.7–1.0	213.8/50, 127.2/1000	1000/-/86.2%/1000	[74]
MgMn2O4 form LiMn2O4 #	1.0 M Mg(NO3)2	4.0–6.0	SCE	−0.6–0.5	42/45.5	20/35/-/45.5	[75]
MgMn2O4 nanoparticles	3.0 M Mg(NO3)2	2.1–3.7	Ag/AgCl	0–1.3	~150/60 μA	20/~90/-/60 μA	[40]
MgMn2O4/rGO nanocomposites	0.5 M MgCl2	~5.5	SCE	−0.5–0.85	211.8/50, 140.1/1000	1000/119.5/85.3%/1000	[33]
MgFe1.33Mn0.67O4	0.5 M MgCl2	~6.0	SCE	−0.8–1.1	136.5/50, ~88.3/1000	1000/~88.3/100%/1000	[80]
MgMn2O4/MWCNTs	0.5 M MgSO4	-	SCE	−0.7–1.0	412.9/50, 318.4/100, 276.3/200, 212/300, 186.8/500, 176.6/800, 163.2/1000	1000/106.3/73.3%/1000	[77]
MgMn2O4/MWCNTs	0.5 M MgSO4	~4.0	SCE	−0.7–1.0	322.3/50, 239.1/100, 207.8/200, 189.8/300, 169.6/500, 152.8/800, 145.8/1000	2000/101/81.8%/1000	[78]
MgMn2O4 hollow sphere/rGO	1 M MgSO4 in water/acetonitrile (1:10 in volume)	-	Ag/AgCl	−0.4–1.2	305/300	100/290/-/300	[76]
MgMn2O4/PANI	0.5 M MgSO4	~0.2–1.0	Ag/AgCl	−0.6–1.2	232/500	250/170/73%/500	[79]
MgMn1.5Ni0.5O4@PPy	0.5 M MgCl	-	SCE	−0.6–0.9	273.3/50, 150.2/1000	2000/133/90.8%/1000	[82]
MgMn1.5Ni0.5O4	0.5 M MgCl	-	SCE	−0.6–0.9	221.8/50, 186.5/100, 162.2/200, 142.5/300, 136.8/500, 121.8/800, 115.2/1000	500/104/92.9%/1000	[81]
MgMn2O4 with Oxygen defects	0.5 M MgSO4	1.68	SCE	−0.2–1.0	230.8/100, 84/2000	550/~110/-/500	[26]
Mg2MnO4	1.0 M MgSO4 + 0.1 M MnSO4	1.0–2.0	Ag/AgCl	−0.2–0.8	116.1/160, 71.7/800, 60.8/3200	50/71.7/-/800	[85]
Milling-Mn3O4	2.0 M MgSO4	~5.0	Ag/AgCl	−0.2–1.0	105.8/100, 93.2/500, 79.8/1000, 68.5/2000	300/81/-/500	[89]
Mn3O4-A	1.0 M MgSO4	-	SCE	0–1.3	98.9/200, 73.8/500, 61.5/1000, 50.0/2000, 33.1/5000	2000/~98.3/99.4%/200	[91]
CADDs-Mn3O4	0.5 M MgSO4	~1.4	Hg/HgO	−0.6–1.0	392.93/100, 181.24/1000	2000/181.24/100%/1000	[41]
Cs-doped Mn3O4	0.5 M MgSO4	~1.3	Hg/HgO	−0.8–1.1	214.44/100, 179.60/200, 154.90/300, 131.30/500, 116.20/800, 110.14/1000	1500/-/95.38%/1000	[90]
Mn3O4 nanoparticles	1 M MgSO4 + 0.01 M MnSO4	-	SCE	−0.5–1.4	311/200, 105.6/5000	2000/-/94.9%/1000	[18]
Mn2O3@TiO2@MXene	0.5 M MgSO4	~1.3	Hg/HgO	−0.8–1.1	241.5/100, 180.5/200, 154.6/300, 138.3/500, 114.4/800, 106.7/1000	1350/-/85.7%/1000	[92]
EDA-Mn2O3	0.5 M MgSO4	-	Hg/HgO	−0.8–1.1	188.97/100, 160.47/200, 143.02/300, 128.47/500, 110.79/800, 97.24/10,000	1200/-/93.63%/1000	[32]
DID-MnS/MnO	0.5 M MgSO4	~1.3	Hg/HgO	−0.8–1.1	237.9/100, 161.8/200, 141.3/300, 117.8/500, 101.7/800, 95.9/1000	1200/90.1/94.7%/1000	[93]

^# We believe that the unit of current density in the original paper is wrong, it should be mA/g instead of A/g.

Table 3.

Electrochemical properties of AMIB full cells constructed with Mn-based oxide cathode materials.

Full Battery	Electrolyte	Voltage Range (V)	Average Discharge Voltage (V)	Capacity (mAh g−1)/Current Density (mA g−1)	Cycle Number/Capacity/Retention Rate/Current Density (mA g−1)	Energy Density (Wh kg−1)/Power Density (W kg−1)	Ref.
FeVO4/C||Mg-OMS-1 (τ-MnO2)	1.0 M MgSO4	0.0–1.8	~0.8	78.2/50, 58.9/100, 49.5/200, 38.9/300, 32.1/500	100/-/97.7%/100	70.4/-	[105]
FeVO4·0.9H2O/graphene||Mg-OMS-1 (τ-MnO2)	1.0 M MgSO4	0.0–1.8	~0.6	65.1/50, 53.1/100, 42.3/200, 32.0/300, 24.4/500	100/-/97.2%/100	58.5/50.8	[104]
VO2||δ-MnO2@CMS	1.0 M MgSO4	0.0–1.8	~0.9	112.2/50, 94.6/100, 78.7/200, 69.5/300, 59.1/500	100/-/46.9%/-	-	[65]
VO2||Ni-doped τ-MnO2 #	1.0 M MgSO4	0.0–1.8	~0.9	198.2/50, 125.1/100, 102.7/200, 95/300, 81.9/500	100/-/94.6/100	-	[106]
VO2||δ-MnO2/MWCNT	0.5 M MgSO4	0.0–2.0	~0.7	108.4/50,81.5/100	500/20/-/1000	-	[57]
VO2||Li-δ-MnO2	0.5 M MgSO4	0.0–1.6	~0.7	74.8/20, 54.7/50, 46.4/100, 34.0/200, 32.1/300	50/46.2/80.5%/100	-	[61]
VO2||MgMn2O4/MWCNTs	0.5 M MgSO4	0.0–1.9	~0.8	115/50, 75.8/100, 54.0/200, 41.5/300, 30.7/500, 21.4/800, 17.6/1000	500/18.1/~100%/1000	-	[78]
VO2||δ-MnO2 #	4 M MgCl2	0.0–1.8	~0.7	228.5/100 (25 °C), 97.9/100(−20 °C), 37.1/100 (−50 °C)	1000/-/74%/2000 (25 °C), 1000/-/90%/2000 (−20 °C), 1000/-/~100%/2000 (−50 °C)	-	[108]
VO2||δ-MnO2 #	PAAm/MgCl2 hydrogel electrolyte	-	-	178.5/100 (25 °C), ~80/100(−20 °C)	500/-/88%/2000 (25 °C), 500/-/~100%/2000 (−20 °C)	-	[108]
VO2/V6O13||V6O13@Mg-MnO2 #	2 M Mg(NO3)2 + 25 mM MgCl2	−0.45–1.4	~0.3	120.62/50	100/117.55/97.45%/50, 200/86.8/71.96%/50, 700/45.21/37.52%/50	-	[107]
VO2||MgMn1.5Ni0.5O4 *	0.5 M MgCl	0.0–1.7	~0.7	85.8/50, 24.8/1000	400/22.1/82.1%/1000	-	[81]
VO2||nano-Mg/Ca-δ-MnO2 *	2 M Mg(NO3)2 + 0.01 M MgSO4	−0.45–1.4	~0.3	113.2/50, 94.85/100, 74.6/250, 59.1/500, 45.8/1000	1000/63.39/72.5%/50, 500/-/87.7%/500	-	[88]
VOx||MnO2 *	4.5 M MgCl2	~0.0–1.8	~0.8	144.9/130, 73.9/260, 59.9/400, 54.2/670, 47/1300, 40.3/2600	1000/-/84.5%/2600	-	[111]
V6O13−x||δ-MnO2 *	1.0 M Mg(Ac)2	0.01–1.6	~0.7	55.1/200, 131.2/500, 122.6/700, 111.5/1000, 92.1/2000	200/123/~80%/200, 3000/45/50%/2000	-	[97]
PTCDA||ε-MnO2 #	1.0 M MgCl2	~0.0–1.7	~0.7	246.6/500, 160/1000, 63.3/6000	800/-/72.6%/1000	98.6/-	[52]
PI||Mg2MnO4 #	1 M MgSO4 + 0.1 M MnSO4	0.0–1.6	~0.8	93.6/360, 70.7/18,000	10,000/60.7/89%/18,000	60.1/15,300	[85]
PI||ECMB *	0.5 M Mg(ClO4)2	0.0–2.2	~0.9	65/100, 33/1000	500/-/99%/500	65.2/96	[58]
PTCDI||MgMn2O4 #	0.5 M Mg(NO3)2	0.01–2.10	~0.7	194/100, 125/1000	5000/124.8/80.9%/1000	80.9/648	[84]
PTCDI||Mg-intercalated α-MnO2 *	1 M MgCl2	-	-	94.8/200, 82.6/500, 77.3/800, 74.8/1000, 69.5/2000, 64.4/4000	400/-/103.44%/-	-	[49]
PTCDI||δ-MnO2/rGO *	2 M MgSO4 +2 M MgOAc	0.0–2.0	~0.8	36.2/1000, 35.2/2000, 33.8/3000, 32.1/4000, 30.6/5000	3000/-/76.4%/2000	29.8/822.5, 25/4098.36	[67]
PTCDI||MnO2−xFy	1.0 M Mg(NO3)2	0.0–2.0	~1.0	190.9 F g−1/500, 66.1 F g−1/10,000	10,000/-/88.1%/5000	106.1/0.5, 15.9/20	[69]
PTCDI||Mn3O4 *	1 M MgSO4 + 0.01 MnSO4	0.0–2.7	~1.0	73.3/200, 63.2/500, 52.4/1000, 44.8/2000	1000/-/88.8%/1000	75.9/211.6, 42.8/1772.7	[18]
PTCDI||Od-WMO *	1.0 M MgSO4	0.0–1.5	~0.3	106.6/100, 96.5/200, 82.6/500, 70.6/1000, 60.2/2000	1500/-/83%/500	80.4/75	[29]
Mg||MnO2 #	Saturated MgCl2 + 1 M MnCl2 in H2O:PEG = 1:1 (SIW-2)	1.4–2.8	~2.5	~500/5C (1 mA cm−2)	1000/-/~99%/5C (1 mA cm−2)	-	[102]

Calculations of capacity and energy density are based on the total active mass of the cathode and anode, unless otherwise specified. ^# Specific capacities or energy density/power densities are calculated based on the mass of cathode material. * The mass on which the calculation of specific capacity or energy density/power density is based is not mentioned.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

Not applicable.

Conflicts of Interest

The authors declare no conflicts of interest.

Funding Statement

This work was supported by the Tenth Young Elite Scientists Sponsorship Program by CAST and City University of Hong Kong Donation Research Grants (Nos. DON-RMG 9229021 and 9220061).