Challenges and Perspectives for Doping Strategy for Manganese-Based Zinc-ion Battery Cathode

Abstract

As one of the most appealing options for large-scale energy storage systems, the commercialization of aqueous zinc-ion batteries (AZIBs) has received considerable attention due to their cost effectiveness and inherent safety. A potential cathode material for the commercialization of AZIBs is the manganese-based cathode, but it suffers from poor cycle stability, owing to the Jahn–Teller effect, which leads to the dissolution of Mn in the electrolyte, as well as low electron/ion conductivity. In order to solve these problems, various strategies have been adopted to improve the stability of manganese-based cathode materials. Among those, the doping strategy has become popular, where the dopant is inserted into the intrinsic crystal structures of electrode materials, which would stabilize them and tune the electronic state of the redox center to realize high ion/electron transport. Herein, we summarize the ion doping strategy from the following aspects: (1) synthesis strategy of doped manganese-based oxides; (2) valence-dependent dopant ions in manganese-based oxides; (3) optimization mechanism of ion doping in zinc-manganese battery. Lastly, an in-depth understanding and future perspectives of ion doping strategy in electrode materials are provided for the commercialization of manganese-based zinc-ion batteries.

1. Introduction

Due to the excessive consumption of fossil fuels and the growing problem of climate change caused by environmental pollution, renewable energy research and development, such as solar, wind, and tidal energy, has attracted worldwide attention [1,2,3,4]. However, renewable energy power generation is often intermittent and unpredictable, which requires large-scale energy storage systems to effectively buffer such fluctuations to achieve stable energy output for smart grid [5,6,7]. As an efficient and flexible energy storage device, lithium-ion batteries (LIBs) have not only been successfully applied in electronic consumer products such as cellphones and laptops, but are also expanding their range to electric vehicles and other fields [8,9,10,11]. However, the high production cost, limited lithium resource reserves, and the use of toxic and flammable organic electrolytes make lithium-ion batteries expensive, hazardous, and environmentally polluting, strongly impeding their further development and application in grid-scale energy storage [12,13,14]. Therefore, researchers are seeking new energy storage battery systems to replace LIBs in terms of cost, safety, and sustainability.

Among those energy storage devices, aqueous electrolytes have been widely employed in secondary battery systems for Na⁺, K⁺, Mg²⁺, Al³⁺, Ca²⁺, Zn²⁺, etc. in recent years because of their safety, high ionic conductivity, and ease of operation [15,16]. Among aqueous multivalent ion batteries, ZIBs stand out due to the following characteristics: (1) Zn metal reserves are abundant and the manufacture process of ZIBs occurs in an air environment, making it cost-effective; (2) Zn metal anode has a low redox potential of −0.76 V with respect to a standard hydrogen electrode and high theoretical gravimetric/volumetric capacity (820 mAh·g⁻¹/5855 mAh·cm⁻³); (3) Zn metal can be directly applied as an anode due to its excellent electrochemical stability and reversibility in water; (4) ZIBs are highly safe because of the application of nontoxic aqueous electrolyte [17,18]. However, considering the large ionic radius of hydrated zinc (5.5 Å vs. 0.74 Å for Zn ion), the intercalation of hydrated zinc ions would either require large spacing to accommodate the large ions or withstand a large desolvation penalty for smaller dehydrated ions to intercalate, imposing a great challenge in the development of suitable cathode materials [19]. At present, manganese-based oxides, vanadium-based oxides, and Prussian blue analogs are mainly developed as cathode materials for ZIBs [20]. Among them, manganese-based oxides are widely recognized as candidates for the commercialization of ZIBs because of their mature synthesis process, abundant resources, lack of pollution, high specific capacity, and high operating voltage [21]. However, the further application of Mn-based cathodes is hindered by two major issues. The redox reaction involving Mn⁴⁺ is usually accompanied by the Jahn–Teller effect and leads to the formation of Mn²⁺, which tends to dissolve into the electrolyte and lead to irreversible capacity loss. On the other hand, the poor ion/electric conductivity of the transition metal oxide would sacrifice the rate capability of the zinc battery [9,22,23]. At present, various strategies such as nanostructure engineering [24,25,26], conductive agent coating [27], and ion doping are widely adopted to tackle the above problems [21,28,29]. Among the various strategies, ion doping involves a small number of guest ions being preinserted into the manganese-based oxide framework and interacting with the host atoms to achieve an inherent structure optimization, which significantly enhances the electrochemical performance from a fundamental thermodynamics and dynamics aspect. This approach is recognized as an efficient and straightforward optimization strategy, breaking through the limitations of the inherent crystallographic structure [30].

In this review, considering the current research progress of ion doping optimization strategies in manganese-based oxides, we first briefly summarize the synthesis strategy, discuss the dependence of the valence states of doping ions on the manganese-based oxide structure, and highlight the optimization mechanism of ion doping in terms of the electrochemical performance. Lastly, we provide unique insights and prospects for ion doping strategies in manganese-based oxides for the application of zinc batteries for large grid-scale energy storage devices.

2. Synthesis Strategy for Ion-Doped Manganese-Based Oxides

Manganese oxides are highly dependent on dopant ions due to their effect on the crystalline phase, crystal structure, and average valence of the manganese oxides [31,32,33]. Daniil et al. built an ab initio model using the SCAN function to reveal the effect of doping different guest ions on the formation of manganese dioxide with different phases, as depicted in Figure 1a [31]. As shown in Figure 1b, the doping of Na⁺, K⁺, and Ca²⁺ was more likely to form α-MnO₂, whereas Li⁺ and Mg²⁺ favored the formation of γ-MnO₂, and δ-MnO₂ was easily stabilized by Na⁺. The mechanistic basis was that distinct metal ion doping resulted in varied formation free energies required for the different phases of manganese dioxide. Hu et al. demonstrated that partial Mg²⁺ intercalation resulted in tunnels of various sizes, such as 3 $\times$ 3, 4 $\times$ 3, and 5 $\times$ 3 tunnels in T-MnO₂, not just 3 $\times$ 3 tunnels [32]. Furthermore, the synthesis methods and synthesis circumstances also have a significant impact on the electrochemical performance of manganese oxides. Up to now, various synthesis methods have been widely applied in ion doping, including the hydrothermal method, ion penetration/exchange method, electrodeposition method, and calcination treatment.

2.1. Hydrothermal Method

Because of its simple controllability over the diverse crystal phases of manganese-based oxides, hydrothermal synthesis is the most frequently used approach for ion doping. Ion doping can be controlled by the addition of the different ions in the raw solutions before hydrothermal treatment. Zhang et al. prepared α-K_0.19MnO₂ nanotubes via decomposition of KMnO₄ combined with carbon nanofibers as templates in 2019 [34]. Under the hydrothermal conditions of 140 °C for 10 h, a typical K⁺-doped tunnel structure was achieved. The reaction mechanism was as follows: K⁺ + MnO₄⁻ + C + H₂O → δ-K_0.19MnO₂·nH₂O + CO₃²⁻ + HCO₃⁻. Through a similar route, MnO₂H_0.16(H₂O)_0.27 nanolayers were synthesized via the reaction of KMnO₄ with acetylene black at 120 °C for 24 h by Pan et al. [35]. Interestingly, this new phase exhibited excellent rate capability (115.1 mAh·g⁻¹ at 10 °C) with robust structural stability even with an interlayer spacing of less than 0.3 nm. Moreover, they developed a layered K_0.36H_0.26MnO₂·0.28H₂O via the neutralization reaction of KMnO₄ and MnSO₄ with K₂SO₄ additive as an excellent cathode in ZIBs [36]. The layer-type structure of monoclinic birnessite phase was obtained following a hydrothermal reaction of 120 °C for 12 h with a large interlayer spacing (7.12 Å). Shi et al. synthesized a cathode (K_0.29MnO₂·0.67H₂O) with a larger interplanar spacing (7.4 Å) through a hydrothermal potassium insertion strategy [37]. Peng et al. developed a Na⁺-incorporated layered δ-MnO₂ by subjecting K⁺-containing δ-MnO₂ to hydrothermal treatment using a 0.5 mol·L⁻¹ Na₂SO₄ solution at 180 °C for 3 h [38]. Layered Ca_0.28MnO₂·0.5H₂O was synthesized by Tao et al. through a hydrothermal method at 160 °C for 12 h using CaCl₂, KMnO₄, and MnSO₄ as reactants with a molar ratio of 1:6:1 [39]. This work demonstrated that divalent alkaline earth metal ions could also support layered manganese oxides, resulting in excellent electrochemical performance. Li et al. successfully incorporated Ni²⁺ into α-MnO₂ for boosting the diffusion kinetics of protons in the tunnels, which proved the substitution of divalent metal ions for Mn sites in tunnel-type manganese based oxides [40]. Du et al. found that the addition of Ce³⁺ ions during hydrothermal synthesis could induce a phase transition of MnO₂ from β to α, which resulted in a larger tunnel structure (2.3 $\times$ 2.3 Ų vs. 4.6 $\times$ 4.6 Ų) [41]. Wang et al. developed a Bi³⁺-doped α-MnO₂ cathode with an enlarged lattice spacing [42]. First, Bi(NO₃)₃ and MnSO₄ were mixed uniformly; then, KMnO₄ was added and stirred for 2 h, before being transferred to a 100 mL autoclave and reacting at 120 °C for 12 h. Yan at al. also designed Al-intercalated α-MnO₂ using a hydrothermal approach with a narrower electronic bandgap [43]. Interestingly, Al-doped MnO₂ exhibited a sea urchin-like morphology with a size of 4.5–5.0 µm and an enlarged interlayer spacing (0.24 nm vs. 0.29 nm). Xiong et al. reported that Al-doped α-MnO₂ coated with lignin was formed through a hydrothermal reaction involving KMnO₄, NH₄F, Al₂(SO₄)₃, and lignosulfonate at 200 °C, as illustrated in Figure 2a [44].

2.2. Ion Penetration/Exchange Method

The use of manganese oxides as precursors and the subsequent introduction of guest ions through a post-treatment process are more straightforward ideas for the ion doping strategy. Dai et al. developed a porous H_xMn₂O₄ cathode using a cation exchange strategy, which exhibited a novel crystal structure with an excellent cycle stability (1000 cycles at 1 A·g⁻¹) [45]. First, they used ZnSO₄ as the zinc source, MnSO₄ as the manganese source, and NH₄HCO₃ for the coprecipitation reaction to obtain ZnCO₃-MnCO₃ composites, followed by high-temperature treatment at 600 °C for 3 h to obtain the ZnMn₂O₄ precursor, which was finally dispersed into 0.5 M H₂SO₄ ion exchange solution for 12 h to get H_xMn₂O₄. The mechanism of zinc-ion extraction by H⁺ was the disproportionation of Mn³⁺ and [ZnO₄] tetragonal distortion. This distinctive spinel-type cathode offered new opportunities for long-life span ZIBs. Banerjee et al. reported that a Cu-intercalated MnO₂ layered cathode was attained by mixing the prepared MnO₂ powder with a 1 M CuSO₄ solution for 48 h [46]. The penetration of Cu²⁺ into δ-MnO₂ resulted in an enlarged lattice spacing, thereby lowering the charge transfer resistance. Furthermore, they exploited the redox potential of Cu for full capacity using two electrons. A cobalt-modified δ-MnO₂ with a redox-active surface showed superior self-recovery capability, as reported by Shao et al. [47]. As shown in Figure 2b, a molten-salt method was adopted to synthesize δ-MnO₂ using MnSO₄·H₂O and NaNO₃ as the reactants, and then δ-MnO₂ powder was mixed with 1 M CoCl₂ aqueous solution by constant stirring for 8 h at room temperature. The deposition–dissolution mechanism was proven by the electrochemical performance (over 500 mAh·g⁻¹), and Co²⁺ played a catalytic role in the electrochemical deposition of Mn²⁺. In addition, a Mn²⁺ additive was introduced into the electrolyte for enhanced cycle-stability.

2.3. Electrodeposition Method

Electrodeposition methods often involve depositing an electrolyte onto a conductive substrate by applying a certain current or voltage, which has the advantage of outstanding conductivity because of the highly conductive substrate. Dai et al. prepared a Na⁺-doped MnO₂@GCF cathode via the electrodeposition of 0.1 M Na₂SO₄ and 0.05 M MnSO₄·H₂O onto a graphene-like carbon film (GCF) [48]. The cathode was synthesized through a two-step procedure. As illustrated in Figure 2c, they first transformed the raw graphite paper into GCF by electrodepositing it into H₂SO₄ electrolyte, which had a 2D–3D hybrid network composed of graphene sheets. Then, the H₂SO₄ electrolyte was replaced by NaSO₄ and MnSO₄·H₂O to obtain Na⁺-doped MnO₂@GCF. The prepared cathode achieved excellent energy density (511.9 Wh·kg⁻¹ at 137 W·kg⁻¹). A Co–MnO₂ membrane was electrodeposited onto N-decorated carbon cloth (N-CC) by Nakayama et al. [49] in 2020, using an electrolyte consisting of MnSO₄, ZnSO₄, and CoSO₄. Furthermore, the cathode delivered an impressive capacity of 280 mAh·g⁻¹, even at 1.2 A·g⁻¹. Wang et al. reported the electrodeposition synthesis of multivalence cobalt-doped Mn₃O₄ (Co-Mn₃O₄) [50]. Similarly, a pretreated carbon cloth was applied as the substrate, while the cobalt and manganese sources were Co(CH₃COO)₂·4H₂O and Mn(CH₃COO)₂·4H₂O, respectively. Moreover, cobalt was present in multiple valence forms in the manganese oxides and played different roles, resulting in improved charge/ion transport and enhanced structure stability.

2.4. Calcination Treatment

Calcination treatment can provide high kinetics for guest ion intercalation, which is also applicable for manganese-based oxide doping strategies. Low-bandgap Ni_xMn_3−xO₄ nanoparticles were synthesized by Guo et al. through different calcination processes using manganese acetate as the manganese source and nickel acetate as the additive [51]. A Ni–Mn-layered double hydroxide-derived Ni-doped Mn₂O₃ (NM) was developed by Huang et al. [52]. First, the precursor Ni–Mn-LDH was formed by adding ammonia to a mixture of Ni(NO₃)₂·6H₂O, MnSO₄·H₂O, and NH₄F for coprecipitation at room temperature, and then Ni-doped Mn₂O₃ was obtained by calcining the precursor at 450 °C, as depicted in Figure 2d. A metal–organic framework template strategy was adopted by Sun et al. to synthesize a N-doped Mn-based cathode (MnO_x@N-C) [53]. Firstly, MnO₂ was generated by decomposing potassium permanganate under an acidic environment, and then MnO₂ was mixed with PVP, Zn(NO₃)₂·6H₂O, and 2-methylimidazole at room temperature to produce PVP-modified MnO₂@ZIF-8. Finally, MnO_x@N-C was obtained via calcination of MnO₂@ZIF-8 at 700 °C. Xia et al. fabricated a N-doped MnO_2–x cathode by calcining MnO₂ at 200 °C under an NH₃ atmosphere [54]. MnO₂ was deposited on TiC/C via KMnO₄ decomposition, while N doping was processed by NH₃ treatment at low temperature. Li et al. designed a N-doped Na₂Mn₃O₇ (N-NMO) in combination with sodium pre-intercalation and nitrification strategies [55]. In the first step, they used a chemical reaction involving KMnO₄, C₆H₁₂O₆, and NaKC₄H₄O₆ to synthesize rugby-type MnCO₃ particles as precursors. Next, Na₂Mn₃O₇ (NMO) was obtained by calcining MnCO₃ and NaNO₃ with a molar ratio of 3:2 at 600 °C for 4 h. Finally, N was introduced into NMO via further calcination under an ammonia atmosphere. Sun et al. reported that sulfur-doped MnO₂ (S-MnO₂) nanosheets were obtained using a two-zone furnace for application as a high-performance cathode [56]. The S powder was placed on the upstream side under a temperature of 450 °C, while MnO₂ was placed on the downstream side under a temperature of 250 °C. This process was maintained for 1 h under Ar atmosphere.

2.5. Other Methods

In addition to the synthesis methods summarized above, several other strategies have been reported. Zn²⁺-doped MnO₂ was also constructed by Wang et al. using KMnO₄ as the oxidant and zinc powders as the reducing agent [57]. Lu et al. adopted a facile one-step liquid coprecipitation to compose La–Ca co-doped ε-MnO₂ using CaCl₂ and La(NO₃)₃·6H₂O as the additives combined with a neutralization reaction of potassium permanganate and manganese sulfate [58]. K_0.41MnO₂·0.5H₂O was synthesized via a conventional sol–gel method by Tao et al. [59]. Mai et al. reported that Ti–MnO₂ was obtained by calcining core–shell MnO₂@TiO₂ nanowires at 450 °C in air for 4 h. MnO₂@TiO₂ nanowires were synthesized with MnO₂ as the substrate and TiCl₄ as the additive in a homemade atomic layer deposition system [60]. Cu–MnO was prepared by dissolving δ-MnO₂ and Cu(CH₃CN)₄PF₆ in acetone, followed by refluxing at 45 °C for 4 h by Wu et al. [61]. Barpanda et al. combined simple sonochemical and high-temperature calcination techniques to convert MnCO₃ into a novel K_1.33Mn₈O₁₆ [62]. Liu et al. reported a novel Mn₃O₄@NC cathode, which showed superior cycling stability for ZIBs [63]. As shown in Figure 2e, they first prepared the MnOOH precursor via a KMnO₄ decomposition reaction, and then the precursor @PPy was obtained by mixing MnOOH with C₁₈H₂₉SO₃Na and ammonium persulfate in a cold bath (0 °C). Finally, the as-prepared sample was sintered at 400 °C for 2 h under Ar flow.

3. Valence-Dependent Dopant Ions in Manganese-Based Oxides

Due to the variable electronegativity of the dopant ions in different valence states, the intercalation sites in the host structure are distinct, performing diverse modifying functions. In this section, pertinent publications about the ion doping of manganese-based oxides reports are discussed and summarized considering the multiple valence states of the doped ions, with the goal of delving into various manganese-based ion doping rules and gaining novel insights. The valence states of dopant ions can be roughly divided into the following categories: (1) monovalent cation doping, (2) multivalent cation doping, and (3) anion doping. In order to facilitate an understanding of the role of different dopant ions in various synthesis strategies, we present their electrochemical performance in

Table 1. Summary of the synthesis strategies and electrochemical performance of various dopants for manganese-based zinc-ion battery cathodes.

Cathode Material	Synthesis Method	Capacity (mAh·g−1/mA·g−1)	Rate Performance (mAh·g−1/mA·g−1)	Cycle Performance (X %/cycles)	Ref.
MnO2–birnessite	Other	266/100	134/2000	83.7%/2000	[64]
δ-Na0.44Mn2O4·1.5H2O	Other	278/300	106/6000	98%/10,000	[65]
Na:MnO2/GCF	Electrodeposition	381.8/100	175/2000	75.2%/1000	[48]
Na+–δ-MnO2	Hydrothermal	335/500	194/4000	93%/1000	[38]
Na0.1MnO2·0.5H2O	Other	270/30	100/3000	98%/5000	[66]
Na0.44MnO2	Other	301.3/100	80/1000	69.3%/800	[67]
N–Na2Mn3O7	Calcination treatment	300/200	100/10,000	78.9%/550	[55]
α-K0.19MnO2	Hydrothermal	266/300	113/6000	90%/400	[34]
K1.33Mn8O16	Other	312/30	80/1500	80%/650	[62]
K0.8Mn8O16	Hydrothermal	300/100	154/1000	100%/1000	[68]
K0.36H0.26MnO2·0.28H2O	Hydrothermal	329.8/30	100.1/3000	93.4%/3000	[36]
K0.29MnO2·0.67H2O	Hydrothermal	300/200	158/2000	98%/12,000	[37]
Ag1.5Mn8O16	Other	240/50	-	-	[69]
MnO2H0.16(H2O)0.27	Hydrothermal	275.6/30	115.1/3000	79%/2000	[35]
H1.57Mn2O4	Ion exchange	281/100	133.4/1000	93.5%/1000	[45]
Cu–MnO	Other	320/150	156/900	70%/1000	[61]
Ocu–Mn2O3	Other	246/50	112/1000	88%/600	[70]
NixMn3−xO4	Calcination treatment	139.7/50	98.5/1200	93.5%/800	[51]
Ni-doped Mn2O3	Calcination treatment	252/100	132/1000	85.6%/2500	[52]
Ni0.052K0.119Mn0.948O2·0.208H2O	Hydrothermal	303/30	175/1200	71.4%/2000	[40]
Zn2+-δ-MnO2	Other	358/300	124/3000	80%/2000	[57]
Co-Mn3O4	Electrodeposition	362/100	90/4000	80%/1100	[50]
Ca0.28MnO2·0.5H2O	Hydrothermal	298/175	124.5/3500	100%/5000	[39]
Ce-doped MnO2	Hydrothermal	270/150	134/1500	70%/100	[41]
La3+-δ-MnO2	Other	278.5/100	121.8/1600	71%/200	[71]
Bi-α-MnO2	Hydrothermal	325/300	150/5000	90.9%/2000	[42]
Al0.35Mn2.52O4	Hydrothermal	302/100	147/1500	95%/1000	[72]
Al-intercalated MnO2	Hydrothermal	401/100	229/4000	94.5%/2000	[43]
Fe/α-MnO2@PPy	Other	270/100	96/800	75%/1000	[73]
Ti–MnO2	Other	259/100	179/1000	80%/4000	[60]
V-doped MnO2	Other	266/66	67/1064	-	[74]
La–Ca co-doped ε-MnO2	Other	297.3/200	161/1600	76.8%/200	[58]
ZnNixCoyMn2−x−yO4@N-rGO	Other	200.5/10	93.5/1500	79%/900	[75]
N–CC@MnO2	Other	353/500	201/6000	93.6%/1000	[76]
N–C@MnOx	Calcination treatment	305/500	100/2000	100%/1600	[53]
Mn3O4@NC	Other	280/100	97/1000	100%/700	[63]
ZnMn2O4/N	Other	221/100	110/1000	97.4%/2500	[77]
MnO2–NHCS	Other	349/100	100/2000	78.7%/2000	[78]
N–MnO2–x	Calcination treatment	285/200	155.2/2000	85.7%/1000	[54]
P–MnO2−x@VMG	Other	302.8/500	150.1/10,000	Above 90%/1000	[79]
S–MnO2	Calcination treatment	324/200	205/2000	95%/1000	[56]

.

3.1. Monovalent Cation-Doped Manganese-Based Oxides

Monovalent ions can often be easily embedded into the host structure because of their low charge density and their large ionic radius tending to enlarge the interlayer spacing after inserting, thereby improving the electrochemical performance and the stability of manganese-based oxides. The monovalent ions used to modify manganese-based oxides are mainly sodium and potassium ions, while other monovalent ions remain largely unexplored.

In 2018, Wang et al. synthesized poorly crystalline MnO₂–birnessite at room temperature with a good rate performance (110 mAh·g⁻¹ at 10 °C and 308 mAh·g⁻¹ at 1 °C) and outstanding cycle performance (2000 cycles at 2 A·g⁻¹). However, the roles of Na⁺ and H₂O were not clearly illustrated [64]. In 2019, Zhi et al. discovered that the doping of Na⁺ ions and water molecules could actively improve the zinc storage capability of δ-MnO₂, which delivered excellent electrochemical performance (106 mAh·g⁻¹ at 6 A·g⁻¹ after 10,000 cycles) [65]. Furthermore, they succeeded in making self-healing Zn–δ-MnO₂ cells by via with a self-healing substance for electrodes, which could be repaired and restored even after repeated destructive cutting. Carboxylated polyurethane (CPU) and polyacrylamide (PAM) hydrogel electrolytes were used to design flexible batteries, whereby the former was used as the substrate for the electrodes and the latter was used as the quasi-solid-state electrolyte. The assembly structure of the self-healing device and the self-healing process are illustrated in Figure 3a. The self-healing ability was mainly due to carboxyl groups and hydrogen bonds of the CPU substrate being rebuilt after contact, as shown in Figure 3b. Dai et al. fabricated a Na-MnO₂@GCF cathode via electrochemical deposition [48]. DFT was used to investigate the effect of sodium ion doping on layered manganese dioxide. The introduction of sodium ions could reduce the binding energy between zinc ions and manganese dioxide, allowing more zinc ions to be combined. The electron domains of oxygen atoms around the zinc ions became more concentrated, further indicating that the electrode material was better for zinc absorption due to the improved electrochemical performance, as illustrated in Figure 3c,d. In 2020, Peng et al. reported the fabrication of Na⁺ incorporated into layered δ-MnO₂ with a uniform array arrangement and high conductivity by adopting nickel foam as a substrate, which delivered an excellent rate performance (131 mAh·g⁻¹ at 6 A·g⁻¹) [38]. Zhu et al. proposed that regulating the coordination of solvent water molecules with Zn²⁺ could significantly improve the performance of a birnessite cathode (Na_0.1MnO₂·0.5H₂O) [66]. Urea was applied to control the coordination number and geometry of Zn²⁺ because of its strong coordination, preventing the dissolution of Mn via the interface effect between the cathode and the electrolyte. Wang et al. employed a mechanical cell grinder for doping Na⁺ into MnO₂ and obtained a Na_0.44MnO₂ cathode, which had a large specific surface area [67]. In 2022, Li et al. reported a N-doped Na₂Mn₃O₇ cathode that displayed a capacity of 300 mAh·g⁻¹ at a current density of 0.2 A·g⁻¹, which even retained a capacity of 100 mAh·g⁻¹ at 10 A·g⁻¹ [55]. The modification principle was that the doping of sodium ions could enlarge the interlayer spacing and serve as a pillar to stabilize the host structure. Nitrogen doping could accelerate electron conduction in the electrodes and suppress the dissolution of Mn during repeated cycling.

In addition to Na⁺, K⁺ is usually used for manganese-based oxide doping. In 2019, α-K_0.19MnO₂ was synthesized through a self-sacrificing template method by Zhang et al. [34]. In order to determine the relationship between the potassium content of the host structure and the electrochemical performance, a concentrated HNO₃ treatment was employed to adjust the content of K ions in MnO₂. The results showed that the cathode with high potassium content possessed better electrochemical performance, shown in Figure 4a. Moreover, a K-salt was added to the electrolyte to prevent the deintercalation of K⁺ from the host material, further enhancing the battery performance (Figure 4b). Barpanda et al. reported that K_1.33Mn₈O₁₆ was prepared via a simple sonochemical technique, adding a new member to the family of manganese-based oxides [62]. Liang et al. reported a K⁺-stabilized K_0.8Mn₈O₁₆ with oxygen defects, whereby the incorporation of K⁺ for suppressing manganese dissolution was confirmed by ICP-OES (Figure 4c), and the oxygen defects improved the ability of electron transfer and opened the [MnO₆] octahedral walls for ion diffusion, as illustrated in Figure 4d,e [68]. In 2021, Pan et al. provided a new viewpoint that generating hybrid phases could enhance electrochemical performance, which was confirmed by a layered K_0.36H_0.26MnO₂·0.28H₂O (K36) cathode. Via a proton and Zn²⁺ co-intercalation mechanism, this cathode underwent a gradual phase transition from original K36 to hybrid layer-type K_xH_yZn_zMnO₂·nH₂O and spinel-type ZnMn₂O₄ nanocrystals after several cycles [36]. In 2022, Shi et al. reported a K_0.29MnO₂·0.67H₂O cathode with ultralong cycles, whose capacity remained at 158 mAh·g⁻¹ after 12,000 cycles at a high current density of 2000 mA·g⁻¹ [37].

Additional monovalent ion doping work has also been reported. In 2019, Ag_1.5Mn₈O₁₆ nanorods were prepared via a facile reflux method and successfully applied as a cathode in ZIBs by Takeuchi et al. [69]. A structurally stable newfangled phase of manganese-based oxide MnO₂H_0.16(H₂O)_0.27 (MOH) was designed by Pan et al. [35]. As shown in the chemical formula, the intercalation of protons and water molecules into the manganese oxides offered a possibility for improved performance. In 2022, a manganese dioxide (H_xMn₂O₄) with protons inserted was derived from ZnMn₂O₄ templates via a cation exchange method by Dai et al., which delivered a high capacity of 281 mAh·g⁻¹ at 100 mA·g⁻¹, as well as a good rate performance of 133.4 mAh·g⁻¹ at 1 A·g⁻¹ for 1000 cycles [45].

3.2. Multivalent Cation-Doped Manganese-Based Oxides

Multivalent ions are widely applied in the doping of manganese-based oxides because of their wide variety and variable valence state, providing more possibilities for doping modification. They are classified as cations or anions and summarized according to the presence of single-ion doping or multi-ion co-doping, where the valence state of dopant ions ranges from low to high.

In 2017, Banerjee et al. developed a Cu²⁺ plug-in Bi–δ-MnO₂ that delivered near-full dual-electron capacity after cycling more than 6000 times. The excellent electrochemical performance was attributed to the valence state changes of Cu and Bi during charge and discharge, as shown in Figure 5a [46]. In 2020, Cu–MnO and Cu–Mn₂O₃ were respectively reported by Hwang et al. and Zhang et al., both of which improved performance by introducing oxygen defects through copper doping [61,70]. Following the reports of Ni_xMn_3−xO₄ by Guo et al. [51] in 2019 and Ni-doped Mn₂O₃ by Huang et al. [52] in 2021, Pan et al. proposed a Grotthuss proton transport mechanism for the first time, which was verified by Ni doping of substitute Mn in α-MnO₂ [40]. According to this mechanism, a greater disordered degree of tetragonal–orthorhombic (TO) states results in better diffusion kinetics of protons in the tunnels. Ni dopants could promote the TO distortion of the lattice during discharge, resulting in the improvement of electrochemical performance, as illustrated in Figure 5b. A Zn²⁺-stabilized layer-type MnO₂ structure was reported by Wang et al. in 2019 [57], where hydrated Zn²⁺ worked as pillars to enhance structural stability. In 2020, Shao et al. reported a cobalt-modified δ-MnO₂ cathode that assisted the deposition–dissolution mechanism, and they proposed that cobalt doping played a role in catalysis [47]. In 2021, Wang et al. explained the effect and mechanism of Co with different valence states in Mn-based materials for zinc storage, which was confirmed by constructing a multivalent cobalt (Co²⁺, Co³⁺)-doped Mn₃O₄ nanosheet [50]. Tao et al. designed a Ca_0.28MnO₂·0.5H₂O cathode material in 2020, which exhibited an ultralong life of 5000 cycles with no apparent capacity decay [39]. In 2019, Du et al. found that cerium doping could transform β-MnO₂ to α-MnO₂, which was accompanied by a significant improvement in conductivity and stability [41]. In the same year, birnessite δ-MnO₂ nanoflowers with La³⁺ doping were reported by Lu et al. [71]. The intercalation of La³⁺ into δ-MnO₂ (LMO) could enlarge the interlamellar spacing and lower the resistance of Zn²⁺ (de)insertion, thereby remarkably enhancing the performance of LMO. A Bi-doped α-MnO₂ electrode was reported by Wang et al. in 2021 [42]. They insisted that Bi doping could catalyze the conversion of MnOOH to Mn²⁺, which improved the deposition and dissolution efficiency of manganese, attaining a more stable second discharge platform. Mesoporous Al_0.35Mn_2.52O₄ (AMO) was prepared through Al substitution in Mn₃O₄ by Shi et al. in 2021 [72]. As shown in Figure 5c, Al was first introduced into Mn₃O₄ to replace Mn sites through a simple hydrothermal method. Subsequently, cation defects were obtained by selectively etching Al away from Al_0.35Mn_2.52O₄. Thus, the electronic conductivity and diffusion kinetics of H⁺ and Zn²⁺ were both significantly improved as a result of the high specific capacity (302 mAh·g⁻¹ at 0.1 A·g⁻¹) and outstanding cycling performance (147 mAh·g⁻¹ after 1000 cycles at 1.5 A·g⁻¹). Yan et al. constructed an Al-intercalated MnO₂ cathode, proposing the critical role of Al doping, which was confirmed by DFT calculations and in situ Raman investigation [43]. Recently, an Fe-doped α-MnO₂ coated with polypyrrole was created by Kong et al. [73]. The doping of Fe³⁺ could effectively enlarge the lattice spacing of α-MnO₂ and improve the diffusion kinetics of Zn²⁺, which led to a better electrochemical performance. In 2019, Ti doping was introduced to build an oriented electric field by Mai et al., achieving impressive performance [60]. As early as 2017, V⁵⁺ doping was adopted by Kim et al. to improve the conductivity of manganese-based oxides [74].

While the single-ion doping modification strategy has achieved good results, dual-ion doping has also emerged. In 2020, a La–Ca co-doped ε-MnO₂ cathode was obtained through a simple one-step liquid coprecipitation process by Lu et al. [58]. Their experimental results revealed that both La³⁺ and Ca²⁺ played a role in enhancing capacity and reversibility. Wang et al. reported that Ni and Co co-substitution of spinel ZnMn₂O₄@N-rGO was an effective approach to suppress the Jahn–Teller distortion of Mn³⁺, thereby stabilizing the structure [75].

3.3. Anion-Doped Manganese-Based Oxides

In addition to multivalent cation doping, anion doping is regarded as an effective strategy, e.g., N, P, and S. N is always doped in a carbon-containing substrate and then compounded with manganese oxides to improve performance. In 2017, Li et al. reported a 3D porous MnO₂ supported by an ideal N-doped carbon cloth substrate, which delivered a high capacity of 353 mAh·g⁻¹ and remarkable cycle capability (93.6% retention after 1000 cycles) [76]. In 2018, a porous N-C@MnO_x cathode derived through MOF-modified manganese oxides was reported for the first time by Sun et al. [53]. The cathode material exhibited a superior electrochemical performance of 305 mAh·g⁻¹ at 500 mA·g⁻¹ after 1600 cycles at 2000 mA·g⁻¹, which benefited from several factors. On the one hand, the onion-shaped N-doped amorphous carbon layers established a high-speed path for electron conduction and maintained the stability of the material host structure. On the other hand, the porous structure of N-C@MnO_x accelerated electrolyte infiltration and ion transfer. In 2019, Mn₃O₄@N-doped carbon matrix composite nanorods and a ZnMn₂O₄/N-doped graphene nanocomposite were reported by Kong et al. and Yang et al., respectively [63,77]. In 2020, ultrathin MnO₂ nanoflakes grown on N-doped hollow carbon spheres were prepared by Yang et al., which displayed high capacity (349 mAh·g⁻¹ at 0.1 A·g⁻¹) and long-term cycling performance (2000 cycles at 2 A·g⁻¹) [78]. Compared with N doping of carbon-based modification layers, the introduction of N into the bulk phase for modification is relatively rare. In 2019, Xia et al. fabricated N–MnO_2−x by introducing N into the bulk phase of MnO₂ via ammonia calcination at 200 °C, achieving excellent electrochemical performance [54]. In 2020, oxygen vacancies and P ion doping were generated simultaneously by Zheng et al. in a P–MnO_2-x@VMG cathode via phosphorization [79]. The oxygen vacancies could increase the electrical conductivity of MnO₂, while the P ions were able to expand its interlayer spacing, thereby accelerating ion transport. Recently, sulfur-doped MnO₂ (S-MnO₂) nanosheets were applied as a cathode for ZIBs by Sun et al., which showed a high capacity (324 mAh·g⁻¹ at 0.2 A·g⁻¹) and excellent rate performance (205 mAh·g⁻¹ at 2 A·g⁻¹) [56].

4. Optimization Mechanism of Ion Doping in Zinc–Manganese Battery

Ion doping alters the behavior of electrode materials in a variety of ways. It is vital to produce a complete overview to develop better electrode materials and identify knowledge gaps for in-depth research in the future. As far as the current research progress is concerned, the positive effects of ion doping can be roughly divided into three categories: (1) enlarged interlayer spacing for improved ion diffusion kinetics, (2) defect engineering for enhanced electrical conductivity, and (3) pillar effect for enhanced stability of the host structure.

4.1. Enlarged Interlayer Spacing for Improved Ion Diffusion Kinetics

Theoretically, Zn²⁺ ions have a small ionic radius (0.74 Å) and high ionic conductivity in aqueous solution (~1–10 mS·cm⁻¹) [80]; however, in practice, due to their high charge density, Zn²⁺ ions combine with water molecules to form hydrated [Zn(H₂O)₆]²⁺, leading to an increment in ionic radius to 5.5 Å, slowing down the diffusion of Zn²⁺. Furthermore, the solid electrostatic effect between Zn²⁺ and the host structure of the cathode material also causes sluggish Zn²⁺ intercalation [81,82]. The diffusion rate of carriers has a linear negative relationship with the electrostatic repulsion (ƒ) between the carriers and the host structure. According to the formula $\mathrm{ƒ}\propto \frac{1}{{\epsilon }_r{r}_0^2}$, where ${\epsilon }_r$ is the permittivity and r₀ is the distance between Zn²⁺ and the closest ions, a larger value of r₀ means faster diffusion kinetics [21,83]. In other words, a larger layer spacing leads to better diffusion dynamics. Ion doping is an efficient strategy to expand the layer spacing of the cathode material, thus enhancing performance.

Kim et al. reported that V-doped MnO₂ (VMO) could enhance zinc storage properties by expanding the layer spacing [74]. The (211) peak in the X-ray diffraction (XRD) patterns of VMO showed a minor shift toward lower scanning angles, as shown in Figure 6a, confirming anisotropy of the unit cell parameters, which would facilitate the insertion of zinc ions. Lu’s group obtained a cathode with a larger interlamellar spacing by doping La³⁺ into δ-MnO₂ (LMO), which showed lower resistance of Zn²⁺ (de)insertion and better structural stability [71]. The rate performance of LMO was significantly improved (121.8 mAh·g⁻¹ at 1.6 A·g⁻¹) compared to pristine δ-MnO₂ (only 3.4 mAh·g⁻¹ at 1.6 A·g⁻¹). Zheng et al. reported that phosphate ion-doped MnO₂ could expand the interlayer spacing of the (001) plane from 0.68 nm to 0.70 nm, accelerating ion transfer. Simultaneously, oxygen vacancies were introduced via phosphorization, enhancing the electrical conductivity of MnO₂ [79]. Wang’s work revealed that the pre-intercalation of Bi³⁺ into α-MnO₂ could effectively enlarge the lattice spacing and have a positive effect on the ion diffusion rates, resulting in a superior rate performance with a capacity retention of 150 mAh·g⁻¹ at 5 A·g⁻¹ [42]. K_0.29MnO₂·0.67H₂O (KMO) with an interplanar spacing of 7.4 Å was synthesized via a simple hydrothermal strategy by Shi et al. [37], as shown in Figure 6b, exhibiting high capacity (300 mAh·g⁻¹ at 0.2 A·g⁻¹) and an ultralong cycle performance (158 mAh·g⁻¹ after 12,000 cycles at 2 A·g⁻¹). According to the XRD pattern in Figure 6c, this work calculated that the interlayer spacing corresponding to the (001) plane increased from 6.8 Å to 7.4 Å according to Bragg’s rule. The diffusion energy barriers of H⁺ and Zn²⁺ in MnO₂ and KMO were explored using density functional theory (DFT)-based first-principles calculations, and the results revealed a lower value of KMO (0.11 eV and 0.19 eV) than MnO₂ (0.32 eV and 0.49 eV), as shown in Figure 6d, indicating that the increased interlayer spacing indeed accelerated ion transfer. The kinetic behavior of the KMO sample was further investigated using the galvanostatic intermittence titration technique (GITT). As shown in Figure 6e,f, KMO displayed a smaller overpotential and higher diffusion coefficient than MnO₂ during the discharge process, which indicated that the doping of K⁺ indeed promoted ion diffusion kinetics.

4.2. Defect Engineering for Enhanced Electrical Conductivity

For secondary batteries, electron transfer between the cathode and anode is an integral part of completing the whole electrochemical reaction; hence, the electrical conductivity of the cathode plays an important role in the electrochemical performance [84]. However, manganese-based oxides are usually semiconductors with poor electrical conductivity [85]. The strategy of complexing with conductive agents is generally adopted to accelerate electron transfer, while ion doping is another method to enhance the electronic conductivity of cathode materials [86]. For example, a distinctive N-doped MnO_2−x cathode with numerous oxygen defects was prepared through NH₃ treatment at 200 °C by Xia et al. in 2019 [54]. Oxygen vacancies were introduced at the same time as N doping, which increased the electron density and lowered the bandgap of manganese dioxide, resulting in a better electronic conductivity and activity. As shown in Figure 7a, the position of the absorption edge corresponding to the oxidation of N-doped MnO_2−x in the XANES spectrum presented a shift toward a lower energy, indicating higher average electron density. On the other hand, the FT spectrum implied that N-doped MnO₂ did not change phase but increased its level of disorder. The DFT calculation results (Figure 7b) showed that N-doped MnO_2–x possessed a much smaller bandgap (0.12 eV) than pure MnO₂ (1.83 eV), revealing a significant enhancement of electronic conductivity. Excellent electrochemical performance was achieved that 285 mAh·g⁻¹ at 0.2 A·g⁻¹ with 85.7% retention after 1000 cycles at 1 A·g⁻¹. In the same year, Ti–MnO₂ with oxygen vacancies was reported by Mai’s group [60], indicating that the replacement of manganese with titanium and the introduction of oxygen vacancies could break through the manganese–oxygen octahedral walls, resulting in heterogeneous charge distribution. As revealed by the EIS spectrum (Figure 7c), Ti-doped MnO₂ exhibited lower charge migration resistance, confirming that the unbalanced local electric field in the host structure could boost the mobility of ions/electrons. Furthermore, according to the DFT calculations (Figure 7d), the electron cloud of Ti substitution and its derived oxygen vacancies could balance the disordered interfacial electric field, allowing electron transit through the [MnO₆] octahedral walls. Liang et al. proposed a K⁺-stabilized Mn-based cathode with rich oxygen defects (K_0.8Mn₈O₁₆ with oxygen defects), which exhibited impressive stability over 1000 cycles with no obvious fading [68]. As described in Figure 4d,e, oxygen defects could accelerate H⁺ diffusion by opening the [MnO₆] octahedral walls from the ab-plane. Moreover, the oxygen defects could reduce the energy for electron and charge transfer during the redox reaction, as illustrated in Figure 7e, KMO showed smaller overpotential gaps than pure MnO₂ (1.399/1.614 V vs. 1.389/1.612 V). More recently, Zhang et al. designed a cathode (O_cu–Mn₂O₃) by replacing sites of trivalent manganese with divalent copper ions to create oxygen defects in Mn₂O₃ for better electronic conductivity [70]. Long et al. fabricated a low-bandgap cathode (Ni_xMn_3−xO₄) via the replacement of Mn with Ni. The DOS indicated that Ni-doped Mn₃O₄ exhibited a narrower bandgap than pure Mn₃O₄ (1.20 eV), thereby significantly enhancing the electronic conductivity [51].

4.3. Pillar Effect to Stabilize the Host Structure

Manganese-based oxide cathode materials inevitably suffer from the dissolution of manganese caused by the Jahn–Teller effect, disproportionation, and irreversible phase transformation during the charge/discharge process [74,87,88], resulting in structural instability. This is essentially due to the unavoidable presence of some trivalent manganese during the charge/discharge cycle, and Mn³⁺ is extremely unstable because of its high-spin d⁴ (= t_2g³ − e_g¹) electronic configuration in a symmetric octahedron [21,89,90], tending to transform into Mn⁴⁺ and Mn²⁺. Compared to the Mn⁴⁺–O bond length (1.93 Å), the Mn²⁺–O/Mn³⁺–O bonds are longer (2.23 Å/2.045 Å), which weakens the Mn–O bonding strength and increases the disorder degree of the crystal structure [91]. After the insertion of Zn²⁺, the [Mn³⁺O₆] octahedrons or [Mn²⁺O₄] tetrahedrons can be easily extruded by the strong electrostatic repulsion between the carrier ions and host framework, resulting in manganese dissolution or irreversible phase transition, which is further reflected in the rapid capacity decay [21]. Ion doping is considered as an effective strategy to prevent material frameworks from collapsing, whereby guest ions can behave as pillars to help support the host structure.

Zhi et al. pre-intercalated Na ions and water molecules in δ-MnO₂ (δ-NMOH) to obtain a stable cathode [65]. As shown in Figure 8a, the layered structure of δ-NMOH consisted of two parts; one part was composed of manganese–oxygen octahedra connected by edge-sharing to form a layer, and the other part was composed of sodium ions and water molecules acting as pillars in the middle of the layer. Due to the stabilizing effect of Na⁺ and H₂O molecules, when H⁺ and Zn²⁺ were inserted/extracted, the host structure could only expand or shrink the interlayer spacing, whereas no change in crystal structure would occur, resulting in better structural stability. An impressive cycle life of 10,000 cycles with 98% retention was measured (Figure 8b) [65]. Wang et al. constructed Zn²⁺ ion-stabilized MnO₂ (ZMO) nanospheres for a cathode with a long lifespan. The ZMO electrode manifested a rate capacity of 124 mAh·g⁻¹ at 3.0 A·g⁻¹, attributed to its high surface area and unique mesoporous texture for sufficient active sites and electrolyte permeation. Furthermore, an outstanding cyclability over 2000 cycles was achieved by the layered-type structure filled with zinc ions for enhanced stability [57]. Zhang et al. reported a K⁺-doped α-MnO₂ (α-K_0.19MnO₂) cathode with a high K content, which exhibited high rate performance (20 °C, 113 mAh·g⁻¹) [34]. As shown in Figure 8c, K⁺ was still located in the α-MnO₂ tunnel after the water-solvated H⁺ was inserted into the host framework; while the subsequent intercalation of Zn²⁺ caused a transformation of MnO₂ from the tunnel to the layer, the pre-intercalation K⁺ remained in the interlayer to stabilize the structure and expand the channel for zinc-ion migration. In 2021, a multivalent cobalt-doped Mn₃O₄ (Co–Mn₃O₄) with an outstanding cyclability over 1000 cycles at 2000 mA·g⁻¹ was reported by Wang et al. [50]. Co ions with different valence states played important roles in the phase change charge product (Co–MnO₂) and discharge products (Co–ZnMn₂O₄ and Co–MnOOH). The Co²⁺ ions doped between the [MnO₆] octahedrons on both sides of the δ-MnO₂ layers acted as an interlayer pillar due to their stronger adsorption energy, as shown in Figure 8d.

5. Summary and Perspectives

AZIBs have great application prospects in the field of large-scale energy storage due to their safety, environmental friendliness, and low cost. Among various cathode materials, manganese-based compounds are most likely to be commercialized due to their high capacity and suitable potential. However, the low electrical conductivity and structural instability of manganese-based materials caused by the Jahn–Teller effect greatly limit their development. It is well known that the structure of a material determines its properties. Due to modification of the crystal structure, the ion doping optimization strategy is regarded as the most effective approach for the construction of high-performance manganese-based cathodes. In this review, on the basis of the redox mechanism, we divided the synthetic principles of ion doping strategies into four categories and summarized them in detail so as to inspire the design of synthetic routes. According to the summary of the different valence states of the doping ions, we believe that, even though both monovalent ion and multivalent ion doping can improve the performance of manganese-based materials, it has been reported that only multivalent ion doping can replace manganese sites or play a catalytic role in the process of zinc storage [47,52,60,72], which shows that multivalent ions present more possibilities than monovalent ions. We also focused on the effect of ion doping strategies on the electrochemical performance of manganese-based cathode materials and provided a detailed summary of their research progress. Regardless of the type of dopant ion, such as cation vs. anion or monovalent vs. multivalent ion, the interlayer spacing can be expanded, the ion transport rate can be improved, defects can be introduced for high electronic conductivity, and a pillar can be established to stabilize the structure, thereby enhancing the cycle performance and rate capability of the ZIBs. Despite many successful cases demonstrated in previous studies, challenges and opportunities for the ion doping modification strategy still exist, which deserve further in-depth research and discussion.

Firstly, the identification of the exact potion of the intercalated ions in the MnO_x crystal structure is crucial in understanding the chemical state of Mn, as well as the effect of the intercalated ions on the zinc/proton intercalation into the host structure.

Secondly, a deep understanding of the influence of the pre-intercalated ions in the manganese-based oxides is required through in situ characterizations, including in situ XRD, in situ TEM, in situ Raman spectroscopy, and in situ FT-IR spectroscopy, to reveal the fundamental mechanism.

Thirdly, although ion doping greatly improves the electrochemical performance of cathode materials, the current status of manganese oxide cathodes is still in its infancy, operating at relative low mass loading (below 1 mg/cm², corresponding to an areal capacity of 0.1–0.2 mAh/cm²); hence, we are far from the realizing the commercialization of zinc-ion batteries, requiring an areal capacity of over 2 mAh/cm². In this sense, a mass loading above 10 mg/cm² is highly needed to achieve a practical zinc-ion battery. Additionally, in terms of a high-mass-loading cathode, the utilization of a zinc metal anode is another typical challenge to achieving a high energy density of the zinc anode.

Lastly, most synthesis methods are too complex and expensive for large-scale industrial production; therefore, feasible and low-cost doping methods should be urgently developed on the basis of the actual conditions.