Nanostructured MnO₂ as Electrode Materials for Energy Storage

Abstract

Manganese dioxides, inorganic materials which have been used in industry for more than a century, now find great renewal of interest for storage and conversion of energy applications. In this review article, we report the properties of MnO₂ nanomaterials with different morphologies. Techniques used for the synthesis, structural, physical properties, and electrochemical performances of periodic and aperiodic frameworks are discussed. The effect of the morphology of nanosized MnO₂ particles on their fundamental features is evidenced. Applications as electrodes in lithium batteries and supercapacitors are examined.

1. Introduction

In recent years, intense Research & Development input on nanotechnology has delivered nano-objects (particles with size ≈ 100 nm or less) that possess a rich combination of physical properties, inasmuch as they differ from those of the bulk and depend on polymorphism, morphology, size of particles, size distribution, coating, and the precursor used in the synthesis [1]. With the need for renewable energies, these nano-substances have undergone extensive research, in order to develop new systems that can be used for energy storage and/or conversion. Among them, transition-metal oxides including TiO₂, MnO₂, V₂O₅, etc. are stable and robust materials with tunable properties offering large surface areas.

In the early ages, mineral manganese dioxides (MDOs) were used as black pigments for rock-art painting in paleolithic caves of the Magdalenian culture [2]; they can be considered as the first nanomaterials used up until then by human civilization. Today MnO₂ is an important functional metal oxide, which is technologically attractive for applications in different fields such as catalysts [3,4,5], absorbent of toxic metals [6], ion-sieves, molecular-sieves [7], artificial oxidase [8], component of the dry cell (Leclanché cell) [9], inorganic pigment in ceramics, electrodes for electrochemical batteries (lithium, magnesium, sodium) [10,11,12,13], and electrodes for supercapacitors [14,15]. MnO₂ has also been widely used in Duracell (alkaline) based barriers, photocatalytic activities, and electrolysis. Owing to its ability to absorb toxic ions, MnO₂ has been also found to have important applications in water-cleaning [16,17]. MDOs are non-stoichiometric compounds, because of inevitable structural water molecules that are physisorbed.

The engineering of manganese oxides used for energy storage and conversion has become more and more important to the point where a huge number of works is devoted to these materials. The great interest of MnO₂ as an inorganic material in the battery industry is due to its theoretical capacity (308 mAh·g⁻¹ comparable to 270 mAh·g⁻¹ of LiCoO₂) and capacitance (1370 F·g⁻¹), natural abundance, low cost, and low toxicity [1]. MDOs crystallize with various morphologies and crystallographic forms including the α-, β-, γ-, δ- ε-, λ- and R-polymorphs, which are naturally occurring minerals such as hollandite (2 × 2), pyrolusite (1 × 1), nsutite (1 × 1)/(1 × 2) with hexagonal (hex.) structure, birnessite (1 × ∞), akhtenkite (dense stack), spinel (1 × 1), and ramsdellite (1 × 2), respectively, where (m × n) denotes the tunnel dimension. The polymorphism is due to the different ways of linking the MnO₆ octahedral architectonic units through corner- or edge-sharing that show variations in the chain and tunnel structures (see Figure 1) [18]. For example, the (3 × 3) tunnel structure of todorokite-type MnO₂ (octahedral molecular sieve labeled OMS-1) has a pore size of about 6.9 Å [7]. Note that the hollandite, cryptomelane, and coronadite minerals that are structurally related to (2 × 2) tunnel materials have water and different cation contents such as Ba²⁺, K⁺, and Pb²⁺, respectively [19]. Consequently, properties of MDOs depend strongly on the crystalline structure, particle size, and morphology. The crystallographic data of some MDO compounds are summarized in

Table 1. Summary of crystallographic data of some manganese dioxide (MDO) compounds.

Compound	Mineral	Crystal Symmetry	Lattice Parameters (Å)	Features
α-MnO2	hollandite	tetragonal (I4/m)	a = 9.96; c = 2.85	(2 × 2) tunnel
R-MnO2	ramsdellite	orthorhombic (Pbnm)	a = 4.53; b = 9.27; c = 2.87	(1 × 2) tunnel
β-MnO2	pyrolusite	tetragonal (P42/mnm)	a = 4.39; c = 2.87	(1 × 1) tunnel
γ-MnO2	nsutite	complex tunnel (hex.)	a = 9.65; c = 4.43	(1 × 1)/(1 × 2)
δ-MnO2	birnessite	rhombohedral (R-3m)	ahex = 2.94; chex = 21.86	(1 × ∞) layer
Mg-Bir	Mg-birnessite	monoclinic (C2/m)	a = 5.18; b = 2. 84; c = 7.33	(1 × ∞) layer
Na-Bir	Na-birnessite	monoclinic (C2/m)	a = 5.17; b = 2.85; c = 7.32	(1 × ∞) layer
ε-MnO2	akhtenkite	hexagonal (P63/mmc)	a = 2.85; c = 4.65	dense stack
λ-MnO2	spinel	cubic (Fd3m)	a = 8.04	(1 × 1) tunnel
ψ-MnO2	psilomelane	monoclinic (P2/m)	a = 9.56; b = 2.88; c = 13.85	(2 × 3) tunnel
T-MnO2	todorokite	monoclinic (P2/m)	a = 9.75; b = 2.85; c = 9.59	(3 × 3) tunnel

.

Since MnO₂ was introduced as a depolarizing element in zinc-alkaline cells, a major step was realized with the utilization of natural and synthesized materials such as electrochemically (EMD), chemically (CMD), and heat-treated (HTMD) prepared MnO₂. All these frameworks are generally classified as members of the nsutite group: the γ-MnO₂ and ε-MnO₂ phases, distinguished by the quality of their X-ray diffraction diagrams [9]. Direct synthesis of EMD or CMD results in structural defects: (i) the “De Wolff defects” (denoted Pr) are intergrowth of pyrolusite in the ramsdellite matrix and (ii) the micro-twinning defects (denoted Tw). Figure 2 shows the degree of micro-twinnings as a function of the pyrolusite intergrowth in synthesized CMD, EMD, and HTMD manganese dioxides. In commercial EMD powders, Pr and Tw are close to 45–55% and 80–100%, respectively. Being relatively cheap, EMD materials are widely employed as electrodes in primary alkaline batteries and supercapacitors. EMD powders with manganese-cake microstructure (predominantly γ-MnO₂ phase) deliver a specific discharge capacity of 280 mAh·g⁻¹ using 9 mol·L⁻¹ KOH electrolyte [20]. An extensive review devoted to electrolytic MnO₂ has been published recently [21]. Therefore, we simply direct the readers to this review for specific properties of EMD, and attention in the present work is thus focused on other syntheses and related MnO₂ in the other α- and β-phases [21]. For the same reason, we did not detail any discussion on the MnO₂-based supercapacitors, because a 43-page review on them, including experimental aspects and discussion, prospective, has been recently published [22]. Consequently, we made the choice to focus the present work on the other major application of the MDOs, namely their use as active cathode elements of Li-ion batteries, in relation to the synthesis, structural, and morphological aspects of the nano-sized particles. Special attention has been paid to the synthesis aspect of the MDOs because all their electrochemical properties strongly depend on the preparation process that also determines the morphology of the nanostructure. Some properties of the MnO₂-based supercapacitors are also discussed in this context, and we direct the reader to the review [22] for more details.

In this review article, we investigate the properties of MnO₂ nanomaterials with different morphologies. Their structural and physical properties are reported. Periodic (α-, β-, and R-MnO₂) and aperiodic (γ-MnO₂) structures are considered. In Section 2, we briefly discuss the beneficial effect of nanosizing. In Section 3, we summarize the techniques used for the synthesis of nanomaterials as the structure and morphology of MnO₂ are related to the synthesis conditions (reagents, temperature, pH, etc.). Section 4 is devoted to the electrochemical features of bulk MDOs. In the following Section 5, Section 6, Section 7 and Section 8, we report the properties of the various nanostructures (nanourchins, doped MnO₂ nanomaterials, polypyrrole-coated MnO₂, nanocomposites). For each material, the effects of the morphology of nano MDOs on their physical and electrochemical performance are evidenced. Applications such as electrodes in lithium batteries and supercapacitors are examined.

2. Beneficial Effect of Nanosizing on Transport Properties

One major reason for the use of nanosized particles of materials for energy storage comes from their poor transport properties that imply poor rate performance of these electrochemical devices. This is the case for oxides used as electrodes in batteries and supercapacitors. For example, the electronic conductivity of MnO₂ is ≈10⁻⁸ S·cm⁻¹ at room temperature, which requires some sophisticated technology such as the use of slurry containing carbon (carbon “Super P”, acetylene black, rGO, CNTs, etc.) or deposition at the surface of the grains for enhanced charge carrier transport [23]. Achieving high rate capability depends ultimately on the geometry of the active objects building the positive (cathode) and negative (anode) electrodes. The performance of an electrode is governed by the transport of both electrons and ions; consequently, the ionic and electronic conductivity of the materials must be considered.

Let us examine the ionic and electronic transport properties of particles as a function of size L (Figure 3a). As the motion of ions is a diffusion process, the characteristic time τ for an ionic species i (in practice Li⁺ ions in the present case) to reach the surface of any active particle of dimension L is given by the second Fick’s law that applies the chemical diffusion coefficient with D* of moving ions [2]. In the case when the chemical reaction proceeds by a single phase (sp) process, i.e., within a solid solution, τ is given by:

$${\tau }_{sp}=\frac{L^2}{4\mathsf{\pi }D*},$$

(1)

In the case of a two-phase (tp) process, for which there is a separation between a Li-rich and Li-poor phase instead of a solid solution, the chemical reaction proceeds by nucleation of the phases and motion of the propagation of the boundaries that separate the two phases. In this case, the characteristic time is given by:

$${\tau }_{tp}=\frac{F^2}{2V_m}\frac{L^2}{\langle {\sigma }^i\rangle \Delta {\mu }^i}$$

(2)

where V_m is the molar mass of the active compound, σⁱ the ionic conductivity, and Δμⁱ the difference of the chemical potential of ions between the two phases. Note that in both cases τ is proportional to L². Therefore, by reducing the size L of the active particles of electrodes from micrometer to nanometer, one reduces τ for the diffusion of ionic species in the solid-state phase by a factor of 10⁶. A decrease of τ corresponds to a minimization of the charge duration of the battery. For example, let us consider the case of the Li_xMnO₂ electrode in which Li⁺ ions are moving in the tunnel of an EMD framework. The chemical diffusion coefficient of the lithium in this case is D* = 3.4 × 10⁻¹² cm²·s⁻¹ at room temperature with an activation energy of 30 kJ [24], so that, at a fast rate of extraction of Li⁺ ions from the host lattice, the charge is achieved in 27 h for a 10-µm particle size. However, this time will be reduced to 4 min for a 500-nm particle size. Thus, 500-nm sized particles can be fully charged/discharged even at 4C rate (0.7 A·g⁻¹). However, in addition to the effects of size and distribution of particles on the insertion reaction mechanism, the effect of the high specific surface area of nanoparticles (safety problems) and the minimization of the volume expansion upon Li-ion insertion should be considered.

The second transport parameter is the electronic conductivity σ_e of the particles, which monitors the rate capability of the electrode, i.e., overpotential. The electronic current flowing through a particle is a function of electron diffusivity expressed by the Einstein relation:

$$D_e=\frac{k_{B}T}{\left|e\right|}{\mu }_e,$$

(3)

where µ_e is the electron mobility, T the absolute temperature, k_B the Boltzmann constant, and |e| the elementary charge. Let us consider the case of electrons flowing through one particle between the current collector and the neighboring one with a small contact area ε_c (Figure 3b). According to the ohmic law, the resistance of the particle R_e is given by

$$R_e=\frac{L}{{\sigma }_e{\epsilon }_c}$$

(4)

Consequently, the resistance can be reduced by decreasing the particle size to increase the surface area of the material, provided that the whole surface area of the particles can be considered as the contact area with the collector. To reach approximately this goal, the MnO₂ particles can be coated with a conductive material [25,26]. This will be discussed in Section 7 and Section 8.

3. Synthesis of MnO₂ Nanomaterials

Several routes are currently used for the synthesis of MnO₂: electrochemical methods [21,27,28] and eco-friendly wet-chemical [29,30] techniques have been reported. As already mentioned in the introduction, samples prepared by the electrochemical method lead to γ- and ε-MnO₂ and their synthesis has been reported in a recent review [20]. We thus focus attention in this section on the other synthesis routes. Various strategies have emerged as new methods to synthesize nanostructured MnO₂ samples with different controlled-morphologies (shape and size). The synthetic methods for nanostructured MnO₂ include simple reduction, coprecipitation, sol-gel, thermal decomposition, and the hydrothermal synthesis molten-salt method (see [31] for a review). Various nano-structured objects investigated over the years include nanowires [32,33,34], nanorods [35,36], nanoflowers [37], nanosheets [38], nanoflakes [39,40], nanotubes [41,42], nanourchins [43,44], nanospheres [45], nanobelts [46], nanodisks [47], and nanofibers [48]. The various morphologies of nanostructure MnO₂ materials are illustrated by the SEM images in Figure 4. On the basis of classical synthesis, MDOs are classically prepared by oxidation of aqueous Mn²⁺ solution using various oxidants such as MnO⁴⁻, S₂O₈²⁻, H₂O₂, O₃, ClO³⁻, Cr₂O₇²⁻, etc. It has been experimentally shown that the size and morphology of particles depend on the nature of the oxidant and the pH of the mixture. For example, the crystallization domains of the MnO₂ structures in the pH/synthesis temperature diagram are shown in Figure 5 [49]. In the following, we report the different routes used for the synthesis of nano-MnO₂ including examples of the literature.

3.1. Redox Reaction

Several processes, in which Mn⁷⁺ is reduce to Mn⁴⁺, were used to grow α-MnO₂ nanocrystals. However, the preparation of polymorph MnO₂ from reactions of MnO₄⁻ and Mn²⁺ are known to be critical on the nature of precursors [50]. The most popular route consists of the reduction of KMnO₄ by salts or organic substances. As an example, with manganese acetate as oxidant of potassium permanganate, the simple redox reaction can be expressed by:

3Mn(CH₃COO)₂ + KMnO₄ + 2H₂O → 5 MnO₂ + 4CH₃COOH + 2CH₃COOK.

(5)

Ragupathy et al. [45] prepared MnO₂ nanospheres by reduction of KMnO₄ by aniline. The mole ratio of KMnO₄ to aniline was 2:1. The as-synthesized amorphous MnO₂ converts into crystalline α-form upon annealing at temperatures <400 °C. Hashem et al. synthesized two polymorph MnO₂ nanorods using redox reaction between (NH₄)₂S₂O₈ and MnSO₄·4H₂O for α-MnO₂ and (NH₄)₂S₂O₈ and Mn(NO₃)₂·4H₂O for β-MnO₂ [51]. Rod-shaped structures of α-MnO₂ and β-MnO₂ are shown in Figure 6. The diameters of these rods are in the range of 15–20 nm for α-MnO₂ and β-MnO₂ samples, respectively.

Single-crystalline α-MnO₂ nanorods were synthesized by a hydrothermal method based on the redox reactions between the permanganate anion MnO₄⁻ and H₂O in a mixture containing KMnO₄ and HNO₃ [52]. These results are consistent with the investigations of Yin et al. [53] who studied the effects of metal cations and protons on the structures and morphologies of MnO₂. K⁺ and H⁺ are competitive in solution to form: (i) cryptomelane α-MnO₂ is formed when the amount of K⁺ is higher than the amount of H⁺; (ii) for the growth of the pyrolusite structure β-MnO₂ occurs at a higher quantity of H⁺; (iii) the layered phase δ-MnO₂ is obtained at a concentration of K⁺ much greater than that of H⁺. Liu et al. [54] successfully prepared nanosheets (typical thickness 1 nm) using a slow redox reaction between KMnO₄ and sodium dodecyl sulfate (SDS) in acidic medium (diluted H₂SO₄), in which SDS served as the precursor to reduce KMnO₄. Jeong and Manthiram [55] mentioned the preparation of MnO₂ by the reduction of KMnO₄ using various inorganic reducing agents such as potassium borohydride, sodium dithionate, and sodium hypophosphite. MnO₂ and Pb, Ni-mixed MnO₂ were prepared at room temperature by the reduction of KMnO₄ with Mn/Ni/Pb acetate solutions [56]. The solid-state reaction between Mn⁷⁺ and Mn²⁺ in high-energy ball milling was successfully applied to grow α-MnO₂ nanorods doped with different metal M²⁺ cations (M = Cu, Co, Ni, and Zn). The synthesized samples exhibited excellent textural characteristics, i.e., BET surface area of ~128 m²·g⁻¹ and pore size of ~8 nm [57].

3.2. Thermal Decomposition

Lee et al. prepared K_xMnO_2+δ·nH₂O and amorphous MnO₂ materials by direct thermal decomposition of finely ground KMnO₄ powders at T in the range 350–100 °C which contained a large amorphous/crystalline ratio [58]. Komaba et al. [59] fabricated K_xMnO₂ powders (δ-structure, x = 24.7 wt %) synthesized by simple decomposition of KMnO₄ at 300–800 °C in air. Further washing in 1 mol·L⁻¹ HCl aqueous solution reduced the potassium content to x = 0.26 wt %. Layered-type δ-MnO₂ nanoflake-like particles were prepared via the thermal decomposition of KMnO₄ according to the reaction:

5KMnO₄ → K₃MnO₄ + K₂MnO₄ + 3MnO₂ + 3O₂,

(6)

The MDO sample was obtained after heating at 350 °C for 5 h [60].

3.3. Hydrothermal Route

The nature of the MDO nanomaterials formed by this method depends on temperature, fill level in the pressure vessel, and solvent. By simply tuning of the hydrothermal reaction time of the decomposition of KMnO₄ and MnSO₄·H₂O in heated aqueous solution, Subramanian et al. [61] obtained different nanoarchitectures of MnO₂ particles by changing the hydrothermal time from 1 to 18 h. For the mixture heated at 140 °C for 1 h, they reported the formation of flowerlike nanowhiskers of MnO₂, which transformed to α-MnO₂ nanorods after 12 h. Xiao et al. [62] prepared three types of MnO₂ nanostructures: microsphere/nanosheet core–corona hierarchical architectures, one-dimensional (1D) nanorods, and nanotubes, employing a simple hydrothermal process in an autoclave heated at different temperatures (100–200 °C) for the same duration (12 h). In a typical synthesis, the hydrothermal decomposition of single KMnO₄ to produce the α-MnO₂ phase occurs in acidic conditions in the absence of templates or surfactants: concentrated H₂SO₄ or HCl (37 wt %) were added to deionized water [63]. The nanosized birnessite-type δ-MnO₂ (monoclinic, C2/m) was formed at 100 °C, while pure α-MnO₂ nanorods (tetragonal, I4/m) crystallized at 120 °C, transforming to α-MnO₂ 1D nanotubes at 140 °C. Single-crystal β-MnO₂ nanotubes (200–500 nm diameters) were prepared by a simple hydrothermal method by oxidizing MnSO₄ with NaClO₃ in the presence of poly(vinyl pyrrolidone) (PVP) [42]. Wang et al. [63,64] synthesized various nanostructured MnO₂ polymorphs (α-, β-, γ-, and δ-forms) using a common hydrothermal method with pH and NH₄⁺ cation concentration adjustment. All nano-samples have a similar formation of layered δ-MnO₂. β- and γ-MnO₂ grown as nanowires/nanorods (Figure 7), while α-MnO₂ and todorokite-type MnO₂ have fiber or needle morphologies [43]. Well-crystallized nanorods of α-MnO₂ (12 nm diameter) were prepared by the hydrothermal route in the presence of poly(sodium 4-styrene-sulfonate) using KMnO₄ and MnSO₄ mixed in a solution of water and ethanol (4:1 in volume) [65]. A polyethylene glycol (PEG) polymer-precursor route was employed to prepare γ-MnO₂ nanowires/nanotubes. In a typical synthesis, MnSO₄ aqueous solution was mixed with PEG-6000 dissolved in aqueous methanol solution forming the precursor to which NaOH was added. The nanotubes were obtained after treatment in an autoclave at 120 °C for 20 h [66]. Ma et al. [67] synthesized layered MnO₂ nanobelts by the hydrothermal treatment of Mn₂O₃ powders in an aqueous solution of NaOH at 170 °C for >72 h. The structure of nanobelts is characterized by a basal spacing of ~7.1 Å indicating the transformation of Mn₂O₃ to δ-MnO₂.

Cheng et al. [68] synthesized α-MnO₂ nanowires on the basis of the hydrothermal reaction between KMnO₄ and MnSO₄·H₂O in aqueous solution at 140 °C for 12 h, and γ-MnO₂ nanowires from the mixture of MnSO₄ and (NH₄)₂S₂O₈ treated at 90 °C for 24 h. Cryptomelane-type manganese dioxide (α-K_xMnO₂) nanofibers with typical diameters of 20–60 nm and lengths of 1–6 μm were grown by reacting KMnO₄ with MnSO₄ under hydrothermal conditions (140 °C for 12 h) [69]. The nanofibers crystallize in a body centered tetragonal structure (space group I4/m) with unit cell parameters a = 9.8241(5) Å and c = 2.8523(1) Å. Their actual composition is K_0.11MnO_2.07.

3.4. Refluxing Route

The refluxing method allows in situ sample crystallization and requires ambient conditions of atmospheric pressure and temperature below the boiling point of solvent. This method requires heating a solution with an attached condenser preventing loss of reagents. However, many organic compounds have low boiling points and will vaporize upon exposure to such high heat [70]. Reflux treatment is convenient for large-scale preparation. The one-step direct refluxing route to synthesize α-MnO₂ consists of the reduction of KMnO₄, NaCr₂O₇ or KClO₃ using inorganic or organic acid as additive [71]. Wang et al. [72] prepared λ-MnO₂ nanodisks by the refluxing technique using Mn(Ac)₂·4H₂O and polyvinyl pyrrolidone (PVP) in dimethyl sulfoxide (DMSO) solution. The success of this synthesis is due to the synergic control of the surfactant (PVP) and the solvent (DMSO) that promote MnO₂ nanoparticles. A reflux treatment of KMnO₄ and MnSO₄ in HNO₃ acidic solution was used to synthesize single-crystalline β-MnO₂ nanorods. Cui et al. [36] reported that the dimensions depend on the acidity of the solution: nanorods exhibited diameters of 20–50 nm and lengths that ranged from approximately 0.5 to 2.0 μm with decreasing HNO₃ concentrations from 0.8 to 0.1 mol·L⁻¹. Cryptomelane-type α-MnO₂ nanofibers with particle sizes as small as 6 nm were synthesized on the basis of the reduction of KMnO₄ by H₂O₂ under acidic conditions followed by reflux. The particle size and crystallite size were adjusted by varying the pH of the mixture using an acetate-containing buffer solution and HNO₃ [73].

3.5. Catalytic Reaction

A homogeneous catalyst can reduce the potential energy of a chemical reaction (for example, oxidation of MnSO₄) and control the growth of oxide materials [74]. The synthesis of birnessite-type MnO₂ electrode for supercapacitors was realized by in situ electrochemical oxidation of Mn₃O₄ films composed of nanowall arrays with porous structure [75]. The preparation of a core-shell structure (α- and β-MnO₂ forms) with spherically aligned nanorods by a simple room-temperature solution-based catalytic reaction using AgNO₃ was reported. The catalyst dissolved in aqueous solution was added to the aqueous solution of MnSO₄·H₂O and (NH₄)₂S₂O₈ with concentrated sulfuric acid (98%). A suitable amount of acid promoted the formation of intermediate Mn³⁺ that disproportionated to α-MnO₂ and Mn²⁺ [76].

3.6. Sol-Gel Route

In this method, the gel is generated through a redox reaction between KMnO₄ and a carboxylic acid, i.e., citric, tartaric, fumaric acid, etc. [44]. The formation of MnO₂ tunnel structures is known to be controlled by adjustment of the pH, with H₃O⁺ and/or H₂O [77,78]: the growth of α-MnO₂ is favored in aqueous concentrated acid [72], whereas δ-MnO₂ is formed in aqueous concentrated base [73]. Oxidation of Mn²⁺ cations (in MnSO₄) by S₂O₈²⁻ anions (in (NH₄)₂S₂O₈) in aqueous solution without catalysts is achieved according the chemical reaction:

MnSO₄ + (NH₄)₂S₂O₈) + 2H₂O → MnO₂ + (NH₄)₂S₂O₄ + 2H₂SO₄

(7)

Nanowires are obtained by this technique [32]. An alternative method uses manganese sulfate MnSO₄ and potassium peroxodisulfate K₂S₂O₈ as starting materials [79]. Oaki and Imai [80] prepared δ-MnO₂ nanosheets (10 nm thick) using a sol-gel method assisted by ethylene diamine tetra-acetate (EDTA) as chelating agent. The reaction started upon mixing solutions containing Mn²⁺/EDTA and NaOH (basic solution). A low-temperature sol-gel process using manganese acetate tetra-hydrate (CH₃COO)₂Mn·4H₂O and concentrated nitric acid (67 wt %) associated with different surfactants in ethanol solvent was applied. MnO₂ nanowires and nanorods were formed with the assistance of cetyltrimethyl ammonium bromide and polyvinyl pyrrolidone, respectively [81]. Figure 8 shows the typical X-ray diffraction pattern of the α-K_xMnO₂ structure synthesized by the sol-gel route [43].

3.7. Co-Precipitation Method

This technique, which offers advantages such as simple and rapid preparative synthesis, as well as easy control of particle size and composition can be achieved without surfactant. Currently, the co-precipitation process is performed by using manganese salts with two different anions in equal concentration, such as manganese(II) sulfate and manganese oxalate for example. The pH must be adjusted to 12 by addition of NaOH to obtain brown precipitates that are MnO₂ precursors [82]. Nanoneedles were obtained using MnCl₂ mixed with isopropanol heated at ≈80 °C in a refluxing process and KMnO₄ dissolved in distilled water [83]. A simple co-precipitation of MnO₂ was achieved by mixing aqueous solutions of KMnO₄ and (MnSO₄·H₂O), where KMn^VIIO₄ is used as the oxidizing agent for Mn^IISO₄ in distilled water [84]. The KMnO₄:(MnSO₄·H₂O) molar ratio of 2:3 leads to a dark brown precipitate with the final chemical formula K_0.02MnO₂H_0.33·0.53H₂O [85].

Nanostructured α-MnO₂·nH₂O (BET = 303 m²·g⁻¹) obtained by precipitation of KMnO₄ and Mn(II) acetate in aqueous solutions was also reported [86]. MnO₂ nanosheets used as artificial enzymes (nano-oxidases) were obtained by exfoliation of bulk material prepared by precipitation of MnCl₂·4H₂O and a mixture of tetramethylammonium hydroxide (TMA·OH) and H₂O₂ in aqueous solution [8]. α-MnO₂ nanosheets were also formed by co-precipitation of MnCl₂·4H₂O and MnC₂O₄·2H₂O in aqueous solution with addition of sodium hydroxide to control the pH of the solution [87]. Hydrothermal synthesis was applied to grow two types of MnO₂ nanorods: (i) rutile-type β-MnO₂ due to a redox reaction (135 °C for 12 h) between manganese(II) sulfate with ammonium persulfate and (ii) hollandite-type α-K_xMnO₂ (x = 0.15 and 0.18) due to the decomposition of potassium permanganate obtained in the presence of sulfuric acid added to water after stirring to form a solution (150 °C for 8 h) [88]. The EPR spectra of α-K_xMnO₂ nanorods contain two signals. One is attributed to Mn⁴⁺; the other one to manganese in mixed-valence Mn⁴⁺/Mn³⁺ environment close to K⁺ ions [89]. Urchin-like α-MnO₂ materials were also prepared through a simple precipitation reaction of H₂SO₄ and KMnO₄ in aqueous solution heated at 85 °C [90]. A two-step green precipitation route was used for the reduction of KMnO₄ in the presence of natural extracts such as extracts of grape stems and apple peels to initiate the nucleation process [91]. TEM images showed the presence of short and long nanorods with diameters in the range 28–70 nm and lengths in the range 85–180 nm, which form dense agglomerates. Another green synthesis pathway consists of the mixture of manganese acetate salt as precursor and methanolic extract of phyllanthus amarus plant as reducing agent stabilized by curcumin extracted from turmeric [92].

3.8. Oxidation Reaction in Alkaline Conditions

Jana et al. [93] reported the fast synthesis of rod-shaped MnO₂ nanoparticles by the reaction between MnCl₂·4H₂O and sodium dodecylbenzene sulfonate NaC₁₈H₂₉SO₃ at room temperature in aqueous solution in alkaline conditions adding NaOH solution. The MnO₂ nanorods and nanospherical particles were grown from variable surfactant concentrations. These nanorods had additional reaction in an aqueous solution of AgNO₃ to become Ag-doped MnO₂ nanoflowers [37]. Song et al. [44] fabricated urchin-like α-MnO₂ by the reaction of MnSO₄ and KClO₃ using the sodium dodecyl sulfate (SDS)-assisted hydrothermal route.

3.9. Oxidation Reaction in Acidic Conditions

α-K_xMnO₂ was prepared using potassium permanganate KMnO₄ under acid conditions dissolved in distilled water with various acids, for example, hydrochloric acid solution [41]. Kijima et al. [77] synthesized nanoparticles of α- and γ-MnO₂ by an ozone-oxidation method in acidic medium. The α-MnO₂ phase was produced at high H₂SO₄ concentrations and high reaction temperatures using three pairs of manganese-salt-hydrates and acid, i.e., MnSO₄·5H₂O and H₂SO₄, Mn(NO₃)₂·6H₂O and HNO₃, MnCl₂·4H₂O and HCl. In contrast, the γ-MnO₂ phase was obtained by the ozone-oxidation of MnSO₄ dissolved in lower H₂SO₄ concentrations at lower temperatures, again followed by ozone oxidations of Mn(NO₃)₂ dissolved in HNO₃ or MnCl₂ dissolved in HCl.

3.10. Molten Salt Method

Nanowires MnO₂ were formed using KNO₃ as molten salt with NaNO₃ and LiNO₃ applied as the reaction media. [94]. Large scale synthesis of 1D α/β-MnO₂ nanowires was realized by mixing anhydrous MnSO₄ and KNO₃ heated at 380 °C for 3 h, loading of the KNO₃/MnSO₄ weight ratio of 15. Similar procedure was used for the synthesis of β-MnO₂ with a mixture of NaNO₃ and LiNO₃-added MnSO₄. The formation mechanism of α/β-MnO₂ nanostructures was proposed on the basis of the time-dependent experiments due to the kinetics of inserted cations into the tunnel structures: larger cations K⁺ in the (2 × 2) cavities of α-MnO₂ versus smaller cations Li⁺ or Na⁺ in the (1 × 1) cavities of β-MnO₂ [94].

3.11. Witzrmann’s Method

This consists of the reaction between KMnO₄ and small saccharides, i.e., glucose and sucrose or other polyalcohols in non-aqueous sol-gel chemistry (alcoholic solution). Ching et al. developed a modified sol-gel reaction between tetra-alkyl ammonium permanganate and methanol. MnO₂ nanopowders were well crystallized after a calcination at 450 °C for 2 h [95].

3.12. Template Approach

Aerogel MnO₂ with ultralow density (~0.5 mg·cm⁻³) was prepared from nanosheet colloids via an ice-template method for use as an effective absorbent for toxic reducing gas [96].

4. Electrochemistry of Li-MDOs

The various forms of MnO₂ have been intensively studied as electrode materials of primary zinc-alkaline cells, primary lithium cells in aqueous electrolytes, rechargeable lithium batteries, and electrodes of asymmetric supercapacitors. In addition to the interest of the various structures and specific properties of the tunnel framework, these materials possess the advantages of low cost, sufficiently high specific capacitance, and environmentally friendly nature. At the end of this Section, we briefly examine the electrochemical response of some MDO phases in batteries and their properties as supercapacitor electrodes.

The “energy density” is a common measure in evaluating battery systems. Specific energy stored (in Wh·kg⁻¹) in a battery is measured by discharging a battery at an appropriate current:

E_pr = V_oc Q_dis,

(8)

where V_oc is the operating potential in volt (V) obtained from the energy change for the cell reaction and Q_th is the specific capacity in ampere-hour per mass (Ah·kg⁻¹), or equivalently in mAh·g⁻¹. Its theoretical value is obtained from the Faraday law [1]:

$$Q_{\mathrm{th}}=\frac{1000\times nF}{3600\times M_{\mathrm{w}}}=\frac{26.8}{M_{\mathrm{w}}}\times n,$$

(9)

where M_w is the molecular mass of the “limiting” electrode material. With the transfer of 1e⁻ per formula unit, the theoretical specific capacity of MnO₂ (M_w = 92.93 g mol⁻¹) is 308 mAh·g⁻¹.

The three forms of synthesized MnO₂ (EMD, CMD, and HTMD) considered as excellent cathode materials for zinc-alkaline cells, have a hexagonally close-packed structure closely related to the polymorphs β- and γ-MnO₂ structures. Chabre and Pannetier [9] showed that the electrochemical features of γ-MnO₂ are strongly influenced by the irregular intergrowth of pyrolysite in the ramsdellite framework creating micro-twinnings and De Wolff defects. Using Raman spectroscopy, Julien et al. [97] quantitatively elucidated the structural disorder present in γ-MnO₂. It was found that the usual range of pyrolysite intergrowth, Pr, depends on the synthesis. CMD is currently close to ramsdellite with Pr < 40%, EMD has a structure with 40% < Pr < 60% and HTMD is close to pyrolusite (see Figure 2). Due to their superior structural and electrochemical performances, EMD compounds prepared by electrolysis of an acidic slurry of ground MnO₂ ore at ~95 °C are widely used in the power source industry [66,98]. McBreen [99] reported that the electrochemistry of β-MnO₂ in 7 mol·L⁻¹ KOH alkaline electrolyte differs from γ-MnO₂ showing no lattice dilatation, while the relationship between synthesis conditions and the performances of CMDs and HTMDs was established by Sarciaux et al. [100]. It appears that the maximum reversible Li uptake depends strongly on concentration of Pr and Mt defects. The largest reversible intercalation capacities are obtained from γ-MnO₂ with low Pr and Mt values. For HTMD sample with Pr = 45% and Mt = 8%, ~0.9Li/Mn can be inserted in the potential range 2–4 V at C/6 rate. Malankar et al. [101] reported the discharge characteristics of γ-MnO₂ with De Wolff disorder in the range 0.21 < Pr < 0.32 in alkaline medium. At low discharge rate, two electrochemical steps govern the reduction of MnO₂: (i) the homogeneous phase reduction MnO₂ → MnO_1.5 and (ii) the conversion of MnO₂ to MnOOH due to the increase of the concentration of Mn³⁺ and OH⁻ with the liberation of 1e⁻. The discharge performances of nanostructured α-, β- and γ-MnO₂ were tested in alkaline Zn-MnO₂ cells using electrodes made of 85% active material, 10% acetylene black, and 5% PTFE; these cells deliver a discharge capacity of 235, 140, and 267 mAh·g⁻¹ to an end potential of 0.8 V at constant current of 40 mA·g⁻¹, respectively [63].

The commercial DURACELL^® alkaline MN1500 (cylindrical “AA” size), nominal voltage 1.5 V, lasts 140 h with a load of 32 Ω (constant current of 20 mA), which is reduced to ~8 h with a 3.9 Ω loading at higher current of 275 mA (capacity of 2200 mAh). The reactions of discharge are a reduction of the oxygen-rich MnO₂ and the oxidation of the zinc, while the transport of ions occurs through the conductive alkaline electrolyte according the simplified cell reaction:

Zn + 2MnO₂ → ZnO + Mn₂O₃,

(10)

where the cathode is a mixture of EMD and graphite, while the anode is composed of high purity Zn powder held by a synthetic gel providing an open-circuit voltage (OCV) = 1.6 V [102]. Cheng et al. [66] improved the commercial AA (LR6)-type Zn-MnO₂ cells using nanostructured (nanowires/nanotubes) γ-MnO₂ synthesized via a polymer (polyethylene glycol, PEG) route. Figure 9a compares the voltage profiles for the modified and the commercial (Duracell MN1600) cells discharged at current of 0.1 A. The modified alkaline cell exhibits a similar discharge shape but delivers a significantly higher capacity of 3.0 Ah against 2.3 Ah for the commercial battery. Extensive works on MnO₂ materials for both batteries and supercapacitors in aqueous solutions can be found in Refs. [98,103,104,105,106,107,108,109,110].

Due to their basic tetragonal structure formed by double (2 × 2) and (1 × 1) tunnels, cryptomelane and hollandite (α-MnO₂ polymorph) have shown possible applications as cathode materials for rechargeable Li-MnO₂ cells. Cheng et al. [68] compared the discharge features of α-MnO₂ nanowires with that of γ-MnO₂ nanorods and concluded that this latter cathode exhibits better electrochemical performance than the former one, i.e., 220 vs. 204 mAh·g⁻¹ with the same conditions. The hollandite-type MnO₂ (HMDO) synthesized by reaction of MnSO₄ in concentrated H₂SO₄ in the presence of bubbling O₂ and ozone blended gas (50% in volume) was formed of coral-like particles and delivered a specific capacity of 165 mAh·g⁻¹ after 20 cycles in the potential range 1.8–4.0 V [106]. Ma et al. [67] carried out electrochemical measurements on δ-MnO₂ nanobelts prepared by the hydrothermal method that reversibly hosted lithium as cathode in the potential range 1.0–4.8 V vs. Li⁺/Li. Specific capacity of 220 mAh·g⁻¹ was delivered over 45 cycles (uptake of 1.3 Li per formula unit) at current density of 20 mA·g⁻¹.

The discharge curves of different forms of manganese dioxide in Li-MnO₂ batteries, i.e., single-phase α-MnO₂, β-MnO₂, R-MnO₂, and the stabilized phase α/β-MnO₂ are shown in Figure 9b [1]. These results were obtained with cells using lithium counter-electrode, same electrolyte (1 mol·L⁻¹ LiPF₆ in ethylenecarbonate (EC)/dimethylcarbonate (DMC) 1:1) and the same separator (Whatmann^®-GF/D 70 mm Ø, Darmstadt, Germany), to make possible quantitative comparison. These data show that the stabilized two-phase α/β-MnO₂ sample delivers higher discharge capacity than the single-phase α-MnO₂. Furthermore, ramsdellite (R-MnO₂) and pyrolusite (β-MnO₂) display the highest discharge capacities. These materials present flat discharge curves while the hollandite structure shows an “S”-shaped discharge curve. On the initial discharge process the stabilized α/β-MnO₂ material delivers a specific capacity of 230 mAh·g⁻¹. This electrode shows good rechargeability with a capacity retention of 150 mAh·g⁻¹ after 20 cycles. The initial capacity loss of 33% suggests that about 0.3 mol of inserted lithium ions are used [87]. Thackeray et al. [107] indicate that the rechargeability of R-MnO₂ is poor for the electrode cycled on deep discharge. The initial capacity 230 mAh·g⁻¹ declines to 115 mAh·g⁻¹ after 10 cycles at a current of 0.4 mA. This capacity loss is concomitant to structural changes upon the deep discharge process, i.e., large variations of the unit cell volume.

Several doping methods have been tested to improve the performance of MnO₂ in lithium cells. The Bi-doped MnO₂ was investigated by Bach et al. [108] suggesting that due to the presence of interlayer Bi³⁺ ions, a pillaring effect minimizes the structural modifications. Other examples of doping are treated in the following. Yang et al. [109] compared two commercial MnO₂ grades, i.e., EMD and CMD to hydrothermally synthesize α-MnO₂ composed of nanorods ~20 nm diameter in lithium cells. The results demonstrated that, at a discharge current 50 mA·g⁻¹, α-MnO₂ material delivers a specific capacity of 189 mAh·g⁻¹ against 134 and 148 mAh·g⁻¹ for, CMD and EMD type, respectively. Despite the favorable tunnel structure of γ-MnO₂, this phenomenon can be well understood in terms of the shortened diffusion path for Li⁺ ions in low dimensional α-MnO₂.

The charge storage mechanism in a MnO₂ electrode for a supercapacitor performed in aqueous electrolyte has been described with two mechanisms [14]. The first one involves the insertion of alkali ions (Na⁺, K⁺ or Li⁺) or protons (H⁺) in the empty sites of the MnO₂ bulk upon the reduction reaction:

MnO₂ + H⁺ +e ⁻ ↔ MnOOH,

(11)

or

MnO₂ + Li⁺ +e ⁻ ↔ MnOOLi.

(12)

The second mechanism consists of adsorption of electrolyte cations (Li⁺) on the surface of MnO₂

(MnO₂)_surface + Li⁺ +e ⁻ ↔ (MnO₂⁻Li⁺)_surface.

(13)

Typical galvanostatic charge-discharge profiles at various current densities (from 2 to 10 mA·cm⁻²) of an asymmetric supercapacitor using polypyrrole (PPy)/MnO₂ composite material in 1 mol·L⁻¹ Na₂SO₄ aqueous solution as electrolyte are shown in Figure 9c [110]. The pseudo capacitance is attributed to the Mn⁴⁺/Mn³⁺ reversible redox process accompanied by the insertion/deinsertion of alkali Na⁺ cation or H₃O⁺ protons from the electrolyte. A thin layer of PPy electrodeposited for 40 min on 250-nm sized γ-MnO₂ particles provide an energy density of 12.6 Wh·kg⁻¹ and a power density of 34 W·g⁻¹ [110].

5. MnO₂ Nanostructures: Example of Nanourchins

As we shall see in this section, urchin morphology was reported to be the best choice for enhanced electrochemical properties, so we have chosen it as an example. Recently, the synthesis process to grow MnO₂ nanoneedles (NNs) forming nanourchin (NUs) architecture has been investigated by different groups [30,45,46,88]. As shown above, many factors like acidity of the solution, cationic species (nature and concentration), and additive metal ions (Co²⁺, Ni²⁺, Fe³⁺, Al³⁺, etc.) greatly influence both the structure and the morphology of nanoparticles. For example, γ-MnO₂ urchin-like nanostructures are grown using Mn₃O₄ powder as raw material in H₂SO₄ solution; α-MnO₂ urchin-like composed of single nanorods are obtained from KMnO₄ and H₂SO₄ or from MnSO₄·H₂O, K₂S₂O₈ and concentrated sulfuric acid. A redox reaction between MnSO₄ and (NH₄)₂S₂O₈ as an oxidizing agent can also be used [97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117]. Details of the synthesis of urchin-like α-MnO₂ are shown schematically in Figure 10. Using two identical procedure, i.e., oxidation of MnSO₄·H₂O by K₂S₂O₈, Zhang et al. [118] obtained two different phases that are γ-MnO₂ nanoparticles when using neutral (pH ~ 8) conditions, while α-MnO₂ NUs were grown in acidic (pH ~ 1) conditions with small addition of H₂SO₄. Wang et al. [119] synthesized sea urchin-like α-MnO₂ particles by a one-step chemistry route at room temperature using MnSO₄ in combination with KIO₄ as oxidant. Results revealed a product with lattice constants a = 9.840 Å and c = 2.856 Å (I4/m space group) and a Brunauer-Emmett-Teller (BET) surface area of 201 m²·g⁻¹.

Addition of Al³⁺ affects the hydrothermal synthesis of MnO₂ by modifying the chemical potential of the solution. Wang et al. [64] studied the synthesis conditions of various nanostructured MnO₂ polymorphs by tuning the pH and NH₄⁺ cation concentration. It has been shown that NUs are generally synthesized by sol-gel methods in acidic medium that can control the morphology of the nanoparticles by direct redox reaction [43]. The choice of the cation species (K⁺, NH₄⁺, H⁺) also plays a role evidenced in the hydrothermal crystallization [120]. For example, Yu et al. [30] fabricated various MnO₂ nanoneedles structures by redox reaction of K₂S₂O₈ and MnSO₄·H₂O with a concentrated sulfuric acid solution. The precipitate dried at 60 °C for 8 h is an urchins-shaped α-MnO₂ composed of nanorods with tetragonal lattice constants a = 9.826 Å, c = 2.854 Å. Using the same procedure with addition of Fe(NO₃)₃·9H₂O or Al(NO₃)₃·9H₂O, a 3D clew-like ε-MnO₂ nanoarchitecture with hexagonal lattice constants of a = 2.846 Å, c = 3.530 Å was grown. In contrast, Li et al. [121] claimed the need of Ag⁺ ion (via AgNO₃ solution) as catalyst agent for the crystallization of α-MnO₂ core-shell urchin-like structures. Urchin-like α-MnO₂ nanomaterials were also prepared without template or surfactant by a simple precipitation reaction of H₂SO₄ and KMnO₄ in aqueous solution [89]. Urchins are composed of aggregated crystalline nanorods with a mean diameter of 10 nm and a length of 200 nm. Chen et al. [90] suggested the formation mechanism of urchin-shaped α-MnO₂ from analysis of intermediates products formed in the hydrothermal process at different conditions of temperature and time (55 < T < 85 °C, 10 min ≤ t ≤ 12 h). For T = 55 °C, t = 10 min, microspheres with diameter of 1 µm that consist of nanorods are formed; for T = 65 °C, t = 10 min, nanorods are epitaxially grown; for T = 75 °C, t = 10 min, the morphology turned into flower-shaped α-MnO₂; finally, at T = 85 °C the microspheres transform to urchin-like α-MnO₂. This suggests that the nanomaterial morphology is very sensitive to the temperature of the hydrothermal reaction. The synthesis of MnO₂ nanostructures with sea-urchin shapes carried out by a sodium dodecyl-sulfate (SDS)-assisted hydrothermal process was optimized by tuning the reaction time t while maintaining a constant temperature T = 150 °C. The products were composed of aggregated particles for t = 4 h, while the urchin-like morphology was well nucleated at t > 8 h. It was also shown that for low SDS concentration the nanostructure has a loose center [44].

NNs were prepared by a redox reaction of KMnO₄ and ascorbic acid. After stirring for 3 h, the precipitate dried at 100 °C for overnight aggregated with an urchin-like morphology (see Figure 11) [43]. These NNs grew in the tetragonal K_xMnO₂ structure with lattice parameters a = 9.8314(4) Å, c = 2.8586(1) Å and V = 276.31 ų. The mean crystallite size of the K_xMnO₂ nanoneedles was L_c = 14 nm. From both Rietveld refinement and elemental analysis, the K/Mn atomic ratio of ≈4% was evaluated. Generally, the presence of K⁺ ions in the (2 × 2) tunnels of the hollandite lattice has an impeding effect for the chemical diffusion of the Li⁺ ions. This is due to the much bigger ionic radius of K⁺ (r = 1.33 Å) than Li⁺ (r = 0.69 Å), which prevents the facile lithium motion in the MnO₂ framework [122].

The local structure of K_0.04MnO₂ nanoneedles was carefully analyzed by Raman and FTIR spectroscopy and magnetic measurements. Figure 12 shows the plot of the reciprocal magnetic susceptibility H/M vs. absolute temperature of MnO₂ nanoneedles (with H = magnetic field, M = magnetic moment). For T > 150 K, the magnetization M is linear in field H, so that the magnetic susceptibility χ_m is defined unambiguously by χ_m = M/H for NNs samples. The linear behavior of H/M up to 350 °C follows the Curie-Weiss law. This result evidences the strong antiferromagnetic interactions between the Mn moments as θ_p = −215 K. The experimental value of the effective magnetic moment (μ_eff = 3.94 μ_B) is slightly larger than that of Mn⁴⁺ ions (μ_eff(Mn⁴⁺) = 3.87 μ_B), which gives evidence of the presence of Mn³⁺ ions in the high spin state (μ_eff(Mn³⁺) = 4.90 μ_B). Thus, the concentration of Mn³⁺ ions is calculated [43] to be 5.8%. As it is actually larger than the residual concentration of K⁺ ions, the local electrostatic charge neutrality imposes that oxygen vacancies should be responsible for the extra Mn³⁺ ions in the matrix; thus, the nanoneedle chemical formula is K_0.04MnO_1.97.

Figure 13 presents the charge–discharge profiles of MnO₂-NNs//Li cells vs. specific capacity for cycles up to the 100th [43]. The discharge plateau at ca. 2.35 V corresponding to the reduction of Mn^IV ions (lithium insertion) disappears in the first discharge curve and a smooth S-shaped discharge profile is observed in the next cycles. The hollandite α-MnO₂ phase would have a discharge capacity up to 230 mAh·g⁻¹ based on a discharge voltage range of 4.0–2.0 V vs. Li⁺/Li⁰ [81]. The initial discharge capacity is 230 mAh·g⁻¹ for our sample, which means about 0.73 Li ions per formula unit inserted into the MnO₂ framework. Note that Li ions prefer to occupy the off-center 8 h site near the (2 × 2) tunnel of the α-MnO₂ lattice, but shift to the 8h’ site in α-Li_0.75MnO₂ lattice due to the Coulomb repulsion between Li ions.

Feng et al. [79] and Li et al. [123] showed that α-MnO₂ urchin-like has better performance than other α-MnO₂ samples due to the morphological sensitivity on the electrochemical performance. The good results of the sea-urchin shape are attributed to the hierarchical structure combining with 1D nanorods, which minimize the Li diffusion path with a 3D nanostructure that exhibits a high specific surface area (~95 m²·g⁻¹). It is believed that such morphology prevents the formation of a Li-MnO₂ spinel phase. He et al. [124] pointed out similar properties in the case of α-MnO₂ electrodes for supercapacitors, as the MnO₂ nanorods prepared with 0.59 g KMnO₄ delivered the highest capacitance of 198 F·g⁻¹ with 94% retention after 2000 cycles.

6. Doped-MnO₂ Materials

Doping of MnO₂ frameworks has been successfully realized using metal ions such as Ag⁺, Ni²⁺, Cu², Co²⁺, Fe³⁺, Cr³⁺, Mo⁶⁺, V⁵⁺, W⁶⁺, etc. [125,126,127,128,129,130]. Novel morphologies and enhanced electrochemical properties of the cryptomelane structure(α-K_xMnO₂) are currently obtained by a change of the crystal chemistry by exchange of K⁺ ions with protons and/or doping by a single-type metal cation with either low-valence state (3+, 2+, 1+) or high-valence state (5+, 6+). There are two possibilities to insert metal cation as doping element into the a-MnO₂ lattice: (i) substitution of Mn cation in the octahedral framework that implies a six-coordinated cation with a crystal radius similar to that of ^VIMn³⁺ low spin (0.72 Å), ^VIMn³⁺ high spin (0.785 Å), and ^VIMn⁴⁺ (0.67 Å) and (ii) insertion of the dopant cation into the (2 × 2) tunnel, which allows an eight-coordinated cation with crystal radius similar to K⁺ (1.65 Å) [131]. For dopant cations of lower valence, the more negative charge of the lattice favors the incorporation of more K⁺ ions into tunnels that enhances the structural stability. Dopant cations of higher valence create an excess of electrical charge that can be compensated by the creation of vacancies, which results in structural distortion and thermal instability.

6.1. Literature Survey

Pure α-MnO₂ is a semiconductor with bandgap 1.44 eV that has an antiferromagnetic ground state due to the symmetric nature of Mn-O-Mn bonds. When the material is prepared through a redox reaction of KMnO₄, a large concentration of K⁺ ions can be incorporated in the tunnels, which make α-MnO₂ a half-metallic compound; for a potassium content of ≈12 at%, a ferromagnetic-like behavior is observed at low temperatures <5 K [132]. On the other hand, at low potassium content MnO₂ has a poor electronic conductivity (≈10⁻⁸ S·cm⁻¹). Therefore, the control of doping favors the enhancement of ionic and electronic transport. As MDOs crystallize in multiple tunnel structures, MnO₂ shows different electrochemical behaviors with Faradaic reactivity in the sequence δ-MnO₂ > α-MnO₂ > γ-MnO₂ > β-MnO₂, which can tune the ion insertion reactions. It is well-known that the MnO₂ electrode shows gradual capacity fading during long-term cycling due to structural destabilization related to the Jahn-Teller distortion and partial Mn³⁺ dissolution [133]. To overcome this disadvantage of structural destabilization and to enhance the electronic transport in MnO₂ that facilitates the discharge/charge rate of the electrode, doping with various elements, i.e., Ag, Sn, V, Ni, Cu, Al, etc., has been proposed [10,134,135]. The use of selected doping elements allows the properties of α-MnO₂ to be tuned for practical applications. For example, alkali ions such as Li⁺ (0.076 nm ionic radius) are easily housed in the (2 × 2) tunnels (0.48 nm size) and can move freely under electrochemical stimulus. Such physical behavior has been applied to batteries and supercapacitors [1].

The role of doping in α-MnO₂ as oxygen reduction reaction (ORR) electrocatalyst has been widely investigated [136]. It was shown that Ni doping stabilizes the Mn³⁺/Mn⁴⁺ mediating species involved in ORR activity. Hao et al. [137] synthesized Ni-doped α-MnO₂ nanoneedles via a facile hydrothermal method. The role of nickel was to promote the oxygen reduction reaction in alkaline media, i.e., 0.1 mol·L⁻¹ KOH aqueous solution. The electrochemical measurements show that 2.22% Ni-doped MnO₂ has excellent electrocatalytic activity (EA) due to the increment of Mn(III) as electrochemical active sites. High EA was also reported for MnO₂ particles dispersed on high surface area carbon [138]. Davis et al. [139] studied the catalytic activity of Cu-doped α-MnO₂ nanowires. Due to the similarity of the Al³⁺ and Mn⁴⁺ atomic radius, aluminum can substitute Mn or be located in tunnel of α-MnO₂. Hu et al. [140] showed that Al-doped α-MnO₂ nanoneedles prepared by the hydrothermal method using K-free precursors and Al₂(SO₄)₃·18H₂O as dopant reagent were beneficial for pseudocapacitor electrode application with a specific capacitance of 213 F·g⁻¹. Zn-doped MnO₂ nanoparticles (high surface area ~46 m²·g⁻¹) were prepared by precipitation of KMnO₄ and metal acetates with heat treatment of the precipitate at 400 °C for 3 h [141]. Cr³⁺-ion doping induces a phase transition of MnO₂ from β- to α-polymorph. The size of MnO₂ nanorods increased from 20 to 70 nm with the dopant concentration [142]. Note that insertion of Cr³⁺-ion is favored in the α-MnO₂ phase because the ionic radius (0.1 nm) is closed to that of K⁺ (0.118 nm). A high specific capacitance of 583 F·g⁻¹ at current density of 10 A·g⁻¹ was obtained with Cu-doped MnO₂ nanorods prepared by precipitation of KMnO₄ and copper acetate [143].

6.2. Vanadium-Doped MnO₂

Vanadium-doped MnO₂ nanoparticles were prepared by different routes including the redox reaction [134,144,145]. Alfaruqi et al. [135] used a simple redox reaction between Mn(CH₃COO)₂·4H₂O and KMnO₄ in aqueous solution added to a solution containing V₂O₅ to obtain α-MnO₂ nanoparticles after annealing at 450 °C for 5 h. This material was used as electrode for zinc-ion batteries. V-doped γ-MnO₂ used as the cathode in primary lithium batteries was prepared by the redox reaction of KMnO₄ and MnCl₂·4H₂O with V₂O₅ as dopant reagent in a 3:1:0.15 molar ratio. The final products obtained from the precursors annealed at 375 °C for 10 h exhibit an anisotropic expansion that achieved better diffusion coefficient of Li⁺ ions in the (1 × 1)/(1 × 2) tunnel frameworks, i.e., ~2 × 10⁻⁸ vs. ~5 × 10⁻⁹ cm²·s⁻¹ for pure MnO₂ [146,147].

6.3. Titanium-Doped MnO₂

Li et al. [148] synthesized Ti-doped δ-MnO₂ nanoflakes (with thickness of ≈50 nm) via the anion route for the highly catalytic combustion of benzene. Due to the abundant pore structure and the active oxygen induced by Ti doping, these nanoflakes have the highest catalytic oxidation property over benzene. The interlayer spaces of ∼0.7 nm and mesopores of 4–5 nm and 8–9 nm) facilitate gas diffusion and reactions. Ti-containing γ-MnO₂ was prepared in two steps by in situ precipitation technique. First, MnSO₄ and TiOSO₄·xH₂SO₄ with various Ti/Mn atomic ratios were dissolved in aqueous solution along with concentrated nitric acid. The second step consisted of the precipitate of KMnO₄ with the first solution heated at 90 °C (refluxing). Titanium incorporated into the MnO₂ hollow sphere framework favors electrochemical performance with a high specific capacity (2200 mAh·g⁻¹ of carbon) in the Li/air battery and strong oxidative catalytic activity in the toluene oxidation process as well [149]. Nanostructured 5% Ti-doped α-MnO₂ particles were synthesized by hydrothermal methods using two different oxidizing agents, i.e., ammonium persulfate and potassium permanganate for electrocatalytic applications. The doped samples show an efficient oxygen reduction reaction (ORR) activity in alkaline media that leads to a significant shift of the ORR potential (~100 mV) comparable to the well-performing Pd₄₅Pt₅Sn₅₀ material [150].

6.4. Al, Cu, Mg-Doped MnO₂

The advantage of the Mn to Al substitution in γ-Mn_1-yAl_yO_2-δ interconnected nanowires was an increase of the surface area from 17 to 184 m² g⁻¹ for y(Al) = 0.11, which resulted in an increase of the Faradic behavior [151]. Pure MnO₂ and its M-doped MnO₂ (M = Al, Cu, Mg) were prepared by redox reaction of KMnO₄ and fumaric acid (C₄H₄O₄) [152]. Both pure and doped samples show the same characteristic peaks of cryptomelane-MnO₂ (K₂Mn₈O₁₆). No extra peaks related to Al, Cu, and Mg compounds are observed. No diffraction lines associated with doping elements were observed, which can receive three interpretations: (i) transition metal oxides are in low content or have crystalline domains beyond the detection limit of XRD; (ii) transition metals are incorporated in the cryptomelane structure, with the formation of a solid solution; and/or (iii) transition metals are incorporated into the channels of the K_xMnO₂ structure, replacing K⁺ ions [19], All possible reflections of cryptomelane compounds are present in the prepared samples. The MnO₂ lattice of all samples is related to the presence of K⁺ ions inside the (2 × 2) tunnels of the prepared samples as observed. Chemical analysis shows that the amounts of potassium (in %) for P-MnO₂, Al-MnO₂, Cu-MnO₂, and Mg-MnO₂ samples are 0.7, 7.9, 5.6, and 8.9, respectively. Chemical analysis shows also that the percentages of Al, Cu, and Mg in the doped samples are 0.4, 0.6, and 0.3, respectively.

The electronic transport measurements were performed below room temperature (RT) by the Van der Pauw four-point method. All four MDOs show a semi-conducting behavior at RT with a decrease of resistivity by three to four orders of magnitude depending on the dopant. The Al-MnO₂ sample has the lowest resistivity between the four oxides. Above RT, the electrical resistivity ρ was activated, as can be deduced from the linear dependence of the ln (ρ) as a function of 1/T reported in Figure 14. The activated band gaps E_g of the four oxides determined from the slope of these linear curves show that the gap of parent MnO₂ is close to the value of 0.69 eV obtained for γ-MnO₂ and to the value of 0.58 eV obtained for cryptomelane MnO₂ [153]. Doping reduced E_g to a value close to 0.34 eV, close to the value 0.26–0.3 eV reported for β-MnO₂ [154]. The introduction of dopant ions like Al, Cu, and Mg seems to stabilize the MnO₂ structure and hence reduce the capacity fading observed for pure MnO₂. The presence of a low concentration of stabilizing atoms within the (2 × 2) tunnel of a cryptomelane- or hollandite-type framework is required to facilitate the diffusion of Li ions during charge–discharge cycling as observed for doped α-MnO₂ samples [152].

6.5. Tin-Doped α-MnO₂

To prevent the transformation from α-MnO₂ to α-Mn₂O₃ that takes place in the temperature range of 500–600 °C, doping with Sn and Co was proposed by Hashem et al. [155]. Samples were synthesized in an acidic medium using the reduction of KMnO₄ by fumaric acid (10:3 molar ratio) with addition of SnCl₂ as Sn dopant source. Final products were obtained by heat treatment at 450 °C for 5 h. Figure 15 shows the thermogravimetric analysis (TGA) of the pristine and Sn-doped MnO₂ samples heated in the range 30–1000 °C in air. The TGA curve for the undoped sample shows a slight weight loss (ca. 3%) due to the removal of surface and structural water and an abrupt weight loss at ca. 540 °C due to the exothermic reaction of the phase transition from α-MnO₂ to Mn₂O₃ and release of oxygen. TGA curves of α-MnO₂:Sn illustrate the structural stability of the doped samples up to 850 °C. This effect of the introduction of a small concentration ≈5% of Sn into the crystal lattice, is attributed to the fact that the doping maintains the tunnel structure of α-MnO₂. The magnetic susceptibility measurements confirm this stabilization effect. At high temperature T > 150 K, MnO₂ exhibits Curie-Weiss paramagnetic behavior, while a ferromagnetic contribution is observed at low temperature (T < 30 K), due to the 180° Mn³⁺-O-Mn⁴⁺ bridge. The increase of dopant concentration decreases the Mn³⁺ content and reduces the ferromagnetic content.

Figure 16 shows the electrical conductance σ_ac vs. frequency for α-MnO₂ samples with different Sn or Co dopant concentrations. Results show the typical features of a semiconducting material with a frequency dependence that obeys the power law:

σ_ac = σ_dc + Aωⁿ,

(14)

where the low-frequency value corresponds to the direct-current conductivity σ_dc, n is the power exponent, and A is a constant. An increase in the electrical conductivity is clearly observed in the presence of dopant in comparison with the pristine α-MnO₂ material. It is believed that free electrons of Co(II) and Sn(II) contribute to the increase of the conductivity of the doped samples.

Hashem et al. [10] investigated the electrochemical performance of Sn-doped α-MnO₂ nanorods-like particles. The specific discharge capacities vs. the cycle number for P-MnO₂ and Sn–MnO₂ are shown in Figure 17. Capacities are ~65 and ~80 mAh g⁻¹ for the P-MnO₂ and Sn-MnO₂ electrodes at the 40th cycle, respectively. These results show that capacity fading of the pristine electrode is much higher than that of Sn-doped MnO₂. The electrochemical performance and the structural stability are attributed to the decrease of Mn³⁺ Jahn-Teller ions upon insertion of Sn ions into the (2 × 2) tunnels. The second reason for the electrochemical degradation of pristine MnO₂ is due to the reduction of Mn²⁺ ions, which dissolve in the electrolyte.

6.6. Ag-Doped MnO₂

Pristine K_xMnO₂, Ag-doped and Ag-coated K_xMnO₂ materials (x ≈ 0.065) were obtained by a simple wet-chemical process. Then the particle size was reduced to ~20 nm by re-stirring the as-prepared oxides in de-ionized water for 24 h at RT [156]. Elemental analyses show a concentration of silver 1.4% and 3.9% in doped and coated K_xMnO₂, respectively. From magnetic measurements the Ag-coated K_xMnO₂ sample shows an increasing Mn⁺⁴/Mn⁺³ ratio and hence a reducing amount of Jahn-Teller Mn⁺³ ions. The net result is a better electrochemical performance of Ag-coated K_xMnO₂ (Figure 18). For the Li//MDO cells cycled up to 40th cycle, the discharge specific capacities are 115, 110, and 90 mAh·g⁻¹ for Ag-coated K_xMnO₂, pure K_xMnO_2, and Ag-doped K_xMnO₂ samples, respectively. The Ag-coated K_xMnO₂ sample showed the best results for capacity retention due to nanosized particles obtained by stirring in deionized water and to enhanced conductivity after Ag coating.

6.7. Co- and Ni-Doped MnO₂

Cobalt-doped α-K_xMnO₂ was synthesized following the same process used for tin-doping, except that the Sn precursor was now replaced by Co(NO₃)₂·6H₂O as Co dopant source in a molar ratio 3:1:0.07, respectively [157]. Table sugar was the source of carbon for coating. Chemical analysis gives the chemical stoichiometry K_0.009MnO₂ for the pure sample and K_0.095Co_0.013MnO₂ for the doped sample. Both electrochemical inactive Co³⁺ and K⁺ ions are trapped inside the large tunnel (4.6 Å width). The additional effect of Co doping and carbon coating results in a good rechargeability and a decrease of capacity fading at the expense of the initial capacity (Figure 19). The carbon layer acts as a protective film surrounding the particles and favors the charge-transfer rate of Li⁺ insertion/extraction reactions. Magnetic properties indicate that the mixed valence state Mn⁴⁺/Mn³⁺ with low concentration of Mn³⁺ decreased after the coating and doping process.

Nanospheres of 5 wt % Co-doped R-MnO₂ (diameters in the range 350–500 nm) composed of nanoflakes 3 nm thick were grown with a yolk-shell structure using a redox reaction of K₂S₂O₈ and MnSO₄·H₂O with added CoSO₄·H₂O [158]. These nanospheres had specific surface area 135 m²·g⁻¹ with pore size of 9 nm. Korosec et al. [159] reported the structural properties and thermal stability of cobalt- and chromium-doped α-MnO₂ nanorods synthesized by decomposition of KMnO₄ in an acidic environment. EXAFS studies showed that both dopant ions (Co²⁺, Cr³⁺) substitute Mn⁴⁺ in the center of an octahedron increasing the negative charge of the lattice compensated by an increase of K⁺ ion concentration in the tunnels. In another work [160], Co/Ni-doped K_0.14MnO₂ tetragonal phase (cryptomelane structure) was synthesized via a common redox reaction with metal sulfates as dopant agents. The mole fractions of Ni²⁺ and Co²⁺ in the final product were 2% and 7%, respectively. The samples were composed of nanowires of diameter 15–20 nm, length of 100–300 nm. The Co/Ni doping did not modify the 1D nanostructure of α-MnO₂, because of the growth mechanism of the dissolution–recrystallization process. Co-doped birnessite (Co-bir) δ-MnO₂ is a catalyst synthesized by a modified sol-gel method for the oxidation of benzylic alcohols to benzaldehydes achieved in heated toluene under oxygen atmosphere [161]. The enhanced electrical conductivity of δ-MnO₂ is attributed to location of Co²⁺ ions in the octahedral lattice.

A different result was reported by Biswal et al. [162] who found different morphologies depending on whether the dopant is Co or Ni. The synthesis process was also different from the previous one: a galvanostatic method, starting from manganese sulfide in sulfuric acid medium. Note this preparation misses the presence of K⁺ ions that was found to be so important in this review to stabilize the material and optimize the electrochemical properties of the α- and β-MnO₂ phases. It is also important that these EMD samples were found in a different phase, namely the γ-MnO₂ phase. One of these samples was prepared with Ni- and the other one with Co- in situ doping. With Co-doping, the EMD was synthesized with the form of cauliflowers, while the EMD with Ni-doping was sea-urchin shaped. In both cases, the doping increased the energy density, but not at the same level: 395 mAh·g⁻¹ for Ni-doping, against 670 mAh·g⁻¹ for Co-doping; on another hand, the cycling life was better with Ni-doping. Therefore, in any phase, the Co or Ni doping increased significantly the conductivity and the electrochemical properties. However, in [162], the increase of conductivity in Co-doped EMD was attributed to the presence of Co₃O₄. It would be of interest to conduct Raman experiments to verify this hypothesis.

6.8. Bismuth-Doping and Additives

The incorporation of Bi³⁺ cations has been known to be beneficial to the electrochemical properties of MDO for many years, irrespective of the crystal phase [163,164,165,166]. This improvement includes an increase of the conductivity like the introduction of the other dopant ions, but in addition, a specific property of the bismuth is that it reduces the formation of the spinel structure [167], which is responsible for irreversibility of the MnO₂ cells, as the Mn₃O₄ spinel is not electroactive. The reason why Bi has such an important effect has been described by Yu [168] who noticed that the ionic radius of Bi³⁺ being much larger than the ionic radii of Mn²⁺ and Mn³⁺, they cannot insert into the spinel lattice, which prevents the formation of the spinel along the chain of reactions during the synthesis of MnO₂. This important role of bismuth was also observed more recently by Im and Manthiram [169] who incorporated Bi³⁺ cations into γ-MnO₂ with the Bi₂O₃ additive in an alkaline electrolyte. Comparing the effect of Ti- and Bi-incorporation on the electrochemical properties of γ-MnO₂, Sundaram et al. [170] found that Ti is even more efficient than Bi in preventing the formation of Mn₃O₄. In addition, they found that even better electrochemical properties were obtained by multiple additives. In particular, the synergetic effect of adding 3 wt % Bi₂O₃ plus 2 wt % TiS₂ led to a superior capacity of 240 mAh·g⁻¹, much larger than the results found with Bi₂O₃ or TiS₂ only.

Other additives that have improved the electrochemical properties of γ-MnO₂ are TiB₂, CeO₂, MgO, and B₄C [171,172,173,174,175]. Several reports have shown the irreversible dissolution of Mn³⁺ ions in alkaline KOH solutions. This reaction leads to the growth of electrochemically inactive phases, for example δ-MnO₂ and Mn₃O₄. Not surprisingly, the additives that can suppress the dissolution of the Mn³⁺ ions are also those which have been shown to prevent the formation of Mn₃O₄ such as TiB₂, Bi₂O₃, and also Ba-containing compounds [176].

7. MnO₂ Polymer Composites

MnO₂ is a material that can be combined with various polymers to make nanoarchitecture hybrids as highly performing electrode materials for pseudocapacitive devices. A blend formed by electrochemical polymerization of pyrrole monomer (Py) on prepared manganese dioxide powders was studied as the electrode for a supercapacitor [110,177]. The capacitance of the MnO₂ electrode is predominantly pseudocapacitive, which is attributed to reversible redox transitions involving exchange of protons and/or cations with the electrolyte. In practice, the MnO₂ specific capacitance is ~200–300 F·g⁻¹ due to its intrinsically poor electronic conductivity, size of particles, and porosity of the oxide [53]. γ-MnO₂, i.e., (1 × 1)/(1 × 2) tunnel structure, was prepared by a precipitation method of MnCl₂·4H₂O and KMnO₄ in distilled water and dried at 110 °C for 10 h.

7.1. Polypyrrole-Coated MnO₂

The electrodeposition of polypyrrole (PPy) was carried out with a chronoamperometry test at a monomer oxidation potential 900 mV vs. SCE. The net effect of the PPy deposit is an increase of the BET surface area for PPy/γ-MnO₂ of 125 m²·g⁻¹ vs. 64 m²·g⁻¹ for γ-MnO₂. Figure 20 shows the SEM images of pristine and γ-MnO₂ particles covered with electrodeposited PPy. We observe that the electrochemical polymerization process does not change the morphology of the MnO₂ grains and that MnO₂ particles synthesized by the precipitation route have a regular shape with an average grain size 250 nm [177].

The MnO₂ and PPy/MnO₂ pseudocapacitance is due to the Mn⁴⁺/Mn³⁺ reversible redox reaction accompanied by a reversible insertion/desinsertion of alkali cation (Na⁺) or protons H₃O⁺ present in the electrolyte:

MnO₂ + Na⁺ + e⁻ → MnOONa.

(15)

The specific capacitance was determined from galvanostatic charge-discharge cycling tests at a constant current density 2 mA·cm⁻² [110]. The asymmetric supercapacitor with PPy/γ-MnO₂) composite cathode and carbon anode has high specific capacitance of ~142 F·g⁻¹ vs. ~74 F·g⁻¹ for γ-MnO₂. Note that the specific capacitance of PPy/MnO₂ materials is proportional to the thickness of the PPy deposit. The performance of the composite material was measured in a constant charging–discharging experiment at a discharge current density 2 mA·cm⁻² over 500 cycles (Figure 21). The stabilization of the specific capacitance indicates that the electrode had regular capacitive behavior and good cycling stability.

Note, however, that the DMO-polymer association does not give the best supercapacitor. For comparison, the aqueous asymmetric capacitor with EMD obtained from a leach liquor derived from manganese ore/residue delivered a capacity of 50 F·g⁻¹. The outstanding performance with respect to the results we reported above, however, is the cycling life, since 100% capacity was retained after 2000 cycles [98].

7.2. Polybithiophene-Coated MnO₂

A new composite formed by polymeric polybithiophene (PBTh) and crystallized MnO₂ was applied as a p-n heterojunction with good photoconducting performance in solar cells. The PBTh/MnO₂ sample was deposited on an indium tin oxide (ITO) substrate. Incorporation of MnO₂ particles into the polymer films greatly increases the generated photocurrent from 5.9 μA·cm⁻² for ITO/PBTh up to 20.6 μA·cm⁻² for the ITO/PBTh-MnO₂ films with 100 mg MnO₂ incorporated [178].

Similarly, a polymer/inorganic composite was used as cathode material in Zn//γ-MnO₂ electrochemical cells. The composite was prepared by electrodeposition of PBTh on MnO₂ particles in 0.01 mol·L⁻¹ PBTh/0.1 mol·L⁻¹ LiClO₄ in acetonitrile (CH₃CN) solution [179]. The performance of MnO₂ electrodes was tested by EIS experiments for both discharged Zn//MnO₂ and Zn//PBTh + MnO₂ cells (see Figure 22). The Nyquist plots display EIS profiles containing a semicircle (high-frequency range) and a quasi-linear line (low-frequency range). The semicircle is due to the charge transfer resistance (R_ct) of the cathode material in relation to the contact between particles. The quasi-linear part at low frequency is the Warburg contribution of proton diffusion through the bulk of the material. The fit illustrated in Figure 22 gives a charge transfer resistance R_tc = 4.49 Ω·cm⁻² for Zn//MnO₂ cell and reduces to R_tc = 3.42 Ω·cm⁻² for the Zn//PBTh + MnO₂ cell.

Figure 23 presents the electrochemical profile of Zn//MnO₂ and Zn//MnO₂ + PBTh cells discharged at current density of j = 2 mA·cm⁻² [17]. A continuous decrease of the cell voltage is observed in the 1.45–0.9 V potential range. No plateau can be observed. The capacity of the Zn//MnO₂ + PBTh cell is 25% higher than that of the Zn//MnO₂ cell. The overall cathodic reaction that reduces MnO₂ to MnOOH, involves a solid-state diffusion process for protons moving from the surface to the interior of the MnO₂ grains, as follows:

$${\mathrm{H}}_2\mathrm{O} (1) \rightleftarrows {\mathrm{H}}^{+} (\mathrm{surface})+\mathrm{O}{\mathrm{H}}^{-} (\mathrm{aq}),$$

(16)

H⁺ (surface) → H⁺ (bulk),

(17)

xMnO₂ + H⁺ (bulk) + e⁻ → (MnO₂ )_x−1 (MnOOH) (s).

(18)

The conducting polymer coat on MnO₂ particles has an important role. First, it favors the diffusion of protons; second, the conducting polymer can be reduced during discharge or can occupy the pores, which results in more active material and a larger effective surface area. Composite materials of conducting polymer and β-MnO₂ were prepared by electrodeposition in CH₃CN/0.1 mol·L⁻¹ LiClO₄ cell of conducting polymer on a β-MnO₂ surface with different monomers: bithiophene (BTh) or pyrrole (Py) in CH₃CN/LiClO₄ (0.1 mol·L⁻¹) [180]. A successful electro-polymerization requires the formation of a layer able to inhibit the dissolution of the oxidant metal. At the same time, however, access of the monomer must be kept allow for its further oxidation. It is known that MDO could catalyze oxygen reduction reaction (ORR), which occurs via a two-electron reduction mechanism in alkaline solution with the formation of hydrogen peroxide ion (HO₂^–). β-MnO₂ was chosen because, with its (1 × 1) tunnel structure (rutile-type), it has the best structural properties among the MDOs.

Figure 24 shows the cyclic voltammograms recorded for O₂ reduction in O₂ saturated 2 mol·L⁻¹ KOH solution (solid line) vs. argon saturated solution (dashed line). The O₂ reduction peak occurs at −506 and −365 mV for PBTh/β-MnO₂ and PPy/β-MnO₂ electrodes, respectively. The enhanced electrocatalytic effect of PPy/β-MnO₂ can be witnessed by a significant positive shift of the O₂ reduction potential from −412 to −365 mV and a decrease in the O₂ reduction peak current from 289 to 83 μA·cm⁻². In addition, PBTh/β-MnO₂ is gifted with very good electrocatalytic activity for ORR owing to more negative onset potential than β-MnO₂.

8. Nanocomposites

An ideal nanocomposite electrode, for supercapacitors, that possesses long cycle stability should contain a high-power density material (carbon-based) associated with a high-energy density compound (oxide). MnO₂ has high theoretical specific capacitance (1380 F·g⁻¹) but its main disadvantage is poor conductivity (10⁻⁵–10⁻⁶ S·cm⁻¹) that can be enhanced by the fabrication of various MnO₂/conductive matrix hybrid materials such as SnO₂/MnO₂ [181], multiwalled carbon nanotube (MWCNT)/MnO₂ [182] and C/MnO₂ nanomaterials [183].

8.1. MnO₂-Carbon Nanocomposite

For example, Chen et al., deposited MnO₂ nanoparticles on graphene oxide (GO) sheets that enhanced the electrochemical properties due to the chemical interaction between MnO₂ and GO [184]. Lv et al. [185] demonstrated the superior cycling performance (97% after 5000 cycles) of the nanocomposite formed by N-doped carbon tubes and Au-doped MnO₂ nanoparticles. Fan et al. [186] proposed a new composite of carbon nanotubes (CNTs)/graphene, composed of CNT pillars sandwiched between the graphene sheets that showed a specific capacitance as high as 385 F·g⁻¹. Graphenes decorated with flower-like MnO₂ nanostructures were fabricated by electrodeposition for electrodes of supercapacitors. The MnO₂ nano-flowers consisted of tiny rods with a thickness of less than 10 nm. The specific capacitance after the MnO₂ deposition was 328 F·g⁻¹ at a charging current of 1 mA with an energy density of 11.4 Wh·kg⁻¹ [187]. Song et al. [188] fabricated a nanocomposite for a supercapacitor composed of needle-like MnO₂ nanowire arrays on graphene synthesized by in-situ growth of MnO₂ nanowires on the surface of graphene nanosheets (GNS). The preparation is a simple redox reaction between KMnO₄ and GNS, which can produce the composite at large scale at low cost. The nanocomposite exhibited high-capacitance performance of 276 F·g⁻¹ at 0.5 A·g⁻¹. MnO₂ nanoparticle enriched poly(3,4-ethylenedioxythiophene) (PEDOT) nanowires were fabricated by simply soaking the PEDOT nanowires in KMnO₄ solution [189]. Due to their extremely high surface area the MnO₂ nanoparticles showed very high specific capacitance (410 F·g⁻¹) as supercapacitor electrode materials, as well as high storage specific capacity (300 mAh·g⁻¹) as cathode materials for the Li ion battery [190].

Long et al. [190] prepared ultrathin polymer coatings (10-nm thick) onto nanostructured birnessite-type MnO₂. The composite formed by electrodeposition of poly(o-phenylenediamine), which preserved the mesoporosity of MnO₂, showed good stability as electrode material in acid electrolytes. Yan et al. [191] investigated the compatibility of MnO₂ nanowires with SnO₂ to make a high performing electrode for supercapacitors. A specific capacitance of 800 F·g⁻¹ was achieved at a current density of 1 A·g⁻¹ in 1 mol·L⁻¹ Na₂SO₄ aqueous solution. MnO₂ nanowires were electrodeposited onto carbon nanotube (CNT) paper by a cyclic voltammetry technique [28]. This MnO₂ nanowire/CNT composite used as a flexible electrode for electrochemical supercapacitors displayed specific capacitances as high as ~167 F·g⁻¹ at a current density of 77 mA·g⁻¹ with faradic efficiency of 88% after 3000 cycles.

8.2. Organo-MnO₂

Composites termed as “organo-MnO₂” can host organic guest species that can capture and detect iodine in organisms and the environment (essential element in thyroid hormones). Such a composite layered δ-MnO₂ structure was synthesized by electrodeposition of MnSO₄·5H₂O and CoSO₄·7H₂O in aqueous solution in the presence of a cationic surfactant, cetyltrimethylammonium (CTA) [192]. Electrodeposition of the films was performed at constant potential of +1 V with a fixed charge of 200 mC·cm⁻². The oxidation of inserted Co²⁺ ions took part of the deposition process of MnO₂ at 70 °C. CTA molecules occupying the MnO₂ interlayer have the role of a sensing element that could extract I⁻ ions for solutions, while the Co-framework ions dopant achieved fast electron kinetics for the oxidation of I⁻ ions [192].

8.3. SnO₂-MnO₂ Composites

To improve significantly the electrochemical performance, MnO₂ particles can be coated with a conductive material, namely a thin SnO₂ layer (~20 nm). In addition, the particles are mixed with a few percent of conductive carbon that percolates through the structure and thus makes electrical contact between the particle and the current collector. Therefore, once an electron of any MnO₂ particle has reached any point of the surface of the particle, it can be driven by the electrical field up to the current collector through the conductive carbon in contact with the conductive SnO₂ layer. On the contrary, in the absence of carbon, even if the particles were coated with a SnO₂ layer, the electron would have to pass from one particle to the other one by the contact point between the particles so that the material would be insulating. This is illustrated in Figure 25.

On the other hand, if the particles were not coated, the electrical charge would not be homogenously distributed on the surface along the direction of the electrical field, so that the effective surface area ε_c would be much smaller that the surface area of the particle. This is also illustrated in Figure 25, where the Nyquist plots recorded by electrochemical impedance spectroscopy (EIS) show a large decrease of charge transfer resistance R_ct occurring after coating. A decrease of R_ct allows high rates for the charge/discharge (deinsertion/insertion) process [25].

To reduce the Mn dissolution in organic electrolyte by surface protection, SnO₂-MnO₂ composite powders were investigated by Hashem et al. [193], who also reported their electrochemical properties. Yan et al. [191] investigated the compatibility of MnO₂ nanowires with SnO₂ to make a high performing electrode for supercapacitors. A specific capacitance of 800 F g⁻¹ was achieved at current density of 1 A·g⁻¹ in 1 mol·L⁻¹ Na₂SO₄ aqueous solution. Nanosized SnO₂-MnO₂ composites were prepared by the wet-chemical method [142] based on a redox reaction between KMnO₄ and Mn(II) acetate. The SnO₂-coated samples were obtained from SnCl₂ as coating agent. The TEM micrographs of α-MnO₂ and SnO₂/α-MnO₂ are shown in Figure 26. These images indicate that the samples had almost spherical morphology [193]. Both α-MnO₂ and SnO₂/MnO₂ samples were nano-sized particles with a narrow size distribution centered at 200 nm.

The elemental analysis (ICP) shows a concentration of potassium ions of 5.5 mol% in SnO₂/α-MnO₂ samples. In α-MnO₂, the presence of Mn³⁺ originates from insertion of K⁺ ions in the tunnel and from oxygen vacancies. Therefore, a small oxygen deficiency results in large crystallite strains and lattice distortions, because Mn³⁺ is a Jahn-Teller ion. This was also evidenced by magnetic susceptibility measurements. For the coated sample, this residual concentration of Mn³⁺ (3d⁴ configuration) in the low-spin state (spin S = 0) is responsible for a decrease of the magnetic moment. The lower value for µ_eff = 3.87 μ_B is due to the presence of K⁺ ions in the (2 × 2) tunnels that have the chemical formula K_yMn⁴⁺_8−yMn³⁺_xO_16−z, where z is the concentration of oxygen vacancies and x = y + 2z the amount of Mn³⁺. The experimental results give x = 8.2% in pristine (P)-MnO₂, decreasing to 6.2% in SnO₂/MnO₂. Since y = 5.5 at%, the concentration of oxygen vacancies is z = 1.3 at% in P-MnO₂ and decreases to z = 0.3 at% in the SnO₂/MnO₂. Therefore, the SnO₂ coating has protected MnO₂ against the loss of oxygen.

Figure 27 shows the discharge curves for Li cells with α-MnO₂ and SnO₂-coated MnO₂ electrode material using 1.0 mol·L⁻¹ LiPF₆ in ethylenecarbonate- diethylcarbonate (EC-DEC) as electrolyte at the 1st and 45th cycle. It is obvious that the SnO₂ deposit acts as a protective layer around MnO₂ particles which prevents dissolution of Mn-ions into the organic electrolyte.

9. Concluding Remarks

The new technologies require urgently flexible and wearable energy storage devices. In this context, a breakthrough has been achieved by Zeng et al. [194], who reported a flexible quasi-solid-state Zn-MnO₂ battery with sandwiched structure consisting of a MnO₂@PEDOT (poly(3,4-ethylenedioxythiophene)) cathode, a Zn nanosheet anode ≈50 nm in thickness homogeneously grown on the carbon fibers without any binder, a separator, and a modified poly(vinyl alcohol) (PVA)/1 mol·L⁻¹ ZnCl₂/0.4 mol·L⁻¹ MnSO₄ gel electrolyte. The PEDOT shell ≈9 nm in thickness was used as a protection layer to increase the cycling life. This as-fabricated flexible quasi-solid-state battery had an energy density of 505 Wh·kg⁻¹ (34 mWh·cm⁻³) and peak power density 8.6 kW·kg⁻¹. It maintained 78% of its initial capacity (282.4 mAh per gram of MnO₂, 91.6 per gram of the total cathode) and nearly 100% coulombic efficiency after 300 cycles at current density 1.86 A·g⁻¹. An amount of 61.5% of the initial capacity was retained after 1000 cycles. This performance outperforms the most recently reported quasi-solid-state batteries. The great potential of Zn-MnO₂ as a flexible energy storage device combining low cost, safety, high energy density, and environmental friendliness is confirmed by the results obtained by Qiu et al. [195] using MnO₂ nanorod arrays and Zn nanoparticles uniformly deposited on N-doped porous carbon cloth as the free-standing cathode and anode, respectively. Using the same electrolyte, the device achieved an energy density of 440 Wh·kg⁻¹ and power density of 7.9 kW·kg⁻¹.

Until recently, only 10% of the theoretical capacity (617 mAh·g⁻¹) was accessible in rechargeable alkaline batteries. However, an important improvement has been recently achieved with a class of Bi-birnessite (a layered manganese oxide polymorph mixed with bismuth oxide (Bi₂O₃)) cathodes intercalated with Cu²⁺ that deliver near-full two-electron capacity reversibly for more than 6000 cycles [196]. The addition of MnSO₄ to the electrolyte is known to passivate the electrode surface [197]. Recently, however, a chemical conversion reaction mechanism between α-MnO₂ and H⁺ was evidenced when mild aqueous ZnSO₄-based solution was used as the electrolyte in a Zn/MnO₂ battery [198]. As a result, the capacity of this battery was 285 mAh·g⁻¹ (MnO₂), with capacity retention of 92% over 5000 cycles. These recent results show that the problems of the Zn/MnO₂ battery (limited cycling life and power) have now been entirely solved. Therefore, the alkaline Zn/MnO₂ batteries now outperform the Li-ion batteries not only by the lower price, but also by their performance, both in energy and in power density.

The future of MnO₂ as a supercapacitor element looks also bright. A composite with a core made of N-doped hollow carbon spheres (NHCS) and a shell composed of hierarchical birnessite-type MnO₂ nanoflakes was used as an anode for an asymmetric supercapacitor equipped with an NHCS cathode [199]. This supercapacitor in 1 mol·L⁻¹ Na₂SO₄ aqueous electrolyte operated in the voltage range 0–1.8 V, delivering an energy density of 26.8 Wh·kg⁻¹ at a power density of 233 W·kg⁻¹ with minimal capacitance drop (4.8%) and 100% coulombic efficiency over 4000 cycles. Comparable results were found with an asymmetric supercapacitor based on a MnO₂ nanoflake/carbon nanotube core-shell particle composite [170]. We can thus conclude that MnO₂ now has great potential for energy conversion and storage applications.

Still, further improvements are predictable in the near future. It will be difficult to increase the energy density since we have already mentioned that its experimental value is close to theoretical for the most advanced electrodes [200], but the power density can still be improved by working on the porosity of the electrode material. For this purpose, synergetic effects between the research on batteries and the research on supercapacitors is expected as porosity increases the effective surface area with the electrolyte, which implies an increase of the power density, provided that the solid-electrolyte interface (SEI) is well controlled; porosity is also desired for capacitive electrodes [201,202,203]. A capacitance larger than 200 F·g⁻¹ has already been obtained with mesoporous MnO₂ nanosheets [204]. Electrodes with optimized porosity should then also be built to increase the power density.

As we just mentioned, however, this increase of the effective surface area in contact with the electrolyte will be efficient only if the SEI is well controlled. Therefore, further studies should be devoted to the investigation of the interface between MnO₂ and its environment. The coating of the particles, which can be used to increase the electrical conductivity and to protect the particle, is presumably perfectible. We have persisted in this review with structural properties that are so important to determine the rechargeability of the MnO₂-based devices. In particular, we have persisted with the role of K⁺ ions on the physical and chemical properties of the DMO. In this context, further investigations should be useful to understand the role of different dopants on the morphology and the structural properties of MnO₂.

In their review on EMD batteries, Huang et al. [22] reported that engineers in the USA built a prototype EMD battery that can be rechargeable up to 90% within two minutes, and envisioned applications in electric cars. Actually, it is even overperforming for this purpose, because the grids cannot deliver such high power to recharge batteries. The maximum that the grids can deliver for private use without destabilization of the whole grid supply is the order of 50 kW·h. However, these high-power densities can be useful to solve the intermittence problems for integration of renewable energies on the grids.

All these improvements in the MnO₂ electrochemical properties and the performance already achieved make them already attractive and quite competitive with others and opens the route to high production, which should sustain further research along the lines mentioned above.