Closed Battery Systems

Abstract

Battery technologies are expected to strongly contribute to the global energy storage industry and market. Among the several promising battery technologies, Li-metal batteries, all-solid-state Li batteries, and beyond-lithium systems are discussed in this chapter. Li metal represents a key anode material for boosting the energy density of batteries, but the formation of Li dendrites limits a safe and stable function of the system. The use of solid-state electrolytes allows a safer battery operation, by limiting the electrolyte flammability and dendrite formation, yet the performance is insufficient because of slower kinetics of the lithium ion. Possible solutions against these critical problems, especially through the discovery of new materials, are here discussed. Moreover, other innovative technologies based on Na, Ca, and Mg, so-called beyond-lithium batteries, are presented. Insights into these emerging battery systems, as well as a series of issues that came up with the replacement of lithium, are described in this chapter. Focus is particularly placed on development of battery materials with different perspectives, including performance, stability, and sustainability.

1 Lithium Metal Batteries

Electrochemical plating/stripping of lithium metal at the negative electrode of aprotic batteries can in principle provide significant functional improvements compared to Li ion intercalation electrodes, in terms of the outstanding performance of the Li metal electrode (theoretical capacity = 3860 mAh g⁻¹ and E° = −3.04 V vs. SHE, standard hydrogen electrode, cf. 372 mAh g⁻¹ and ≈ − 2.9 V vs. SHE for graphite) [1]. A high energy density of 400–600 Wh kg⁻¹ can be achieved when lithium metal anode is combined with intercalation-type cathode materials such as LiFePO₄, LiCoO₂, lithium–nickel–manganese–cobalt oxide (NMC), and lithium–nickel–cobalt–aluminum oxide (NCA). The use of Li metal anode is also proposed in Li–sulfur (Li–S) batteries and Li–oxygen (Li–O₂) batteries, thus enabling further increase in the energy density of these next-generation battery chemistries. However, and despite the extraordinary research efforts made in the last five decades by researchers all over the word, lithium metal electrodes in secondary, i.e., rechargeable, batteries are still far from being commercially feasible. In fact, there are fundamental limitations that hinder the reversibility of lithium metal stripping/deposition in aprotic electrolytes based on liquid, polymer, or crystalline components [2].

In the lithium metal battery (LMB) technology, the reduction potential of Li is below that of conventional organic electrolytes. Therefore, lithium metal electrodes show Coulombic efficiencies upon cycling lower from unity in batteries. This fact makes the electrolyte a critical component in the development of LMB. In addition, the uneven deposition morphology of Li and inhomogeneous electrochemical dissolution lead to dendrite and dead lithium formation, lowering the Coulombic efficiencies [3]. A wide variety of approaches has been proposed to tackle these drawbacks, rooted in thermodynamics and kinetics of the lithium stripping/deposition. Overall, the electrochemical reversibility of lithium metal electrodes can be enhanced by the following four main strategies [4]:

• 1. Optimization of the electrolyte formulation to control the solid electrolyte interphase (SEI) formation over the electrodes

• 2. Pre-deposition of an artificial SEI over the lithium metal electrode

• 3. Exploitation of open three-dimensional scaffolds with controlled meso- or nano-morphologies to buffer volume expansions and self-heal the formation of dead lithium

• 4. Substitution of liquid electrolytes with solid polymer and inorganic materials, and their combination into hybrid electrolytes, to limit side reactions at the lithium metal surface

1.1 Mechanism of the Electrochemical Lithium Stripping/Deposition in Liquid Electrolytes

Lithium metal is an extraordinary reducing agent with the redox potential of E° = −3.04 V vs. SHE. Almost all aprotic liquid solvents with polar groups suitable for the formulation of secondary battery electrolytes are thermodynamically unstable in contact with lithium. This thermodynamic constraint causes an immediate degradation of solvent molecules in contact with the surface of lithium metal through a pseudo-corrosion mechanism [5]. The degradation is initiated by the irreversible reduction of one solvent molecule:

$$ \mathsf{solvent}+\mathsf{Li}\to {\mathsf{solvent}}^{\bullet -}+{\mathsf{Li}}^{+}\ \left(\mathsf{solv}\right) $$

and this reaction unavoidably creates pitting holes over the lithium metal through the release of solvated Li⁺ ions in the electrolyte, simultaneously generating the unstable radical solvent^{• −} anion in a doublet state [6]. This radical molecule easily undergoes further reactions with other solvent molecules:

$$ {\mathsf{solvent}}^{\bullet -}+\mathsf{solvent}\to {\left(\mathsf{solvent}\hbox{--} \mathsf{solvent}\right)}^{\bullet -} $$

to form a short-chain radical polymer (solvent–solvent)^{• −}. The radical can also react with the salt anion (e.g., LiPF₆) to form insoluble lithium salts like lithium fluoride:

$$ {\mathsf{solvent}}^{\bullet -}+{\mathsf{LiPF}}_{\mathsf{6}}\ \left(\mathsf{solv}\right)\to \mathsf{LiF}\ \left(\mathsf{solid}\right)+{\left({\mathsf{PF}}_{\mathsf{5}}\hbox{--} \mathsf{solvent}\right)}^{\bullet -}. $$

In turn, unstable alkyl carbonate–fluorophosphate adducts, i.e., (PF₅–solvent) ^{• −}, are formed that easily further degrade through irreversible multistep processes, leading to a precipitation of the insoluble species and/or an alteration of the electrolyte composition. Another degradation process of the solvent^{• −} radical anion is its intramolecular breakup promoted by a further reduction, leading to the formation of gaseous species (i.e., molecular hydrogen, ethylene, and other small volatile organic molecules like formaldehyde, acetylene, etc.) [7]. These multifarious radical chain reactions, initiated by the inevitable reduction of solvent molecules, lead very rapidly to the precipitation of insoluble inorganic and organic by-products (e.g., Li₂CO₃, Li₂O, lithium alkyl carbonates, esters, etc.) over the surface the lithium metal, thus forming a passivation film (i.e., a natural solid electrolyte interphase, namely, n-SEI) [8]. This n-SEI completely passivates the surface of lithium metal, within the pitting holes as well as in the outer areas, but the pseudo-corrosion is not remarkable in the latter. Therefore, the n-SEI is unavoidably inhomogeneous in the nanometer scale, with significant compositional and morphological differences. Furthermore, in the initial electrochemical lithium deposition or oxidation of the lithium electrode, further electrolyte degradation occurs ending in the complete passivation of the surface by a thick passivation n-SEI film that is reminiscent of the original natural SEI [9].

The compositional inhomogeneity of the n-SEI originates from the so-called weak points, where the accumulation of organic by-products makes the local electric resistance larger compared to the surrounding areas of the electrode. These weak points become the most favorable nucleation sites for the electrochemical deposition of lithium metal and the electrochemical stripping of lithium ions. Overall, the surface of the lithium metal electrodes upon cycling will alter, driven by the local fluctuation of the n-SEI transport properties, thus leading to the aggravation of the original morphological inhomogeneities with formation of dendrites and pitting holes [10].

Once formed, dendrites increase remarkably the volume of the lithium metal electrode and grow preferentially toward the counter-electrode, following the electric field lines. This growth leads to internal short circuits and catastrophic cell failures. Moreover, dendrite formation exposes additional fresh lithium surface to the electrolyte, thus leading to further electrolyte degradation and accumulation of n-SEI. This mechanism is particularly unfavorable as it drives the formation of dead lithium upon stripping. This is because dendrites can be electrochemically dissolved during discharge, and therefore it is possible that a portion of the metallic lithium dendrite, passivated by the unfavorable n-SEI, may lose an electronic contact with the bulk of the electrode. Unavoidably, these isolated lithium dendrites cannot undergo further electrochemical reactions leading to the loss of active material [11].

In general, the n-SEI grown on lithium metal before and upon cycling should meet several requirements to mitigate the dendrite growth, enhance the reversibility of plating/stripping, and suppress the formation of dead lithium as schematically summarized in Fig. 10.1.

Fig. 10.1

Optimal properties of SEI layers over metallic lithium [12]

Unfortunately, the n-SEI layer formed in typical LIB electrolytes is unable to fulfill these demands, and, as already mentioned, different strategies have been considered to tackle this challenge, in particular forming artificial SEI layers or altering the nature of the n-SEI. Another radical approach is to replace liquid electrolytes moving to solid-state (electrolyte) batteries.

1.2 SEI Modulation via In Situ Formation

The formation of the SEI can be modulated by electrolyte additives that can alter the degradation mechanism to form more homogeneous and stable passivation films over the surface of the lithium metal electrodes. These sacrificial chemical species initially consume some metallic lithium of the electrode, but this consumption is generally limited and will end once the lithium electrode surface is passivated in situ by a modified solid electrolyte interphase (m-SEI). These electrolyte additives belong to two main groups depending on their main effect, i.e., reduction-type and reaction-type additives [13].

Reduction-type additives have a relatively high redox potential and are reduced upon lithium deposition prior to any unfavorable effect on the electrolyte [14]. Their decomposition products form an insoluble film, protecting the electrode/electrolyte interface, with enhanced transport properties compared to the n-SEI formed spontaneously without the additive. Reduction-type additives are divided into two subgroups. The first subgroup consists of reactive compounds containing unsaturated bonds like vinylene carbonate (VC) or fluoroethylene carbonate (FEC) and promotes the accumulation of polymerized organic species over the surface of the lithium metal electrodes. The second subgroup includes reductive agents aiding the SEI formation. This class of reductive additives is electrochemically reduced before the electrolyte decomposition by-products precipitate onto the lithium surface. These additives can also react with radical species generated during the initial solvent reduction process, thus terminating the radical chain reaction that leads to the uncontrolled n-SEI formation. The most common chemical species that belong to this class are sulfur-containing chemical like sulfolane, ethylene sulfite, or dialkyl-sulfone [15]. Their degradation ends with the formation of insoluble lithium sulfites or alkyl lithium sulfites.

Reaction-type additives are typically molecules with scavenging abilities. These species can react with intermediates of the natural degradation radical chain reaction of the electrolyte in contact with lithium surfaces, altering the final composition of the SEI by promoting the accumulation of electrochemically inert species with favorable transport properties. (Trimethylsilyl)isothiocyanate can easily scavenge PF₅ as well as phosphite-containing compounds, thanks to their strong nucleophilic character [16].

Besides reaction-type and reduction-type additives, it is important to mention the possible use of co-salts of multivalent cations (e.g., Mg²⁺, Ca²⁺, Zn²⁺ Fe²⁺, In³⁺, and Ga³⁺). Many of these cations can alloy with lithium upon reduction worsening the ionic transport properties across the entire electrode surface. This effect leads to a smaller ionic conductivity but mitigates the uncontrolled growth of lithium dendrites [17].

1.3 Ex Situ Deposition of Artificial SEI

The second general strategy to improve the reversibility of lithium stripping/deposition is the ex situ preformation of an artificial SEI layer (a-SEI) on the surface of the metallic electrode. This strategy cannot fully prevent the spontaneous reactivity of the electrolyte with the lithium surface but can strongly mitigate its occurrence and therefore strongly reduce the chemical and morphological inhomogeneity of the lithium electrode surface. In general, a-SEIs over lithium metal electrodes can be synthesized by atomic-layer deposition (ALD), aeration, or chemical coating in a liquid precursor solution [18].

ALD is an advanced thin-film fabrication technique capable of producing homogeneous and ultrathin films at room temperatures or slightly above [19]. The final surface film is extremely thin, thanks to the careful control of the mass-loading of the a-SEI allowed by the technique. These ultrathin films are typically constituted by Al₂O₃ that, during the electrochemical deposition of lithium, firstly converts into the highly conductive lithium aluminate Li_xAl₂O₃ that stabilizes further the a-SEI and allows a remarkable limitation of the dendrite growth [20].

Aeration is a chemical deposition method from the gas phase, in which the lithium metal surface reacts with almost permanent gaseous species like N₂ or O₂ to form homogeneous nitride or oxide layers. Since Li₃N is an electrochemically stable and fast ion conductive chemical species, the precipitation of a continuous and homogeneous thin film of lithium nitride over the surface of lithium metal acts to prevent side reactions between lithium metal and the electrolyte without hindering the lithium ion mobility [21].

The a-SEI, starting from selected liquid precursor solutions, can be obtained by using the so-called dip-coating or drop-casting method. A suitable chemical composition of the precursor solutions promotes the formation of a-SEI with controlled compositions and surface moieties. In fact, the constituent species of the precursor solution degrades due to the direct chemical reaction with the metallic lithium forming a preliminary passivation film [22]. This a-SEI can mitigate the unavoidable reactivity of the metallic lithium surface with the electrolyte upon cycling. The composition of the precursor solutions can modulate the composition of the resulting a-SEI allowing the control of the transport and mechanical properties of the a-SEI. An almost innumerable number of precursor solutions have been proposed in the literature, all leading to an improvement in the overall reversibility of the lithium plating/stripping. Among the many possibilities, one that should be mentioned is the use of polyphosphoric acid solution in organic solvents like dimethyl sulfoxide that promotes the formation of a Li₃PO₄-rich a-SEI, with excellent chemical stability, high Young’s modulus (10–11 GPa), and high lithium ion conductivity [23].

1.4 3D Engineering of the Electrode Morphology

The third possible approach is to modify the morphology of the surface where Li is plated (either Li metal or directly the current collector) by an engineering at the nanoscale. Mechanically robust and chemically inert three-dimensional scaffolds or generic frameworks are used to facilitate homogeneous deposition of lithium metal, thus indirectly limiting the degradation of the electrolyte over fresh lithium surface [24]. The coating of lithium metal electrodes with very thin amorphous carbon-based hollow nanostructures can modulate the Li⁺ ion transport pathways toward the metal surface, leading to the nucleation and growth of the lithium deposits either over Li or the current collector. Furthermore, lithium metal can deposit within the hollow nanostructures with a minimal contact with the electrolyte, thus mitigating any further degradation. Also, the use of two-dimensional materials like graphene or hexagonal boron layer can modulate the lithium stripping/deposition reaction by inducing specific growth morphologies and the formation of a tailored a-SEI [25].

The use of open nickel foams as a current collector and support for the lithium deposition is also an effective strategy to accommodate the huge volumetric changes experienced by conventional host-less flat electrodes to lead homogeneous lithium deposition. In fact, the 3D open foam can be infused with molten lithium by capillary force forming a composite scaffold with a minimized interfacial resistance and empty spaces to buffer volumetric changes upon stripping and deposition [26].

2 All-Solid-State Lithium Metal Batteries

The use of commercially available organic liquid electrolytes in LMBs, as well as in Li ion batteries (LIBs), unavoidably poses major safety issues because of their flammability and strongly exothermic reactivity with lithium that can easily self-heat the battery in case of accidents, leading to catastrophic explosions. The replacement of liquid electrolytes with a solid electrolyte (SE) allows to tackle this serious drawback drastically and reduces the risk of fire and explosion even in case of major thermal, mechanical, or electrical abuse. As a further advantage, the use of SEs as an electrolyte having a function of separator increases the overall volumetric energy density of the batteries, thus improving the overall performance [27].

The ideal set of properties of SEs is [28, 29]:

• 1. High ionic conductivity (>10⁻⁴ S cm⁻¹, better >10⁻³ S cm⁻¹ in the case of thick composite electrolytes) with a high Li⁺ transference number (t_Li+), especially at low temperatures

• 2. Good chemical compatibility with other battery components including lithium metal

• 3. Good thermal and electrochemical stability at a wide range of temperatures and voltages for the constant operation of cells, including the thermal stability also in terms of the absence of phase transition so as to avoid the formation of low conductive solid phase

• 4. Minimal interfacial resistance between SE and electrodes

• 5. High electronic area-specific resistance, resulting in conductivities <10⁻¹² S cm⁻¹, to prevent self-discharge

• 6. Appropriate mechanical strength to resist dendrite growth

• 7. Good affordability in terms of less environmental impact, allowing for simple and low cost fabrication of both SE itself and devices

All-solid-state batteries with lithium metal electrodes (ASS-LMB) are classified as “generation 4b batteries” by the EU commission. However, despite the excellent functionality of SEs, several drawbacks hinder their development beyond the lab-scale and the final commercialization. The most relevant challenges are the unsatisfactory ionic conductivity compared to classical liquid electrolytes, the large impedance at the electrode–electrolyte interfaces, and the electrochemical instability against lithium metal of SE constituents.

2.1 Classification of Solid Electrolytes

Solid-state electrolytes can be either inorganic solid electrolytes (ISEs), solid polymer electrolytes (SPEs), or their composites. The former electrolyte can further be divided into oxide types, sulfide types, and others including hydride, borate, and phosphate types. The oxide-type ISEs have good chemical and electrochemical stability. The design of a material with a good air stability and low toxicity is possible. They exhibit a fast ionic conductivity in bulk, up to 10⁻³ S cm⁻¹ at room temperature; however, to achieve this ionic conductivity, a sintering procedure at high temperatures, conventionally at around 1000 °C, is necessary to reduce grain boundary resistance between electrolyte particles. In contrast to this, sulfide-type ISEs have a low oxidation stability and high reactivity in the presence of moisture but exhibit a high ionic conductivity merely by pressing the materials at room temperature (so-called low-temperature sintering). This is because of a larger size S²⁻ which broadens ion conduction pathways in the electrolyte structure and the higher polarizability of S²⁻ which weakens the interaction with Li⁺. On the other hand, SPEs have several advantages in manufacturing processes, such as simple production of large-area films, easy formation of a seamless interface with the electrodes, and possible handling of SPE in ambient air. However, it is not easy to find a suitable combination of a polymer matrix and additive salt that can exhibit ionic conductivity higher than that of ISEs, except for the combinations forming gel polymer electrolytes.

2.2 Ion Conduction Mechanisms of Solid Electrolytes

ISEs normally consist of two sublattices: a crystalline framework composed of immobile ions and a sublattice of mobile ions [30]. Therefore, Li⁺ ions move within an essentially static framework through ion hopping (i.e., the Grotthuss mechanism), which is favorable with regard to a faster ion conduction. The mobile species in a crystalline solid need to pass through periodic bottleneck points, which defines migration energy. To reduce the energy barrier and achieve faster ion conduction, (1) the number of mobile ions and their hopping sites, as well as the vacancy of the mobile ions, (2) the size of bottleneck points, (3) the degree of structural order in a mobile ion sublattice, and (4) the presence of highly polarizable anions in sublattices are critical factors [31]. In glassy inorganic materials, the ion conduction mechanisms are quite like those in crystalline structures. The ions at local sites, being excited to neighboring sites, diffuse collectively on a macroscopic scale. In contrast to these, in SPEs, the motion of Li⁺ ions is mediated by the dynamics of the host polymer, i.e., the vehicular mechanism, thereby restricting the ion conduction to a relatively slow speed. The segmental motions of polymer chains create free volumes that allow for the migration of ions coordinated by the polymer polar groups, and ions migrate from one coordinate site to another, promoted by the segmental motions. The ionic conductivity, in this case, is strongly dependent on the crystallinity of the host polymer.

2.3 Development of Inorganic Solid Electrolytes

The diffusion of ions in crystalline solids is significantly affected by both ion valence and ionic radius of the migrating species because these influence electrostatic interactions between mobile ions and cations forming the structural skeleton [31]. In addition to this, the natures of the ligands and metals composing the skeleton of the host framework have large impacts on the performance of ISEs because they determine the channel size for ion migration. In the case of Li⁺ diffusion, since controlling the bottleneck size for the ion diffusion has been successful in enhancing Li⁺ conductivity and reducing the activation energy, structural tuning by cation substitutions within a given structural framework has been intensively studied. Among those investigated, the most attractive Li⁺ conductors, based on oxides such as natrium super ionic conductor (NASICON) type, garnet type, and perovskite type, as well as sulfides such as lithium super ionic conductor based on sulfides (thio-LISICON) type, Li₁₀MP₂S₁₂ (LGPS) type, argyrodite type, and Li₇P₃S₁₁ type, are summarized in this section and Table 10.1.

Table 10.1 Properties of different kinds of ISEs

NASICON-Type (e.g., LATP/LAGP)

NASICON-type Li⁺ conductors are represented by the general formula of AM₂(PO₄)₃, in which A = Li and M = Ge, Ti, Zr, etc., forming a three-dimensional framework of MO₆ octahedra alternatively connected with PO₄ tetrahedra by sharing their vertices [31]. Lithium ions occupy two different sites in the structure, so-called A1 and A2 sites, and their migration occurs via ion hopping between these sites. The limitation for the ion motion comes from triangular oxygen windows, which separate the A1 and A2 sites, known as the bottleneck [38]. Therefore, optimization of the bottleneck size effectively improves the ionic conductivity and reduces its activation energy.

The first strategy for this issue is the use of larger M ion in AM₂(PO₄) structure [33]. For example, the ionic conductivity of LiGe₂(PO₄)₃ is only 6.62 × 10⁻⁹ S cm⁻¹ [32]. In the case of LiTi₂(PO₄)₃ and LiZr₂(PO₄)₃, where Ge⁴⁺ is replaced with larger ions, their ionic conductivities improve to 2 × 10⁻⁶ S cm⁻¹ and 3.8 × 10⁻⁵ S cm⁻¹ at room temperature, respectively. The second strategy is the aliovalent substitution, resulting in A_1 + xM′_xM_2–x(PO₄)₃, in which M′ = Al, Ga, La, etc. This increases the Li⁺ concentration and also its mobility [33]. The concentration of M³⁺ needs to be limited to ∼15% (x = 0.3) to avoid the formation of a secondary phase due to an ionic radius mismatch [31]. Al-doped NASICON-type materials, such as Li_1 + xAl_xTi_2 − x(PO₄)₃ (LATP) and Li_1 + xAl_xGe_2–x(PO₄) (LAGP), are known to exhibit ionic conductivities of ~3 × 10⁻³ S cm⁻¹ (when x = 0.3) [31] and 1.0 × 10⁻³ S cm⁻¹ (when x = 0.5), respectively [33]. Both LATP and LAGP are stable at relatively high potentials and in air, while they are unstable at low potentials because Ti⁴⁺ is easily reduced. The formation of a protective layer based on polymers, LiPON, or other ISEs allows the use of this kind of materials with lithium metal anodes [39].

Garnet-Type (e.g., LLZO)

The general formula of garnet-type materials is A₃B₂(XO₄)₃, in which A = Ca, Mg, Y, La, etc., B = Al, Fe, Ga, Ge, Mn, Ni, V, etc., and X = Si, Ge, Al, etc., forming eightfold, sixfold, and fourfold coordinated structures centered on A, B, and X cations, respectively [32]. Li⁺-conductive garnets have the general composition of A₃B₂(LiO₄)₃. The Li⁺ content can be increased by the aliovalent doping, and several materials with different stoichiometry such as the Li3 series, Li₃Ln₃Te₂O₁₂ (Ln = Y, Pr, Nd, etc.), Li5 series, Li₅La₃M₃O₁₂ (M = Nb, Ta, Sn, etc.), Li6 series, Li₆ALa₂M₂O₁₂ (A = Ca, Sr, Ba, etc., and M = Nb, Ta, etc.), and L7 series, Li₇La₃M₂O₁₂ (M = Zr, Sn, Hf, etc.), have been reported [32]. Garnets with Li⁺ concentration > 3, e.g., Li₇La₃Zr₂O₁₂ (LLZO), are identified as stuffed lithium garnets [40]. In these stuffed lithium garnets (i.e., Li5-7 series), Li⁺ occupies both tetrahedral and highly distorted octahedral coordination sites, while in the conventional garnets (i.e., Li3 series), Li⁺ occupies only a tetrahedral coordination. Li ions in octahedral site migrate easily, while those in the tetrahedral site are not.

In the case of Li7 series garnet-type materials, including LLZO, the material can possess a cubic phase featuring a disordered lithium arrangement in addition to a tetragonal phase featuring a fully ordered lithium occupancy [32, 35]. The formation of a cubic structure is critical to achieve a high level of ionic conductivity. Aluminum doping is known to stabilize the cubic phase of LLZO and enhance Li⁺ conductivity in one to two orders of magnitude compared to undoped tetragonal LLZO [35]. Furthermore, the ionic conductivity can be increased to 1.32 × 10⁻³ S cm⁻¹ by doping Ga³⁺ [35]. Regarding the electrochemical stability, garnet-type electrolytes are favorable in terms of showing a stability against lithium metal. In the case of LLZO, however, the material is lithiophobic [41], resulting in a formation of the interphase with a high charge transfer resistance. Softening of the lithium at 170–175 °C (about 5–10 °C below Li melting temperature) [1] and the formation of lithiophilic coating such as those based on ionic liquids or polymers [42, 43] are known to be an effective solution for this problem. Instability at the interface with cathode has also been reported. The coating formation is effective also in this case in preventing oxidations of the electrolyte.

Perovskite-Type (e.g., LLTO)

Perovskite structures are described as follows: ABO₃, in which the A-site ions (typically divalent alkaline earth metal ions) locate at the corners of a cube structure, B ion (typically transition metal ions, such as Ti⁴⁺) at the center, and oxygen atoms at the face-centered positions. Based on this framework, 12-fold coordination is formed around A sites and sixfold coordination (BO₆) around B sites by a sharing corner with each other [31]. A Li⁺ conductor can be obtained by replacing the divalent alkaline earth metal ions at A sites with the trivalent rare-earth element such as La³⁺ and monovalent Li⁺. Just like the other ISEs, the bottleneck size for the ion conduction needs to be controlled to improve ionic conductivity, and this has been done by introducing large rare-earth or alkaline earth metal ions in the A site [31]. The presence of vacancy in the material, more specifically Li_3xLa_2/3–x□_1/3–2xTiO₃ (LLTO, 0 < x < 0.167, □ represents vacancy), for example, also allows for the enhanced bulk conductivity of ∼10⁻³ S cm⁻¹ at room temperature. In such materials, La-rich and La-poor (i.e., Li-rich) layers are formed and alternately stacked. Within this structure, Li⁺ transport is restricted to the Li-rich layers, and minimal transport occurs between these layers, which limits Li⁺ transport to be two-dimensional. In polycrystalline structures, Li-rich layers in each grain are misaligned, and this reduces the ionic conductivity of the materials one to two orders of magnitude lower than that of the bulk [44]. As observed in other materials containing Ti⁴⁺ with less resistance to reduction, LLTO is not stable in a low-potential region, making it incompatible with the lithium metal anodes. Therefore, the partial or complete substitution of Ti⁴⁺ with Sn⁴⁺, Zr⁴⁺, Mn⁴⁺, and Ge⁴⁺ has been reported. These modifications need to be carried out so as not to form a secondary phase and not to decrease the Li⁺ concentration in the materials [45].

Thio-LISICON-Type and LGPS-Type

Thio-LISICON-type electrolytes can be obtained by substituting O²⁻ by S²⁻ in a LISICON structure, specifically from Li_xM_1–yM′_yO₄ to form Li_xM_1–yM′_yS₄ (M = Si or Ge; M′ = P, Al, Zn, Ga, etc.) [29]. The thio-LISICON-type electrolytes exhibit higher ionic conductivity compared with the LISICON electrolytes [31]. This is because of a larger size and higher polarizability of sulfide ions as discussed before for the classification of SEs. Depending on the valence of the cations, lithium vacancies in the structure can be controlled which affect Li⁺ conduction significantly [46]. The maximum conductivity of the thio-LISICON family can be as high as 2.2 × 10⁻³ S cm⁻¹ in the case of Li_3.25Ge_0.25P_0.75S₄ [36]. Thio-LISICON-type electrolytes tend to be unstable faced with lithium metal. However, their high ionic conductivity established this class of electrolytes as potential ISEs for the future of ASS batteries.

Li₁₀MP₂S₁₂ (M = Si, Ge, or Sn) and Li₁₁Si₂PS₁₂ are frequently considered as one of the thio-LISICON-type electrolytes [31]. However, thio-LISICON-type and these electrolytes have different crystalline structure, specifically based on orthorhombic and tetragonal crystal systems, respectively [47], and therefore can be regarded as different class of materials, namely, LGPS type. The electrolyte consists of (Ge/P)S₄ tetrahedra alternating with LiS₆ octahedra, forming a one-dimensional chain along the c-axis by sharing a common edge. These linear chains are connected to each other by PS₄ tetrahedra by sharing a vertex with LiS₆ octahedra [36], forming a backbone structure. Lithium ions locate in the empty space of the backbone structure by forming LiS₄ tetrahedra and create a one-dimensional LiS₄ chain along the c-axis, which is available for Li⁺ conduction. In addition to this conductive channel, Li⁺ diffusion is possible between neighboring conduction paths through the LiS₆ tetrahedra. Aliovalent doping enhances ionic conductivity, and, in particular, Li_9.54Si_1.74P_1.44S_11.7Cl_0.3 exhibits ionic conductivity of 2.5 × 10⁻² S cm⁻¹ [32].

Argyrodite Type

Lithium argyrodite, Li₆PS₅X (with X = Cl, Br, or I), is also one of the sulfide-based Li⁺ conductors, having a face-centered cubic lattice of X⁻, in which PS₄ tetrahedra locate as if isolated by the anions [31]. Li⁺ ions form Li₆S octahedra and distribute over the remaining tetrahedral interstices [31]. In the case of Li₆PS₅Cl and Li₆PS₅Br, the halide ions can partially occupy the place of S²⁻ in the Li₆S octahedra due to an X⁻/S²⁻ site disorder, which promote Li⁺ mobility significantly. As a result, annealed Li₆PS₅Cl and Li₆PS₅Br exhibit high ionic conductivities of 1.9 × 10⁻³ S cm⁻¹ and 6.8 × 10⁻³ S cm⁻¹, respectively [36]. In contrast to these, the site disorder does not occur in the case of Li₆PS₅I due to a large size of I⁻, and annealed Li₆PS₅I shows the low ionic conductivity of 4.6 × 10⁻⁷ S cm⁻¹ [36].

Li₇P₃S₁₁ Type

Li₇P₃S₁₁ is glass-ceramic-type sulfide-based electrolytes, which can be obtained through the crystallization of glassy electrolytes formed by a binary mixture of Li₂S–P₂S₅. Binary mixtures of Li₂S–MS_n, such as Li₂S–P₂S₅, Li₂S–B₂S₃, and Li₂S–SiS₂, mixed by a mechanical milling technique exhibit glassy phase with conductivities of 10⁻⁵–10⁻⁴ S cm⁻¹ at room temperature. The conductivity can be improved in the order of 10⁻³–10⁻² S cm⁻¹ by heat treatments because the glassy electrolyte softens and reduces the grain boundary resistance during the crystallization process [36]. Particularly, Li₇P₃S₁₁, obtained in 70% Li₂S–30% P₂S₅ in molar ratio, exhibits the ionic conductivity of ∼1.7 × 10⁻² S cm⁻¹ [37]. Li₇P₃S₁₁ consists of P₂S₇⁴⁻ di-tetrahedra and slightly distorted PS₄ tetrahedra. An inherent flexibility of readily fluctuating P₂S₇⁴⁻ polyhedra enables the fast Li⁺ migration in the material [48]. The ionic conductivity of Li₇P₃S₁₁ phase is higher than those of highly conductive thio-LISICONs, which mainly contain PS₄³⁻ [45].

Overall, sulfide-type ISEs possess considerable high ionic conductivities. However, their drawback is their instability. Several Li₂S–P₂S₅ binary systems are known to generate H₂S upon an exposure to moisture [45]. In addition, at the interface with lithium metal, multiple solid phases including Li₂S, Li₃P, Li₁₇Ge₄, and polyphosphide are usually formed, and the most common solution for this problem is the formation of a-SEI on lithium metal [1]. Also, for the cathode side, the formation of a protective layer, so-called a buffer layer, usually based on lithium metal oxide (e.g., LiNbO₃, Li₂ZrO₃) or lithium borate (e.g., Li₃B₁₁O₈), is effective [1].

2.4 Solid Polymer Electrolytes

Non-swollen dry polymer solid electrolytes, so-called SPEs, with suitable physicochemical and mechanical properties can be easily processed into thin separators. The advantages of SPEs as compared to the ISEs are easy fabrication, better scalability, high levels of safety, and flexible shapes. Some polymers with a high polarity, such as poly(ethylene oxide) (PEO) and poly(ethylene carbonate) (PEC), dissolve a large variety of lithium salts, such as lithium triflate (LiCF₃SO₃), lithium bis(fluorosulfonyl)imide (LiFSI), and lithium bis(trifluoromethanesulfonyl)imide (LiTFSI). The cations and anions dissociated in the polymer matrices are both mobile, and therefore SPEs act as a dual-ion conductor. Since the ion motion in the polymer matrices is assisted by the segmental motion of polymer chains, it mostly occurs within the amorphous fraction of the polymer matrix above its glass transition temperature (T_g). In this temperature range, the SPEs exhibit an ionic conductivity of around 10⁻³–10⁻⁴ S cm⁻¹. To reduce T_g, but to still retain the solid state, copolymerization, insertion of a branch structure, and cross-linking have been undertaken [49]. In addition, the use of inorganic additives such as Al₂O₃-, ZrO₂-, TiO₂-, SiO₂-, and Li⁺-conductive ceramic fillers (i.e., ISEs) has been confirmed to improve ionic conductivity and, mainly, enhance the mechanical strength of the resulting polymer membranes [50].

Despite these promising features, many drawbacks hinder the application of SPEs in LMB such as:

• 1. Small t_Li+, generally lower than 0.5

• 2. Occurrence of lithium dendrite growth and dead lithium formation, leading to poor efficiency and limited reversibility upon cycling

• 3. Low electrochemical anodic limit, around 4–4.1 V vs. Li⁺/Li, that limits the choice of cathode.

Many efforts are presently devoted to tackle these issues. The fixation of an anion structure such as carboxylate, sulfonate, sulfonylimide, etc., onto a polymer backbone allows the SPEs to have the t_Li+ of unity. In this case, the effective dissociation of the immobilized anions and Li⁺ is essential for the ion conduction. Recently, fixation of sulfonylimide on SPEs is dominating the material development because a delocalized negative charge of the anion is favorable for the ion dissociations [51]. In addition, since an enlarged conjugation structure adjacent to the anion structure can reduce the ionic interactions, several polystyrene derivatives with functional structures at the para position have been designed as Li⁺-conductive SPEs [51].

Besides PEO, also PEC has been proposed and demonstrated as a potential polymer matrix having the favorable combination of good physicochemical properties and satisfactory conductivity even at low temperatures. PECs show also a better anodic stability compared to PEO, thus allowing the use of high-voltage cathode such as layered oxides. On the other hand, the mechanical properties of PECs counterbalance the benefits, and relevant dendrite growth occurs at the lithium metal side [52]. In such cases, the use of dual or multiple polymer layers, having different functions, may help the homogeneous lithium metal plating/stripping, thus suppressing the dendrite growth.

3 Sodium and Sodium Ion Batteries

3.1 Battery Technologies Based on Na Metal

Among the emerging battery technologies alternative to LIBs, sodium-based batteries, in particular sodium metal batteries (NMBs), sodium ion batteries (NIBs), all-solid-state sodium metal battery (ASS-NMB), Na–S, and Na–O₂, show attractive benefits such as sustainable precursors, secure raw material supplies, and, in principle, low costs [53,54,55,56].

Figure 10.2 shows a schematic view of the most common Na batteries [55]. NIBs are represented by a porous Na-free negative electrode, a liquid electrolyte, and a Na-rich positive electrode; the full battery is assembled in the discharged state (Fig. 10.2a) like LIBs. The most common negative electrodes are composite films constituted by insertion-type active material, but also conversion and alloying materials have been proposed and tested. On the other side of the battery, the most common positive electrode materials tested successfully are layered ternary oxides that exploit the reversible insertion/de-insertion of Na⁺ ions driven by the redox reactions such as Mn⁴⁺/Mn³⁺, Ni⁴⁺/Ni³⁺, or Co⁴⁺/Co³⁺. Regarding liquid electrolytes, formulation closely matching those of LIBs based on organic carbonate solutions has been demonstrated for the utilization in NIBs by simply substituting lithium salts with sodium salts. An appealing advantage of NIBs compared to LIBs is the possible use of Al as the negative electrode’s current collector instead of Cu, thanks to the inability of Al to alloy with Na above the Na⁺/Na plating/stripping redox potential [57].

Fig. 10.2

Schematic view of the most common Na battery concepts, (a) Na ion battery, (b) solid-state Na metal battery, (c) Na/O₂, and (d) Na/S [55] (License n. 5,396,641,494,902)

ASS-NMBs are based on positive and negative electrodes, like that of NIBs, whereas the electrolyte is a solid composite (Fig. 10.2b). ASS-NMBs are recognized as a promising future battery technology because of their highly improved thermal stability, although there is room for performance improvement. Other advanced systems are Na–O₂ and Na–S batteries. The former exploits the plating/stripping reaction of a Na metal at the anode side, whereas porous conducting carbon-based composites in contact with O₂ are used as the cathode, as depicted in Fig. 10.2c. The electrolyte can be either aqueous or aprotic. The latter, Na–S batteries, has very similar configuration compared to Na–O₂ ones, the most relevant difference being the presences of a sulfur–carbon composite at the positive electrode and the exclusive use of either liquid or solid aprotic electrolyte.

This section addresses an overview of state-of-the-art Na batteries, which represent an attractive solution almost suitable to replace Li ion technology in many future applications [53,54,55,56].

3.2 Toward Sodium Ion Batteries: Cathode Materials

Research on active materials for cathodes in NIBs can take advantages from all similar previous studies for LIBs. Table 10.2 visually summarizes the relative merits of many of the most promising cathodes for NIBs, namely, O3-NaMO₂ and P3-Na_xMO₂, Prussian blue analogs (PBA), sodium vanadate phosphates (NVP), and sodium vanadate fluorophosphates (NVPF), belonging to layered oxides (O3-NaMO₂ and P3-Na_xMO₂), hybrid materials (PBA), and polyanionic phases (NVP and NVPF) [55, 58,59,60,61].

Table 10.2 Comparison of performance for the most promising NIB’s cathodes. The different colors of the symbols represent high (green), average (orange), and poor (red) performance. Acronyms of the compounds are described in the main text [55] (License n. 5,396,641,494,902)

Layered Oxides

These materials have a two-dimensional layered structure, in which the Na ions intercalate and deintercalate reversibly. The general formula of the most advanced materials within this class is Na_xMM’M”M”‘O₂, where M, M’, M”, and M”’ = transition metals, Sn, Al, Mg, etc. This family of materials has a small molecular weight and therefore a large theoretical capacity; their stoichiometries and redox mechanisms are comparable to those of their Li analogs successfully used in LIBs [62]. Differently from the lithium layered oxides, Na ions can be coordinated by both prismatic (P) and octahedral (O) sites in the layered oxide structure, giving rise to a large variety of phases with different transition metals and stacking structures. From the performance point of view, capacities can reach as large as 200 mAh g⁻¹ [62, 63]. The most critical drawback of these materials is the anticorrelation between Na content and cycling stability. In fact, materials with a Na/M ratio close to unity show large initial specific capacities but suffer from a remarkable capacity fading. On the other hand, layered phases with Na/M ratios close to 0.7 show excellent stability for prolonged cycling at the expense of the overall capacity [62, 63]. Presently, research efforts are placed on adding extra-capacity to layered phases arising from excess sodium and anionic redox activity. However, true Na-rich phases, i.e., Na_1 + xM_1-xO₂, similar to the so-called lithium-rich layered oxides, are still elusive due to dimension of the Na atom, being larger than the Li one.

Polyanionic Compounds

Polyanionic compounds are also under study as positive electrodes for NIBs. Their working potentials can be easily tuned by changing the composition of the cations and polyanions in the structure, thanks to the close interplay between the variable intensity of the inductive effect provided by different polyanionic groups and the redox activity of the different transition metals [64, 65]. The most common polyanionic compounds for NIBs crystallize in typical prototype lattices, like tavorite, alluaudite, olivine, and NASICON. The crystalline structure of these materials is more stable compared to layered oxides and allows a full intercalation/de-intercalation of one or two Na ions. The best performances reported in a full cell are delivered by V³⁺-containing compounds [65]. In particular, the NASICON-like Na₃V₂(PO₄)₃ (NVP) [66] and Na₃V(PO₄)₂F₃ (NVPF) in a full cell with hard carbon (HC) as the negative electrode and a liquid aprotic electrolyte can deliver specific capacities as large as 120 mAh g⁻¹ at 3.4 V vs. Na⁺/Na with a good capacity retention even at high current rates [65, 66].

Prussian Blue Analogs (PBAs)

PBAs are very attractive materials for positive electrodes in NIBs. Their general stoichiometry is A_xM[M’(CN)₆]_1-y•zH₂O, where A is an alkali metal, M and M’ are a transition metal ion, and y is the fraction of vacancies in the crystal structure [67, 68]. Like in the case of layered oxides and polyanionic structures, also for PBA the choice of specific transition metal blends allows to tune the redox potential of NIBs [58, 69], allowing the materials to work either as cathode or anode. PBAs have been proposed by the US company Natron that demonstrated the ability of PBAs to operate simultaneously as positive electrodes in the form of sodium hexacyanoferrate and as negative electrodes in the form of hexacyanomanganate, reaching very promising performance [70].

3.3 Toward Sodium Ion Batteries: Anode Materials

A visual comparison among the promising materials for negative electrodes in NIBs is summarized in Table 10.3. Materials can be grouped depending on the reaction mechanism upon Na⁺ incorporation/de-incorporation: insertion (HC, graphite, TiO₂, Na_xTiO₃), alloying (Sn/C, P/C, Sb/C), conversion (MoS₂/C), and mixed alloy conversion (Sb₂O₃/C).

Table 10.3 Comparison of performance for the most promising NIB’s anodes. The different colors of the symbols represent high (green), average (orange), and poor (red) performance [55] (License n. 5,396,641,494,902)

Insertion Materials

The most promising insertion-based materials are the so-called hard carbons (HCs) having their specific capacities in the range of 300–350 mAh g⁻¹, with a Coulombic efficiency in the first cycle as high as 80% [71] and a working potential below 1 V vs. Na⁺/Na. On the other hand, soft carbons (SCs) show a much smaller capacity compared to HCs, i.e., 200–250 mAh g⁻¹, but at lower working potential 0.5–0.6 V vs. Na⁺/Na. Overall, despite the capacity retention in prolonged cycling being better for SCs compared to HCs, the latter phases (i.e., HCs) are the most promising active materials due to the much smaller accumulation of irreversible capacity compared to SCs, both in the first cycle and cycle-by-cycle, thus minimizing the waste of charge and the sacrificial positive electrode masses in a full NIB configuration. To improve the performance of HC, two main strategies are presently explored [72]. One is the optimization of the HC microstructure and surface composition to enhance the specific capacity through pseudo-capacitive and pseudo-plating sodium incorporation mechanisms [73, 74], whereas the second one is the incorporation of additional and different sp² carbon structures (like SCs or graphene) in HC-based carbon electrodes [75]. Both strategies can increase the capacity unfortunately at the expense of the Coulombic efficiency [75]. Beyond the improvement of performance, keeping sustainability and low environmental impact of these kinds of materials is important because these are critical characteristics. With this respect, it is also crucial to choose the correct precursors for these materials to keep the low costs and the sustainable philosophy of HC [76].

Alloying Materials

This type of anode material shows a higher specific capacity compared to HC and operates at low voltages [77]. The alloy-forming materials gain attention due to the ability to incorporate/de-incorporate more than one Na atom per redox atom, but as a trade-off, they suffered from huge volumetric changes upon cycling. Different strategies have been proposed to address this issue, such as minimization of active material particles into a nanometric scale to reduce the mechanical strain per one particle, the formation of core–shell materials to physically restrict the volume change, and the utilization of carbon-based additives as a buffer space [78]. At the same time, however, it is necessary to evaluate whether these improvements are essential from different perspectives. For instance, carbon-based additives are added as stabilizing components for alloy-forming materials, but it also commits as “dead weight” and affects practical capacity of total electrode mass.

The alloy-forming materials are classified into group 4 elements (Si, Ge, Sn), group 5 elements (P, As, Sb, Bi), and their binary or ternary alloys. Among the group 4 elements, Si and Sn are attractive materials because of their abundance. Despite the successes of Si in LIBs, Si has not been that successful in NIBs. In the case of Sn, there are a number of similarities between the Sn in NIBs and LIBs. Among group 5 elements, P is the element with highest theoretical capacity 2596 mAh g⁻¹ arising from the formation of Na₃P [79], but due to its high flammability, its practical applications have been hindered [80]. Most of developments are centered on P/carbon composites due to large volume changes during cycling as well as low electric conductivity. P/graphene composites are known to show outstanding performances [81], as well as Sb/carbon composite which can reversibly deliver up to 3 Na per atom and reach a specific capacity of 610 mAh g⁻¹ with 95% of capacity retention over 100 cycles [82]. Unfortunately, however, due to the high cost, toxicity, and low sustainability of the latter material, its practical application is not recommended.

Metal Oxides

Like in the case of LIBs, a variety of metal oxides have been investigated as anode materials for NIBs, and they are known to show distinctive mechanisms of anode functionality. Among these materials, the most representative one is titanium oxide, which operates through an insertion mechanism changing its structure from crystalline to amorphous [83]. Despite the high operating voltage and the low Coulombic efficiency, this material is very promising because of the low cost, low toxicity, and availability [83]. Since the conductivity of titanium oxide is low, the use of dopant has been proposed to increase its electronic conductivity. To this end, the materials such as TiO₂ nanosheets/graphene oxide composites have been developed and demonstrated to show a capacity above 175 mAh g⁻¹ at 1C for 200 cycles and 90 mAh g⁻¹ at 20C for 10,000 cycles [84].

Conversion Materials

The reaction of this type of materials with Na is known to form a new phase which is structurally very different from the starting one, specifically M_aX_b + (bc)·Na ⇄ aM + bNa_cX, in which M is a metal (Cu, Fe, Sn, etc.) and X is an anion (O, S, P, etc.). Despite their high theoretical capacity, there are different issues related to the conversion materials, such as a low first cycle Coulombic efficiency, an insulating property, a voltage hysteresis, and a large volumetric expansion upon cycling together with consequent electrode degradation in long term due to the deep structural transformation. The electrode composition design and the right choice of the electrolyte are two investigated strategies to improve and solve the issues listed above. MoS₂ represents a typical conversion-type material with a theoretical capacity of about 670 mAh g⁻¹ [85]. The electrochemical performance and its stability can be improved by engineering the materials such as the formation of different heterostructures or the combination of MoS₂ layers with different carbon-containing species [86]. Yet, due to the complexity to scale up a controlled synthesis of MoS₂, the large-scale application of this material has not been successful.

New emerging candidates are the metal phosphides [87] because they sustain the conversion reaction of the phosphorus within a metal atom network which provides electric conductivity. Moreover, in these materials, the volumetric expansion and long-term degradation are mitigated [88].

3.4 Liquid Electrolytes and SEI Formation in NIBs

In the case of LIBs, there are various standard electrolytes, which are also commercially available. In contrast to this, the optimal electrolyte of NIBs is still under development [89]. Like the electrolytes for LIBs, the performance of electrolytes for NIBs is strongly dependent on the electrode chemistry and electrode combination, making the optimization of an electrolyte composition complex. Especially in the case of NIBs, the ability of electrolyte to form SEI at the anode side is crucial because the SEI based on sodium salts has a higher solubility compared to that based on lithium salts; thus, the SEI undergoes continuous creation and destruction [90].

The main salts used in literature are sodium perchlorate (NaClO₄), sodium hexafluorophosphate (NaPF₆), sodium triflate (NaCF₃SO₃), and sodium bis(trifluoromethanesulfonyl)imide (NaTFSI), which are combined with ether or carbonate solvents [89]. NaClO₄ was chosen in the early stage of NIB research, but it is not ideal because of its explosive nature. As an alternative salt, NaPF₆ has been employed most frequently, although its performances are strongly affected by the presence of impurities [91].

The main solvents used together with these salts are the same for LIB technology, such as propylene carbonate (PC), ethylene carbonate (EC), and dimethyl carbonate (DMC), or their mixtures, which are known to form a stable SEI in the case of LIBs [91]. Since, in the case of NIBs, there is a problem in the stability of SEI as mentioned before, the use of additives has been recommended to form a compact, uniform, and more stable SEI, such as FEC and VC [92], which is the same additive used to create m-SEI in LIBs. In addition to these, sodium difluoro(oxalato)borate (NaDFOB), succinonitrile (SN), and 1,3-propane sultone (PS) have been reported as potential additives. It should be noted that in the case of NIBs, also ether-based electrolytes show very good performances for different active materials [93] including high surface carbons [94]. The use of ether-based electrolytes allows the co-intercalation of solvated Na⁺ into graphite and enables the formation of stable SEI with enhanced Na plating/stripping efficiency.

Several ionic liquids and aqueous-based electrolytes are also proposed as more sustainable and stable electrolytes [89, 95]. The use of ionic liquids has positive effect on the stability of SEI. Because of this, very promising results have been obtained, for example, when ionic liquids are combined with TiO₂ anode material [96]. Due to the high cost of this class of electrolytes, practical application is not yet realized. In contrast to this, aqueous electrolytes have advantages in terms of their low cost, good sustainability, and safety [60]. However, the structural stability of the electrodes in water and the possible side reactions need to be assessed carefully [60]. Recently, the concept of the “water-in-salt” (WiS) electrolytes has been proposed. This class of electrolytes is defined as the concentrated aqueous solution of salt, in which the number of water molecule per ion is far below the solvation number [60]. These electrolytes consist of contact ion pairs (CIPs) and aggregated cation–anion pairs, which decrease the availability of water and improve its electrochemical stability. By employing the WiS electrolyte, symmetric NIBs based on a dual V³⁺/Ti⁴⁺ NASICON-structured Na₂VTi(PO₄)₃@C bifunctional electrode were successfully investigated, showing a stable performance over 2500 cycles at 10C [97]. This result demonstrated that the fluorine-rich SEI can suppress the electrode dissolution [97]. Important results of the advances in electrolyte/electrode optimization, related to formulation of advanced electrolyte system, are well summarized in the review of Chen and co-workers [98].

3.5 Next-Generation Sodium Batteries

An important step is being taken toward an all-solid-state configuration of the NIB to eliminate the serious problems related to the flammability of liquid electrolytes [99]. The first solid-state sodium ion conductor dates to the 1960s, when a fast two-dimensional sodium-ion-transport phenomenon was discovered in β-alumina (Na₂O•11Al₂O₃). In the same period, NASICON-type compounds were first studied leading to the development of Na_1 + xZr₂Si_xP_3-xO₁₂ (0 ≤ y ≤ 3). Many efforts have been made to elucidate the mechanism of Na⁺ transport and to achieve the optimal compositions. Recently, the inclusion of NaF at the synthesis step of NASICON was found to be effective for the fast ion transport (resulting in a conductivity of about 4 × 10⁻³ S cm⁻¹). In addition, different kinds of solid-state electrolytes have been reported, such as those based on SPEs as well as oxide and sulfide ISEs, as summarized by Zhao et al. [100]. The ionic conductivity and electrochemical stability of Na⁺-conductive solid-state electrolytes are still low, and further development is necessary for the commercial use [99].

Other state-of-the-art systems for future sodium technologies are Na–S and Na–O₂ batteries [101]. As a closed system, in this section, focus is placed on Na–S batteries only, but both are promising systems. Na–S batteries were patented by Ford Motor Company in the 1960s, which were made of a β-alumina ceramic electrolyte and cycled at high temperature around 300 °C [102]. At this temperature, Na and S are both molten and need to be separated by the solid electrolyte. For a practical application, the working temperature of Na–S batteries must be reduced, and thus room temperature or intermediate temperature Na–S batteries are being reconsidered [103]. The important difference between Li–S and Na–S batteries is the thermodynamically stable species of alkali metal polysulfides, which should be considered for further development, possibly being both advantageous and disadvantageous.

4 Battery Technologies Based on Alkaline Earth Metals

4.1 Rechargeable Magnesium Ion Batteries

Magnesium ion batteries (MIBs) have attracted intensive attention due to their high capacity, high security, and low-cost properties. However, the performance of MIBs is seriously hindered by the intense polarization and slow diffusion kinetics of Mg²⁺ ions. To solve these issues, numerous efforts based on both experimental and theoretical studies have been proposed [104, 105]. In this section, the latest advancement in anode and cathode materials as well as electrolytes for MIBs is summarized and discussed.

4.1.1 Negative Electrode Materials for MIBs

Metallic Magnesium Anodes

The volumetric energy density of Mg is higher than that of Li (3833 vs. 2046 mAh cm⁻³), which is beneficial for the energy storage systems [106]. Different from the lithium metal, Mg was considered for a long time as a metal whose plating is homogeneous without formation of any dendritic structures [107, 108]. This, together with the low reduction potential of Mg²⁺ (−2.37 V vs. SHE), makes Mg metal an ideal anode. Despite these merits, recently some authors pointed out the potential hazards related with a formation of dendritic forms or needlelike structures due to uneven deposits on magnesium metal [109,110,111]. Moreover, surface side reactions are known to occur on the anode side resulting in the accumulation of passivation films during the initial cycles of batteries. Unlike the formation of the SEI on the lithium metal, many reports assume that the passivation layers on the metallic magnesium have low ionic conductivity for magnesium cations [112]. Related to the formation and passivation properties of the interphase, the Coulombic efficiency of magnesium stripping and plating is the most critical parameter, and this is strongly affected by the electrolyte, its chemical properties, purity, and concentration of salt(s). Early generation of nucleophilic electrolytes [113, 114] in ether-based solvents is known to keep magnesium surface at least partially active (non-passivated). Acting on the metal anode itself, Liang et al. synthesized ultrasmall Mg particles with a diameter of ∼2.5 nm [115]. The ultrasmall nanoparticles reduced the thickness of the passivation film, which improved the deposition of Mg. Chemical modification is another feasible way to control the reactivity of Mg metal. Lv et al. used a SnCl₂–dimethoxyethane (DME) solution to treat Mg foil and obtained a modified Mg anode with a tin-based artificial layer [116]. The Mg anode can maintain stable plating/stripping for more than 4000 cycles at a high current density of 6 mA cm⁻². In the same manner, by adding GeCl₄ into Mg(TFSI)₂–DME electrolyte, the Ge-based artificial layer was also formed on the Mg surface, showing a self-repairing process [117].

Alloying Anodes

Some elements from block p are known to electrochemically react with magnesium ions to form an alloy. Alloying materials, such as bismuth (Bi), tin (Sn), and phosphorus (P), are promising alternatives to Mg metal anode [118]. Alloy materials may have synergistic effects in MIBs, bringing new properties that single Mg metal does not have. Bi anode (theoretical volumetric capacity is 3783 mAh cm⁻³) [119] can rapidly insert and extract Mg²⁺, with the ion dynamics being related to defect chemistry. In general, gravimetric and volumetric capacities related to the electrochemical alloying are very high (theoretically up to 900 mAh g⁻¹_(Sn) and 6570 mAh cm⁻³_(Sn) for the formation of Mg₂Sn). Synergistic effects have been expected by combining multiple elements such as Bi–Sb (solid solution), Bi–Sn (composite), or intermetallic compositions of InBi, SnSb, or InSb [120,121,122]. Like the alloying-type electrode materials in LIBs, the volumetric changes due to the alloying and de-alloying process hamper the long-term cycling of the electrode. However, with a good electrode formulation, high reversible cycling has been demonstrated especially with Bi-based electrodes (Fig. 10.3) [123]. It is also important to consider that alloy electrodes might be easier to produce than magnesium metal electrode foils. Indeed, alloys can be synthetized in the form of powder by ball-milling or high-temperature reactions, which can be easily integrated in the battery industry. Moreover, alloys are certainly less surface sensitive than magnesium metal and might be of interest for protecting the Mg surface from the dendritic growth.

Fig. 10.3

Electrochemical behavior of Bi nanowires in a Bi|Mg half-cell at C/2 with an organohaloaluminate electrolyte [123] (License n. 1,322,765)

Intercalation-Type Anodes

Although graphite has been widely employed in conventional LIBs, it is difficult for Mg²⁺ to be inserted into graphite due to its high ionization potential (7.65 eV for Mg and 5.39 eV for Li) [124]. However, Kim et al. proved that the co-intercalation of Mg²⁺ and linear ether solvents is possible through the density functional theory (DFT) calculations [125]. The Mg²⁺ storage properties of carbon nanotubes (CNTs) have also been studied by the DFT. Aslanzadeh et al. studied the influence of CNTs’ diameter on the voltage of MIBs by calculating the adsorption energies of zigzag CNTs [126] and demonstrated that the cell voltage increased with the increase of CNTs’ diameter. Recently, research interest was shifted to graphene to explore its potential as the anode material for MIBs. Graphene with 25% double-vacancy defects can achieve a Mg²⁺ capacity of 1042 mAh g⁻¹ [127]. In addition to these, many other two-dimensional materials have been studied, such as transition metal carbides (MXene) [128] and borides (MBene) [129]. Regarding metal oxide insertion materials, TiO₂ is a very common anode for various secondary batteries. However, the limited capacity of 110 mAh g⁻¹ at 0.1 C seriously obstructs its application in MIBs [130]. According to the research of Luo et al., proton charge compensation in Ti-deficient TiO₂ (B) nanowires ensures more thermodynamic feasibility and sufficient intercalation sites for Mg²⁺, thereby increasing the capacity.

4.1.2 Positive Electrode Materials for MIBs

The cathode material, a key component of MIBs, predominantly determines the energy density of batteries. However, most of the cathode materials of MIBs show small capacity and poor rate capability, which seriously hinder the battery performance. Most studies on Mg batteries focus on the combination of Mg metal with inorganic cathodes based on transition metal redox centers. As of today, inorganic oxide, polyanionic, and sulfide compounds are the focus of attention of the research community, but they all present their pros and cons. Moreover, research on inorganic cathodes for magnesium batteries is sometimes quite perplexing as the lack of highly oxidative stable Mg electrolytes prevents the evaluation of Mg insertion reactions at high voltages.

Chevrel Phase Mo₆S₈

Research on the cathode materials for Mg insertion was accelerated by the seminal work of Aurbach et al. with a prototype full Mg cell based on the Chevrel phase Mo₆S₈ [114] and opened a large avenue of research on chalcogenides. These structures are very promising due to the weak electrostatic interactions between Mg²⁺ ions and the sulfide/selenide-based anion framework. The low operating voltage (~1.1 V vs. Mg²⁺/Mg) and low specific capacity (~100 mAh g⁻¹) of Mo₆S₈, related to an incomplete de-insertion of Mg ions due to the trapping of partial charges at room temperature, are limiting factors for commercial application of the given technology. Substituting sulfur by selenium allows for a 100% capacity usage [131] but at the expense of the specific capacity value. Although the voltage and capacity values are too modest to obtain energy densities competitive with Li ion batteries, Chevrel structures remain the benchmark electrodes for Mg batteries as they offer remarkable insertion/de-insertion kinetics and good reversibility.

Layered Cathode Materials

Layer materials possess two-dimensional transmission channels, which enable rapid Mg²⁺ migration. The reported layered cathode materials for MIBs include both transition metal oxides and transition metal sulfides, such as V₂O₅, MnO₂, MoS₂, and TiS₂ [132]. As a representative, V₂O₅ has attracted great attention due to its high theoretical capacity (∼295 mAh g⁻¹ for MgV₂O₅) and working voltage (∼2.35 V vs. Mg²⁺/Mg). Compared with transition metal oxides, the relatively low ionization degree of S in transition metal sulfide weakens the electrostatic interaction between Mg²⁺ and negative charge, which is favorable for the migration of Mg. Yang et al. studied the diffusion kinetics of Mg²⁺ in MoS₂ (theoretical capacity of 223.2 mAh g⁻¹) [133].

Polyanionic Cathode Materials

Polyanionic compounds have been widely used in MIBs because of their versatile variety, stable structure, and strong inductive effect [134]. In MgMSiO₄ (M = Mn, Co, Fe, etc.), Mg²⁺ diffuses from the octahedral (O) site to the tetrahedral (T) site [135], with an energy barrier of 740÷770 meV [136]. Despite olivine FePO₄ performing well in LIBs, its capacity in MIBs is only ∼13 mAh g⁻¹ at 20 mA cm⁻² [137], which can be attributed to the amorphous phase produced on the material surface during the discharge process, preventing Mg²⁺ from entering the bulk phase of FePO₄.

Organic Cathode Materials

Although inorganic materials have widely dominated the field of rechargeable batteries, a great amount of attention has been recently placed on organic materials because they possess many crucial advantages, like safety, sustainability, green, low cost, and high theoretical capacity [138]. Carbonyl conjugates are a large group possessing many C=O functionalities, essentially determining properties such as diversity, fast reaction kinetics, and high specific volume. Therefore, compared to other types of organic cathode materials, carbonyl-conjugated compounds are expected to develop as the next generation of cathode materials for Mg batteries. Quinone-based monomers are particularly suitable as an active storage unit, and most of recent reports [139] provided highly attractive properties in terms of energy and power density as well as cyclability. Other reports showed applicability of imides and radical organic compounds [140], which in theory are less attractive in terms of energy density. As well known, electrochemical characteristics depend also on type of electrolyte, where both salts and solvents play important roles [141]. It has been demonstrated that the use of Mg(TFSI)₂ salt can increase the capacity utilization of organic electrodes, while the use of AlCl₃ can upshift the potential of redox active compounds [142]. Although redox active organics in the form of different polymers show the best properties in terms of cycling and rate capability, there are still some important challenges that need to be addressed before their potential commercialization.

4.1.3 Electrolytes for MIBs

To make MIBs commercially available, further breakthroughs in the electrolyte chemistry development are needed. As for other battery technologies, the electrolytes for MIBs should fulfill multiple criteria such as low toxicity, low cost, wide electrochemical stability window, and high ionic conductivity. The following main types of the electrolytes have been explored in MIB technologies: (i) liquid-organic solvent electrolytes, (ii) solid-state electrolytes, (iii) polymer electrolytes, and (iv) ionic liquid-based electrolytes [104].

Liquid electrolytes based on organic solvents or ionic liquids are up to now the best performing materials for Mg deposition/stripping processes. The former electrolytes present a wider electrochemical stability window, while the latter group shows a lower overvoltage in the Mg deposition process, a higher thermal and chemical stability, and a negligible flammability.

For the electrochemical reactivity at the Mg metal anode, even traces of water or other oxygenated coordination ligands must not be present in the electrolyte, to avoid any compromise in the anodic reversibility, cyclability, and current density [143]. Actually, oxygenated species cause the formation of a compact MgO_x(OH)_y layer on the magnesium metal anode, which hinders the anode–electrolyte charge exchange processes promoting on its surface the growth of dendrites. On the other hand, as demonstrated by Novak et al. in the early 1990s [144, 145], water is necessary in the cathodic side to efficiently and reversibly exchange magnesium ions between the solid cathode active material and the electrolyte during insertion and de-insertion processes. This is because water molecules can exfoliate the cathode layered structure and enhance the Mg²⁺ diffusion into the bulk cathode materials by solvating the ions and facilitating their solid-state migration phenomena [146, 147]. Taking all together, developing electrolytes for MIBs able to address the so-called “devil” (at anode) and “holy” (at cathode) water dilemma is a very difficult target.

4.2 Calcium Batteries

Calcium (Ca) batteries are emerging as a promising next-generation electrochemical energy storage system, due to the abundant reservation of calcium and the competitive redox potential of Ca²⁺/Ca. However, the practical realization of rechargeable Ca and Ca ion batteries (CIBs) still relies on the identification of suitable electrodes and electrolytes. Despite reversible calcium plating–stripping being recently demonstrated [148], efforts are still needed to improve both kinetics and efficiency and to allow a wider range of electrolyte formulations. Widening the electrochemical stability window of the electrolyte is crucial to lead the development of positive electrodes operating at high potential [149, 150].

4.2.1 The Benefits of Calcium Batteries

Calcium is a divalent alkaline earth metal with an extraordinarily strong oxidative ability in consideration of the −2.87 V vs. SHE redox potential for the Ca²⁺/Ca couple [151, 152], to be compared to the −3.04 V vs. SHE of the lithium metal electrode. In comparison to other elements under development for battery applications, calcium is the multivalent metal with the most negative redox potential and an ionic radius of 114 pm, very similar to Na⁺, that is the cation easily intercalated/deintercalated in/from a variety of materials [153]. The theoretical properties of calcium metal electrodes, in terms of gravimetric and volumetric specific capacities, surpass those of potassium, sodium (both gravimetric and volumetric), and zinc (gravimetric) and are like that of lithium (volumetric), thanks to the favorable combination of intermediate atomic weight and density [149]. In addition, compared to aluminum and magnesium, calcium has a larger ionic radius and a smaller electronegativity [154, 155], thus suggesting on the one hand a lower coordination binding in the liquid phase, thanks to the smaller charge density, and on the other hand a less covalent bonding in solid lattices. Furthermore, it is highly abundant on Earth’s crust and industrially inexpensive, much more than Li, Na, K, Mg, and Zn [150, 156, 157]. Even though calcium has an atomic weight seven times larger than lithium, the specific capacities of the calcium-based oxides vary in the 100–250 mAh g⁻¹ range, comparable with lithium intercalation positive electrodes, also for what concerns the thermodynamic potentials. Similar to the anode materials available for LIBs, silicon, phosphorus, and carbon negative electrodes can reach large theoretical capacities at relatively low redox potentials also in the case of CIBs. On the other hand, calcium carbide is expected to suffer from large kinetic limitations and overpotentials being the crystal structures of all CaC₂ polymorphs remarkably different from graphite [158]. Moreover, outstanding theoretical performance is achievable for Ca–O₂, as well as for Ca–S, leading to the formation of calcium peroxide and calcium sulfide.

A tentative evaluation of the relative merit of various calcium-based batteries in comparison with LIBs can be made using the theoretical performance of the LiCoO₂/C LIB as the baseline (360 Wh kg⁻¹, calculated assuming a ΔE° = 3.6 V and a specific capacity normalized by the sum of both positive and negative electrode materials) [159]. The CIBs constituted by CaMn₂O₄ and Si can disclose a theoretical energy density of about 520 Wh kg⁻¹, being superior to the benchmark LIB (mentioned above) and approaching the desirable figures of the LiMn_1.5Ni_0.5O₄/Si and LiFePO₄/Li configurations [159]. Both sulfur battery chemistries, i.e., Ca–S and Li–S, have a comparable theoretical performance (approximately 1800–2000 Wh kg⁻¹) that is, in both cases, exceeded by the Ca–O₂ and Li–O₂ ones (~2400 and ~ 3500 Wh kg⁻¹, respectively).

Overall, calcium-based battery chemistries can theoretically achieve performance of interest, not far from those of lithium-based ones, but research in the field is still in its infancy. To date, the serious exploitation of cathode materials for CIBs just started because of the pioneering work of Ponrouch et al., demonstrating the successful Ca plating/stripping using conventional organic solvents in 2016 [148]. However, the field being in its early stage, the number of reports on positive electrode materials in the literature is still limited [159,160,161,162,163,164,165].

4.2.2 Challenges in Developing Calcium Batteries

The operating principle of typical calcium batteries is depicted in Fig. 10.4, using inorganic framework as cathode and metallic Ca as anode. In this process, a series of limiting steps occurs [166]. Firstly, the slow diffusion of Ca²⁺ in the positive electrode is inevitable. Although the charge density of Ca²⁺ is small (52 C mm⁻³, the same as Li⁺), its diffusion rate is significantly affected by its large ion radius (0.99 Å, compared to 0.76 Å for Li⁺), which is a dominant limiting factor. Secondly, Ca²⁺ forms solvated calcium in the electrolyte. In some cases, the formed solvated group is relatively large, making it difficult to migrate in the electrolyte. Moreover, when the solvated cluster moves to the electrodes, it needs to be de-solvated. A higher voltage beyond the reorganization energy is needed to break the interatomic bonds in the solvation cluster. This voltage needs to be within the electrochemical stability window of the electrolytes; otherwise, it is easy to decompose the electrolyte. Thirdly, the shuttle of de-solvated calcium ions in SEI is also a matter to be discussed. SEI is a thin film with a thickness of several nanometers between the electrolyte and the anode. This layer can protect the anode from continuous damage and the electrolyte from side reaction with the anode. Ideally, the SEI film is a good ion conductor but needs to be an electronic insulator. As for CIBs, the main components of SEI are CaCl₂, Ca(OH)₂, CaCO₃, CaF₂, CaH₂, etc. [167]. These substances undergo structural changes during the calcium ion diffusion process. Due to the strong interatomic bond, many components are thought to be unable to transmit Ca²⁺ (such as CaCl₂, Ca(OH)₂, CaCO₃, etc.). Also, continuous insertion/de-insertion of Ca²⁺ induces a large volume change of SEI, which may rupture the thin film. Once Ca²⁺ passes the SEI and arrives at the anode, the intercalation, nucleation, or plating of Ca happens [168]. Owing to these complex steps, it is extremely challenging for calcium to deposit.

Fig. 10.4

The diffusion, migration process of Ca²⁺ in a typical calcium battery [166] (License n. 5,487,100,248,324)

The main issues and their current state, to be carefully addressed to allow the calcium technology being widely applicable, are summarized as follows:

• The Ca²⁺ ion hosting materials that must be stable enough during the Ca²⁺ diffusion and insertion/de-insertion: Prussian blue analogs, transition metal compounds, and organic compounds are known as potential materials for this requirement [169, 170].

• The reversible stripping/plating of Ca metal at moderate temperature: the proof-of-concept batteries, as of today, showed an average Columbic efficiency of 96% for 50 cycles, which is not sufficient for the practical application [160, 171].

• The efficient solvation/de-solvation process of Ca²⁺-based electrolytes: different from Li⁺-based electrolytes, which can form Li⁺-conducting SEI, the reactions between electrolytes and calcium tend to generate Ca²⁺-blocking phases when contacting with organic electrolytes [172].

Overall, although great efforts have been made to change the electrolyte and modify the anode–electrolyte interface, the reversible plating/stripping of Ca metal is still tricky and complicated matter. The complexity of working directly with a Ca metal anode drives the search for alternative anodes, which can be categorized into three types based on the calcium ion storage mechanism, including alloying anodes, intercalation anodes, and organic anodes [166, 173], showing the direction toward CIBs.

5 Conclusions

Among the emerging battery technologies, Na ion batteries as well as alkali metal batteries, probably solid state, are those expected to become commercially available in the near- (2–5 years) to mid-term (5–10 years) future. Na ion batteries are already being developed by several industries following the CATL announcement made at the end of 2021. Polyvalent metal-based battery chemistries appear to be in the early stage of development with a few hurdles to be addressed in the next few years.