Recent research progress on iron- and manganese-based positive electrode materials for rechargeable sodium batteries

Na-containing 3d-transition metal layered oxides are widely studied as positive electrode materials for NIBs. The most common layered structures are built up from a sheet of edge-sharing MeO₆ octahedra. Polymorphisms appear when these sheets are stacked in different ways along the c-axis. Sodium-based layered materials can be categorized into two main groups using the classification proposed by Delmas et al [22]: O3 type or P2 type, in which the sodium ions are accommodated at octahedral and prismatic sites, respectively, as shown in figure 3. Schematic illustrations of crystal structures used in this article were drawn using the VESTA program [23]. O3-type NaMeO₂ consists of a cubic close-packed (ccp) oxygen array, in which sodium and 3d-transition metal ions are accommodated at distinct octahedral sites because the ionic radius of sodium ions (1.02 Å) is much larger than those of 3d-transition-metal ions in the trivalent state (<0.7 Å) [24]. O3-type layered phases are classified as cation-ordered rock-salt superstructure oxides [25]. Edge-shared NaO₆ and MeO₆ octahedra are ordered into alternate layers perpendicular to [111] for the rock-salt lattice (corresponds to [001] for the crystal lattices used for layered structures), forming the NaO₂ and MeO₂ slabs, respectively. As a layered structure, NaMeO₂ is composed crystallographically of three different MeO₂ layers (AB, CA, and BC layers in figure 3, and the layered stacking of MeO₂ is isostructural with that of CdCl₂) to describe the unit cell, and sodium ions are accommodated at the octahedral (O) sites between MeO₂ layers. The structure is therefore classified as an O3-type layered structure. The structure is also classified as 3R phase with a space group (SG) of R $\bar{3}$ m. P2-type Na_x MeO₂ also consists of two MeO₂ layers (AB and BA layers). However, the sodium environment is different from that of the O3 type. Because sodium ions are much larger than lithium ions, sodium ions can also be located at trigonal prismatic (P) sites. When a sodium off-stoichiometry condition (typically 0.6 < x < 0.7 in Na_x MeO₂) is applied to materials synthesis, the P2-type phase is empirically known to be structurally stabilized. The P2-type layered structure is also often classified as 2H phase with an SG of P6₃/mmc. Moreover, a prime symbol (') is added to the abbreviation when the crystal lattice contains in-plane distortion, such as O'3-type NaMnO₂ with a monoclinic lattice (SG C2/m) [26] and P'2-type Na_x MnO₂ with an orthorhombic lattice (SG Cmcm) [27]. These layered oxides can be easily prepared by conventional solid-state reaction from a mixture of a stoichiometric amount of a sodium salt (sodium carbonate, sodium peroxide, etc) with 3d-transition metal oxides (and hydroxides, carbonates etc). See the literature cited in this review article about the detailed synthesis conditions for each sample.

Zoom In Zoom Out Reset image size

Sodium extraction from O3- and P2-type phases generally induces phase transitions. Sodium ions in the O3-type phase are originally stabilized at edge-shared octahedral sites with MeO₆ octahedra. Prismatic sites for Na ions often become energetically stable when sodium ions are partially extracted from the O3-type phase—associated with the formation of vacancies, similar to the P2-type phase. The formation of prismatic sites is achieved by the gliding of MeO₂ slabs without breaking Me-O bonds [18, 22]. As a result, oxygen packing changes from 'AB CA BC' to 'AB BC CA', and this phase is classified as P3-type, as shown in figure 3. It is also possible to directly crystallize this P3-type phase by solid-state reaction without electrochemical sodium extraction in Na cells in some cases. For instance, P3- and P2-type Na_2/3[Ni_1/3Mn_2/3]O₂ are known as low- and high-temperature phases, respectively [28]. The phase transition from P3/O3-type to P2-type phase is, however, impossible in Na cells because its phase transition is achieved only by breaking/re-forming Me–O bonds, and a higher-temperature environment is therefore required.

Instead, the P2 phase changes into an O2-type phase in Na cells (figure 3). Because large prismatic sites in the P2-type phase are energetically stabilized by large sodium ions, MeO₂ slabs move (glide) to form octahedral sites after the extraction of sodium ions [29, 30] (assisted also by Na⁺/Li⁺ ion exchange [31]). Such MeO₂ gliding leads to the formation of a new phase with a unique oxygen packing, 'AB AC AB' (figure 3). This phase contains crystallographically two different MeO₂ layers with AB and AC oxygen arrangements. Vacancies left between AB and AC layers are octahedral sites, that is, the O2-type phase. Both O3- and O2-type phases have close-packed oxygen arrays. In the O3-type phase with the ccp array (ABC-type oxygen arrangement), NaO₂ layers share only edges with MeO₂ layers on both sides. In contrast, the O2-type phase is locally composed of ABA-type and ACA-type (figure 3) oxygen arrangements; namely, local structures are classified as a hexagonal close-packed (hcp) oxygen array. On the basis of oxygen packing, the O2-type structure is classified as an intergrowth structure between ccp and hcp arrays. Note that two glide vectors of MeO₂ slabs, (1/3, 2/3, z) and (2/3, 1/3, z), exist to form the O2 phase. This often induces the formation of staking faults after sodium extraction from the P2 phase [29]. Although only a few studies of the P2-O2 phase transition in Na cells have been reported so far [29, 30], a number of reports on P2–O2 phase transitions via Na⁺/Li⁺ ion exchange, including the formation of a variety of staking faults, are available in the literature [28, 31–36].

Alpha-NaFeO₂ is known as a typical example of O3-type layered structures and is easily prepared by conventional solid-state reaction [37]. In contrast, there are no reports of direct synthesis of LiFeO₂ with O3-type layered structures by solid-state reaction. Because the ionic radii of Li⁺ and Fe³⁺ are similar (table 1), the cation-disordered rock-salt phase is easily obtained by solid-state reaction at high temperatures. However, the single phase of LiFeO₂ has been successfully obtained by hydrothermal treatment at 230 °C with concentrated KOH aqueous solution so that the synthesized LiFeO₂ was electrochemically inactive in the Li cells [38]. Instead, triphylite-type LiFePO₄ has been extensively studied with a view to developing iron-based positive electrode materials for LIBs [39]. Electrochemical reversibility of sodium extraction/insertion processes for a single phase of O3-type NaFeO₂ is shown in figure 4(a). Electrochemical Na extraction from NaFeO₂ using a Li/NaFeO₂ hybrid-ion cell was first demonstrated by Takeda and co-workers [40]. Reversible charge and discharge processes for a Na/NaFeO₂ cell were first reported by Okada and co-workers as previously mentioned [20, 21]. Reversibility in electrode materials is significantly influenced by cutoff conditions upon charge (sodium extraction) as shown in figure 4. Although charging capacity, corresponding to amounts of sodium ions extracted from the crystal lattice, increases as a function of cutoff voltage, reversible capacity obviously decreases when the material is charged beyond 3.5 V. Excellent reversibility with small polarization is observed with the cutoff voltage of 3.4 V. The observed reversible capacity reaches 80 mAh g⁻¹, indicating that approximately 0.3 mole of Na is reversibly extracted from NaFeO₂ and inserted into the Na_0.7FeO₂ host structures. Such deterioration of the electrode properties beyond 3.5 V originates from the irreversible phase transition as suggested by ex situ x-ray diffraction (XRD) [41] and x-ray absorption spectroscopy (XAS) in figures 4(b), (c). Intensity of the pre-edge peak, which is observed at 7114 eV in x-ray absorption near the edge structure (XANES), increases with the increase in the sodium extraction amount. The intensity of the pre-edge peak for Fe³⁺ is often intensified when iron ions are located at tetrahedral sites [42]. In addition, clear changes in the local environment of Fe are noted from the extended x-ray absorption fine structure (EXAFS). Decreases in length for the first coordination shell (Fe–O bond) and in intensity for the second coordination shell (Fe–Fe bond) are also found. These facts suggest that some of the Fe³⁺ ions migrate into the tetrahedral sites. Note that because the XAS spectra were collected after discharge to 2.0 V in Na cells, this Fe³⁺ migration process is irreversible. A proposed degradation mechanism is summarized in figure 4(d). When sodium ions are extracted from the crystal lattice, vacancies are created at tetrahedral sites that are face-shared with FeO₆ octahedra. Because trivalent iron ions are energetically stabilized at tetrahedral sites, iron ions easily migrate to the face-shared sites, similar to LiCo_x Fe_1 − xO₂ [43]. Solid-state diffusion of sodium ions is easily disturbed by iron at tetrahedral sites, leading to the degradation of electrode properties. As a result, the reversible range of Na_x FeO₂ is limited to be narrow. This reversible process in the Na cell is expressed as follows:

$$\begin{eqnarray}&&{\rm NaFe}{{{\rm O}}_{{\rm 2}}}\mathop{\rightleftarrows }\limits_{{\rm reduction}}^{{\rm oxidation}}{{\square }_{0.3}}{\rm N}{{{\rm a}}_{{\rm 0}{\rm .7}}}{\rm Fe}{{{\rm O}}_{{\rm 2}}}+0.{\rm 3N}{{{\rm a}}^{+}}+0.{\rm 3}{{{\rm e}}^{-}}\end{eqnarray} \tag{ 1 }$$

where the open square denotes the vacant sites created in the structure. The energy density available in terms of electrode material is also limited (300 mWh g⁻¹ vs Na). Although the reversible range is narrow, the rather small polarization between charge/discharge processes with sub-micrometer-size particles is attractive in terms of electrode material, as shown in figure 5. The rate capability of Na_x FeO₂ in Na cells is shown in figure 6(a). Such cells can deliver >50% of discharge capacity at 1 C rate. (The C rate is defined as the current that delivers a nominal capacity in 1 h, and the nominal capacity here is defined as one-electron redox of iron, i.e., 241 mA g⁻¹). Another drawback of NaFeO₂ as an electrode material is the Na⁺/H⁺ ion exchange when NaFeO₂ is in contact with water [44]. NaFeO₂ changes into FeOOH and NaOH (Na₂CO₃ and/or NaHCO₃ by uptake of CO₂). Such Na⁺/H⁺ ion exchange is generally observed in O3-type NaMeO₂.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Recently some iron-based layered materials have also been reported with a wider reversible range. In these materials, Ni, Mn, and Co ions are partially substituted for Fe ions [45, 47]. The metal substitution effectively changes the phase transition behavior for NaFeO₂. For the pure iron system, the crystal structure of Na_0.5FeO₂ was reported as an O'3-type phase with monoclinic lattice distortion [48]. O3–P3 phase transition is not evidenced for Na_x FeO₂, and therefore, such iron migration to adjacent face-shared tetrahedral sites may be unavoidable. The O3–P3 phase transition in Na cells, however, occurs for the metal-substituted NaFe_1 − xMe_x O₂ samples. Because iron migration to the large prismatic sites is expected to be unlikely, the formation of the P3 phase can effectively extend reversible ranges for the electrode materials in Na cells. Stabilization of the P3 phase has also been theoretically predicted for Na_0.5MeO₂ (Me = Ni, Mn, and Co) [49]. Among layered materials with 3d-transition metal ions, the P3-type phase is significantly stabilized for Na_x CoO₂ [49], and excellent electrode performance has been reported for NaFe_1/2Co_1/2O₂ [46] and NaNi_1/3Co_1/3Fe_1/3O₂ [50]. These experimental and theoretical approaches are consistent with each other, and the substitution of metal ions for iron ions is an important strategy to further improve the reversibility of iron-based O3-type layered materials.

O'3-type LiMnO₂ is also known as a metastable phase [51–53], similar to O3-type LiFeO₂. The orthorhombic phase (zigzag-type layered phase) crystallizes as a thermodynamically stable phase. Although relatively large reversible capacity is obtained using O3-type and zigzag-type layered phases, lowering operating voltage (and changing the voltage profile) during continuous electrochemical cycling is known to be a disadvantage for electrode materials [54–57]. The voltage profile closely resembles that of spinel-type Li_x Mn₂O₄ after electrochemical cycle tests. These Li–Mn oxides are composed of the common ccp oxygen lattice. Therefore, two polymorphs of LiMnO₂ easily transform into spinel as the energetically favorable phase during electrochemical cycles.

O'3-type NaMnO₂ with a monoclinic lattice (SG C2/m) [26] and P'2-type Na_x MnO₂ with an orthorhombic lattice (SG Cmcm) [27] are prepared as thermodynamically stable polymorphs of layered Na–Mn oxides. Non-distorted P2-Na_xMnO₂ with a hexagonal lattice (SG P6₃/mmc) is also obtained by controlling synthesis conditions [27]. Here only layered Na–Mn oxides are described, and the structural complexity of Na–Mn oxides will be further discussed in a later section. Electrochemical reversibility of Na extraction/insertion from/into O'3-type NaMnO₂ was first reported in 1985 [26]. A very narrow reversible range (x < 0.2 in Na_1 − xMnO₂) was, however, reported in this original publication. Electrochemical properties of O'3-type NaMnO₂ were revisited in 2011 [58], and it was found that O'3-type Na_1 − xMnO₂ shows a wide reversible range (x < 0.8). Although this reversible range is much wider than that of Na_1 − xFeO₂, cyclability in the Na cell is insufficient for electrode materials. Large reversible capacity rapidly declines by subsequent electrochemical cycles. Selected charge/discharge curves of O'3-type NaMnO₂ are shown in figure 5. The Na/NaMnO₂ cell delivers 120–30 mAh g⁻¹ of reversible capacity (approximately 50% of Na ions are extracted/inserted) with small polarization. The operating voltage of Na_1 − xMnO₂ is lower than that of Na_1 − xFeO₂, as shown in figure 5.

Electrode performance of P2-Na_x MnO₂ was also examined in some early studies [26, 59]. Its reversible compositional range for Na extraction/insertion is wider than that of O'3 Na_1 − xMnO₂. More than 150 mAh g⁻¹ of reversible capacity in the Na cell was reported in the early study [59]. The reversible limit is greatly increased (x = ∼0.8 in P2-Na_x MnO₂) when battery-grade (high-purity with < 20 ppm of water) electrolyte solution is used [60]. Approximately 190 mAh g⁻¹ of reversible capacity with good capacity retention is obtained, as shown in figure 5. The electrode performance of P2-Na_x MnO₂ is much better than that of O'3-Na_1 − xMnO₂. Although in-plane conduction of electrons/holes in MnO₂ layers must be the same for both polymorphs, the difference is expected to originate from the manner of in-plane Na-ion conduction between MnO₂ layers. Schematic illustrations of in-plane Na-ion conduction in different phases are compared in figure 7. For the O3-type layered system, because direct hopping from one octahedral site to an adjacent octahedral site requires high activation energy, sodium ions migrate through interstitial tetrahedral sites (similar to O3-type Li_x CoO₂ [61]), which are face-shared sites with MeO₆ octahedra in MeO₂ layers. According to the results of first-principles calculations, the diffusion barrier of sodium ions (vacancies) in the O3-type layered framework structure is relatively small (180 meV for O3-type Na_x CoO₂) [62]. The diffusion barrier of sodium ions is expected to be slightly smaller than that of Li in O3-type Li_x CoO₂ (205 meV) [62]. In contrast with the O3-type layered system, because the P2-type layered framework has an open path for Na ions, a lower diffusion barrier than that of the O3-type phase is expected. Sodium ions migrate from one prismatic site to adjacent sites through open rectangular bottlenecks surrounded by four oxide ions (and thus smaller repulsive interaction from oxide ions is anticipated [63]) because of no interstitial tetrahedral sites dissimilar to the O3 structure. Indeed, higher ionic conductivity of P2-type layered materials than of O3-type layered materials is observed for the samples containing similar chemical compositions for sodium/vacancy [63]. Note that O3–P3 and P2–O2 phase transition may alter the mechanism of in-plane Na-ion diffusion. For instance, according to the results of first-principles calculation, the activation energy of Na-ion diffusion in P2-type Na_x [Ni_1/3Mn_2/3]O₂ (2/3 > x > 1/3) significantly increases (>100 meV) in O2-type Na_x [Ni_1/3Mn_2/3]O₂ (1/3 > x > 0) [64] even though the P2–O2 phase transition is unavoidable for Na_x [Ni_1/3Mn_2/3]O₂ in Na cells. On the other hand, O3-type Na_x [Fe_1/2Co_1/2]O₂ shows excellent rate capability in Na cells, which may originate from the fast ion conduction for the P3 phase formed by the O3–P3 phase transition in the desodiation process [46]. Note that the local environment of Na ions at prismatic sites and the diffusion pathway in the P2 and P3 phases are different, as shown in figure 7. It is expected, therefore, that the mobility of Na ions in both phases may be different. Although an NMR study on P2 and P3 Na_x CoO₂ suggested that the mobility of Na ions in both phases clearly differs on the NMR time scale [65], the difference in Na–ion mobility in terms of electrode materials is not known so far. Further systematic studies of the diffusion process of Na ions in these different phases are necessary.

Zoom In Zoom Out Reset image size

The electrode performance of P2-Na_x MnO₂ is promising based on reversible capacity as shown in the preceding section. The average operating voltage based on the Mn³⁺/Mn⁴⁺ redox couple is, however, limited to less than 3 V versus metallic Na. The average operating voltage of O3-NaFeO₂ with an Fe³⁺/Fe⁴⁺ redox couple is much higher (3.2–3.3 V) than that of P2 Na_x MnO₂, even though the reversible range associated with irreversible iron migration is narrower. From these pieces of experimental evidence, P2-type Na_x FeO₂ is predictably considered to be a possible candidate for high-capacity, high-energy electrode material. All of our trials, however, failed because Fe⁴⁺ cannot be stabilized in the oxide-ion framework under ambient conditions [66]. Therefore, manganese ions were partially substituted for iron to stabilize the P2-type phase, and we finally succeeded in the synthesis of P2-type Fe-based layered oxide, Na_2/3[Fe_1/2Mn_1/2]O₂ [67]. P2-Na_2/3[Fe_1/2Mn_1/2]O₂ can be easily prepared via solid-state reaction of Na₂O₂ (or Na₂CO₃), Fe₂O₃, and Mn₂O₃ at 900 °C for 12 h [67]. The electrode performance of P2-type Na_2/3[Fe_1/2Mn_1/2]O₂ in the voltage range of 1.5–4.2 V is also shown in figure 5. P2-Na_2/3[Fe_1/2Mn_1/2]O₂ delivers large reversible capacity, similar to P2-Na_x MnO₂, in the Na cell. The reversible capacity (reversible range) of P2-Na_2/3[Fe_1/2Mn_1/2]O₂ is much higher (wider) than that of O3-NaFeO₂. The rate capability of P2-Na_2/3[Fe_1/2Mn_1/2]O₂ is also better than that of O3-NaFeO₂ as compared in figure 6. Additionally, the operating voltage is also increased compared with P2-Na_x MnO₂; this will be discussed in a later section.

O3-type Na[Fe_1/2Mn_1/2]O₂ can be also prepared by changing the ratio of sodium/(iron and manganese) [67]. The electrode performance of O3-Na[Fe_1/2Mn_1/2]O₂ in the Na cell is shown in figure 5. Large polarization (>1 V) between oxidation (charge) and reduction (discharge) processes is observed when the Na cell is cycled in the voltage range 1.5–4.2 V. In comparison with the O3 phase, the P2 phase clearly shows larger reversible capacity. Because the P2 and O3 phases have similar particle morphology (figure 5), the kinetic limitation related to the diffusion length is not responsible for the difference in performance. Research interest in the Na–Fe–Mn system as high-capacity electrode material is rapidly increasing, and now the electrode performance for P2- and O3-Na_x [Fe_y Mn_1 − y]O₂ with different chemical compositions has been determined [68–70].

Charge compensation mechanisms in the sodium extraction process were further analyzed by XAS. XANES spectra of P2-Na_x [Fe_1/2Mn_1/2]O₂ at the iron K-edge are shown in figure 8(a). When Na_x [Fe_1/2Mn_1/2]O₂ is oxidized from 3.8 V to 4.2 V, a shift in the iron K-edge absorption spectrum is found in the higher-energy region. The shift may be attributable to complicated situations, including changes in the local structures—for example, sodium extraction from the iron face-shared sites and phase transitions [67]. Therefore, EXAFS spectra during the sodium extraction were further analyzed. The change in the radial distribution around iron during oxidation to 4.2 V is shown in figure 8(b). The radial distribution around iron is not affected by oxidation to 3.8 V, and the interatomic distance of Fe–O remains unchanged (2.00 Å for the as-prepared and 1.99 Å for the 3.8 V charged state). The Fe-O local environment is markedly changed after oxidation to 4.2 V. The intensity for both first and second coordination shells, Fe–O and Fe–Fe(Mn), is reduced, indicating distortion around Fe. This distortion also influences the local structure of neighboring manganese ions [67]. The interatomic distance of Fe–O is clearly shortened after oxidation to 4.2 V (1.90 (1) Å). A large Debye–Waller factor indicates distortion by a non-cooperative Jahn–Teller effect of high-spin Fe⁴⁺ (t_2g³e_g¹). Mössbauer spectroscopy further supports the oxidation of Fe³⁺ to Fe⁴⁺, as shown in figure 8(c). The oxidation of Fe³⁺ was also reported for O3-NaFeO₂ by Mössbauer spectroscopy [20, 21]. The results indicate that Fe³⁺ in P2-Na_x [Fe_1/2Mn_1/2]O₂ is electrochemically active in the sodium system based on the Fe³⁺/Fe⁴⁺ redox. Note that O3-LiFeO₂ is never electrochemically active on the basis of the Fe³⁺/Fe⁴⁺ redox [38], or the reaction seems to be dominated by the oxygen removal process at the solid/electrolyte interface. Indeed, nanometer-sized O3-type LiFeO₂ is electrochemically active in association with the Fe²⁺/Fe³⁺ redox [71]. As the Fe³⁺ 3d-orbital is strongly hybridized with the oxygen 2p orbital in the Li system, oxygen removal is favorable rather than oxidation to Fe⁴⁺, similar to the Li₂MnO₃ system. In general, the sodium system shows a lower redox potential than that of the lithium system, as shown in figure 2. Sodium ions are strongly ionized when compared with lithium ions, resulting in lower covalency with oxygen. Iron and oxygen gain more (net) electrons when compared with the lithium system. As a result, the electrochemical potential of the Fe³⁺/Fe⁴⁺ redox is relatively low, approaching that of Na/Na⁺. Thus, the Fe³⁺/Fe⁴⁺ redox in P2-Na_x [Fe_1/2Mn_1/2]O₂ may be accessible without oxygen loss and may be used as electrode material with relatively high operating voltage in the Na system. Note that tetravalent iron provides the strong hybridization character with the oxygen 2p orbital. Indeed, charge transfer from oxygen to iron has been proposed for SrFeO₃ [72]. The possibility of such a charge transfer process (namely, the formation of a hole in the oxygen 2p orbital instead of the formation of Fe⁴⁺) cannot be negligible for P2-Na_x [Fe_1/2Mn_1/2]O₂. Further systematic study is needed to understand the charge compensation process in greater detail.

Zoom In Zoom Out Reset image size

As described previously, the available reversible capacity of P2-Na_x [Fe_1/2Mn_1/2]O₂ reaches 190 mAh g⁻¹ with an average voltage of 2.75 V versus sodium metal. The energy density is estimated to be 520 mWh g⁻¹ vs Na, which is comparable to that of LiFePO₄ (about 530 mWh g⁻¹ versus Li) and slightly higher than that of LiMn₂O₄ (about 450 mWh g⁻¹). The density of P2-Na_x [Fe_1/2Mn_1/2]O₂ is estimated from x-ray diffraction to be 4.1 g cm⁻³, which is higher than that of LiFePO₄ (3.6 g cm⁻³), and the submicrometer-sized primary particles are electrochemically active without carbon coating. In addition, the many voltage plateaus observed for P2-/O'3-Na_x MnO₂ are not observed for P2-Na_x [Fe_1/2Mn_1/2]O₂ (figure 5). The appearance of voltage plateaus is closely correlated with in-plane sodium/vacancy ordering (presumably coupled with charge ordering in MnO₂ layers). Stabilization of the Na/vacancy ordering also results in bi-phasic reactions (two phases with, generally, different unit cell volume coexist in a single particle) in the region of voltage plateaus. Rate capability in terms of electrode material is expected to be restricted because of the phase boundary movement for two-phase regions [46]. The partial Fe substitution for Mn reduces the tendency of in-plane sodium/vacancy ordering. These characteristics are beneficial for increasing the energy and power density of the electrode materials in sodium batteries.

Although the P2-type Na–Fe–Mn system is a promising candidate as a high-capacity positive electrode material for NIBs, three major drawbacks are known: (1) large volume change (11.3% shrinkage after charge to 4.2 V) during electrochemical cycles, (2) the system's hygroscopic character, which restricts sample handling in moist air (as-prepared P2-Na_2/3[Fe_1/2Mn_1/2]O₂ is somewhat oxidized by water, forming P2-Na_1/2[Fe_1/2Mn_1/2]O₂ and NaOH), and (3) sodium deficiency in the as-prepared sample. In the first discharge process, the sodium ions are in part inserted beyond the starting composition of x = 2/3 in Na_x [Fe_1/2Mn_1/2]O₂. The excess amount of sodium is estimated to be 0.2 (∼50 mAh g⁻¹), forming P'2-Na_0.86[Fe_1/2Mn_1/2]O₂ in the fully discharged state [67]. It is difficult to directly prepare this sodium-rich phase because this phase is metastable. The O3 and/or O'3 phases are thermodynamically stable. Compensation for the deficient sodium ions is, therefore, needed to design the NIBs and to use the full capacity of P2-Na_x [Fe_1/2Mn_1/2]O₂. One idea to compensate for the sodium ions is the addition of sacrificial salts. For instance, NaN₃ electrochemically decomposes into sodium ion and N₂ gas, and this reaction is irreversible. Mixing P2-Na_2/3[Fe_1/2Mn_1/2]O₂ with NaN₃, therefore, increases the initial charge capacity and thus effectively compensates for the sodium deficiency for the P2 phase in the initial state [73]. Further optimization is clearly needed to overcome these disadvantages in terms of electrode material and to develop high-energy NIBs on a large scale in the future.

NaFeO₂ containing trivalent iron ions crystallizes into two different polymorphs. The electrode performance of the low-temperature phase, α-type (O3-type) NaFeO₂, is described in an earlier section (figure 4). Synthesis at higher temperatures (>760 °C) easily results in phase transition into β-type NaFeO₂ as the high-temperature polymorph. The crystal structure of β-NaFeO₂ is related to the mineral wurtzite. The packing of oxide ions is a hexagonal close-packed array, and both sodium and iron ions are located at tetrahedral sites (figure 9(a)) [74]. The electrode performance of β-NaFeO₂ was examined in Na cells. Charge/discharge curves of a Na/β-NaFeO₂ cell are shown in figure 10(a). Beta-NaFeO₂ seems to be electrochemically inactive as electrode material in the voltage range 2.0–4.0 V versus Na metal. From these results, it is speculated that Fe³⁺ at tetrahedral sites is difficult to oxidize to the Fe⁴⁺ state, which is energetically stable only at octahedral sites.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Inverse spinel-type Fe₃O₄ (and γ-type Fe₂O₃) and corundum-type (α-type) Fe₂O₃ were also examined as host materials for sodium insertion. Reduction in particle size to the nanoscale often activates the electrode reversibility of transition metal oxides, which had been thought to be inactive in Li cells [75, 76]. Similar to the Li system, nanosized Fe₃O₄ and Fe₂O₃ are reported to be electrochemically active in Na cells [77, 78], even though micrometer-sized Fe₃O₄ (and other oxides) are electrochemically inactive. The galvanostatic discharge/charge curves of the Fe₃O₄ samples with different particle sizes are compared in figures 11(a)–(c). Electrochemical reversibility in 1.0 mol dm⁻³ LiClO₄ dissolved in propylene carbonate as electrolyte is highly improved as the particle size of Fe₃O₄ decreases. The reversible capacity is less than 10 mAh g⁻¹ for the submicrometer-sized Fe₃O₄, whereas the rechargeable capacity reaches 190 mAh g⁻¹, which corresponds to the 1.6 mole of Li in Fe₃O₄, for the nano-sized Fe₃O₄ particles prepared by a precipitation method (figure 11(i)). Similar to the Li system, the nanocrystalline Fe₃O₄ powder is electrochemically active in the NaClO₄ electrolyte, as shown in figures 11(d)–(f). The 400 nm Fe₃O₄ sample is electrochemically inactive in the Na system, similar to the Li system. For the nanocrystalline Fe₃O₄ powder, the reversible capacity increases to 170 mAh g⁻¹. Although the reversible capacity of the 10 nm sample is slightly smaller than that of the LiClO₄ electrolyte, acceptable capacity retention for the 30-cycle test was observed in the NaClO₄ electrolyte. The electrochemical reactivity of the Fe₃O₄ powder is activated by controlling the size of particles for both Li and Na systems. Recently it was demonstrated that a binder-free nanocomposite with carbon nanotube and nano iron oxides shows excellent electrode performance as electrode material in Na cells [79].

Zoom In Zoom Out Reset image size

Early studies of Na–Mn–O ternary compounds as the sodium insertion host were published in the 1980s [26]. The structural chemistry of Na–Mn–O ternary compounds is complicated, and many different phases, including polymorphs, are obtained as the atomic ratio of Na to Mn changes [80]. Two different Na–Mn layered phases, O'3-type NaMnO₂ and P2-type Na_x MnO₂ (x = ca. 0.7), are described in an earlier section. Two polymorphs are known for NaMnO₂: α-type (O'3-type) NaMnO₂, the low-temperature phase; and β-type NaMnO₂, the high-temperature phase [80]. Beta-NaMnO₂ is prepared at >900 °C, and the crystal structure is different from β-NaFeO₂, with its wurtzite-related structure. The packing of oxide ions for β-NaMnO₂ is the same as for α-NaFeO₂ with the ccp lattice, and distribution of Na and Fe at octahedral sites in the ccp lattice is different. The crystal structure of β-NaMnO₂ is the same as orthorhombic LiMnO₂, the so-called zigzag-type layered phase (figure 9(c)). The electrode performance of β-NaMnO₂ in a Na cell was also reported [26]. The potential difference between charge/discharge processes in a Na cell appears to be relatively small, similar to α-NaMnO₂ (figure 10(b)).

When the fraction of sodium to manganese ions is reduced, two different phases, Na_0.44MnO₂ and Na_0.4MnO₂, are obtained. The Na insertion properties of Na_0.44MnO₂ (Na₄Mn₉O₁₈) were first examined in the 1990s at 85 °C with solid-state polymer electrolyte [81, 82] and revisited in 2007 to be examined at room temperature with aprotic solvent [83]. A crystal structure of Na_0.44MnO₂ is also shown in figure 9. A framework structure of Na_0.44MnO₂ (SG Pbam) consists of four MnO₆ octahedra and one MnO₅ square pyramid. Trivalent manganese ions are energetically stabilized at the square-pyramidal sites. These octahedral and square-pyramidal sites are connected to each other by either edge or corner sharing (vertex sharing), forming the complicated framework structure with two different tunnels along the c-axis direction. Sodium ions are accommodated at three different sites in the framework structure, and the sodium ions in both large and small tunnels are highly mobile [84]. However, trivalent manganese ions at the square-pyramidal sites cannot be oxidized to the tetravalent state [84], and thus 20% of the sodium ions cannot be extracted from the framework structure. During the discharge (reduction) process, Na_x MnO₂ is reversibly uptaking sodium ions to form Na_0.67MnO₂, and approximately 120 mAh g⁻¹ of reversible capacity is obtained as electrode material. Complicated phase transitions occur in the insertion/extraction of sodium ions, as expected from the galvanostatic charge/discharge curves in figure 10(c) [83]. Detailed configurations of sodium ion ordering in Na_x MnO₂ in electrochemical cycles have been studied by first-principles calculations [84]. Recently well-crystallized, uniform nanowires (diameter ∼50 nm and growth orientation along [001]) of Na_0.44MnO₂ were prepared (figure 10(d)) by a polymer-pyrolysis method [85]. The Na_0.44MnO₂ nanowires demonstrate relatively good rate capability and excellent capacity retention, as shown in figure 10(e). Approximately 100 mAh g⁻¹ of reversible capacity is obtained in the Na cells, even after 100 cycles.

Other manganese (di)oxides studied for lithium insertion materials, such as α-MnO₂ and β-MnO₂, have also been tested as sodium insertion host materials [86]. Framework structures of both α-MnO₂ and β-MnO₂ consist of MnO₆ octahedra that are connected by both edge and corner sharing. The structure of β-MnO₂ is classified as rutile-type with relatively small 1 × 1 tunnels along the c-axis direction, which is built up with chains of edge-shared MnO₆ octahedra. Alkali ions, even small lithium ions, cannot be inserted into the small tunnels of bulk β-MnO₂ [87]. However, nanosized β-MnO₂ with ordered mesopores is electrochemically active in Li cells [87]. Large reversible capacity of >250 mAh g⁻¹ is obtained by using nanosized β-MnO₂. Although the crystalline phase is lost (changes into an amorphous phase) following electrochemical cycles, the electrode performance of β-MnO₂ is activated by using such nano-engineered samples. Therefore, it is also expected that this methodology can be used for Na insertion materials, similar to iron oxides, as described in an earlier section. Indeed, recently it was reported that β-MnO₂ nanorods prepared by a hydrothermal method with a diameter of 100 nm, which preferably grow along [001], deliver large reversible capacity (>200 mAh g⁻¹) in a Na cell. Because the radius of sodium ions is much larger than that of lithium ions, phase transitions from the rutile structure are expected to take place in Na cells. Although large reversible capacity is obtained with the β-MnO₂ nanorods, note that the difference in voltage between the charge/discharge in Na cells is much larger compared with Na insertion reactions in highly crystallized particles based on the topotactic reaction as shown in figure 5.

Alpha-type MnO₂, of the group of minerals called hollandite, has been studied in a Na cell. The structure of α-MnO₂ shown in figure 9(e) is built up with edge- and corner-shared MnO₆ octahedra, similar to β-MnO₂. Large 2 × 2 tunnels are formed with double chains of edge-shared MnO₆ octahedra. The open tunnels in α-MnO₂ are stabilized by incorporation of large cations, such as K⁺, and sodium migration in the open diffusion path is expected to be efficient. Although α-MnO₂ nanorods show large reversible capacity of >200 mAh g⁻¹ in a Na cell, polarization between charge/discharge curves appears to be large, similar to β-MnO₂ nanorods [86]. Results of the density functional theory (DFT) calculation suggest that the migration barrier of Na ions in α-MnO₂ is comparable to or slightly smaller than that of Li ions [88].

Manganese oxides with even more open paths for Na migration, 2 × 3 tunnels consisting of double and triple MnO₆ chains, are found in a mineral called psilomelane, as shown in figure 9(f) [89]. Large Ba²⁺ ions and water molecules are incorporated as a template to stabilize such large tunnels in psilomelane. A Na⁺ sample substituted for Ba²⁺ and H₂O in psilomelane, Na_0.4MnO₂, is also prepared by a simple solid-state method [26, 80]. The sample shows rather small polarization in a Na cell, even though the mechanism of sodium insertion into Na_0.4MnO₂ with 2 × 3 tunnels is not fully understood. At least 0.3 mole of sodium ions are reversibly inserted/extracted into/from the framework structure of psilomelane with the large tunnels [26].

Spinel-type LiMn₂O₄ is the one of the most widely studied materials in terms of positive electrodes for LIBs [3–6]. Layered O'3-type LiMnO₂ which is also prepared by a Na/Li ion-exchange method from O'3-NaMnO₂, is isostructural with O'3-NaMnO₂ [51, 52]. Although the electrode performance of O'3-LiMnO₂ in Li cells has also been reported, its major problem as an electrode material is the gradual phase transition to spinel with the common ccp oxygen lattice. Trivalent manganese ions should be highly mobile in the ccp lattice, and therefore, LiMnO₂ easily transforms into spinel as the energetically favorable phase during electrochemical cycles in Li cells.

In contrast, a Na counterpart, O'3-NaMnO₂, can be used as electrode material without such a phase transition into spinel. The difference probably originates from a large gap in size between Na and Mn ions. A three-dimensional framework structure such as that of spinel is stable for the Li system and not for the Na system. The question arises here whether spinel-type manganese oxides are used as electrode materials in Na cells without phase transition [90, 91]. Figure 12 compares charge/discharge curves of stoichiometric LiMn₂O₄ cycled in 1.0 mol dm⁻³ lithium or sodium perchlorate dissolved in propylene carbonate. LiMn₂O₄ shows two voltage plateaus at approximately 3.92 and 4.16 V in a Li cell in a first charge process. Although the almost identical voltage profile is observed in a Na cell, the voltage of the Na cell is 0.34 V lower than that of the Li cell. This observation directly relates to the difference in the standard electrode potential, as similarly shown in figure 2. As shown in figure 12(a), a voltage profile upon the first charge in the Na cell is superimposed on that in the Li cell when voltage is shifted higher by 0.34 V. These results clearly suggest that the same reaction proceeds in both Na and Li systems, i.e., lithium extraction from LiMn₂O₄, forming Li_x Mn₂O₄ (x = ca. 0.1). A clear difference between the two cells is found via a discharge (reduction) process. In the Li cell, Li ions are reversibly inserted into the spinel framework structure with similar voltage profiles on charge. In the Na cell, voltage rapidly decreases to 3.2 V versus Na/Na⁺, at which time a long plateau is observed (approximately 100 mAh g⁻¹ of discharge capacity), and then the voltage gradually decreases to 2.0 V. First-discharge capacity reaches approximately 200 mAh g⁻¹ in this experimental condition. Although lithium ions are extracted in Na cells, the extracted lithium ions are not reversibly inserted into Na cells, and instead Na ions seem to be inserted into Li_x Mn₂O₄ (x = ca. 0.1). Figure 12(b) shows tenth charge/discharge curves of LiMn₂O₄ electrodes cycled in Li and Na cells. Voltage profiles change during the electrochemical cycling in the Na cell. For the tenth cycle, the voltage profiles are completely different for both cells, suggesting a phase transition from the spinel phase to another phase in the Na cell. The changes in crystal structures of stoichiometric LiMn₂O₄ cycled in Na cells were examined by XRD. Figure 13 shows the XRD patterns of LiMn₂O₄ electrodes before and after the electrochemical cycling in Na cells. When the electrodes are cycled in Na cells, the XRD patterns of electrodes drastically change. Peak profiles of all the Bragg lines from LiMn₂O₄ are significantly broadened, which is indicative of the phase transition induced by sodium insertion into the spinel phase. After 5 cycles in the voltage range 2.0–4.0 V versus Na/Na⁺, new peaks appear at 17 degrees (in two theta) with the appearance of some additional peaks. After 10 cycles in the range 2.0–4.0 V, these new peaks become more visible. These new peaks observed after 10 cycles resemble those of O'3-type NaMnO₂. From the XRD patterns, newly formed O3-type Na_y MnO₂ seems to coexist with the peak broadened spinel-type phase. Because some lithium-ions still exist in the spinel phase by electrochemical oxidation to 4.0 V in Na cells, the phase transition to the layered phase may be partially suppressed. Such spinel-to-layered phase transition in common oxygen lattices (schematic illustrations are shown in figure 13) is different compared with the Li system, and the results suggest that the layered sodium manganese phase seems to be energetically stable, as recently supported by the results of first-principles calculation [49]. Stabilization of layered manganese systems is another advantage of the Na system for battery applications.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Because layered lithium iron(III) oxide LiFeO₂ shows unimpressive battery performance based on the Fe²⁺/Fe³⁺ redox couple after oxygen loss in the Li cells as previously discussed, polyanionic compounds of iron have been extensively studied based on the use of the Fe²⁺/Fe³⁺ redox couple [92]. Lithium iron(II) phosphate, LiFePO₄, is the most widely studied polyanionic compound in terms of positive electrode materials for LIBs because of the interest with regard to practical applications [93]. A crystal structure of LiFePO₄ is classified as the triphylite-type structure (SG Pnma) with a distorted hcp oxygen lattice. The structure is closely related to the olivine-type structure, and two cations (lithium and iron ions in the case of LiFePO₄) are located in two distinct octahedral sites in the common framework structure composed of XO₄ (X = Si, P, Mo, etc) tetrahedra. Phosphate ions, (PO₄)³⁻ tetrahedra, share one edge and four corners with FeO₆ octahedra, forming the framework structure of triphylite-type FePO₄, as shown in figure 14(a). Formation of edge-shared sites between FeO₆ octahedral and PO₄ tetrahedral units relatively stabilizes the energy of the Fe²⁺/Fe³⁺ redox couple through the inductive effect, resulting in high voltage (3.45 V versus Li metal) for iron-based compounds based on the Fe²⁺/Fe³⁺ redox [39, 93]. Although the FeO₆ octahedra share corners in the triphylite-type structure (figure 15(a)), electrical conductivity is still too low for electrode materials. The electrode performance of LiFePO₄ is significantly improved by carbon coating [94] and shortening of a Li diffusion path with nanosized particles. Such engineered nanosized particles deliver large reversible capacity in Li cells, which now almost reaches its theoretical limit (170 mAh g⁻¹). Conduction mechanisms of Li ions in triphylite-type LiFePO₄ have been extensively studied [95, 96]. LiO₆ octahedra are connected to one another by edge sharing along, [010] forming 1D chains, as shown in figure 15(c). Li ions migrate with a low diffusion barrier (∼150 meV [62]) along the 1D chains throughout particles if the formation of antisite defects between Li and Fe sites is negligible [96].

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

A sodium counterpart, NaFePO₄, is also studied as an electrode material for NIBs [97]. NaFePO₄ crystallizes into a maricite-type structure as the thermodynamically stable phase. Its crystal structure is also closely related to the olivine-type structure with a distorted hcp oxygen lattice. A large gap in the size of ionic radii between Na and Fe results in significant distortion in the hcp oxygen lattice (figure 15(b)) in comparison to triphylite-type LiFePO₄. Iron ions are located at octahedral sites, and FeO₆ octahedra share edges with one another, forming 1D chains as shown in figure 15(d). The structure of the 1D chains of FeO₆ in the maricite phase is essentially the same as that of the 1D LiO₆ chains in the triphylite phase. Sodium ions are located at large tetrahedral sites (note that such a site is also regarded as an irregular site coordinated by 10 oxygen within 3 Å [98]), which share corners with PO₄ tetrahedra (figure 15(b)). Because sodium sites are isolated in the structure, as shown in figure 15(b), a large barrier to sodium migration is expected in the maricite-type NaFePO₄. Indeed, the reversibility of Na extraction/insertion for maricite-type NaFePO₄ seems to be unacceptable in terms of electrode material [97, 99].

In contrast, triphylite-type NaFePO₄, which is a metastable polymorph of NaFePO₄, is electrochemically active [97, 100]. Triphylite-type NaFePO₄ can be prepared by an ion-exchange method from LiFePO₄. Li ions are chemically (and electrochemically) extracted from triphylite-LiFePO₄ without the destruction of its core structure, forming heterosite-type FePO₄. Chemical (and electrochemical) Na insertion into heterosite-type FePO₄, which possesses the same framework structure as triphylite-LiFePO₄, results in the formation of triphylite-type NaFePO₄. Triphylite-type NaFePO₄ is stable below 480 °C in inert atmosphere and transforms into maricite-NaFePO₄ by further heating above 480 °C [100, 101]. Charge/discharge (oxidation/reduction) curves of triphylite-NaFePO₄ in a Na cell [100] are shown in figure 16(a). Nearly one mole of Na ions are reversibly inserted/extracted into/from triphylite-NaFePO₄. Because the ionic radius of the sodium ion is much larger than that of the lithium ion, the unit cell volume of triphylite-NaFePO₄ is approximately 10% larger than that of triphylite-LiFePO₄. Two voltage plateaus are observed on discharge (figure 16(a)). Open-circuit voltage on the plateaus is observed to be 2.87 V and 2.97 V [100], which is slightly lower than that of ∼3.1 V versus Na as expected from LiFePO₄ (3.45 V vs Li). The voltage difference between Li and Na cells is, however, less significant than that observed for the layered oxides (figure 2). The appearance of two voltage plateaus originates from the formation of an intermediate phase (a sodium-ion ordered phase) as Na_0.4FePO₄ (point B in figure 16(a)). Such an intermediate phase is not generally found in Li_x FePO₄ under equilibrium conditions. Larger repulsive interaction for Na ions, as compared with Li ions, may result in the formation of the intermediate phase [100, 102], similar to layered oxides. According to the DFT calculation, the barrier of sodium migration in narrow 1D chains is much larger (270 meV) compared with Li ions (150 meV) [62]. Diffusion barriers correlate highly with diffusion paths in host structures. Indeed, experimentally measured sodium insertion kinetics in triphylite-Na_x FePO₄ appear to be much slower compared with Li ions [103].

Zoom In Zoom Out Reset image size

The thermodynamically stable phase of NaMnPO₄ is also maricite type [104]. Similar to triphylite-NaFePO₄, it is also expected that lithiophilite-type NaMnPO₄, which is isostructural with triphylite-type Li(Na)FePO₄, is obtained by ion-exchange reaction and/or related soft chemical methods. It has been demonstrated that lithiophilite-type NaMnPO₄ is prepared by topochemical reaction below 100 °C from NH₄MnPO₄·H₂O as the precursor [104]. Electrode performance of triphylite-NaMn_1/2Fe_1/2PO₄ prepared by topochemical reaction has also been reported [104]. Electrode reversibility of lithiophilite-NaMnPO₄ may be limited without the preparation of nano-engineered particles, similar to LiMnPO₄, because of the formation of antisite defects [105, 106] and/or the slow nucleation rate for the delithiated (desodiated) phase [107].

Following reports touting lithium iron(II) pyrophosphate, Li₂FeP₂O₇, as a positive electrode material for LIBs [108, 109], a great deal of literature has been published regarding the pyrophosphate system. The use of pyrophosphate, instead of phosphate, as the framework structure has been extended to the Na system, i.e., Na₂FeP₂O₇ [110–112]. Na₂FeP₂O₇ is isostructural with one of the polymorphs of Na₂CoP₂O₇, with an SG of P1 [113]. The crystal structure of Na₂FeP₂O₇ contains corner-shared FeO₆ octahedra (Fe₂O₁₁ units), which are connected by P₂O₇ units, forming the 3D framework structure with open sodium diffusion paths (figure 14(c)). Na ions are located at large distorted square-pyramidal sites. One Na ion is reversibly extracted from Na₂FeP₂O₇ based on the Fe²⁺/Fe³⁺ redox couple, and reversible capacity reaches 90 mAh g⁻¹ (theoretical capacity: 97 mAh g⁻¹) as shown in figure 16(b). Although the available energy density is lower than that of triphylite-NaFePO₄, direct synthesis of Na₂FeP₂O₇ is possible by conventional solid-state methods. Additionally, rate capability seems to be much better for Na₂FeP₂O₇, even though FeO₆ octahedra and FeO₅ square pyramids are isolated by the pyrophosphate ions in the framework structure. This result is probably because of the presence of open diffusion paths formed by pyrophosphate ions. Na diffusion barriers in Na_x FeP₂O₇ have been also calculated by first-principles calculation [112]. Sodium ions can diffuse through 1D paths along [011] with a relatively low migration barrier (∼480 meV), and all Na sites are interconnected by 1D/2D paths with a diffusion barrier below 540 meV.

Recently it has been reported that one of the Na₂MnP₂O₇ polymorphs, β-Na₂MnP₂O₇, which is isostructural with Na₂FeP₂O₇, also shows good electrode performance in Na cells [114]. Although the polarization seems to be large, Na₂MnP₂O₇ is potentially usable as a 3.6 V–class electrode material. Off-stoichiometric synthesis for sodium metal pyrophosphates induces the formation of Na-rich phases [115]. Thus the prepared sample shows better electrode performance than does the stoichiometric sample without nanosizing and carbon coating. Pyrophosphates that include other transition metal elements as electrode materials have been reviewed in the literature [116].

An interesting characteristic regarding the design of electrode materials for NIBs is the high structural flexibility of sodium-based compounds, e.g., the P2-type layered structure as mentioned in an earlier section. It is impossible to prepare a P2-type layered structure for the Li system. Similarly, a unique framework structure, with mixed polyanion groups of phosphate (PO₄)³⁻ and prylophosphate (P₂O₇)⁴⁻ ions, is found in the Na system (and is not known in the Li system). Recently Na₄Fe₃(PO₄)₂(P₂O₇) was synthesized as the first mixed polyanion compound containing Fe(II) [117]. The structure is isostructural with Na₄Me₃(PO₄)₂(P₂O₇) (Me = Co, Mn, and Ni), with an SG of Pn2₁a [118]. FeO₆ octahedra are connected to on another by edge and corner shares, and PO₄ tetrahedra also share one edge and two corners with the FeO₆ octahedra, forming a layer-like unit along the b-c direction. These layer units, consisting of FeO₆ and PO₄, are further connected by P₂O₇ pyrophosphate ions, forming the crystal structure of Na₄Fe₃(PO₄)₂(P₂O₇), as shown in figure 14(d). Na₄Fe₃(PO₄)₂(P₂O₇) delivers approximately 100 mAh g⁻¹ of reversible capacity, which is higher than that of Na₂FeP₂O₇ (figure 16(c)). Three Na ions are reversibly extracted from Na₄Fe₃(PO₄)₂(P₂O₇) based on the Fe²⁺/Fe³⁺ redox. The sample shows good capacity retention as electrode material in Na cells [117].

The mixed-anions system containing fluoride and phosphate ions is also used as electrode material for NIBs. The layered fluorinated iron phosphate Na₂FePO₄F has been widely examined as a positive electrode material for rechargeable batteries [119, 120]. The crystal structure of Na₂FePO₄F is isostructural with Na₂CoPO₄F-type structures [121] (or Na₂FePO₄(OH)-type structures) as two-dimensional layered fluorophosphates (figure 17(a)). Na ions are accommodated between FePO₄F layers, in which FeO₄F₂ octahedra share edge and corners. According to the results of first-principles calculation, sodium ions migrate between FePO₄F layers through a two-dimensional path [122]. The crystal structure of Na₂MnPO₄F is different from that of Na₂FePO₄F [123]. All MnO₄F₂ octahedra share each corner and form 1D Mn₂F₂O₈ chains. The chains are connected by PO₄ tetrahedra via corner sharing, thus forming a less dense 3D framework structure (figure 17(b)). Na₂[Fe_x Mn_1 − x]PO₄F (0 < x < 1.0) samples have also been synthesized, and it was found that the Na₂MnPO₄F-type 3D structure is thermodynamically stable at x < 0.75 in Na₂[Fe_x Mn_1 − x]PO₄F [120].

Zoom In Zoom Out Reset image size

The electrochemical behavior of carbon-coated Na₂FePO₄F and Na[Fe_1/2Mn_1/2]PO₄F in Na cells is compared in figure 18 [124, 125]. Na₂FePO₄F demonstrates a relatively high operating voltage (ca. 3.0 V versus Na) and fair reversibility as an electrode material. The carbon-coated Na₂FePO₄F sample delivers approximately 110 mAh g⁻¹ of reversible capacity, which corresponds to approximately 90% of theoretical capacity based on the one-electron redox of iron (Fe²⁺/Fe³⁺). Two well-defined voltage plateaus, which are centered at 3.06 and 2.91 V with small polarization, are observed in figure 18(a). The Na/Na₂[Fe_1/2Mn_1/2]PO₄F cell also shows a reversible capacity of 110 mAh g⁻¹ (figure 18(b)). In addition, from the operating voltage, the reaction in the Na cell consists of three different regions. Differential capacity dQ/dV plots of both samples are also shown in figures 18(a), (b) (insets). In the discharge process for the Na/Na₂[Fe_1/2Mn_1/2]PO₄F cell, three peaks are clearly observed at 3.36, 3.04, and 2.86 V. A redox couple centered at 3.53 V is found in the dQ/dV plot of Na₂[Fe_1/2Mn_1/2]PO₄F, which may be attributable to a Mn²⁺/Mn³⁺ redox reaction. The polarization of the Mn redox is larger than that of Fe²⁺/Fe³⁺, similar to the iron/manganese phosphates. Although the layered framework structure seems to be preferable for the sodium migration process in the fluorophosphates system, carbon-coated, nanosized Na₂[Fe_1/2Mn_1/2]PO₄F samples show good reversibility as electrode material in Na cells [125].

Zoom In Zoom Out Reset image size

Recently synthesis and electrode performance have been reported for sodium-based carbonophosphates, Na₃MePO₄CO₃, which are the new series of mixed-anion compounds used as electrode materials, similar to Na₄Fe₃(PO₄)₂(P₂O₇) [126, 127]. Na₃MePO₄CO₃ contains two different anions of carbonate (CO₃)²⁻ and phosphate (PO₄)³⁻ ions, and its Mn compound, Na₃MnPO₄CO₃, is found in the natural mineral sidorenkite [126]. The crystal structure of Na₃MnPO₄CO₃ is shown in figure 17(c). FeO₆ octahedra and PO₄ tetrahedra share each corner and CO₃ triangles share one edge with the FeO₆ octahedra, resulting in a large distortion of the FeO₆ octahedra. Na₃MnPO₄CO₃ can be prepared by hydrothermal reaction at 120 °C [116]. Two Na ions are extracted from Na₃MnPO₄CO₃ based on the Mn²⁺/Mn⁴⁺ two-electron redox reaction, resulting in large initial charge capacity (figure 18(c)). Although the initial charge capacity is large (∼200 mAh g⁻¹), it has been reported that the sample shows relatively large irreversible capacity (∼75 mAh g⁻¹) in the initial cycle, with acceptable capacity retention in subsequent continuous cycles [126].

Metal fluorosulfates are also used as the framework structure for the Na insertion host. Recently sodium iron fluorosulfate, NaFeSO₄F, was successfully prepared and its electrode performance was examined in Na cells. The crystal structure of NaFeSO₄F is isostructural with the mineral tavorite, LiFe(PO₄)OH [128]. Although the results of first-principles calculation suggest diffusion of sodium ions in the 1D path along [101] [129], reversibility of NaFeSO₄F in terms of electrode material is limited to a narrow range [128].

Because trivalent iron Fe(III) is, in general, stable under ambient conditions, many Fe(III)-containing phosphates are found in natural minerals. Some Fe(III)-containing phosphates are known to be electrochemically active as sodium insertion hosts, based on the Fe²⁺/Fe³⁺ redox. The simplest compounds are FePO₄ with polymorphs, including an amorphous phase. A thermodynamically stable phase is isostructural with the mineral berlinite, AlPO₄ [130], in which FeO₄ and PO₄ tetrahedra share corners with each other, forming relatively open 3D channels for the insertion of alkali ions (figure 19(a)). The electrode performance of amorphous and crystalline FePO₄ (berlinite-type) is reported by Okada and co-workers [131]. Figure 20(a) shows the electrode performance of amorphous FePO₄ in Li and Na cells. Approximately 80 mAh g⁻¹ of reversible capacity is obtained using amorphous FePO₄ in the Na cell. Crystalline FePO₄ obtained by heating the amorphous phase is also used as an electrode material [131]. FePO₄ polymorphs can be prepared by topochemical reaction from hydrates, and one of them consists of corner-shared FeO₆ octahedra and PO₄ tetrahedra [132]. These polymorphs are electrochemically active in Li cells [132], but electrode performance in Na cells has not been reported so far.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

According to a study on a Na₃PO₄–FePO₄ binary phase diagram [133], at least three different phases exist: Na₃Fe(PO₄)₂, Na₃Fe₂(PO₄)₃, and Na₃Fe₃(PO₄)₃. A structural type of the most Na-rich phase, Na₃Fe(PO₄)₂, was reported to be related to the mineral glaserite, K₃Na(SO₄)₂ [134]. FeO₆ octahedra and PO₄ tetrahedra share corners, forming 2D Fe₂(PO₄)₂ layers. Sodium ions are accommodated between the Fe₂(PO₄)₂ layers. Similar to Na₃Fe(PO₄)₂, Na₃Fe₂(PO₄)₃ also consists of corner-shared FeO₆ octahedra and PO₄ tetrahedra, but Na₃Fe₂(PO₄)₃ has a 3D framework structure. Such a structure of Na₃Fe₂(PO₄)₃ is known to be related to the Na⁺-ion super ionic conductors (NASICON) type of structure (figure 19(b)), and the polymorphic characteristics of Na₃Fe₂(PO₄)₃ have been extensively studied [135]. Ion-exchanged Li₃Fe₂(PO₄)₃ from Na₃Fe₂(PO₄)₃ is known to be electrochemically active in Li cells [39]. Approximately 0.4 mole of Na is reversibly inserted/extracted into/from Na₃Fe₂(PO₄)₃ in Na cells [136], even though ionic conductivity of Na₃Fe₂(PO₄)₃ is reported to be very low (1.2 × 10⁻⁷ S cm⁻¹ at 50 °C) [137].

Na₃Fe₃(PO₄)₄ is also classified as a layered-type phosphate [133]. FeO₆ octahedra and PO₄ in Na₃Fe₃(PO₄)₄ tetrahedra share both edge and corners, forming 2D Fe₃(PO₄)₄ layers (figure 19(c)). This type of layered structure is found in K₃Fe₃(PO₄)₄ • H₂O [138]. The electrode performance of Na₃Fe₃(PO₄)₄ in Li and Na cells has been reported [139, 140]. Approximately 2 moles of Li and Na ions are reversibly inserted into Na₃Fe₃(PO₄)₄. Polarization in the Na cell is much smaller than in the Li cell, as shown in figure 20(b). Although highly crystallized and microsized particles show good electrode performance, the operating voltage is relatively low—2.5 V versus Na, based on the Fe²⁺/Fe³⁺ redox.

Binary phosphates containing Fe(III) and Mn(II) are also found in natural minerals. One example is the mineral alluaudite [141]. The crystal structure of alluaudite contains 1D chains made from edge-shared Fe(Mn)O₆ octahedra with distortion. These 1D chains are connected with PO₄ tetrahedra by corner sharing, forming the 1D tunnel sites for Na ions (figure 19(d)). Na ions are located at distorted octahedral sites and large eight oxygen coordinated sites. Electrode performance of alluaudite-type NaMnFe₂(PO₄)₃ has been tested in Na cells [140]. Na ions (ca. 0.5 mole) can be partially extracted from NaMnFe₂(PO₄)₃ prepared using a sol-gel method, and the reversible capacity reaches 60 mAh g⁻¹ in a Na cell (figure 20(c)). Much better reversibility was found in Na₃Fe₃(PO₄)₄ without Mn(II) ions.

The electrode performance of the fluorophosphate containing Fe(III), Na₃Fe₂(PO₄)₂F₃, was recently reported [142]. The crystal structure of Na₃Fe₂(PO₄)₂F₃ is isostructural with Na₃Fe₂(PO₄)₂(OH)₂F [143]. The framework structure of Na₃Fe₂(PO₄)₂F₃ consists of FeO₄F₂ octahedra and PO₄ tetrahedra, all of which share corners, forming Fe₂(PO₄)2F₃ layers. A two-dimensional migration path for sodium ions along the a–b plane is formed by corner sharing for the FeO₄F₂ octahedra in the Fe₂(PO₄)₂F₃ layers. Na₃Fe₂(PO₄)₂F₃ delivers 40 mAh g⁻¹ of reversible capacity in a Na cell [142].