Introduction

Sodium-ion battery (SIB), or Na-ion battery (NIB) is a rechargeable battery that employs the reversible electrochemical reaction of sodium ions to store electrical energy. The ions reversibly move between the electrodes, creating a potential difference that gives a voltage to the cell.

Na-ion batteries operate on the same working principle as Li-ion batteries (LIBs); therefore, the study of NIBs began with the LIBs (1980s). However, after 1991 the focus shifted almost exclusively to LIBs due to the faster development of this technology and bigger commercial success. As sustainability and scarcity concerns arose in recent years, the Na-ion battery research experienced a renaissance. ^[1]

^[2][3]

Operating principle

A sodium-ion battery consists of a cell that uses two sodium insertion materials at positive and negative electrodes and an electrolyte. As battery is charged and discharged, the Na+ ions shuttle between the cathode and the anode. This is also can be described as a ‘rocking chair’ battery because the sodium ions “rock” to and from the electrode during the cycling.

Figure 1. A schematic illustration of a Na-ion battery. (Figure: Amina Alimbekova)

The reaction in the NIB cell that consists of, for example, layered carbon (C₆) anode and M as a layered d-metal oxide (like the LiC₆/Li_1−xMO₂ setup) can be seen in Fig 1. The reaction is thus (discharging reaction is from left to right): \[ { Na_{x}C_{6} + Na_{1-x}MO_{2} <=> NaMO_{2} + C_{6} } \] The half reactions are:

for positive electrode  \( { Na_{1-x}MO_{2} + xNa^{+} + xe^{-} <=> NaMO_{2} } \) ,

for negative electrode  \( Na_{x}C_{6} <=> C_{6} + xNa^{+} + xe^{-} \) .

Cathode Materials

Since the underlying reaction is reliant on intercalation reaction, highly reversible materials that allow enough interstitial Na⁺ ions are necessary. Such materials usually fall under categories of oxides, phosphates, fluorosulfates, etc. The structures of these compounds ideally should not change upon reversible intercalation.

In the case of oxides, layered oxides in particular, have been extensively studied for NIBs as well as for LIBs. The common layered oxides are MO₆ (M – 3d transition metal) octahedra that share the edges and stacked vertically. Depending on the orientation of the stacking O (octahedral)-type or a P (prismatic)-type phases can be distinguished (Fig. 2)

Figure 2. A schematic illustration of a crystal structures: a) side (above) and top (below) view of (00l) layer of the O3 phase; b) side (above) and top (below) view of (00l) layer of the P2 phase, where green sphere is face shared Na1
and blue one is edge sharing Na2 atoms; c) side (above) and top (below) view of (00l) layer of the P3 phase. (Figure: Amina Alimbekova)

O-type (O – for octahedral) of NaMO₂ can be described as alternate layers of NaO₂ and MO₂, where sodium is at the octahedral sites between the layers of MO₂. P-type (P – for prismatic) also appears as layers of NaO₂ and MO₂, however the sodium is embedded in trigonal ‘prismatic’ sites. For O2, O3, P2, P3 the number specifies the packing number of Na ions. Also, there could be monoclinic distortion of distinct phases, which is described by the prime ( ‘ ). The oxides are not restricted to a single phase, rather they experience transition according to the amount of sodium within layers. O3 and P2 type structures can be seen for sodiated oxides, Na_1-xMO₂ (where x~0). When sodium ions are removed electrochemically, the structures transform reversibly, for example O3 ↔ O’3 ↔P3 ↔ P’3 by sliding MO₂ layers. When Na⁺ are rare within the oxide, the crystal structure energetically prefers prismatic phase, as it creates vacancies. Additionally, the interlayer distance increases as oxygens cause strong repulsion. Thus, in P’3 phase the diffusion of incoming Na⁺ ions happens quicker.^[4][5]

NaFeO₂

One of the ‘classic’ cathode materials is a O3 type layered structure of α-NaFeO₂ (Fig. 3). Its lithium counterpart, LiFeO₂, forms a disordered rock-salt phase as ionic radii of Li⁺ and Fe³⁺ are similar (Li⁺ - 0.76 Å, Fe³⁺ - 0.645 Å (high spin)^[6]). This material was found to be inactive electrochemically due to Fe³⁺ cations blocking Li⁺ intercalation and deintercalation^[7]. On the contrary, the NaFeO₂ can be prepared by solid state reactions, where oxides are heated up for a fixed amount of time. The resulting structures can be α-NaFeO₂, or β-NaFeO₂. The structure of α-NaFeO₂, with space group R3m can be described as a ‘rock-salt’ structure, where oxygen is a cubic close packed structure with iron and sodium are in the octagonal holes. Also, the oxygen-iron layers are separated by O^2- – Na⁺ – O^2-sandwiches (Fig. 3a). β-NaFeO2 structure is a wurtzite-like structure with a Pna2₁ space group. Here, the oxygens are hexagonal close packed, and Na and Fe reside in a half tetrahedral site in a random fashion (Fig. 3b).^[8]

Figure 3. A schematic illustration of a crystal structures: a) side and top view of (00l) layer of α-NaFeO₂; b) side and top view of (00l) layer of the β-NaFeO₂. O atoms are red, Na polyhedra green, Fe polyhedra pink. (Figure: Amina Alimbekova). Crystallography data reference^[9][10]

The electrochemical reversibility of sodium ion intercalation/deintercalation for the α-NaFeO₂ while possible, is accompanied by irreversible transformation of crystal structure to monoclinic P3 phase. As the structure changes, it is signified by iron ion migration, which hinders Na⁺ ion intercalation upon discharge. Here a part of Fe³⁺ ions move into the now vacant tetrahedral sites causing the structure change (Fig. 4c). This can be confirmed with the XRD (Fig. x). As x = 0.5 for Na_1-xFeO₂, the position of the principal reflections shift as well as some peaks split. 

Figure 4. a) side view of (00l) layer of α-NaFeO₂; b) side view of (00l) layer of α-Na_0.5FeO₂; c) comparison of crystal structures of unit cells of α-NaFeO₂ (hexagonal, solid) and α-Na_0.5FeO₂ (monoclinic, dotted); c) comparison of XRD of α-NaFeO₂ and α-Na_0.5FeO₂. (Figure: Amina Alimbekova. ICSD: 75588,14779)

Anode Materials

Anodes for NIBs can be grouped into the following types: carbon-based materials, oxides and polyanionic compounds that are able to change phases, certain metals and alloys^[11]. As compared previously to LIBs, graphitic carbon can be used as an insertion host for sodium ions with the resulting stoichiometry NaC₆. One of the problems with this anode material, however, is the inefficient intercalation of Na⁺ into the graphitic interlayer due to the larger size of the ion. One of the solutions is to widen the layer distance by oxidation-reduction steps of graphite, as reported by Wen et al. (2014)^[12]. 

Electrolytes

Electrolytes in NIBs, like in LIBs or others, are required to be chemically, electrochemically and thermally stable, electronically insulating and ionically conductive, with low toxicity and minimal costs. Electrolytes usually consist of a salt and a solvent, so the salt should be soluble stable in relation to redox reactions, and to the cell components. The solvent should be polar, less viscous, inert in relation to the cell components and have a high dielectric constant, a low melting point and a high boiling point. Not a wide range of compounds can check these requirements simultaneously, but there are electrolyte salts that can fit better than others. NaClO₄ or NaPF₆ can be used as the salts in carbonate-ester binary/ternary mixtures, for example dimethyl ether (DME), propylene carbonate (PC), or ethylene carbonate (EC). ^[13],^[14]

Na-ion vs Li-ion

Na is the seventh most abundant element in the earth’s crust (22 700 ppm) and can be easily accessed in the ocean, while Li is considerably scarcer in the crustal rocks (18 ppm)^[15]. Thus, using sodium is cheaper to use for battery production and due to wide distribution around the world, makes the NIBs less affected by supply risks.

Due to similar working principles of NIBs to LIBs, there is a ready groundwork for cell production scale-ups, as necessary equipment, materials, facilities, and personnel can be used.^[1]

Although sodium-ion batteries appear to be a promising alternative, improvements to the cell design and more efficient electrode materials should be investigated.

References