eScience 2 (2022) 10–31

Contents lists available at ScienceDirect

eScience
journal homepage: www.keaipublishing.com/en/journals/escience

Review

Advanced characterizations and measurements for sodium-ion batteries

with NASICON-type cathode materials
Yukun Liu a, 1, Jie Li a, 1, Qiuyu Shen a, Jian Zhang a, Pingge He b, *, Xuanhui Qu a,
Yongchang Liu a, *
a
Beijing Advanced Innovation Center for Materials Genome Engineering, Institute for Advanced Materials and Technology, State Key Laboratory for Advanced Metals and
Materials, University of Science and Technology Beijing, Beijing, 100083, China
b
Department of Chemistry and Biochemistry, University of California, 1156 High Street, Santa Cruz, CA, 95064, United States

H I G H L I G H T S G R A P H I C A L A B S T R A C T

The advanced characterizations and

measurements for SIBs with NASICON-
type cathodes have been comprehen-
sively summarized.

The detection principles of advanced
characterization/measurement tech-
niques have been deeply discussed.

Both operando and ex situ techniques
have been highlighted to understand the
structure-mechanism-performance
correlations.

Challenges and promises in developing
high-performance NASICON-cathodes
with advanced techniques are outlined.

A R T I C L E I N F O A B S T R A C T

Keywords: NASICON (Na superionic conductor)-type cathode materials for sodium-ion batteries (SIBs) have attracted
Sodium-ion batteries extensive attention due to their mechanically robust three-dimensional (3D) framework, which has sufficient open
NASICON-Type cathodes channels for fast Naþ transportation. However, they usually suffer from inferior electronic conductivity and low
Characterization techniques
capacity, which severely limit their practical applications. To solve these issues, we need to deeply understand the
Electrochemical measurements
Theoretical computations
structural evolution, redox mechanisms, and electrode/electrolyte interface reactions during cycling. Recently,
rapid developments in synchrotron X-ray techniques, neutron-based resources, magnetic resonance, as well as
optical and electron microscopy have brought numerous opportunities to gain deep insights into the Na-storage
behaviors of NASICON cathodes. In this review, we summarize the detection principles of advanced character-
ization techniques used with typical NASICON-structured cathode materials for SIBs. The special focus is on both
operando and ex situ techniques, which help to investigate the relationships among phase, composition, and
valence variations within electrochemical responses. Fresh electrochemical measurements and theoretical com-
putations are also included to reveal the kinetics and energy-storage mechanisms of electrodes upon charge/
discharge. Finally, we describe potential new developments in NASICON-cathodes with optimized SIB systems,
foreseeing a bright future for them, achievable through the rational application of advanced diagnostic methods.

* Corresponding authors.
E-mail addresses: [email protected] (P. He), [email protected] (Y. Liu).
1
These authors contributed equally to this work.

https://doi.org/10.1016/j.esci.2021.12.008
Received 13 August 2021; Received in revised form 8 November 2021; Accepted 24 December 2021
Available online 29 December 2021
2667-1417/© 2021 The Authors. Published by Elsevier B.V. on behalf of Nankai University. This is an open access article under the CC BY-NC-ND license (http://
creativecommons.org/licenses/by-nc-nd/4.0/).
Y. Liu et al. eScience 2 (2022) 10–31

1. Introduction materials [22]. Among these, polyanionic compounds are under the
research spotlight due to their high operating potential, stable framework
Rechargeable batteries play a key role in modern society on ac- structure, and high safety [23–30].
count of their scalability, cost-effectiveness, and high energy storage NASICON (Na superionic conductor)-type cathodes, a classic group of
efficiency [1]. Since the first recyclable lead-acid battery appeared on polyanionic materials, are well known for their high Na-ion conductivity
the scene in 1895, numerous battery technologies have been invented and stable 3D framework [31,32]. Represented as NaxM2(XO4)3 (where
and developed. Among them, lithium-ion batteries (LIBs) are at the x ¼ 0–4; M ¼ V, Mn, Ti, Fe, Ni, Co, Cr, etc.; X ¼ P, S, Si, As, etc.),
forefront and have made excellent commercial gains in portable NASICON-type cathodes usually exhibit a thermally stable rhombohedral
electronic devices and electric vehicle applications, owing to their high structure (space group: R 3 c) and a lantern-like framework unit, wherein
energy density and long service life [2]. However, the ever-increasing each MO6 octahedron shares corners with three XO4 tetrahedra,
demand for LIBs has caused a surge in the price of lithium due to providing abundant open Naþ diffusion channels. In the crystal structure
limited availability, undermining its feasibility for large-scale energy are two kinds of Naþ ions: Na1 sites show sixfold coordination (6b) along
storage. To address this issue, sodium-ion batteries (SIBs) have been the c-axis, and Na2 sites possess eightfold coordination (18e) along the
extensively investigated, since sodium is an abundant element with b-axis (inset of Fig. 1) [33]. Due to their relatively weaker bonding en-
uniform global distribution [3–6]. Nevertheless, Na ions are larger ergy, the Naþ ions in Na2 sites are more easily extracted than those in
(1.02 Å versus 0.76 Å) and heavier (23 g mol–1 versus 6.9 g mol–1) Na1 sites. Moreover, both the cation M and anion X in the NaxM2(XO4)3
than Li ions, resulting in poor phase stability, sluggish ion-transport structure can be regulated and replaced, thus offering opportunities for
dynamics, and undesirable interphase formation in Na host materials the structural engineering [34,35] of these cathode materials.
[7–9]. Hence, the rational design of effective and durable electrodes Nonetheless, NASICON-type cathodes still face traditional challenges,
with outstanding electrochemical performance, based on a compre- such as poor electronic conductivity and low specific capacity. To solve
hensive understanding of their presently elusive reaction mechanisms, these issues, numerous modification methods have been employed, such
is highly necessary for developing high-performance SIBs. as carbon coating, nanosizing, manipulation of redox couples, and het-
As an essential component of SIBs, cathode materials provide active eroatom doping [36–42]. However, puzzles continue to emerge about the
Naþ ions, determine the operating voltage, and account for the largest activation of high-potential redox centers, electrolyte decomposition,
proportion of SIB cost (up to 32%). Various cathodes have been devel- and complicated phase transitions. State-of-the-art characterization
oped recently, including transition metal oxides [10–15], polyanionic techniques are therefore much needed to gain deep insights into the re-
compounds [16–19], Prussian blue analogues [20,21], and organic action mechanisms of NASICON-type materials. Investigating the

Fig. 1. Summary of advanced characterizations and measurements to probe NASICON-type cathode materials for SIBs.

11
Y. Liu et al. eScience 2 (2022) 10–31

transport properties of Naþ ions and the pseudocapacitive contribution in battery performance in practical applications. In situ/operando XRD and
NASICON-type cathodes also requires the assistance of modern electro- ex situ XRD are very helpful for revealing structural changes and phase
chemical measurements and advanced computational techniques to gain transitions in electrode materials during the charge/discharge process.
a better understanding of the kinetic behaviors of NASICON-type The terms “in situ” and “operando” refer to conducting a measurement
cathodes. when the battery is operating [51,52], while ex situ experiments require
To date, numerous novel technologies and kinetic analysis methods disassembling the cell in a glovebox and cleaning the electrode with
for characterizing NASICON-type cathodes have emerged, such as electrolyte solvent. Hence, in situ/operando experiments effectively
neutron diffraction (ND), in situ Raman spectroscopy, X-ray absorption reduce the impact from the external environment and increase the reli-
spectroscopy (XAS), and the galvanostatic intermittent titration tech- ability of results.
nique (GITT) [43–46], among others. A systematic summary of these In the diffraction patterns of electrodes in different charge/discharge
techniques will be very helpful for future research on NASICON cathodes states, reversible position changes in diffraction peaks are usually related
to enable the further development of high-performance SIBs. Some re- to a solid-solution reaction, and two-phase transformations can be
views have examined structure–composition–performance links, as well confirmed by the splitting, appearance, and disappearance of peaks. For
as design principles and engineering strategies to improve the electro- example, Lim et al. verified the reversible two-phase reaction of NVP
chemical performance of NASICON materials [35,47,48]. However, the during Naþ extraction/insertion by applying ex situ XRD [53]. There-
literature is lacking in summarizing the novel diagnostic methods for use after, Hu's group conducted in situ XRD to analyze the sodiation/deso-
with NASICON materials to better understand the correlations between diation mechanism of NVP, reaching the same conclusion of two-phase
and among the structure, morphology, composition, electrochemical transformation between NVP and NaV2(PO4)3 (Fig. 2a). In addition, the
performance, and dynamics mechanism of NASICON cathodes for SIBs. superior cycling performance of NVP electrodes was demonstrated to be
Herein, our review mainly focuses on advanced characterizations and due to a tiny volume change of 8.26% upon electrochemical charge/di-
measurements that can be used to study the NASICON-type cathodes of scharge [54].
SIBs, presenting a combination of detection principles and typical ex- The in situ XRD method is expected to reveal more in-depth infor-
amples and illustrating what valuable information can be gained through mation about NVP cathode materials at various temperatures and current
these techniques. Both in situ/operando and ex situ technologies have densities. Mai and co-workers investigated the phase transition process of
proven to be superior tools for exploring the reaction mechanisms in NVP at different temperatures (273 and 293 K) and scan rates (0.5, 2, and
sodiation/desodiation. The penultimate section of the paper covers 5 mV s–1) using in situ XRD. Fig. 2b shows that at a high scan rate
theoretical computations, which can be helpful for accelerating the (5 mV s–1), the phase transformation became asymmetric: it turned out to
development of high-performance electrodes for SIBs (as illustrated in be a single-phase solid-solution reaction during the anodic period but a
Fig. 1). At the end of the review, we offer further perspectives on pro- typical two-phase evolution in the cathodic process, which was different
moting the commercial application of NASICON-type cathodes for SIBs. from the symmetric two-phase reaction at a low scan rate. These findings
provide useful guidance for the application of SIBs under various working
2. Material characterization techniques conditions [55].
As vanadium (V) is expensive and toxic, substituting other metals for
The crystal structure, morphology, microstructure, and electronic the V in NVP is opening up a new research direction. In the meantime, the
structure of electrode materials have crucial roles in investigations of working voltage of SIBs can be enhanced by activating the V4þ/V5þ redox
Naþ transport pathways, structural evolution, and charge compensation couple or introducing other high-potential redox pairs [56,57]. Mas-
mechanisms during the Naþ (de)intercalation process, which are closely quelier and co-workers prepared a high-capacity Na4MnV(PO4)3 (156
associated with the Na-storage properties of batteries. Therefore, it is mAh g–1) cathode with various redox couples of V3þ/V4þ (3.38 V),
very important to combine multiple diagnostic techniques to acquire a Mn2þ/Mn3þ (3.58 V), and V4þ/V5þ (3.86 V). However, the operando
comprehensive knowledge of the structure of NASICON cathode mate- XRD results revealed irreversible structural evolution during the first
rials for the development of SIBs. cycle, which caused large capacity loss [58]. Jiao's group then prepared a
Na4MnV(PO4)3@C@graphene aerogel (GA) material. The ex situ XRD
patterns showed that the diffraction peaks of (202) and (226) split into
2.1. Crystal structure-related technologies two peaks in the voltage range of 3.55–3.8 V upon charging, then
returned to single peaks after fully discharging (Fig. 2c), suggesting a
2.1.1. X-ray diffraction single-phase reaction based on V3þ/V4þ redox, followed by a biphase
X-ray diffraction (XRD) is a well-established tool for identifying the reaction based on Mn2þ/Mn3þ redox [59].
crystal structure and phase evolution of electrodes during charge/ Drawing upon research on symmetric batteries built on Na2TiV(PO4)3
discharge cycling. Diffraction patterns containing structural information [60], Masquelier's group studied the oxidation/reduction mechanism of
are generated by X-ray scattering of the long-range ordered atoms in a Na2TiV(PO4)3 through operando XRD. They revealed that a fully
crystal. Rietveld refinement is widely applied in XRD data fitting [49] to reversible biphasic reaction occurred at 3.33 V (V3þ/V4þ) during
provide massive information on lattice parameters, atomic position, oxidation from Na2TiV(PO4)3 to NaTiV(PO4)3, while reduction to
atomic occupancies, and the stoichiometry of electrode materials. For Na4TiV(PO4)3 involved several mono- and bi-phasic steps that included
instance, Hu and co-workers successfully obtained structural information the formation of intermediate Na3TiV(PO4)3 [61]. Li et al. designed a
about the lattice parameters and Na occupation positions of Na3V2(PO4)3 vacancy-contained NASICON cathode through the substitution of Mo6þ
(NVP) and sodium-extracted NaV2(PO4)3 using Rietveld-refined XRD. in the V sites of NVP. Ex situ XRD revealed structural transformation at
Compared with NVP, the sodium-extracted NaV2(PO4)3 showed various states of the charge/discharge process. The appearance of a series
decreased cell parameters but still maintained a rhombohedral structure of new peaks at point III indicated the phase changed from Na2.9V1.98-
with the R 3 c space group. Moreover, two different occupancy sites of Mo0.02(PO4)3 to Na1.9V1.98Mo0.02(PO4)3 (Fig. 2d). The XRD patterns
Naþ in NVP were revealed — the 6b site (Na1) and 18e sites (Na2). Only showed the recovered NASICON structure at the end of discharging,
the Na1 site was occupied in NaV2(PO4)3, indicating that Naþ had been demonstrating the minimal influence of Mo6þ doping on the structural
extracted from the 18e (Na2) sites during the charging process [50]. evolution of the cathode materials [62].
Solid-solution and two-phase evolution during electrochemical cycles
2.1.1.1. In situ/operando XRD and ex situ XRD. The study of phase evo- of Na3MnTi(PO4)3 were both observed by Mai's group using in situ XRD
lution in NASICON cathodes during cycling is vital for figuring out the [63]. Subsequently, Wang et al. synthesized a NASICON-type compound,
sodium insertion/extraction mechanism, which will help optimize Na4MnCr(PO4)3 (NMCP), with a high discharge capacity of 130 mAh g–1

12
Y. Liu et al. eScience 2 (2022) 10–31

Fig. 2. (a) In situ XRD patterns of NVP (◆ NVP, ♣ NaV2(PO4)3). Reproduced from Ref. [54] with permission, copyright 2012, Wiley. (b) In situ XRD patterns of NVP
from 2.3 to 4.2 V at 293 K under 0.5, 2.0, and 5.0 mV s–1, respectively. Reproduced from Ref. [55] with permission, copyright 2017, American Chemical Society. (c) Ex
situ XRD patterns of Na4MnV(PO4)3@C@GA. Reproduced from Ref. [59] with permission, copyright 2018, Wiley. (d) Ex situ XRD patterns of Na2.9V1.98Mo0.02(PO4)3.
Reproduced from Ref. [62] with permission, copyright 2018, Royal Society of Chemistry. (e) In situ XRD pattern of Na4MnCr(PO4)3 in the voltage range of 1.5–4.5 V.
Reproduced from Ref. [64] with permission, copyright 2020, Wiley.

at C/20 (1 C ¼ 110 mA g–1), corresponding to 2.35-Na insertion. The high-quality data obtained with SXRD and single crystal diffraction, the
structural evolution of NMCP cathodes during cycling was investigated structure of NASICON-type cathodes was effectively and accurately
by in situ XRD, and the results revealed a solid-solution reaction from the determined under various conditions [66].
Na4MnCr(PO4)3 phase to the Na3.5MnCr(PO4)3 in the first stage, followed Recently, considerable attention has been paid to the multielectron
by a typical two-phase transfer between x  3.5 and x  3.2, finally reaction in NASICON-type cathodes, in the pursuit of higher energy
achieving a single-phase process until the formation of density. The structural transitions during the first cycle of 1.5-electron
Na1.27MnCr(PO4)3 at 4.5 V (Fig. 2e) [64]. V3þ/V4þ/V5þ reactions in Na3VCr(PO4)3 (NVCP) were revealed by Liu
et al. through in situ SXRD. The results showed that the phase evolution
2.1.1.2. Synchrotron X-ray diffraction (SXRD). The structural complexity during cycling was irreversible (Figs. 3a and b), as the SXRD patterns of
of electrode materials is making increasing demands on high-resolution the cathode in the second charge state were quite different from those in
and high-flux techniques. Synchrotron X-ray diffraction (SXRD) has the first charge state [67].
been developed to study the fine structure of battery materials, owing to As a cathode for SIBs, Na3V2(PO4)2F3 has attracted significant
its high spatial and temporal resolution [65]. Chotard et al. used SXRD to attention, as it delivers higher voltage and theoretical capacity than NVP
reveal the reaction of β → α-Na3V2(PO4)3 from 110 to 10  C and iden- but undergoes a more complicated electrochemical mechanism. Through
tified the different crystal structure of α-Na3V2(PO4)3 below 12.2  C. The in situ SXRD experimentation, Talaie et al. found that unlike the one-step
SXRD pattern of γ-Na3V2(PO4)3 at 200  C was indexed to the rhombo- solid-solution reaction, four intermediate phases were formed during the
hedral R 3 c space group, while the single crystal diffraction of transformation of Na3V2(PO4)2F3 to NaV2(PO4)2F3. Two intermediate
α-Na3V2(PO4)3 at –10  C corresponded to the monoclinic C2/c space phases (Na2.4V2(PO4)2F3 and Na2.2V2(PO4)2F3) were detected during the
group (considered to be an ordered form of γ-Na3V2(PO4)3). Due to the extraction of 1-Naþ from Na3V2(PO4)2F3 (Fig. 3c). Upon further

13
Y. Liu et al. eScience 2 (2022) 10–31

extraction of Naþ, Na2V2(PO4)2F3 disappeared and NaxV2(PO4)2F3 (1.8 investigated the substitution mechanism as well as structural parameters
 x  1.3) was formed (Fig. 3d). NaV2(PO4)2F3 in a Cmc21 space group via NPD. The results revealed that a small amount of F (15%) replaced the
was finally obtained through another biphasic reaction. The precision of partial oxygen at two 36f sites, leaving a 5% deficiency in Na ions. The
SXRD was key for deducing the (de)intercalation mechanisms of chemical composition obtained from the data fitting was consistent with
Na3V2(PO4)2F3 [68]. the stoichiometric composition, demonstrating the high accuracy of the
NPD technique [69].
2.1.2. Neutron diffraction (ND) Notably, ND needs a longer data collection time than XRD, and there
Neutron diffraction (ND) can provide complementary structural in- are only a few neutron research centers across the world, limiting its
formation about battery materials when (S)XRD is not well suited. It is wider application. Both (S)XRD and ND can only determine the long-
able to identify light atoms in the presence of heavier atoms and distin- range crystal structure by the Rietveld method, but many types of ma-
guish between elements with close atomic numbers, based on the in- terials are quite disordered, and their local atomic arrangements are
teractions between neutrons and atomic nuclei, rather than the important for an in-depth understanding of their structures. In compar-
interactions between X-rays and atoms in the electron cloud, as (S)XRD ison, pair-distribution function (PDF) analysis can provide insights into
does. In addition, ND can investigate the magnetic structures and thermal the structures of low-crystalline and disordered materials, but it has not
displacement parameters of materials. yet been applied to investigate NASICON-type cathodes.
Since Mn and Cr atoms possess close atomic numbers, our group
employed neutron powder diffraction (NPD) combined with SXRD to 2.2. Microstructure and morphology-related technologies
study the structure of NASICON-type Na4MnCr(PO4)3/C (NMCP/C) in
depth. The Rietveld refinement of the NPD pattern suggested that NMCP 2.2.1. Nuclear magnetic resonance (NMR)
with a rhombohedral structure (R-3c) exhibited larger lattice parameters Solid-state nuclear magnetic resonance (ss-NMR) aims to probe the
than those of NVP [43]. Elsewhere, Muruganantham and colleagues ionic dynamics and local/electronic structures of electrodes based on
designed Na2.85V2(PO3.95F0.05)3 cathodes with improved electrochemical nuclear spin under an external magnetic field. In one study of the Naþ
properties through substitution at the inactive anion sites of NVP and kinetics in NaxV2(PO4)3, NMR analysis revealed that Naþ at the M1 site in

Fig. 3. (a) In situ SXRD patterns and (b) corresponding unit cell parameters of Na3VCr(PO4)3 at room temperature (first charged to 4.3 V, then discharged to 2.5 V, and
further charged to 4.0 V). Reproduced from Ref. [67] with permission, copyright 2017, American Chemical Society. (c) SXRD patterns during the extraction of 1-Naþ
from Na3V2(PO4)2F3 and (d) from Na2V2(PO4)2F3. Reproduced from Ref. [68] with permission, copyright 2015, American Chemical Society.

14
Y. Liu et al. eScience 2 (2022) 10–31

NVP was immobile, whereas Naþ at the M2 sites could move, thus ss-NMR spectra of NVP, indexing to Naþ at the M1 and M2 sites,
indicating the conduction pathway was directly from M2 to M2 [50]. Liu respectively, and a stronger signal for Naþ at the M1 site was recorded.
et al.'s work on NVCP revealed the activation process of V4þ/V5þ using ex Both peaks shifted slightly after Fe doping, confirming Fe had been doped
situ 51V NMR. As shown in Fig. 4a, only the noise spectrum existed when into the bulk structure. In addition, major changes in the 13C spectrum
NVCP was charged to 3.8 V, suggesting that V did not transfer to V5þ at suggested that elemental Fe significantly affected the electron cloud
3.8 V. However, upon charging to 4.3 V, the resonance peak at –847 ppm structure of carbon [70].
was obvious, indicating the presence of V5þ in samples. When the elec- Hadouchi et al. achieved deep insight into the sodium extraction
trode was discharged to 3.8 V, the intensity of the resonance at –847 ppm mechanism behind the Na3.41£0.59FeV(PO4)3 (£ represents Naþ vacancy)
plummeted, demonstrating the reduction of V5þ to a lower valence state framework through 23Na NMR, demonstrating that Na2 was completely
[67]. extracted while Na1 partially remained in the structure [71]. Liu et al.
Zhong and colleagues conducted 23Na and 13C ss-NMR to reveal the studied the reaction mechanism of the Na3V2(PO4)2F3 cathode through
Fe-doping mechanisms in the bulk structure and carbon layer of ex situ 23Na NMR spectra, finding that the charge process was divided
Na3V2-xFex(PO4)3@C. There were two peaks, at –9 and 6 ppm, in the 23Na into three reaction stages, as shown in Fig. 4b. There were two major

Fig. 4. (a) Ex situ 51V ss-NMR of Na3VCr(PO4)3. Reproduced from Ref. [67] with permission, copyright 2017, American Chemical Society. (b) Ex situ 23Na NMR
spectra and (c) 23Na quantification analysis of the Na1, Na2, and total Na sites upon charging of Na3V2(PO4)2F3. Reproduced from Ref. [72] with permission, copyright
2014, American Chemical Society. (d) FTIR spectroscopy of Na3Fe2(PO4)3-C. Reproduced from Ref. [74] with permission, copyright 2017, Wiley. (e) In situ Raman
spectra of a Na3V2(PO4)3/CNFs cathode from 0.5 to 2.0 V. Reproduced from Ref. [44] with permission, copyright 2018, Royal Society of Chemistry. (f) STEM HAADF
images of (i) NVP and (ii) NaV2(PO4)3 (the green and red balls represent P and V atoms, respectively); STEM ABF images of (iii) NVP and (iv) NaV2(PO4)3 (the yellow
and blue circles are Na atoms at M1 and M2 sites, respectively; the blue arrow points to Na at the M2 site). Reproduced from Ref. [50] with permission, copyright
2014, Wiley. Aberration-corrected STEM images and crystal structures of (g) Na4MnV(PO4)3 and (h) Na3FeV(PO4)3 viewed from the [210] and [101] crystallographic
directions, respectively (Na: green-blue; P: red; Mn/V: green; Fe/V: yellow). Reproduced from Ref. [80] with permission, copyright 2016, American Chemistry Society.

15
Y. Liu et al. eScience 2 (2022) 10–31

resonance peaks with an intensity ratio of 1:2 at 92 and 146 ppm for the and Raman spectroscopy. They showed that Na2Mg2(SO4)3 could only be
pristine sample, which were assigned to Na2 and Na1, respectively. In obtained by quenching at high temperature, otherwise phase separation
stage I, the intensities of the Na1 and Na2 resonances decreased simul- into Na2Mg(SO4)2 and Na2Mg3(SO4)4 would result. The unusual lang-
taneously, and the peak positions shifted slightly. Additionally, quanti- beinite structure for Na2Mg2(SO4)3 at elevated temperature was only
tative analysis of the NMR results (Fig. 4c) suggested that the rate of Na1 formed when large alkali metal ions were present, and the local strain
extraction was double that of Na2. Thus, by combining the results on the caused by the smaller Naþ in the langbeinite framework resulted in
intensity ratio and extraction rate, it was determined that the Na ions complex changes in structure/composition during heating/cooling [78].
were removed randomly (nonselective extraction of Na ions) from two To date, extensive applications of in situ Raman spectroscopy can be
different Na sites. As more Naþ were removed, the faster motion of Naþ found in the electrochemical research, but the required sample prepa-
as more Na vacancies were formed caused the two 23Na resonance peaks ration process remains demanding.
to coalesce into a single, broad resonance peak when a total of 0.9-Na was
extracted. Subsequently, the single 23Na resonance gradually shifted to- 2.2.3. SEM/(HR)TEM
ward 0 ppm with decreased intensity, indicating a single-phase reaction Scanning electron microscopy (SEM) and transmission electron mi-
process. In stage III, a new resonance peak appeared (at 87 ppm) and croscopy (TEM) are widely performed to monitor the morphology, size,
quickly dominated, then it moved toward low frequency at a faster rate and phase composition of materials. Both are based on the interaction
than in stage II. The rapid Na transfer significantly contributed to the between electrons and materials, the main difference being that SEM
outstanding Na storage performance, and with the help of NMR, a clear images are constructed by either backscattered or secondary electrons,
picture of the Na transfer process was obtained [72]. while TEM relies on transmission electrons or elastic scattering electrons,
Although the NMR technique is growing in popularity, challenges still so TEM requires ultrathin samples (about 100 nm). High-resolution TEM
remain for in situ NMR. In situ experiments possess a much lower spectral (HRTEM) is used to obtain more information about the lattice spacing of
resolution than those done with ex situ NMR, because the rotation of materials.
samples under ex situ conditions contributes to high-resolution spectra,
whereas a static sample within in situ experiments tends to induce a 2.2.4. Scanning transmission electron microscopy (STEM)
broad resonance signal, leading to low resolution. Furthermore, specially Combining the principles of both SEM and TEM, scanning trans-
designed cells are needed for in situ NMR experiments. First, the mate- mission electron microscopy (STEM), particularly aberration-corrected
rials used in batteries must be nonmagnetic and chemically stable with STEM, is considered an effective technique for characterizing the mi-
the electrodes and electrolytes. Also, since the metal parts in a battery crostructures of battery materials at the atomic level. It includes two
usually shield the effects of the applied radio frequency pulses on elec- modes: high-angle annular-dark-field (HAADF) and annular-bright-field
trodes, it is better to use metal mesh as a current collector rather than a (ABF). The former is powerful for heavy (transition-metal) atoms, and
dense metal foil to reduce the shielding effect. Moreover, the battery the latter can make light and heavy elements visible simultaneously [79].
should be impermeable against ambient air, because even a small amount Work on NVP has revealed the atomic position of Na using aberration-
of moisture will dramatically degrade the electrochemical performance corrected STEM. The phosphorous and vanadium atoms of NVP and
of in situ cells [73]. NaV2(PO4)3 were clearly shown in HAADF-STEM images (Figs. 4f-i and
ii), indicating that the NASICON structure remained stable after Na
2.2.2. FTIR/Raman spectroscopy extraction. However, no light Na atom could be observed in the images.
Fourier transform infrared (FTIR) spectroscopy and Raman spec- The ABF-STEM images in Figs. 4f–iii and iv show that the M1 sites were
troscopy can be used to analyze the molecular structures of NASICON- occupied by Na atoms in both samples, but the M2 sites were only
type cathodes for SIBs. FTIR is based on the absorption of infrared light occupied in NVP, further proving that Na atoms were extracted from M2
that then undergoes a Fourier transform, yielding information about the sites when NVP transferred to NaV2(PO4)3 [50].
bonding strength in functional groups and the molecular structures of Goodenough and colleagues designed a series of NASICON-
materials. The operating principle of Raman spectroscopy is the inelastic NaxMV(PO4)3 (M ¼ Fe, Mn, Ni) structures and subsequently employed
scattering effect of molecules on photons, which can reveal the atomic aberration-corrected STEM to gain a direct view of the atomic arrange-
environments and local structures of molecules or crystals [74]. These ment in Na4MnV(PO4)3. The high intensities in the STEM images
two methods can be combined, since FTIR can detect symmetrical vi- revealed the presence of Mn/V and P atoms, while Na and O atoms
brations, while Raman can evaluate asymmetrical vibrations [75]. showed weaker intensities due to their lower atomic numbers. Along the
The electrochemical properties of SIBs, such as capacity, cycling, and [210] direction, the Na atom columns appeared near the P atom columns.
rate performance, are strongly dependent on the bonding properties of Furthermore, a framework structure with void space for Naþ migration
their electrode materials. Consequently, having a detailed understanding could be clearly seen (Fig. 4g). In the case of Na3FeV(PO4)3, aberration-
of the bonding environments in cathode materials is essential for opti- corrected STEM showed it had a monoclinic C12c1 structure (Fig. 4h)
mizing electrochemical performance. Rajagopalan et al. utilized FTIR to [80].
understand the functional mechanism of carbon on Na3Fe2(PO4)3. The STEM is an efficient approach for unraveling the atomic-scale struc-
presence of O-C-O and C¼O in the FTIR spectrum indicated that carbon tures of electrode materials and can help us gain an in-depth under-
was bonded to the Na3Fe2(PO4)3 (Fig. 4d) [76]. Currently, in situ FTIR standing of the electrochemical reaction mechanisms of rechargeable
technology for sodium-storage materials is not yet mature, but the batteries. However, it cannot directly reveal the inherent dynamics of a
application of in situ FTIR will attract increasing attention in the future. battery system, and the electron beam can damage samples.
Raman spectroscopy also plays a critical role in the study of NASICON
cathodes for SIBs. Yu et al. used in situ Raman to study the effect of 2.3. Composition and valence-related technologies
carbon nanotube fabrics (CNFs) on an NVP cathode during the electro-
chemical process. The negligible changes in the characteristic Raman 2.3.1. ICP/TGA and DSC/EDS
peaks of CNFs confirmed that CNFs in the self-supporting electrodes Inductively coupled plasma optical emission spectroscopy (ICP-OES)
provided fast channels for electron migration and worked as a flexible is used for measuring sample compositions and accurately determining
matrix for NVP, rather than as hosts for Naþ (Fig. 4e) [44]. Bih and the elemental contents of materials. A Na4MnV(PO4)3 electrode synthe-
co-workers revealed the phase evolution of A3Fe2(PO4)3 (A ¼ Li and Na) sized by Jiao and co-workers was identified using X-ray photoelectron
through the γ, β, and α phases in a temperature range of 25–400  C by spectroscopy (XPS) and ICP technologies. The wide-range spectrum of
demonstrating a continuous change in Raman bands [77]. Trussov et al. XPS indicated the presence of Na, Mn, V, P, O, and C in the material, and
highlighted the thermal evolution of a Na2Mg2(SO4)3 system using XRD the ICP verified the anticipated atomic ratio, demonstrating the high

16
Y. Liu et al. eScience 2 (2022) 10–31

purity of the cathode material (Fig. 5a) [59]. Similarly, Wu's team Energy dispersive spectroscopy (EDS) is another technique for iden-
measured the stoichiometric content of Na3.1V2(PO4)2.9(SiO4)0.1 using tifying the concentration and distribution of elements in a micro-region
ICP, achieving a Na:V ratio of 1.58:1 (molar ratio), which was close to the of a material. Jiao's team prepared Na3MnTi(PO4)3 (NMTP) particles
theoretical value. Because SiO2– 4 has a larger ionic radius than PO4 ,
3–
encapsulated in interpenetrating graphene with a carbon-shell covering
Na3.1V2(PO4)2.9(SiO4)0.1 cathodes widened the diffusion channels for material (rGO@NMTP-C) by a sol-gel method to improve the Coulombic
Naþ and thereby improved their Na storage [81]. efficiency (CE) of batteries and realized around 100% CE at 4.3–1.5 V
Thermogravimetric analysis (TGA) and differential scanning calo- cut-off voltages with three-sodium (de)intercalation. The EDS results
rimetry (DSC) are powerful tools for evaluating the thermal properties of showed the sample's high purity, indicating an atomic ratio of Na:Mn:Ti
electrodes. Using TGA technology, Lim et al. proved that NVP lost only ¼ 3.08:1.14:1 and a uniform distribution of all elements throughout the
1.6% of its original mass until the temperature reached 450  C, and selected region (Fig. 5c) [82].
showed that its decent thermal stability could be attributed to strong P-O
bonding [53]. Chotard et al. used DSC to demonstrate for the first time 2.3.2. Electron paramagnetic resonance (EPR)
that NVP underwent three distinct reversible phase transitions: α ↔ β, β Electron paramagnetic resonance (EPR) or electron spin resonance
↔ β0 , and finally β0 ↔ γ, in a temperature range of –30 to 225  C (Fig. 5b) (ESR) spectroscopy is a fresh tool to study the surface defects and elec-
[66]. Later, Mai's group conducted DSC experiments to study the phase tronic structures of materials. The basic concepts of EPR are analogous to
evolution mechanism of NVP and NaV2(PO4)3 between 213 and 373 K. those of NMR, except that excited electron spins are used in EPR instead
Combining DSC data with the in situ XRD patterns of NVP and of the atomic nuclei spins in NMR. Similar to the previously described
NaV2(PO4)3, it was revealed that in both materials, the phase trans- Na4V2(PO4)3 [83], Na4Ti2(PO4)3 with additional Naþ inserted into the
formation from α to β occurred at ~283 K. The XRD refinements of the M(2) vacancies of Na3Ti2(PO4)3 was first studied by Senguttuvan and
two samples at 273, 283, and 293 K further implied that the high-site colleagues. In that work, EPR was used to prove the existence of Ti(II) in a
symmetry phase with the R/3c space group turned to a low-site sym- reduced Na4Ti2(PO4)3 electrode. Na3Ti2(PO4)3 was chosen as a reference,
metry phase of C2/c at approximately 283 K [55]. and its EPR result displayed a line width of 500 G centered at g ¼ 1.84,

Fig. 5. (a) XPS spectra (inset: ICP) of Na4MnV(PO4)3@C@GA. Reproduced from Ref. [59] with permission, copyright 2018, Wiley. (b) DSC measurement of NVP from
–30 to 225  C. Reproduced from Ref. [66] with permission, copyright 2015, American Chemical Society. (c) EDS mapping images of rGO@NMTP-C. Reproduced from
Ref. [82] with permission, copyright 2020, Elsevier. (d) EPR spectra of pristine Na3Ti2(PO4)3 and electrochemically reduced Na4Ti2(PO4)3. Reproduced from Ref. [84]
with permission, copyright 2013, American Chemical Society. (e) TOF-SIMS profiles and 3D view of the featured species within a Na4MnCr(PO4)3 electrode after 50
cycles from 2.5 to 4.6 V. Reproduced from Ref. [86] with permission, copyright 2021, American Chemical Society.

17
Y. Liu et al. eScience 2 (2022) 10–31

which was typical of the Ti(III) species. By comparison, Na4Ti2(PO4)3 charge/discharge process due to electrolyte decomposition. Generally,
exhibited a narrower signal of 200 G centered at g ¼ 1.94, and quanti- an SEI features a heterogeneous bilayer structure, where the inner
tative analysis revealed that nearly 50% of the Ti(III) still remained in the inorganic-rich layer (Na2O, NaCl, NaF, Na2CO3, etc.) strengthens the
reduced Na4Ti2(PO4)3, demonstrating that half of the Ti(III) converted to mechanical property of the SEI layer and provides fast Naþ ion diffusion
Ti(II) during the sodiation process (Fig. 5d) [84]. However, it is still channels; the organic outer layer (sodium alkyl carbonates (ROCO2Na),
difficult to apply this advanced technology under in situ conditions. sodium alkoxides (RONa), etc.) ensures the structural flexibility and high
chemical stability of the SEI film [92,93].
2.3.3. Time-of-flight secondary-ion mass spectrometry (TOF-SIMS) Since SEI composition is closely related to the electrolyte compo-
Since the surface chemistry of materials and the electrode interfaces nents, attempts have been made to optimize electrolyte formulations and
significantly influence the overall performance of batteries, under- thereby achieve stable SEI films with high ionic conductivity. Ponrouch
standing interface phenomena is highly necessary. Time-of-flight sec- et al. reported an optimum electrolyte composition by combining
ondary-ion mass spectrometry (TOF-SIMS) has proven to be a suitable EC0.45:PC0.45:DMC0.1 (EC: ethylene carbonate; PC: propylene carbonate;
tool to investigate the surface chemistry of materials, with a concentra- DMC: dimethyl carbonate) with 1 M NaClO4. This electrolyte was
tion sensitivity of ppm or even lower [85]. It utilizes sputtering ions to matched with a Na3V2(PO4)2F3 cathode and a hard carbon anode to
excite a small number of secondary ions on the surface of samples and achieve superior rate capability and outstanding cycling performance in
evaluates the ion mass according to the ion flight time. Goodenough's batteries. XPS investigations suggested that stable EC decomposition
team conducted TOF-SIMS analysis to study the cathode electrolyte products were the main components of the SEI, while the addition of
interphase (CEI) composition in a Na4MnCr(PO4)3 cathode system. The DMC did not change the SEI composition but improved the ionic con-
species in the CEI presented a triple-layered architecture with a spatial ductivity due to a decrease in viscosity [94].
distribution (Fig. 5e). C2HO– (related to organic carbonates) dominated Although XPS has been widely applied in SIB systems, there is the risk
the outer layer; NaF2– and NaC2–, corresponding to NaF and Na2CO3, that radiation will damage samples or cause changes in the chemical state
respectively, contributed to the intermediate layer, while the interior of materials due to low-energy electrons.

layer consisted of C10 enriched species (associated with conductive car-
bon). In sharp contrast, the interior layer of the bulk electrodes was 2.3.5. Electron energy-loss spectroscopy (EELS)
composed of CrO– and MnO– species. These findings confirmed the me- Compared with EDS, electron energy-loss spectroscopy (EELS), which
chanical robustness of the NASICON structure and revealed that the detects the compositions of materials by analyzing received energy-loss
decomposition of liquid electrolyte at high potentials leads to poor electrons, shows superior performance in identifying light elements.
cycling stability for the electrode [86]. For example, strong signals detected with EELS confirmed the existence
of Mn, V, and O in Na4MnV(PO4)3 and Fe, V, and O in Na3FeV(PO4)3
2.3.4. X-ray photoelectron spectroscopy (XPS) [80]. Information about valence- and conduction-band electronic prop-
X-ray photoelectron spectroscopy (XPS) is another efficient technique erties, surface natures, and chemical bonding can also be probed via EELS
for surface analysis, with a probing depth of approximately 2–5 nm, technology. Burova et al. prepared Na3V2O2x(PO4)2F3-2x (0 < x  1)
yielding information about the composition (qualitative), elemental cathodes with controllable vanadium oxidation states, O/F ratios, as well
content (half-quantitative), and chemical state of materials under ultra- as final product morphologies by adopting various reducing agents. EELS
high vacuum [87]. For example, Bi et al. utilized an XPS wide-scanning was applied to assess the valence state of vanadium in this work. The
spectrum to investigate the surface chemical composition of a V-L2,3 edge in Na3V2O2(PO4)2F resembled the V-L2,3 edge in KVOPO4
Na3V1.98La0.02(PO4)3/C cathode. The signals originating from Na, V, La, (V4þ), indicating that the vast majority of V was in a tetravalent state,
P, O, and C coincided well with the elemental mapping results, which although the L3 edge peak at 517.8 eV showed that a small amount of V3þ
showed that Na, V, La, P, O, and C were homogeneously distributed in the also existed. Spatial STEM-EELS mapping of V4þ and V3þ further sug-
material [88]. gested that the inner part of the material mainly contained V4þ, while
XPS can also be used to study the valence changes of redox centers V3þ was present on the surface. The V-L2,3 edge in Na3V2O0.8(PO4)2F2.2
during the sodium (de)intercalation process. Balaya and colleagues used ex was similar to the V-L2,3 edge in KVPO4F (V3þ) but with a higher energy
situ XPS to explore the Na insertion/extraction mechanism in the transition loss of 0.4 eV, confirming a mixed V3þ/V4þ state (Fig. 6b) [95].
between NVP and NaV2(PO4)3, revealing that V3þ in the as-prepared ma- Chou and colleagues synthesized N-doped graphene oxide-wrapped
terial was oxidized to V4þ during the charge process, then eventually fully Na3V(PO3)3N with a carbon coating layer and studied the valence vari-
reduced to V3þ after discharge [89]. Tirado's team first studied the effects of ations of V using EELS. The V-L3 edge shifted from 520.57 to 519.77 eV
chromium substitution in NVP through XPS. The XPS spectra of when discharged from 4.25 V to 3 V, indicating charge transfer in the 3d
Na3V1.9Cr0.1(PO4)3 demonstrated that an obvious plateau at ~4 V could be orbital of V. The valence of V changed by about V4.2þ/V3.2þ, according to
ascribed to the V4þ/V5þ redox pair, whereas Cr3þ still remained when the the in situ X-ray absorption near-edge structure (XANES) spectra, which
electrode was charged to 4.1 V [90]. Likewise, Chou and colleagues differed from the supposed V4þ/V3þ redox. The N K-edge without vari-
employed ex situ XPS to explore the sodium-storage mechanism of ations in the EELS spectra suggested that N atoms did not participate in
Na4MnCr(PO4)3. Initially, pristine Na4MnCr(PO4)3 showed a binding en- the electrochemical reactions [96].
ergy at 641.19 eV for Mn 2p3/2, corresponding to Mn2þ. When charged to
3.9 V, the peak shifted to 641.69 eV, suggesting the formation of Mn3þ. 2.3.6. X-ray absorption spectroscopy (XAS)
Upon further charging to 4.3 V, the binding energy at 642.10 eV verified the X-ray absorption spectroscopy (XAS) is a bulk-sensitive technique
appearance of Mn4þ. However, the XPS spectra of Cr 2p remained un- (micrometer size or larger) and provides information about the phase
changed throughout the entire cycling process. These results demonstrated evolution and valence variation of an electrode material during the
that the two charging plateaus at 3.66 and 4.20 V were caused by the electrochemical reaction. As a synchrotron radiation source character-
Mn2þ/Mn3þ and Mn3þ/Mn4þ redox pairs, respectively [91]. ization technology, it includes pre-edge, X-ray absorption near-edge
XPS has allowed the chemical states of Mn and Ti in the compound structure (XANES) in a low-energy range of 5–150 eV, and extended X-
rGO@Na3MnTi(PO4)3-C to be determined, and has been used to verify ray absorption fine structure (EXAFS) in a range of 150–2000 eV [97]. In
the three-step sodium extraction process from fully sodiated Na4Mn- the spectrum, the XANES region reflects the electronic states, while
Ti(PO4)3, corresponding to the transformation of Ti3þ/Ti4þ, Mn2þ/Mn3þ, EXAFS provides information on the atomic bonding. Like most of the
and Mn3þ/Mn4þ, respectively, within a voltage range of 1.5–4.3 V above-mentioned characterization techniques, in situ XAS also needed to
(Fig. 6a) [82]. In addition, XPS is used to investigate the composition of be further developed. Notably, the higher energy of the edge indicates
solid electrolyte interphase (SEI) film, which forms during the first the higher valence of the absorption element.

18
Y. Liu et al. eScience 2 (2022) 10–31

Fig. 6. (a) Ex situ XPS spectra of Na3MnTi(PO4)3. Reproduced from Ref. [82] with permission, copyright 2020, Elsevier. (b) EELS spectra of Na3V2O2(PO4)2F and
Na3V2O0.8(PO4)2F2.2, along with the reference spectra of KVOPO4 and KVPO4F. Reproduced from Ref. [95] with permission, copyright 2019, Royal Society of
Chemistry. (c) XAS spectra of Na4MnCr(PO4)3. Reproduced from Ref. [64] with permission, copyright 2020, Wiley. (d) 57Fe M€ ossbauer spectra and ex situ XRD
patterns of NaxFe2(SO4)3. Reproduced from Ref. [104] with permission, copyright 2018, Royal Society of Chemistry.

Pivko and colleagues investigated changes in the vanadium oxidation V3þ/V4þ and Mn2þ/Mn3þ were activated in Na3.75V1.25Mn0.75(PO4)3
state and the local environment around the redox center in NVP using in under 2.75–3.8 V, further transferring to the V4þ/V5þ and Mn3þ/Mn4þ
situ XAS from 3.0 to 4.0 V. As they elucidated, the oxidation state of redox states upon charging to 4.2 V [100].
vanadium changed from þ3 to þ4 upon charging, and in the process of To enhance the cycle life of Na3Cr2(PO4)3 cathodes [101], Yamada's
Naþ extraction/insertion, the average bond length of V-O remained group partially substituted Cr with Ti, expecting the d0 electronic
almost unchanged, contributing to the outstanding stability of NVP [98]. configuration of Ti4þ to tolerate local distortions during Naþ uptake/r-
Kang's group realized multielectron V2þ/V3þ, V3þ/V4þ, and V4þ/V5þ emoval owing to the second-order Jahn–Teller effect. However, the Ti L2,
redox reactions in a Na3V1.5Cr0.5(PO4)3 cathode over a wide voltage 3-edge XAS of Na2CrTi(PO4)3 did not change significantly in terms of
range of 1.0–4.2 V. No significant shift in the operando Cr K-edge XANES spectra shape upon charging, and the ex situ XRD result demonstrated
spectra was found during the entire charge/discharge process, indicating that the Cr/TiO6 octahedra scarcely varied, even after Naþ extraction.
Cr was not involved in the electrochemical reaction. In contrast, the shift These findings indicated that the distortions of the transition-metal sites
of the V K-edge spectra toward the high-energy side upon sodium were negligible, so the improved stability might have been associated
extraction suggested the increased valence state of V, while electron loss with decreases in the amount of Naþ and Cr4þ/Cr3þ redox within re-
in the VO6 octahedra as well as the shortening of V-O bonds led to a actions, rather than the effect of Ti4þ [102].
decrease in peak intensities in the V K-edge region at the end of the The redox mechanism of three distinct plateaus in a Na4MnCr(PO4)3
charge [99]. (NMCP) cathode during charge/discharge was also studied via ex situ
To gain more insights into the redox processes of Na3þγV2-γMnγ(PO4)3 XANES. As shown in Fig. 6c, both the K-edges (i and ii) of Mn and Cr
cathodes, two samples — γ ¼ 0.25 and 0.75, were subjected to XAS shifted during charging and were nearly restored to their initial positions
studies by Senguttuvan et al.. For the Na3.25V1.75Mn0.25(PO4)3 cathode, after full discharge, manifesting the high electrochemical activity and
progressive oxidation of V3þ to V4þ and V5þ occurred during charging to reversibility of Mn and Cr in cathodes. Because Mn was activated under
3.8/4.2 V, while Mn2þ cations remained unchanged. Nonetheless, both 4.3 V and Cr showed a higher activation potential than Mn, they assumed

19
Y. Liu et al. eScience 2 (2022) 10–31

that the first two plateaus were attributable to Mn2þ/Mn3þ and Mn3þ/ curves of a Na4VMn(PO4)3/C-rGO cathode in the 1st and 500th cycles,
Mn4þ, respectively, and the third was due to Cr3þ/Cr4þ. EXAFS was confirming negligible voltage degradation after 500 cycles. In addition,
conducted to gather details of the Mn valence status at various charge by comparing the proportions of the capacity provided by the high-
states, with the first peak (Figs. 6c–iii) reflecting the Mn-O bond length. potential discharge plateau to the total discharge capacity before and
The results unveiled the oxidation of Mn2þ to Mn3þ when charged to 3.9 after 500 cycles, they found that the main capacity loss upon cycling was
V, and the reduced intensity reflected the Jahn–Teller distortion of Mn3þ; caused by degradation of the high-voltage Mn2þ/Mn3þ redox [106]. The
upon further charging to 4.3 V, Mn3þ was fully oxidized to Mn4þ. The Cr cycling life and rate performance of NASICON-type cathodes are usually
oxidation state was also identified from the Cr pre-edge spectra evaluated through long-term cycling and adjusting the current density
(Figs. 6c–iv). The emerging new peak when charged to 4.3 V demon- during the cycling process, respectively.
strated the existence of Cr4þ [64].
Currently, the design of cells for in situ XAS experiments is relatively 3.2. Electrode process kinetics-related measurements
complex, because most components in cells can absorb incident X-rays,
leading to weakened and distorted signals. Cell design for in situ XAS NASICON-type materials have attracted a great deal of attention due
tests merits further consideration. to not only their structural stability but also their fast ion-diffusion ki-
netics upon sodiation/desodiation. The Naþ transport process can be
2.3.7. M€ ossbauer spectroscopy evaluated using the chemical diffusion coefficient (DNa ). A higher diffu-
Unlike electron-based characterization techniques such as XPS and sion coefficient indicates faster charging and discharging, suggesting
XAS, M€ ossbauer spectroscopy involves the resonant absorption of gamma better rate performance of the electrode materials. Four methods are
rays from the excited states of a nucleus. It is a highly sensitive nuclear commonly used to measure the DNa : GITT, the potentiostatic intermittent
technology for investigating the local atomic environments of materials, titration technique (PITT), CV, and electrochemical impedance spec-
including redox ratios, spin and valence states, and the occupation sites troscopy (EIS). The theories behind these methods have been systemat-
of atoms [103]. ically reported elsewhere, so here we mainly discuss experimental
57
Fe M€ossbauer spectroscopy has been widely applied to investigate operations and summarize the calculations for determining the chemical
the valence state changes of iron during charge/discharge processes. diffusion coefficient [107–109]. Notably, it can be divided into two
Rhombohedral NASICON-structured Fe2(SO4)3 was recently evaluated types, based on physical definition and phase transformation: one is
by Yamada's group as a cathode for SIBs due to the strong inductive effect called the true diffusion coefficient, corresponding to the solid-solution
of SO2– 2þ
4 , which was expected to improve the redox potential of Fe /Fe .
3þ
reaction, while the other is the apparent diffusion coefficient, origi-
Under galvanostatic intermittent titration technique (GITT) test condi- nating from a two-phase reaction [110].
tions, they found that about 93.2% of the Fe3þ was reduced to Fe2þ
during discharge, but 22.8% of the Fe2þ remained unreacted when 3.2.1. GITT/PITT
charged to 4.0 V, indicating irreversible transformations even at a slow In a GITT experiment, a three-electrode system composed of metallic
rate during GITT. They then used ex situ XRD and M€ ossbauer spectros- sodium (counter and reference electrode), electrolyte, and positive
copy to gain more information about the reaction process. As shown in (working) electrode is constructed. During the GITT process, the cell is
Fig. 6d, during discharge, a maximum of 1.84-Na could insert, and the charged or discharged at a constant current pulse for a period of time; this
M€ ossbauer spectrum showed that 92.2% of Fe was in an oxidation state of is followed by a relaxation time for the voltage to reach equilibrium. This
þ2, agreeing well with the value determined from electrochemistry. At procedure is repeated within the whole working-voltage window until
the end of charging, 0.46-Na did not extract. Moreover, the sulfate was the potential limit is reached. It is necessary to choose a suitable interval
smaller than a phosphate, leading to a slow migration rate and limited and relaxation time for the test. Otherwise, either the experimental time
sites for Naþ insertion [104]. will be extremely long or the electrode will not reach equilibrium. The
Very recently, Masquelier and co-workers prepared pure-phase GITT method is a combination of steady-state technology and transient
Na4FeV(PO4)3 through the electrochemical sodiation of Na3FeV(PO4)3, technology, which eliminates the ohmic drop problem commonly found
and the reduction of FeIII to FeII was demonstrated using M€ ossbauer in constant-potential technology. It is often used to estimate the true
spectra. A trace of FeIII signal was also observed at the fully sodiated state, diffusion coefficient, using Equation (1) [109–111]:
which was probably due to the slight oxidation of FeII during the long
 2  2  
acquisition time required for M€ ossbauer spectroscopy (over 72 h) [105]. 4 mB V M ΔEs L2
DNa ¼ τ≪ (1)
Notably, the specific need for a gamma-ray source in M€ ossbauer spec- πτ MB A ΔEτ DNa
troscopy limits its ability to investigate different isotopes. Although
several isotopes can theoretically be investigated using this technique, where τ is the duration time of the current pulse, mB is the mass of the
most research has been restricted to 57Fe M€ ossbauer experiments. active material, MB is the molecular weight, VM is the molar volume, A is
In summary, Table 1 lists the functions and typical examples of the the total contact area between electrode and electrolyte, ΔEs is the dif-
abovementioned characterization techniques. ference between two consecutive stable voltages at the end of the
relaxation period, ΔEτ is the potential difference between the equilibrium
3. Electrochemical measurements potential and the potential maximum at the end of the current pulse, and
L is the thickness of the electrode. The diffusion coefficient can be
3.1. Electrochemical performance-related measurements calculated when the change in the transient voltage ΔEτ during a single
titration shows a linear relationship against τ1=2 , as displayed in Fig. 7a.
Electrochemical charge/discharge is the essential test to evaluate the B€ockenfeld et al. evaluated the sodium-ion diffusivity of carbon-
electrochemical performance of NASICON-type cathode materials. In the coated Na3V2(PO4)3 for the first time using GITT and obtained values
half-cell test, galvanostatic cycling within a certain voltage range is from 6  10–13 cm2 s–1 to 2  10–15 cm2 s–1. The variation in DNa showed
usually carried out, although other testing conditions may be more a V-type trend (Figs. 7b and c), and the minimum value at the voltage
appropriate in specific cases. The information presented in the charge/ plateaus was an indication of the two-phase transition [112]. Jiao's group
discharge curve includes specific capacity, plateaus of redox reactions, synthesized Na4MnV(PO4)3@C@GA with a high energy density of 380
and polarization. The redox reaction can also be distinguished using a Wh kg–1 at 0.5 C (1 C ¼ 110 mA g–1) and a high capacity retention of
cyclic voltammogram (CV) and a differential capacity (dQ/dV) plot (a 68.8% after 4000 cycles at 20 C. The outstanding performance was
derivate of the charge/discharge curve). For instance, Ma's group ascribed to fast sodium migration, with large DNa values ranging from
compared the charge/discharge profiles and corresponding dQ/dV 10–10 to 10–12 cm2 s–1 [59].

20
Y. Liu et al. eScience 2 (2022) 10–31

Table 1 Table 1 (continued )

Advanced characterization techniques in the study of NASICON-type cathodes Items Techniques Capability Examples
for SIBs.
spectrometry
Items Techniques Capability Examples (TOF-SIMS)
Crystal In situ/ Structural Na3V2(PO4)3 [53–55] X-ray I. Monitor Na3V1.98La0.02(PO4)3/C
structure- operando XRD changes and Na4MnV(PO4)3 [58] photoelectron composition [88]
related and ex situ phase evolution Na4MnV(PO4)3@C@GA spectroscopy and electronic Na3V2(PO4)3 [89]
technologies XRD [59] (XPS) structure Na3V1.9Cr0.1(PO4)3 [90]
Na2TiV(PO4)3 [61] II. Surface Na4MnCr(PO4)3 [91]
Na2.9V1.98Mo0.02(PO4)3 analysis rGO@Na3MnTi(PO4)3-C
[62] [82]
Na3MnTi(PO4)3 [63] Na3V2(PO4)2F3 [94]
Na4MnCr(PO4)3 [64] Electron I. Monitor NaxMV(PO4)3 (M ¼ Fe,
Synchrotron X- Fine crystal- Na3V2(PO4)3 [66] energy-loss composition Mn, Ni) [80]
ray diffraction structural Na3VCr(PO4)3 [67] spectroscopy II. Valence Na3V2O2x(PO4)2F3-2x (0
(SXRD) information Na3V2(PO4)2F3 [68] (EELS) electronic < x  1) [95]
Neutron I. Crystal- Na4MnCr(PO4)3 [43] properties Na3V(PO3)3N [96]
diffraction structural Na2.85V2(PO3.95F0.05)3 X-ray I. X-ray near- Na3V2(PO4)3 [98]
(ND) information [69] absorption edge structure Na3V1.5Cr0.5(PO4)3 [99]
II. Sensitive to spectroscopy is sensitive to Na3þγV2-γ Mnγ (PO4)3
light elements (XAS) oxidation state [100]
and elements II. Extended X- Na2CrTi(PO4)3 [102]
with close ray fine Na4MnCr(PO4)3 [64]
atomic structure can
numbers provide
Microstructure Nuclear I. Local/ NaxV2(PO4)3 [50] information
and magnetic Electronic Na3VCr(PO4)3 [67] about bond
morphology- resonance structure Na3V2-xFex(PO4)3@C length and
related (NMR) II. Ionic [70] coordination
technologies dynamics Na3.41£0.59FeV(PO4)3 number
[71] M€ossbauer Local atomic NaxFe2(SO4)3 [104]
Na3V2(PO4)2F3 [72] spectroscopy environment Na4FeV(PO4)3 [105]
FTIR I. Local Na3Fe2(PO4)3 [76]
spectroscopy environments
(characteristic To overcome the inferior electron/ion transport dynamics of
functional Na3MnTi(PO4)3, interpenetrating graphene-encapsulated Na3Mn-
groups) Ti(PO4)3 particles with a carbon-shell covering material (rGO@NMTP-C)
II. Surface-
sensitive
were synthesized by Li et al. and the DNa values calculated by GITT
Raman I. Local Na3V2(PO4)3/CNFs [44] ranged from 10–9 to 10–12 cm2 s–1, which were larger than those of most
spectroscopy environments A3Fe2(PO4)3 (A ¼ Li and Na-storage cathode materials [82]. Chen et al. reported a NASICON-type
(bonding Na) [77] Na4Fe3(PO4)2(P2O7)/C nanocomposite with tunable structures. The
information) Na2Mg2(SO4)3 [78]
corresponding diffusion coefficients were calculated to be 10–10 to 10–13
II. Both surface
and bulk cm2 s–1, highly comparable to those of reported NASICON-type cathode
SEM/(HR)TEM Morphology, materials [113]. Ben Yahia et al. successfully synthesized the mixed
size, phase polyanionic cathode NaFe2(PO4)(SO4)2, and their GITT experiments
composition indicated that the diffusion coefficients were in the range of 10–11 to
Scanning I. High-angle Na3V2(PO4)3 [50]
transmission annular dark NaxMV(PO4)3 (M ¼ Fe,
10–12 cm2 s–1, suggesting that ion diffusion could be a rate-limiting factor
electron field mode is Mn, Ni) [80] at high rates and that the electrochemical performance could be
microscopy sensitive to improved with further work [114].
(STEM) heavy atoms PITT is similarly used to examine the true diffusion coefficient by
II. Annular-
applying an impulse potential to the battery to make the electrode reach
bright-field
mode is able to equilibrium and measuring the change in current. Specifically, a constant
detect both potential pulse is applied to the electrode and maintained for a period of
light and heavy time, then removed for relaxation. This procedure is repeated until the
elements potential limit is reached. At the same time, the corresponding transient
Composition ICP Compositions Na4MnV(PO4)3 [59]
and valence- and elemental Na3.1V2(PO4)2.9(SiO4)0.1
current It plotted against time (t) is recorded, and the value of the
related contents [81] diffusion coefficient is calculated based on the slope of this curve, ac-
technologies TGA & DSC Thermal Na3V2(PO4)3 [53,55,66] cording to Equation (2) [109]. Notably, side reactions such as the
properties nucleation of new phases can be avoided in PITT if the voltage is
EDS Concentration rGO@Na3MnTi(PO4)3-C
controlled within the stability range of the single phase. However, the
and [82]
distribution of ohmic voltage drop varies with time and cannot be easily eliminated
elements from the imposed voltage difference. Compared with GITT, PITT only
Electron I. Sensitive to Na3Ti2(PO4)3 [84] needs to measure the thickness of the electrode rather than the true
paramagnetic surface defects electrode/electrolyte contact area, which is hard to determine. Never-
resonance II. Monitor
(EPR) electronic
theless, the test's complexity limits its wider application for investigating
structure the reaction dynamics of NASICON cathodes.
Time-of-flight Surface Na4MnCr(PO4)3 [86]  
secondary-ion analysis dlnðIt Þ 4L2 L2
DNa ¼ t≫ (2)
mass dt π2 DNa

21
Y. Liu et al. eScience 2 (2022) 10–31

Fig. 7. GITT tests of Na3V2(PO4)3: (a) linear relationship between potential response and τ1=2 , as well as voltage profiles during titration and the change in diffusion
coefficients with voltage during (b) charge and (c) discharge. Reproduced from Ref. [112] with permission, copyright 2014, Springer. CV measurements of
Na4MnCr(PO4)3/C: (d) CV curves at scan rates from 0.2 to 1.0 mV s–1; (e) linear relationships between v1=2 and ip , and (f) log ip versus log v plots; (g) pseudocapacitive
contribution (red region) at 0.6 mV s–1. Reproduced from Ref. [43] with permission, copyright 2020, Wiley. EIS tests of Na3V2-xMgx(PO4)3/C (x ¼ 0–0.1): (h) Nyquist
plots, with corresponding equivalent circuit in the inset, and (i) relationship between Z 00 and ω1=2 at low frequencies. Reproduced from Ref. [120] with permission,
copyright 2015, Royal Society of Chemistry.

ip ¼ 2:69  105 n3=2 ADNa CNa
1=2 * 1=2
3.2.2. Cyclic voltammetry (CV) v (3)
CV is commonly used in electrochemical measurements to gather
information on electrode stability, reversibility, and phase trans- where ip is the peak current, n is the number of electrons transferred per
formation during the ion insertion/extraction process. The peaks in CV molecule, A is the active area between electrode and electrolyte, CNa *
is
curves indicate the occurrence of redox reactions, and the corresponding the concentration of sodium ions in the cathode, and v is the scan rate.
potentials are usually in good agreement with those of the char- Notably, this only works under the assumption that the charge transfer at
ge–discharge plateaus. More importantly, for a reversible system the interface is fast enough and the sodium ion concentration remains
controlled by the diffusion step, CV is a simple but powerful technique to constant [110].
quantitatively estimate both true and apparent diffusion coefficients. As Yu's group reported a Na3V2(PO4)3@C@rGO cathode with a diffusion
shown in Fig. 7d, with an increase in scan rate, the current intensity of the coefficient calculated to be ~10–11 cm2 s–1 for the anodic and cathodic
redox peaks of Na4MnCr(PO4)3/C increased and the potential gap be- processes. Such a fast DNa theoretically supported the possibility of
tween the anodic and cathodic peaks grew [43]. This was due to polar- achieving superior rate performance (up to 100 C). The material's 3D
ization, since sodium ions cannot completely extract/insert from/into an hierarchical architecture was responsible for the fast diffusion across the
electrode within a short time interval at a relatively large scan rate. As surface [36]. Li et al. introduced Mo6þ into Na3V2(PO4)3 to significantly
shown in Fig. 7e, when the peak current presented a linear relationship improve its rate and cycling performance. CV tests confirmed that the
with the square root of the scan rate, the diffusion coefficient could be diffusion coefficients were two orders of magnitude higher than those of
determined from the slope of this linear curve using the Randles–Sevcik the undoped samples, indicating that high-valence ion doping is an
equation (3) [110]: effective route to enhance rate performance [62]. Ji's group found that

22
Y. Liu et al. eScience 2 (2022) 10–31

the DNa values of Na3V2(PO4)2F3 at its high-voltage anodic and cathodic Hence, the pseudocapacitance behavior can be manipulated by
peaks were larger than those in the low-voltage region, implying that the morphological and microstructural design, which thus improves the re-
introduction of fluorine not only elevated the working voltage but also action kinetics and rate performance of batteries.
accelerated the reaction kinetics [115].
To overcome the challenge of the high cost of vanadium, Liang et al. 3.2.3. Electrochemical impedance spectroscopy (EIS)
introduced Mn into NVP and obtained a carbon-coated A low impedance is beneficial for achieving high rate performance,
Na3.5Mn0.5V1.5(PO4)3/C (NMVP/C) cathode with hydrangea-like struc- high Coulombic efficiency, and long cycle life in electrodes, and it is
tures, which exhibited rapid ionic diffusion and great rate performance. necessary to evaluate the impedance at each stage during cycling to
The diffusion coefficient of NMVP/C calculated from CV was 10–9 cm2 s–1, optimize the electrochemical performance of electrodes. EIS can measure
higher than that of NMVP [116]. Deng's group prepared crosslinked variations in impedance by recording the current response to an applied
Na2VTi(PO4)3/porous carbon nanofibers as a bifunctional electrode in potential at various frequencies. For example, the EIS results shown in
aqueous SIBs. The average diffusion coefficient calculated by CV was about Nyquist plots (Fig. 7h) usually consist of one or more semicircles at high
one order of magnitude higher than that of a randomly arranged sample frequencies, followed by an approximately 45 line at low frequencies,
and over two orders higher than that of a low-carbon aggregated powder, where the former semicircles represent the ohmic resistance between
indicating fast sodium transport in the crosslinked nanofiber structure electrodes and electrolyte ðRs Þ and the charge transfer resistance ðRct Þ,
[117]. while the latter slope refers to the “Warburg” impedance, which is related
The CV data at different scan rates can be used to calculate the to the Naþ diffusion behavior in electrodes [120]. Notably, the x and y axes
pseudocapacitive contribution in electrodes. Generally, the charge stor- in Nyquist plots must be symmetric, and all the parameters reflecting re-
age process comes from the Faradaic contribution of diffusion-controlled action dynamics are calculated based on an equivalent electrical circuit
ion intercalation, redox reactions of pseudocapacitance, and the non- model (as shown in the inset of Fig. 7h), which is comprised of several
Faradaic process due to the double-layer effect. Pseudocapacitance usu- basic electrical elements, such as capacitors, resistors, and inductors. The
ally occurs on or near the surface of a material, and it cannot be value of Rct is easily obtained by measuring the diameter of the semicircle
neglected, especially when carbon coating is used to enhance the elec- on the x-axis to evaluate the conductivity, which is one advantage of this
tronic conductivity. The contribution of pseudocapacitance can be method [120]. The sodium-ion diffusion coefficient of the electrode can be
determined using the following equations (4)–(7) [118]: calculated based on the information from the straight line measured in the
low-frequency region, according to Equations (8) and (9) [121]:
ip ¼ avb (4)
R2 T 2
DNa ¼ (8)
ip ¼ k1 v þ k2 v1=2
(5) 2A2 n4 F 4 C 2 σ 2

log ip ¼ b log v þ log a (6) Z ¼ Rs þ Rct þ σω1=2 (9)

where R represents the gas constant, T is the absolute temperature, A is

ip v1=2 ¼ k1 v1=2 þ k2 (7) the contact area of electrode with electrolyte, n is the number of trans-
ferred electrons per molecule during electrochemical reaction, F is the
where ip is the peak current, v is the scan rate, a and b are adjustable
Faraday constant, C is the concentration of sodium ions, and σ is the
parameters, and k1 v and k2 v1=2 represent the capacitive and diffusion- 0
Warburg factor, which is the slope of the line between Z and ω1=2
controlled currents, respectively. Typically, for a diffusion-dominated
(Fig. 7i) and can be determined using Equation (9).
reaction, the b value in Equation (4) is close to 0.5; when the b value is
NASICON-type materials usually suffer from inherently low elec-
greater than 0.5 or close to 1, the surface-controlled capacitance partic-
tronic conductivity, and the EIS test is widely used to investigate their
ipates in the charge storage process. Fig. 7f shows the log ip versus log v
internal and diffusion resistances, providing instructive information for
plots at the oxidation and reduction peaks of a Na4MnCr(PO4)3/C cath-
structural optimization. Wu and his co-workers developed a series of Mg-
ode. The values for each peak are larger than 0.5, indicating the pseu-
doped Na3V2-xMgx(PO4)3/C (x ¼ 0–0.1) materials. In EIS tests, all the
docapacitance effect, which leads to fast reaction dynamics [43]. By
doped samples exhibited a fast charge transfer process, with higher DNa
measuring the peak currents at the same potential at each given scan rate,
values than that of pristine Na3V2(PO4)3, indicating the enhanced elec-
the k1 and k2 values can be calculated via Equation (7). When obtaining a
tronic and ionic conductivities induced by Mg doping [120]. Moreover,
series of k1 values at different potentials, the capacitive contour can be
they reported a Na-rich Na3þxV2-xNix(PO4)3/C (x ¼ 0–0.07) cathode
depicted by plotting the capacitance-dominated current (k1 v) versus
material, in which the Ni2þ went to the V site, and extra Naþ were
potential (V). As shown in Fig. 7g, the capacitance contribution was
introduced to maintain the charge balance. The sodium diffusion coef-
determined by comparing the fitted area with the total stored charge
ficient of the Ni-doped NVP/C (9.39  10–13 cm2 s–1) was higher than
[43].
that of an undoped sample (1.21  10–14 cm2 s–1), demonstrating that Ni
Jiao's group calculated the pseudocapacitance contribution at
doping improved the reaction kinetics during the charge/discharge
different scan rates of rGO@Na3MnTi(PO4)3-C. The results showed it to
process [122].
be much larger than that of NMTP-C, especially at high rates, indicating
Recently, Fe-based NASICON-type cathode materials have gained
its high rate capability. This work confirmed that reduced particle size
wide attention, as Fe is nontoxic, cost-effective, and has abundant
and enlarged specific surface area can enhance the pseudocapacitance of
valence states. Rajagopalan et al. first reported a Na3Fe2(PO4)3 material
electrodes [82]. Our group reported a novel Na3.5Mn0.5V1.5(PO4)3 cath-
wrapped by carbon that accessed stable Fe3þ/Fe4þ redox reactions and
ode material for SIBs with an extraordinary rate capability of 92.7 mAh
showed a higher working potential than traditional NASICON sodium
g–1 at 60 C and an impressive capacity retention of 87.2% after 4000
iron phosphates based on Fe2þ/Fe3þ. The Rct measured by EIS was less
cycles at 20 C (1 C ¼ 110 mA g–1). The pseudocapacitive contribution
than half that of pure Na3Fe2(PO4)3 (193 Ω vs. 410 Ω), indicating the
was calculated to be 63% at 0.5 mV s–1, ascribable to the large active
significantly improved kinetics due to the conductive carbon network
surfaces/interfaces arising from the favorable heterostructure of
[76]. Nevertheless, rate and cycling performances still need to be
Na3.5Mn0.5V1.5(PO4)3/C, along with the nanostructure, carbon coating,
improved, and finding an appropriate electrolyte for use in high-voltage
and high porosity [119]. In the work of Na4MnCr(PO4)3 reported by
battery systems remains a challenge.
Chou's group, the pseudocapacitive contribution was 69.2% at 0.4 mV s–1
Cao et al. synthesized a K-doped K0.24Na2.76Fe2(PO4)3 cathode via a
and reached 81.2% when the scan rate was increased to 1 mV s–1 [91].
solid-state reaction method. According to the EIS results, the Rct was lower

23
Y. Liu et al. eScience 2 (2022) 10–31

and the DNa was an order of magnitude higher than those of pure closely related to the electronic conductivity of electrodes. The bond
Na3Fe2(PO4)3, a finding attributed to the expansion of diffusion channels valence sum (BVS) method is also a well-established tool for the pre-
after doping with large-radius Kþ [123]. Recently, Goodenough and his liminary examination of ionic states and diffusion pathways [113,
co-workers reported a cathode of Na3V1.5Cr0.5(PO4)3 with reversible 125–127].
three-electron redox reactions after one Naþ insertion, successfully Ion doping is a popular strategy to improve the electrochemical
achieving high energy and power densities in batteries. In situ EIS mea- performance of NASICON-type cathodes, and DFT can help to investigate
surements were carried out to reveal the sodium storage behavior during the underlying mechanism. As shown in Fig. 8a, to understand the in-
the first cycle. Rct is known to vary with changes in the occupation of Na fluences of Mg2þ doping on Na3V2(PO4)3/C, Wu's group used first-
sites, which decreases during charging and increases upon discharge, principles calculations to determine the most preferred doping sites.
especially for Na4V1.5Cr0.5(PO4)3, which exhibited the largest Rct owing to They found that Mg2þ tended to substitute for the V ion, and extra Naþ
the full occupation of Na sites [124]. Mai's group developed Na3MnTi(- was introduced to retain the charge balance. Moreover, according to the
PO4)3/C (NMTP/C) hollow microspheres that realized three-electron calculated chemical stability and the volume difference between the
redox reactions with a high specific capacity of 160 mAh g–1 and calculated and XRD Rietveld refinement results, Mg2þ was demonstrated
outstanding cycling stability. EIS tests were carried out on NMTP/C sam- to be unevenly distributed within the material [42]. Zhao et al. investi-
ples prepared with different annealing temperatures, and NMTP/C-650 gated the effects of Al doping on the electronic structure and ion dy-
(thermally treated at 650  C) showed the lowest charge transfer resis- namics of Na3V2-xAlx(PO4)3, finding that the substitution of V by Al in
tance and the highest sodium-ion diffusion coefficient [63]. Na3V2(PO4)3 decreased the band gap (Fig. 8b) and changed the elec-
In general, data processing in EIS tests is relatively simple. Since the tronic structure from an indirect band gap to a direct one (Fig. 8c),
result will be influenced by the material structure, electrode preparation leading to enhanced electronic conductivity [128].
process, experimental conditions, and accuracy of the equivalent elec- DFT is highly helpful for designing novel electrode materials with
trical circuit model, it is usually used in combination with other methods. outstanding electrochemical performance. As presented in Fig. 8d, Ceder
To further understand the kinetics of electrode processes, EIS tests are and co-workers used DFT calculations to compare the fabrication feasibility
performed at different temperatures to calculate the apparent activation of a series of NaxMnM(PO4)3 (M ¼ Cr, Ti, and Zr) cathodes and confirm the
energy (Ea ). Given the Rct value, the exchange current (i0 ) and the successful synthesis of Na4MnCr(PO4)3, which was predicted to provide an
apparent activation energy (Ea ) can be calculated based on the Butler- average voltage of about 4 V and a theoretical capacity of 165 mAh g–1 [64].
–Volmer equation (10) and Arrhenius equation (11), respectively [64]: Moreover, Zhou et al. used Na3MnTi(PO4)3 as both cathode and anode to
  construct a symmetric sodium-ion full cell, which exhibited high energy
RT 1 density and long cycling life. DFT investigation in this work was used not
i0 ¼ (10)
nF Rct only to predict the voltage plateaus but also to analyze the structural con-
  figurations with different Naþ occupancies, which provided deep insights
Ea into the charge/discharge mechanism and a possible way to modulate the
i0 ¼ Aexp  (11)
RT voltage plateaus by manipulating the occupation of carriers [129].
To thoroughly understand reaction kinetics, it is of great importance
Ea 1 to figure out Na-ion migration pathways and the corresponding energy
ln i0 ¼ ln A  (12)
R T barriers. DFT calculation was first employed to explore ion-migration
behavior in Na3V2(PO4)3 by Song et al. The results suggested a 3D
where R is the gas constant, T is the absolute temperature, n is charge transport pathway composed of two channels along the x and y di-
transfer number, F is the Faraday constant, and A is a temperature- rections, and one possible curved route along the z axis; the corre-
independent coefficient. The value of i0 can be used to describe an sponding migration energies were 0.0904, 0.11774, and 2.438 eV,
electrode's ability to gain or lose electrons and can reflect the difficulty of respectively (Fig. 8e) [130]. Traditionally, there are Na1 and Na2 sites in
the electrode reaction. Typically, a larger i0 value indicates higher elec- the structure of Na3V2(PO4)3, where Na1 is usually inactive and Na2 is
trode activity. After a series of equivalent transformations to Equation more active and responsible for Naþ intercalation/deintercalation.
(11), i0 can be expressed by Ea , according to Equation (12). The values of Recently, Wang et al. employed a series of DFT simulations to investigate
Ea can be determined by the slope of the fitting line between ln i0 and T1 . A the electrochemical behaviors of Na3V2(PO4)3. Based on the calculated
lower Ea means a shorter sodium-ion diffusion path. Note that the elec- energies for three possible diffusion pathways, a concerted ion-exchange
trode potential should be set at the voltage plateau before testing begins, mechanism for Naþ diffusion in Na3V2(PO4)3 was proposed, where both
to ensure that the electrode surface environment and charge transfer Na1 and Na2 were engaged in the migration process, suggesting that Na1
process are in stable states. also participated in ionic transportation [131].
Although research work using EIS to calculate the apparent activation In their work on a Na3MnZr(PO4)3 cathode, Gao et al. put forward a
energy is rarely reported due to its complexity, it still could be a useful detailed desodiation pathway wherein Mn and Zr were disordered within
technique to investigate the reaction kinetics in electrodes, especially by the structure, and they proved the high stability of the Mn3þ phase in
comparing samples prepared under different synthesis conditions and by Na2MnZr(PO4)3 by DFT computations, which explained why the coop-
various modification strategies. erative Jahn–Teller distortion of Mn3þ was suppressed [132]. As shown
in Figs. 8f–h, in the study of Na4MnCr(PO4)3, our group employed DFT to
4. Theoretical computation technologies calculate the energy barrier of Na1-Na2-Na1 channels, finding it to be
0.38 eV, which was comparable to that of Na3V2(PO4)3. The projected
In the development of high-performance NASICON cathode mate- density of states (PDOS) showed that some states cross the Fermi energy,
rials, theoretical calculation is a common method to gain insights. For especially for the spin-up electrons, indicating some metallic properties
example, density functional theory (DFT) calculation based on quantum could be responsible for the fast reaction kinetics [43]. As computational
mechanics has been widely used in the investigation of electrochemical abilities continue to improve and the relevant calculation models are
behavior and the optimization of electrode materials' electrochemical continuously optimized, DFT calculation is able to explain mechanisms in
performance. With calculation models continually being optimized, DFT greater depth, and research on high-specific-energy NASICON cathode
calculation has been adopted to analyze and predict the crystal structure, materials will become easier.
operating voltage, theoretical capacity, and structural stability of elec- In addition to DFT, the BVS method also plays an important role in
trode materials. In addition, electronic features such as charge density, theoretical calculations. For instance, Islam et al. revealed three-
energy bands, and density of states (DOS) can be calculated, which are dimensional diffusion pathways in NASICON materials and proposed a

24
Y. Liu et al. eScience 2 (2022) 10–31

Fig. 8. (a) Schematic illustration of the possible Mg-doping mechanisms in Na3V2(PO4)3, according to DFT calculation. Reproduced from Ref. [42] with permission,
copyright 2018, American Society Chemistry. (b) Total density of states of Na3V2(PO4)3 (top) and Na3V1.75Al0.25(PO4)3 (bottom), and (c) corresponding band
structures. Reproduced from Ref. [128] with permission, copyright 2018, Elsevier. (d) Voltage-composition curves for NaxMnM(PO4)3 (M ¼ Cr, Ti, Zr) cathodes,
predicted by first-principles calculations. Reproduced from Ref. [64] with permission, copyright 2020, Wiley. (e) Schematic diagram of possible Naþ migration
pathways along x, y, and curved z axes in Na3V2(PO4)3. Reproduced from Ref. [130] with permission, copyright 2014, Royal Society of Chemistry. (f) Na-ion migration
pathways in Na4MnCr(PO4)3 and (g) corresponding energy barriers. (h) Density of states of Na4MnCr(PO4)3. Reproduced from Ref. [43] with permission, copyright
2020, Wiley.

continuous exchange of Naþ between Na1 and Na2 sites during the re- show that sodium ions tend to move in a one-dimensional direction in a
action by combining molecular dynamics, bond valence, and maximum Na2Fe2(SO4)3@C@GO cathode, and they proposed that the Naþ ions in
entropy methods. Notably, the only input required for the bond valence Na1 sites were movable, which was slightly different from the result that
energy landscapes method are the crystal structure parameters, so the Yamada's group had reported [133].
computational cost is low [125]. Chen et al. [127] used the recently To sum up, Table 2 lists functions and typical examples for the above-
developed bond-valence map and bond-valence electron voltage map to mentioned electrochemical measurements and theoretical calculations.

25
Y. Liu et al. eScience 2 (2022) 10–31

Table 2
Advanced electrochemical measurements and theoretical computation technologies in the study of NASICON-type cathodes for SIBs.
Items Technologies Capability Examples Results

Electrode process kinetics- Galvanostatic intermittent titration Ionic diffusion coefficient DNa (cm2 s–1)
related measurements technique (GITT) Na3V2(PO4)3 [112] 6  10–13~2  10–15

Na4MnV(PO4)3@C@GA 10–10~10–12
[59]
rGO@Na3MnTi(PO4)3-C 10–9~10–12
[82]
Na4Fe3(PO4)2(P2O7)/C 10–10~10–13
[113]
NaFe2(PO4)(SO4)2 [114] 10–11~10–12
Na4MnCr(PO4)3/C [43] 10–11~10–12
Potentiostatic intermittent titration Ionic diffusion coefficient
technique (PITT)
DNa (cm2 s–1)

Cyclic voltammetry (CV) I. Ionic diffusion coefficient Na3V2(PO4)3@C@rGO [36] ~10–11

II. Pseudocapacitance
contribution

Na3-5xV2-xMox(PO4)3 [62] 10–11~10–12

Na3V2(PO4)2F3 [115] ~10–11
Na3.5Mn0.5V1.5(PO4)3/C ~10–9
[116]
Na2VTi(PO4)3@C [60] ~10–11
Na2VTi(PO4)3@C [117] ~10–11
rGO@Na3MnTi(PO4)3-C 76.6% at 0.5 mV s–1
[82]

Na3.5Mn0.5V1.5(PO4)3 [119] 63% at 0.5 mV s–1

Na4MnCr(PO4)3 [91] 71.7% at 0.5 mV s–1
DNa (cm2 s–1)

Electrochemical impedance Na3V1.95Mg0.05(PO4)3/C 1.0  10–13

spectroscopy (EIS) [120]

I. Charge transfer Na3þxV2-xNix(PO4)3/C 9.39  10–13

resistance [122]
II. Ionic diffusion Na3Fe2(PO4)3 [76]
coefficient
III. Activation energy K0.24Na2.76Fe2(PO4)3 [123] 3.98  10–11
Na3V1.9Cr0.1(PO4)3 [90] 2.12  10–10
Na3V1.5Cr0.5(PO4)3 [124]
Na3MnTi(PO4)3/C [63]
Theoretical calculation Density functional theory (DFT) Na3V2(PO4)3 [130] Migration barriers: 0.0904 and 0.11774
technologies eV

Na3V2(PO4)3 [131] Na1 participates in the ionic

transportation
I. Potential prediction Na3þxV2-xMgx(PO4)3 [42] Mg2þ substitutes on V sites in NVP
II. Migration pathway Na3V2-xAlx(PO4)3 [128] Electronic conductivity is enhanced
III. Diffusion energy barrier Na4MnCr(PO4)3 [43] Metallic properties
IV. Density of states Na4MnCr(PO4)3 [64] Theoretical capacity of 165 mAh g–1 and
average voltage about 4 V
Na3MnTi(PO4)3 [129] Theoretical capacity: 176.1 mAh g–1
Na3MnZr(PO4)3 [132] Evidence for the inhibited Jahn–Teller
distortion of Mn3þ by Zr4þ
Bond valence sum (BVS) I. Atom oxidation state Na2Fe2(SO4)3@C@GO Naþ tends to move in one-dimensional
II. Diffusion pathway [127] direction
Naþ in Na1 is movable

26
Y. Liu et al. eScience 2 (2022) 10–31

Fig. 9. Guideline to the development of NASICON-type cathode materials for SIBs.

Fig. 10. Schematic diagram of in situ devices for characterizing NASICON-type cathode materials for SIBs, from the present to the near future [134–138].

5. Conclusion and outlook out among cathode candidates because their stable 3D framework con-
tains open channels for fast Naþ migration. In recent years, a great
In the past few decades, the academic research on SIBs has bur- number of fresh characterization technologies and kinetic analysis
geoned, and there is an ever-increasing interest in the development of methods have emerged for gaining an in-depth understanding of the
high-performance sodium-storage cathode materials due to their crucial structure–performance relationships and energy-storage mechanisms of
role in determining battery performance. NASICON-type materials stand working electrodes. Four design strategies are commonly used to modify

27
Y. Liu et al. eScience 2 (2022) 10–31

NASICON cathode materials: nanosizing, carbon coating, redox chemis- Advanced techniques such as in situ (S)XRD and XPS/XAS are
try modulation, and foreign ion doping. In most cases, XRD, SXRD, and helpful in structural analysis and redox mechanism investigation.
ND are first applied to determine the NASICON-type structure after More experimental methods and characterization techniques
successful preparation, and information on the corresponding local and should be explored to controllably synthesize and precisely
electronic environments is provided by NMR, FTIR, and Raman spectra. modulate binary/ternary NASICON materials, thus enhancing the
After nanosizing and carbon coating, SEM, (HR)TEM, and STEM are the energy density of NASICON-type cathodes via multielectron
most effective techniques to observe morphological and microstructural reactions.
changes. With carbon coating especially, whether the carbon layer is (4) High working voltage is another factor that needs to be considered
evenly coated will affect the electrochemical performance. In addition, before the commercialization of NASICON-type cathodes. How-
ICP, EDS, TOF-SIMS, TGA, and DSC can provide information on a ma- ever, the decomposition of liquid electrolytes at high potentials
terial's composition and thermal stability. Advanced technologies like severely limits the service life of batteries. As illustrated in the
EPR, XPS, EELS, XAS, and M€ ossbauer spectra can be used to reveal above examples, the triple-layered architecture of the CEI in a
valence variations and electronic structures upon charging/discharging, Na4MnCr(PO4)3 cathode system was observed using TOF-SIMS,
particularly in the study of redox chemistry modulation and foreign ion but electrolyte decomposition led to poor cycling stability for
doping. the electrode. In addition, XPS results illustrated the formation of
Some electrochemical measurements, including galvanostatic an SEI between the anode surface and the electrolyte in full bat-
charge/discharge, cycling, and rate tests, are typically carried out to teries when using NASICON as the cathode. Therefore, extensive
evaluate the electrochemical performance of designed electrodes, but studies on optimizing electrolytes and constructing stable SEI
besides these, GITT, PITT, CV, and EIS tests are more advanced methods layers should be carried out to keep pace with high-potential
for investigating the reaction kinetics by calculating the ionic diffusion NASICON cathodes.
coefficient, pseudocapacitive contribution, and internal/external
impedance. The rapid development of theoretical computation has also In brief, advanced characterizations and measurements applied in the
led to the emergence of a new branch of research for designing high- study of NASICON cathodes deepen our understanding of these materials
energy NASICON cathode materials, which can provide theoretical pre- from multiple angles, and we strongly believe that NASICON-type cath-
dictions, help to analyze electronic structures, and investigate energy- odes will in future achieve broad application in SIBs through more in-
storage mechanisms during the (de)sodiation process (Fig. 9). depth investigations that combine efficient experimental and theoret-
To develop and optimize NASICON-type cathode materials for SIBs ical methods.
with a view to a wide range of future applications, we would like to put
forward a few considerations for further research: Competing financial interests

(1) The above techniques under in situ/operando conditions (e.g., The authors declare no competing financial interests.
XRD, Raman, and XAS) have attracted extensive attention, since
they offer more reliable, real-time information, which is essential
Declaration of competing interest
to develop high-performance cathodes. However, current char-
acterization and measurement techniques still need further
The authors declare that they have no known competing financial
improvement, especially the development of in situ/operando
interests or personal relationships that could have appeared to influence
tools for methods such as FTIR, NMR, and EPR, which are mostly
the work reported in this paper.
used under ex situ conditions. It is also a major challenge to design
appropriate mold cells for the various in situ tests so as to acquire
reliable data. Thus, developing more advanced in situ techniques Acknowledgments
for studying NASICON cathodes is highly desirable.
(2) Further exploration of cutting-edge techniques to study NASICON- Financial support from the National Natural Science Foundation of
type cathodes would be useful. For example, PDF analysis can China (22075016 and 21805007), Fundamental Research Funds for the
reveal the local structures of materials with complex or disordered Central Universities (FRF-TP-20-020A3), and 111 Project (B12015 and
structures, helping to understand their physical properties. In situ B170003) is gratefully acknowledged.
atomic force microscopy (AFM), which is based on the indentation
of a tip-shaped cantilever, is an efficient surface-sensitive tool to References
track surface evolution. In situ TEM could be used to study the
[1] T. Jin, H.X. Li, K.J. Zhu, P.-F. Wang, P. Liu, L.F. Jiao, Polyanion-type cathode
reaction mechanisms behind battery performance, by virtue of its materials for sodium-ion batteries, Chem. Soc. Rev. 49 (2020) 2342–2377.
unique ability to dynamically image, with high resolution, the [2] Y. Lu, J. Chen, Prospects of organic electrode materials for practical lithium
structural evolution, phase transformation, and chemical compo- batteries, Nat. Rev. Chem. 4 (2020) 127–142.
[3] C. Yang, S. Xin, L.Q. Mai, Y. You, Materials design for high-safety sodium-ion
sition changes in electrode materials. It is worth taking into battery, Adv. Energy Mater. 11 (2021), 2000974.
careful consideration whether these techniques could be prom- [4] N. Yabuuchi, K. Kubota, M. Dahbi, S. Komaba, Research development on sodium-
ising for investigating NASICON-type materials, under the ion batteries, Chem. Rev. 114 (2014) 11636–11682.
[5] J.-Y. Hwang, S.-T. Myung, Y.-K. Sun, Sodium-ion batteries: present and future,
appropriate conditions. A schematic diagram of the in situ devices
Chem. Soc. Rev. 46 (2017) 3529–3614.
for characterizing NASICON cathodes in SIBs is presented in [6] P.K. Nayak, L.T. Yang, W. Brehm, P. Adelhelm, From lithium-ion to sodium-ion
Fig. 10, from the present to the near future. batteries: advantages, challenges, and surprises, Angew. Chem. Int. Ed. 57 (2018)
102–120.
[7] Y.C. Liu, X.B. Liu, T.S. Wang, L.-Z. Fan, L.F. Jiao, Research and application
(3) The electrons that participate in the electrochemical reactions of progress on key materials for sodium-ion batteries, Sustain. Energy Fules 1 (2017)
most NASICON-structured electrodes are restricted to two per 986–1006.
formula unit. To further improve the specific capacity of NASICON [8] Y. Wang, Y.K. Liu, Y.C. Liu, Q.Y. Shen, C.C. Chen, F.Y. Qiu, P. Li, L.F. Jiao,
X.H. Qu, Recent advances in electrospun electrode materials for sodium-ion
materials, pursuing more than two-electron reversible redox re- batteries, J. Energy Chem. 54 (2021) 225–241.
actions has attracted extensive interest. Considering that different [9] P. Yu, W. Tang, F.-F. Wu, C. Zhang, H.-Y. Luo, H. Liu, Z.-G. Wang, Recent progress
metal elements feature various redox voltages and valence states, in plant-derived hard carbon anode materials for sodium-ion batteries: a review,
Rare Met. 39 (2020) 1019–1033.
the design of NASICON cathodes with binary/ternary metal ele- [10] Q.Y. Shen, X.D. Zhao, Y.C. Liu, Y.P. Li, J. Zhang, N. Zhang, C.H. Yang, J. Chen,
ments to achieve high energy density is highly advisable. Dual-strategy of cation-doping and nanoengineering enables fast and stable

28
Y. Liu et al. eScience 2 (2022) 10–31

sodium-ion storage in a novel Fe/Mn-based layered oxide cathode, Adv. Sci. 7 porous skeleton-supported Na3V2(PO4)3/carbon composite for high power
(2020), 2002199. sodium-ion battery cathode, Small Methods 3 (2019), 1800169.
[11] N. Ortiz-Vitoriano, N.E. Drewett, E. Gonzalo, T. Rojo, High performance [40] Y.J. Fang, L.F. Xiao, X.P. Ai, Y.L. Cao, H.X. Yang, Hierarchical carbon framework
manganese-based layered oxide cathodes: overcoming the challenges of sodium wrapped Na3V2(PO4)3 as a superior high-rate and extended lifespan cathode for
ion batteries, Energy Environ. Sci. 10 (2017) 1051–1074. sodium-ion batteries, Adv. Mater. 27 (2015) 5895–5900.
[12] S.H. Guo, P. Liu, H.J. Yu, Y.B. Zhu, M.W. Chen, M. Ishida, H.S. Zhou, A layered P2- [41] W.H. Ren, Z.P. Zheng, C. Xu, C.J. Niu, Q.L. Wei, Q.Y. An, K.N. Zhao, M.Y. Yan,
and O3-type composite as a high-energy cathode for rechargeable sodium-ion M.S. Qin, L.Q. Mai, Self-sacrificed synthesis of three-dimensional Na3V2(PO4)3
batteries, Angew. Chem. Int. Ed. 54 (2015) 5894–5899. nanofiber network for high-rate sodium ion full batteries, Nano Energy 25 (2016)
[13] P.-F. Wang, Y. You, Y.-X. Yin, Y.-S. Wang, L.-J. Wan, L. Gu, Y.-G. Guo, Suppressing 145–153.
the P2-O2 phase transition of Na0.67Mn0.67Ni0.33O2 by magnesium substitution for [42] H. Li, H.M. Tang, C.Z. Ma, Y. Bai, J. Alvarado, B. Radhakrishnan, S.P. Ong, F. Wu,
improved sodium-ion batteries, Angew. Chem. Int. Ed. 55 (2016) 7445–7449. Y.S. Meng, C. Wu, Understanding the electrochemical mechanisms induced by
[14] Y.C. Liu, Q.Y. Shen, X.D. Zhao, J. Zhang, X.B. Liu, T.S. Wang, N. Zhang, L.F. Jiao, gradient Mg2þ distribution of Na-rich Na3þxV2-xMgx(PO4)3/C for sodium ion
J. Chen, L.-Z. Fan, Hierarchical engineering of porous P2-Na2/3Ni1/3Mn2/3O2 batteries, Chem. Mater. 30 (2018) 2498–2505.
nanofibers assembled by nanoparticles enables superior sodium-ion storage [43] J. Zhang, Y.C. Liu, X.D. Zhao, L.H. He, H. Liu, Y.Z. Song, S.D. Sun, Q. Li, X.R. Xing,
cathodes, Adv. Funct. Mater. 30 (2020), 1907837. J. Chen, A novel NASICON-type Na4MnCr(PO4)3 demonstrating the energy density
[15] Y.-F. Zhu, Y. Xiao, S.-X. Dou, Y.-M. Kang, S.-L. Chou, Spinel/post-spinel record of phosphate cathodes for sodium-ion batteries, Adv. Mater. 32 (2020),
engineering on layered oxide cathodes for sodium-ion batteries, eScience 1 (2021) 1906348.
13–27. [44] S.C. Yu, Z.G. Liu, H. Tempel, H. Kungl, R.-A. Eichel, Self-standing NASICON-type
[16] F.F. Wang, N. Zhang, X.D. Zhao, L.X. Wang, J. Zhang, T.S. Wang, F.F. Liu, Y.C. Liu, electrodes with high mass loading for fast-cycling all-phosphate sodium-ion
L.-Z. Fan, Realizing a high-performance Na-storage cathode by tailoring ultrasmall batteries, J. Mater. Chem. 6 (2018) 18304–18317.
Na2FePO4F nanoparticles with facilitated reaction kinetics, Adv. Sci. 6 (2019), [45] C.L. Xu, J.M. Zhao, E.H. Wang, X.H. Liu, X. Shen, X.H. Rong, Q. Zheng, G.X. Ren,
1900649. N. Zhang, X.S. Liu, X.D. Guo, C. Yang, H.Z. Liu, B.H. Zhong, Y.-S. Hu, A novel
[17] Y.C. Liu, N. Zhang, F.F. Wang, X.B. Liu, L.F. Jiao, L.-Z. Fan, Approaching the NASICON-typed Na4VMn0.5Fe0.5(PO4)3 cathode for high-performance Na-ion
downsizing limit of maricite NaFePO4 toward high-performance cathode for batteries, Adv. Energy Mater. 11 (2021), 2100729.
sodium-ion batteries, Adv. Funct. Mater. 28 (2018), 1801917. [46] J.R. Hou, M. Hadouchi, L.J. Sui, J. Liu, M.X. Tang, W.H. Kan, M. Avdeev,
[18] J. Zhao, X. Yang, Y. Yao, Y. Gao, Y.M. Sui, B. Zou, H. Ehrenberg, G. Chen, F. Du, G.M. Zhong, Y.-K. Liao, Y.-H. Lai, Y.-H. Chu, H.-J. Lin, C.-T. Chen, Z.W. Hu,
Moving to aqueous binder: a valid approach to achieving high-rate capability and Y.H. Huang, J.W. Ma, Unlocking fast and reversible sodium intercalation in
long-term durability for sodium-ion battery, Adv. Sci. 5 (2018), 1700768. NASICON Na4MnV(PO4)3 by fluorine substitution, Energy Storage Mater. 42
[19] R. Liu, S.Y. Zheng, Y.F. Yuan, P.F. Yu, Z.T. Liang, W.M. Zhao, R. Shahbazian- (2021) 307–316.
Yassar, J.X. Ding, J. Lu, Y. Yang, Counter-intuitive structural instability aroused by [47] K. Chayambuka, G. Mulder, D.L. Danilov, P.H.L. Notten, Sodium-ion battery
transition metal migration in polyanionic sodium ion host, Adv. Energy Mater. 11 materials and electrochemical properties reviewed, Adv. Energy Mater. 8 (2018),
(2020), 2003256. 1800079.
[20] Y.Z. Jiang, S.L. Yu, B.Q. Wang, Y. Li, W.P. Sun, Y.H. Lu, M. Yan, B. Song, S.-X. Dou, [48] G.X. Chen, Q. Huang, T. Wu, L. Lu, Polyanion sodium vanadium phosphate for
Prussian blue@C composite as an ultrahigh-rate and long-life sodium-ion battery next generation of sodium-ion batteries-a review, Adv. Funct. Mater. 30 (2020),
cathode, Adv. Funct. Mater. 26 (2016) 5315–5321. 2001289.
[21] J.F. Qian, C. Wu, Y.L. Cao, Z.F. Ma, Y.H. Huang, X.P. Ai, H.X. Yang, Prussian blue [49] H.J. Yu, S.H. Guo, Y.B. Zhu, M. Ishida, H.S. Zhou, Novel titanium-based O3-type
cathode materials for sodium-ion batteries and other ion batteries, Adv. Energy NaTi0.5Ni0.5O2 as a cathode material for sodium ion batteries, Chem. Commun. 50
Mater. 8 (2018), 1702619. (2014) 457–459.
[22] W. Luo, M. Allen, V. Raju, X.L. Ji, An organic pigment as a high-performance [50] Z.L. Jian, C.C. Yuan, W.Z. Han, X. Lu, L. Gu, X.K. Xi, Y.-S. Hu, H. Li, W. Chen,
cathode for sodium-ion batteries, Adv. Energy Mater. 4 (2014), 1400554. D.F. Chen, Y.C. Ikuhara, L.Q. Chen, Atomic structure and kinetics of NASICON
[23] Q. Ni, Y. Bai, F. Wu, C. Wu, Polyanion-type electrode materials for sodium-ion NaxV2(PO4)3 cathode for sodium-ion batteries, Adv. Funct. Mater. 24 (2014)
batteries, Adv. Sci. 4 (2017), 1600275. 4265–4272.
[24] C. Masquelier, L. Croguennec, Polyanionic (phosphates, silicates, sulfates) [51] C.P. Grey, J.M. Tarascon, Sustainability and in situ monitoring in battery
frameworks as electrode materials for rechargeable Li (or Na) batteries, Chem. development, Nat. Mater. 16 (2017) 45–56.
Rev. 113 (2013) 6552–6591. [52] X.J. Wei, X.P. Wang, Q.Y. An, C.H. Han, L.Q. Mai, Operando X-ray diffraction
[25] Y.J. Fang, J.X. Zhang, L.F. Xiao, X.P. Ai, Y.L. Cao, H.X. Yang, Phosphate characterization for understanding the intrinsic electrochemical mechanism in
framework electrode materials for sodium ion batteries, Adv. Sci. 4 (2017), rechargeable battery materials, Small Methods 1 (2017), 1700083.
1600392. [53] S.Y. Lim, H. Kim, R.A. Shakoor, Y. Jung, J.W. Choi, Electrochemical and thermal
[26] Y. You, A. Manthiram, Progress in high-voltage cathode materials for rechargeable properties of NASICON structured Na3V2(PO4)3 as a sodium rechargeable battery
sodium-ion batteries, Adv. Energy Mater. 8 (2018), 1701785. cathode: a combined experimental and theoretical study, J. Electrochem. Soc. 159
[27] P. Barpanda, L. Lander, S.-i. Nishimura, A. Yamada, Polyanionic insertion (2012) A1393–A1397.
materials for sodium-ion batteries, Adv. Energy Mater. 8 (2018), 1703055. [54] Z.L. Jian, W.Z. Han, X. Lu, H.X. Yang, Y.-S. Hu, J. Zhou, Z.B. Zhou, J.Q. Li,
[28] R. Liu, Z.T. Liang, Z.L. Gong, Y. Yang, Research progress in multielectron reactions W. Chen, D.F. Chen, L.Q. Chen, Superior electrochemical performance and storage
in polyanionic materials for sodium-ion batteries, Small Methods 3 (2019), mechanism of Na3V2(PO4)3 cathode for room-temperature sodium-ion batteries,
1800221. Adv. Energy Mater. 3 (2013) 156–160.
[29] X.X. Cao, J. Zhou, A.Q. Pan, S.Q. Liang, Recent advances in phosphate cathode [55] G.B. Zhang, T.F. Xiong, M.Y. Yan, Y.N. Xu, W.H. Ren, X. Xu, Q.L. Wei, L.Q. Mai, In
materials for sodium-ion batteries, Acta Phys. Chim. Sin. 36 (2020), 1905018. operando probing of sodium-incorporation in NASICON nanomaterial:
[30] H.X. Li, M. Xu, Z.A. Zhang, Y.Q. Lai, J.M. Ma, Engineering of polyanion type asymmetric reaction and electrochemical phase diagram, Chem. Mater. 29 (2017)
cathode materials for sodium-ion batteries: toward higher energy/power density, 8057–8064.
Adv. Funct. Mater. 30 (2020), 2000473. [56] F. Lalere, V. Seznec, M. Courty, R. David, J.N. Chotard, C. Masquelier, Improving
[31] N. Anantharamulu, K.K. Rao, G. Rambabu, B.V. Kumar, V. Radha, M. Vithal, the energy density of Na3V2(PO4)3-based positive electrodes through V/Al
A wide-ranging review on Nasicon type materials, J. Mater. Sci. 46 (2011) substitution, J. Mater. Chem. A 3 (2015) 16198–16205.
2821–2837. [57] A. Inoishi, Y. Yoshioka, L.W. Zhao, A. Kitajou, S. Okada, Improvement in the
[32] Z.F. Dai, U. Mani, H.T. Tan, Q.Y. Yan, Advanced cathode materials for sodium-ion energy density of Na3V2(PO4)3 by Mg Substitution, ChemElectroChem 4 (2017)
batteries: what determines our choices? Small Methods 1 (2017), 1700098. 2755–2759.
[33] M. Sawicki, L.L. Shaw, Advances and challenges of sodium ion batteries as post [58] F. Chen, V.M. Kovrugin, R. David, O. Mentre, F. Fauth, J.-N. Chotard,
lithium ion batteries, RSC Adv. 5 (2015) 53129–53154. C. Masquelier, A NASICON-type positive electrode for Na batteries with high
[34] Z.L. Jian, Y.-S. Hu, X.L. Ji, W. Chen, NASICON-structured materials for energy energy density: Na4MnV(PO4)3, Small Methods 3 (2019), 1800218.
storage, Adv. Mater. 29 (2017), 1601925. [59] H.X. Li, T. Jin, X.B. Chen, Y.Q. Lai, Z.A. Zhang, W.Z. Bao, L.F. Jiao, Rational
[35] S.Q. Chen, C. Wu, L.F. Shen, C.B. Zhu, Y.Y. Huang, K. Xi, J. Maier, Y. Yu, architecture design enables superior Na storage in greener NASICON-
Challenges and perspectives for NASICON-type electrode materials for advanced Na4MnV(PO4)3 cathode, Adv. Energy Mater. 8 (2018), 1801418.
sodium-ion batteries, Adv. Mater. 29 (2017), 1700431. [60] D.X. Wang, X.F. Bie, Q. Fu, D. Dixon, N. Bramnik, Y.-S. Hu, F. Fauth, Y.J. Wei,
[36] X.H. Rui, W.P. Sun, C. Wu, Y. Yu, Q.Y. Yan, An advanced sodium-ion battery H. Ehrenberg, G. Chen, F. Du, Sodium vanadium titanium phosphate electrode for
composed of carbon coated Na3V2(PO4)3 in a porous graphene network, Adv. symmetric sodium-ion batteries with high power and long lifespan, Nat. Commun.
Mater. 27 (2015) 6670–6676. 8 (2017), 15888.
[37] Q. Liu, X. Meng, Z.X. Wei, D.X. Wang, Y. Gao, Y.J. Wei, F. Du, G. Chen, Core/ [61] F. Lalere, V. Seznec, M. Courty, J.N. Chotard, C. Masquelier, Coupled X-ray
double-shell structured Na3V2(PO4)2F3@C nanocomposite as the high power and diffraction and electrochemical studies of the mixed Ti/V-containing NASICON:
long lifespan cathode for sodium-ion batteries, ACS Appl. Mater. Interfaces 8 Na2TiV(PO4)3, J. Mater. Chem. A 6 (2018) 6654–6659.
(2016) 31709–31715. [62] X. Li, Y.Y. Huang, J.S. Wang, L. Miao, Y.Y. Li, Y. Liu, Y.G. Qiu, C. Fang, J.T. Han,
[38] J.X. Zhang, Y.J. Fang, L.F. Xiao, J.F. Qian, Y.L. Cao, X.P. Ai, H.X. Yang, Graphene- Y.H. Huang, High valence Mo-doped Na3V2(PO4)3/C as a high rate and stable
scaffolded Na3V2(PO4)3 microsphere cathode with high rate capability and cycling cycle-life cathode for sodium battery, J. Mater. Chem. A 6 (2018) 1390–1396.
stability for sodium ion batteries, ACS Appl. Mater. Interfaces 9 (2017) [63] T. Zhu, P. Hu, X.P. Wang, Z.H. Liu, W. Luo, K.A. Owusu, W.W. Cao, C.W. Shi,
7177–7184. J.T. Li, L. Zhou, L.Q. Mai, Realizing three-electron redox reactions in NASICON-
[39] E.H. Wang, M.Z. Chen, X.H. Liu, Y.M. Liu, H.P. Guo, Z.G. Wu, W. Xiang, structured Na3MnTi(PO4)3 for sodium-ion batteries, Adv. Energy Mater. 9 (2019),
B.H. Zhong, X.D. Guo, S.L. Chou, S.-X. Dou, Organic cross-linker enabling a 3D 1803436.

29
Y. Liu et al. eScience 2 (2022) 10–31

[64] J.Y. Wang, Y. Wang, D.-H. Seo, T. Shi, S.P. Chen, Y.S. Tian, H. Kim, G. Ceder, [89] K. Saravanan, C.W. Mason, A. Rudola, K.H. Wong, P. Balaya, The first report on
A high-energy NASICON-type cathode material for Na-ion batteries, Adv. Energy excellent cycling stability and superior rate capability of Na3V2(PO4)3 for sodium
Mater. 10 (2020), 1903968. ion batteries, Adv. Energy Mater. 3 (2013) 444–450.
[65] X.Q. Yu, H.L. Pan, W. Wan, C. Ma, J.M. Bai, Q.P. Meng, S.N. Ehrlich, Y.-S. Hu, X.- [90] M.J. Arag on, P. Lavela, G.F. Ortiz, J.L. Tirado, Benefits of chromium substitution
Q. Yang, A size-dependent sodium storage mechanism in Li4Ti5O12 investigated by in Na3V2(PO4)3 as a potential candidate for sodium-ion batteries,
a novel characterization technique combining in situ X-ray diffraction and ChemElectroChem 2 (2015) 995–1002.
chemical sodiation, Nano Lett. 13 (2013) 4721–4727. [91] W. Zhang, H.X. Li, Z.A. Zhang, M. Xu, Y.Q. Lai, S.-L. Chou, Full activation of Mn4
þ
[66] J.-N. Chotard, G. Rousse, R. David, O. Mentre, M. Courty, C. Masquelier, Discovery /Mn3þ redox in Na4MnCr(PO4)3 as a high-voltage and high-rate cathode material
of a sodium-ordered form of Na3V2(PO4)3 below ambient temperature, Chem. for sodium-ion batteries, Small 16 (2020), 2001524.
Mater. 27 (2015) 5982–5987. [92] G.G. Eshetu, T. Diemant, M. Hekmatfar, S. Grugeon, R.J. Behm, S. Laruelle,
[67] R. Liu, G.L. Xu, Q. Li, S.Y. Zheng, G.R. Zheng, Z.L. Gong, Y.X. Li, E. Kruskop, M. Armand, S. Passerini, Impact of the electrolyte salt anion on the solid
R.Q. Fu, Z.H. Chen, K. Amine, Y. Yang, Exploring highly reversible 1.5-electron electrolyte interphase formation in sodium ion batteries, Nano Energy 55 (2019)
reactions (V3þ/V4þ/V5þ) in Na3VCr(PO4)3 cathode for sodium-ion batteries, ACS 327–340.
Appl. Mater. Interfaces 9 (2017) 43632–43639. [93] M.Y. Ma, H.R. Cai, C.L. Xu, R.Z. Huang, S.R. Wang, H.L. Pan, Y.-S. Hu, Engineering
[68] M. Bianchini, F. Fauth, N. Brisset, F. Weill, E. Suard, C. Masquelier, L. Croguennec, solid electrolyte interface at nano-scale for high-performance hard carbon in
Comprehensive investigation of the Na3V2(PO4)2F3-NaV2(PO4)2F3 system by sodium-ion batteries, Adv. Funct. Mater. 31 (2021), 2100278.
operando high resolution synchrotron X-ray diffraction, Chem. Mater. 27 (2015) [94] A. Ponrouch, R. Dedryvere, D. Monti, A.E. Demet, J.M.A. Mba, L. Croguennec,
3009–3020. C. Masquelier, P. Johansson, M.R. Palacín, Towards high energy density sodium
[69] R. Muruganantham, Y.-T. Chiu, C.-C. Yang, C.-W. Wang, W.-R. Liu, An efficient ion batteries through electrolyte optimization, Energy Environ. Sci. 6 (2013)
evaluation of F-doped polyanion cathode materials with long cycle life for Na-ion 2361–2369.
batteries applications, Sci. Rep. 7 (2017), 14808. [95] D. Burova, I. Shakhova, P. Morozova, A. Larchuk, O.A. Drozhzhin, M.G. Rozova,
[70] X.H. Liu, G.L. Feng, E.H. Wang, H. Chen, Z.G. Wu, W. Xiang, Y.J. Zhong, S. Praneetha, V. Murugan, J.-M. Tarascon, A.M. Abakumov, The rapid microwave-
Y.X. Chen, X.D. Guo, B.H. Zhong, Insight into preparation of Fe-doped assisted hydrothermal synthesis of NASICON-structured Na3V2O2x(PO4)2F3-2x (0 <
Na3V2(PO4)3@C from aspects of particle morphology design, crystal structure x  1) cathode materials for Na-ion batteries, RSC Adv. 9 (2019) 19429–19440.
modulation, and carbon graphitization regulation, ACS Appl. Mater. Interfaces 11 [96] M.Z. Chen, W.B. Hua, J. Xiao, D. Cortie, X.D. Guo, E.H. Wang, Q.F. Gu, Z. Hu,
(2019) 12421–12430. S. Indris, X.-L. Wang, S.-L. Chou, S.-X. Dou, Development and investigation of a
[71] M. Hadouchi, N. Yaqoob, P. Kaghazchi, M.X. Tang, J. Liu, P.F. Sang, Y.Z. Fu, NASICON-type high-voltage cathode material for high-power sodium-ion
Y.H. Huang, J.W. Ma, Fast sodium intercalation in Na3.41£0.59FeV(PO4)3 pound: a batteries, Angew. Chem. Int. Ed. 59 (2020) 2449–2456.
novel sodium-deficient NASICON cathode for sodium-ion batteries, Energy [97] E. Talaie, P. Bonnick, X.Q. Sun, Q. Pang, X. Liang, L.F. Nazar, Methods and
Storage Mater. 35 (2021) 192–202. protocols for electrochemical energy storage materials research, Chem. Mater. 29
[72] Z.G. Liu, Y.-Y. Hu, M.T. Dunstan, H. Huo, X.G. Hao, H. Zou, G.M. Zhong, Y. Yang, (2017) 90–105.
C.P. Grey, Local structure and dynamics in the Na ion battery positive electrode [98] M. Pivko, I. Arcon, M. Bele, R. Dominko, M. Gaberscek, A3V2(PO4)3 (A ¼ Na or Li)
material Na3V2(PO4)2F3, Chem. Mater. 26 (2014) 2513–2521. probed by in situ X-ray absorption spectroscopy, J. Power Sources 216 (2012)
[73] X.S. Liu, Z.T. Liang, Y.X. Xiang, M. Lin, Q. Li, Z.G. Liu, G.M. Zhong, R.Q. Fu, 145–151.
Y. Yang, Solid-state NMR and MRI spectroscopy for Li/Na batteries: materials, [99] M.Z. Chen, W.B. Hua, J. Xiao, J.L. Zhang, V.W. Lau, M. Park, G.-H. Lee, S. Lee,
interface, and in situ characterization, Adv. Mater. (2021), 2005878. W.L. Wang, J. Peng, L. Fang, L.M. Zhou, C.-K. Chang, Y. Yamauchi, S.L. Chou, Y.-
[74] Y.F. Deng, S.Y. Dong, Z.F. Li, H. Jiang, X.G. Zhang, X.L. Ji, Applications of M. Kang, Activating a multielectron reaction of NASICON-structured cathodes
conventional vibrational spectroscopic methods for batteries beyond Li-ion, Small toward high energy density for sodium-ion batteries, J. Am. Chem. Soc. 143
Methods 2 (2018), 1700332. (2021) 18091–18102.
[75] V. Stancovski, S. Badilescu, In situ Raman spectroscopic-electrochemical studies of [100] S. Ghosh, N. Barman, M. Mazumder, S.K. Pati, G. Rousse, P. Senguttuvan, High
lithium-ion battery materials: a historical overview, J. Appl. Electrochem. 44 capacity and high-rate NASICON-Na3.75V1.25Mn0.75(PO4)3 cathode for Na-ion
(2014) 23–43. batteries via modulating electronic and crystal structures, Adv. Energy Mater. 10
[76] R. Rajagopalan, B. Chen, Z.C. Zhang, X.-L. Wu, Y.H. Du, Y. Huang, B. Li, Y. Zong, (2020), 1902918.
J. Wang, G.-H. Nam, M. Sindoro, S.X. Dou, H.K. Liu, H. Zhang, Improved [101] K. Kawai, W.W. Zhao, S.-i. Nishimura, A. Yamada, High-voltage Cr4þ/Cr3þ redox
reversibility of Fe3þ/Fe4þ redox couple in sodium super ion conductor type couple in polyanion compounds, ACS Appl. Energy Mater. 1 (2018) 928–931.
Na3Fe2(PO4)3 for sodium-ion batteries, Adv. Mater. 29 (2017), 1605694. [102] K. Kawai, D. Asakura, S.-i. Nishimura, A. Yamada, Stabilization of a 4.5 V Cr4þ/Cr3
þ
[77] H. Bih, L. Bih, B. Manoun, M. Azdouz, S. Benmokhtar, P. Lazor, Raman redox reaction in NASICON-type Na3Cr2(PO4)3 by Ti substitution, Chem.
spectroscopic study of the phase transitions sequence in Li3Fe2(PO4)3 and Commun. 55 (2019) 13717–13720.
Na3Fe2(PO4)3 at high temperature, J. Mol. Struct. 936 (2009) 147–155. [103] C.L. Zhao, Y.X. Lu, Y.M. Li, L.W. Jiang, X.H. Rong, Y.-S. Hu, H. Li, L.Q. Chen, Novel
[78] I.A. Trussov, L.L. Male, M.L. Sanjuan, A. Orera, P.R. Slater, Understanding the methods for sodium-ion battery materials, Small Methods 1 (2017), 1600063.
complex structural features and phase changes in Na2Mg2(SO4)3: a combined [104] S.C. Chung, J. Ming, L. Lander, J.C. Lu, A. Yamada, Rhombohedral NASICON-type
single crystal and variable temperature powder diffraction and Raman NaxFe2(SO4)3 for sodium ion batteries: comparison with phosphate and alluaudite
spectroscopy study, J. Solid State Chem. 272 (2019) 157–165. phases, J. Mater. Chem. A 6 (2018) 3919–3925.
[79] S.D. Findlay, N. Shibata, H. Sawada, E. Okunishi, Y. Kondo, T. Yamamoto, [105] S. Park, J.-N. Chotard, D. Carlier, I. Moog, M. Courty, M. Duttine, F. Fauth,
Y. Ikuhara, Robust atomic resolution imaging of light elements using scanning A. Iadecola, L. Croguennec, C. Masquelier, Crystal structures and local
transmission electron microscopy, Appl. Phys. Lett. 95 (2009), 191913. environments of NASICON-type Na3FeV(PO4)3 and Na4FeV(PO4)3 positive
[80] W.D. Zhou, L.G. Xue, X.J. Lü, H.C. Gao, Y.T. Li, S. Xin, G.T. Fu, Z.M. Cui, Y. Zhu, electrode materials for Na-ion batteries, Chem. Mater. 33 (2021) 5355–5367.
J.B. Goodenough, NaxMV(PO4)3 (M ¼ Mn, Fe, Ni) structure and properties for [106] G.J. Cui, Q.Y. Dong, Z.Z. Wang, X.-Z. Liao, S.Q. Yuan, M.D. Jiang, Y.B. Shen,
sodium extraction, Nano Lett. 16 (2016) 7836–7841. H. Wang, H.Y. Che, Y.-S. He, Z.-F. Ma, Achieving highly reversible and fast sodium
[81] M.-Y. Wang, J.-Z. Guo, Z.-W. Wang, Z.-Y. Gu, X.-J. Nie, X. Yang, X.-L. Wu, storage of Na4VMn(PO4)3/C-rGO composite with low-fraction rGO via spray-
Isostructural and multivalent anion substitution toward improved phosphate drying technique, Nano Energy 89 (2021), 106462.
cathode materials for sodium-ion batteries, Small 16 (2020), 1907645. [107] W. Weppner, R.A. Huggins, Electrochemical methods for determining kinetic
[82] H.X. Li, M. Xu, C.H. Gao, W. Zhang, Z.A. Zhang, Y.Q. Lai, L.F. Jiao, Highly properties of solids, Annu. Rev. Mater. Sci. 8 (1978) 269–311.
efficient, fast and reversible multi-electron reaction of Na3MnTi(PO4)3 cathode for [108] W. Weppner, R.A. Huggins, Determination of the kinetic parameters of mixed-
sodium-ion batteries, Energy Storage Mater. 26 (2020) 325–333. conducting electrodes and application to the system Li3Sb, J. Electrochem. Soc.
[83] L.S. Plashnitsa, E. Kobayashi, Y. Noguchi, S. Okada, J.-i. Yamaki, Performance of 124 (1977) 1569.
NASICON symmetric cell with ionic liquid electrolyte, J. Electrochem. Soc. 157 [109] C.J. Wen, B.A. Boukamp, R.A. Huggins, W. Weppner, Thermodynamic and mass
(2010) A536–A543. transport properties of “LiAl, J. Electrochem. Soc. 126 (1979) 2258.
[84] P. Senguttuvan, G. Rousse, M.E.A.Y. de Dompablo, H. Vezin, J.-M. Tarascon, [110] X.H. Rui, N. Ding, J. Liu, C. Li, C.H. Chen, Analysis of the chemical diffusion
M.R. Palacín, Low-potential sodium insertion in a NASICON-type structure coefficient of lithium ions in Li3V2(PO4)3 cathode material, Electrochim. Acta 55
through the Ti(III)/Ti(II) redox couple, J. Am. Chem. Soc. 135 (2013) 3897–3903. (2010) 2384–2390.
[85] Q.Y. Shen, Y.C. Liu, L.F. Jiao, X.H. Qu, J. Chen, Current state-of-the-art [111] K. Tang, X.Q. Yu, J.P. Sun, H. Li, X.J. Huang, Kinetic analysis on LiFePO4 thin films
characterization techniques for probing the layered oxide cathode materials of by CV, GITT, and EIS, Electrochim. Acta 56 (2011) 4869–4875.
sodium-ion batteries, Energy Storage Mater. 35 (2021) 400–430. [112] N. B€ockenfeld, A. Balducci, Determination of sodium ion diffusion coefficients in
[86] Y.J. Zhao, X.W. Gao, H.C. Gao, A. Dolocan, J.B. Goodenough, Elevating energy sodium vanadium phosphate, J. Solid State Electrochem. 18 (2014) 959–964.
density for sodium-ion batteries through multielectron reactions, Nano Lett. 21 [113] M.Z. Chen, W.B. Hua, J. Xiao, D. Cortie, W.H. Chen, E.H. Wang, Z. Hu, Q.F. Gu,
(2021) 2281–2287. X.L. Wang, S. Indris, S.-L. Chou, S.-X. Dou, NASICON-type air-stable and all-
[87] A. Knop-Gericke, E. Kleimenov, M. H€avecker, R. Blume, D. Teschner, S. Zafeiratos, climate cathode for sodium-ion batteries with low cost and high-power density,
R. Schl€ogl, V.I. Bukhtiyarov, V.V. Kaichev, I.P. Prosvirin, A.I. Nizovskii, H. Bluhm, Nat. Commun. 10 (2019) 1480.
A. Barinov, P. Dudin, M. Kiskinova, X-ray photoelectron spectroscopy for [114] H.B. Yahia, R. Essehli, R. Amin, K. Boulahya, T. Okumura, I. Belharouak, Sodium
investigation of heterogeneous catalytic processes, Adv. Catal. 52 (2009) intercalation in the phosphosulfate cathode NaFe2(PO4)(SO4)2, J. Power Sources
213–272. 382 (2018) 144–151.
[88] L.N. Bi, X.Y. Li, X.Q. Liu, Q.J. Zheng, D.M. Lin, Enhanced cycling stability and rate [115] W.X. Song, X.Y. Cao, Z.P. Wu, J. Chen, Y.R. Zhu, H.S. Hou, Q. Lan, X.B. Ji,
capability in a La-doped Na3V2(PO4)3/C cathode for high-performance sodium ion Investigation of the sodium ion pathway and cathode behavior in Na3V2(PO4)2F3
batteries, ACS Sustain. Chem. Eng. 7 (2019) 7693–7699. combined via a first principles calculation, Langmuir 30 (2014) 12438–12446.

30
Y. Liu et al. eScience 2 (2022) 10–31

[116] X.M. Ma, X.X. Cao, Y.F. Zhou, S. Guo, X.D. Shi, G.Z. Fang, A.Q. Pan, B.A. Lu, Na2Fe2(SO4)3/C cathode composite for long life and high energy density sodium-
J. Zhou, S.Q. Liang, Tuning crystal structure and redox potential of NASICON-type ion batteries, Adv. Energy Mater. 8 (2018), 1800944.
cathodes for sodium-ion batteries, Nano Res. 13 (2020) 3330–3337. [128] L.N. Zhao, H.L. Zhao, Z.H. Du, N. Chen, X.W. Chang, Z.J. Zhang, F. Gao,
[117] J. Dong, G.M. Zhang, X.G. Wang, S. Zhang, C. Deng, Cross-linked Na2VTi(PO4)3@C 
A. Trenczek-Zajac, K. Swierczek, Computational and experimental understanding
hierarchical nanofibers as high-performance bi-functional electrodes for of Al-doped Na3V2-xAlx(PO4)3 cathode material for sodium ion batteries: electronic
symmetric aqueous rechargeable sodium batteries, J. Mater. Chem. 5 (2017) structure, ion dynamics and electrochemical properties, Electrochim. Acta 282
18725–18736. (2018) 510–519.
[118] J. Wang, J. Polleux, J. Lim, B. Dunn, Pseudocapacitive Contributions to [129] Y. Zhou, X.J. Shao, K.-h. Lam, Y. Zheng, L.Z. Zhao, K.D. Wang, J.Z. Zhao,
electrochemical energy storage in TiO2 (anatase) nanoparticles, J. Phys. Chem. C F.M. Chen, X.H. Hou, Symmetric sodium-ion battery based on dual-electron
111 (2007) 14925–14931. reactions of NASICON-structured Na3MnTi(PO4)3 material, ACS Appl. Mater.
[119] J. Zhang, X.D. Zhao, Y.Z. Song, Q. Li, Y.C. Liu, J. Chen, X.R. Xing, Understanding Interfaces 12 (2020) 30328–30335.
the superior sodium-ion storage in a novel Na3.5Mn0.5V1.5(PO4)3 cathode, Energy [130] W.X. Song, X.B. Ji, Z.P. Wu, Y.R. Zhu, Y.C. Yang, J. Chen, M.J. Jing, F.Q. Li,
Storage Mater. 23 (2019) 25–34. C.E. Banks, First exploration of Na-ion migration pathways in the NASICON
[120] H. Li, X.Q. Yu, Y. Bai, F. Wu, C. Wu, L.-Y. Liu, X.-Q. Yang, Effects of Mg doping on structure Na3V2(PO4)3, J. Mater. Chem. A 2 (2014) 5358–5362.
the remarkably enhanced electrochemical performance of Na3V2(PO4)3 cathode [131] Q. Wang, M.Y. Zhang, C.G. Zhou, Y.L. Chen, Concerted ion-exchange mechanism
materials for sodium ion batteries, J. Mater. Chem. A 3 (2015) 9578–9586. for sodium diffusion and its promotion in Na3V2(PO4)3 framework, J. Phys. Chem.
[121] B.H. Li, C.P. Han, Y.-B. He, C. Yang, H.D. Du, Q.-H. Yang, F.Y. Kang, Facile C 122 (2018) 16649–16654.
synthesis of Li4Ti5O12/C composite with super rate performance, Energy Environ. [132] H.C. Gao, I.D. Seymour, S. Xin, L.G. Xue, G. Henkelman, J.B. Goodenough,
Sci. 5 (2012) 9595–9602. Na3MnZr(PO4)3: a high-voltage cathode for sodium batteries, J. Am. Chem. Soc.
[122] H. Li, Y. Bai, F. Wu, Q. Ni, C. Wu, Na-rich Na3þxV2-xNix(PO4)3/C for sodium ion 140 (2018) 18192–18199.
batteries: controlling the doping site and improving the electrochemical [133] P. Barpanda, G. Oyama, S.-i. Nishimura, S.-C. Chung, A. Yamada, A 3.8-V earth-
performances, ACS Appl. Mater. Interfaces 8 (2016) 27779–27787. abundant sodium battery electrode, Nat. Commun. 5 (2014) 4358.
[123] Y.J. Cao, Y. Liu, D.Q. Zhao, J.X. Zhang, X.P. Xia, T. Chen, L.-C. Zhang, P. Qin, [134] M.T. Xia, T.T. Liu, N. Peng, R.T. Zheng, X. Cheng, H.J. Zhu, H.X. Yu, M. Shui,
Y.Y. Xia, K-doped Na3Fe2(PO4)3 cathode materials with high-stable structure for J. Shu, Lab-scale in situ X-ray diffraction technique for different battery systems:
sodium-ion stored energy battery, J. Alloys Compd. 784 (2019) 939–946. designs, applications, and perspectives, Small Methods 3 (2019), 1900119.
[124] Y.J. Zhao, X.W. Gao, H.C. Gao, H.B. Jin, J.B. Goodenough, Three electron [135] Z.B. Wu, W.K. Pang, L.B. Chen, B. Johannessen, Z.P. Guo, In situ synchrotron X-ray
reversible redox reaction in sodium vanadium chromium phosphate as a high- absorption spectroscopy studies of anode materials for rechargeable batteries,
energy-density cathode for sodium-ion batteries, Adv. Funct. Mater. 30 (2020), Batteries Supercaps 4 (2021) 1547–1566.
1908680. [136] C. Yang, M.N. Han, H.H. Yan, F. Li, M.J. Shi, L.P. Zhao, In-situ probing phase
[125] Y. Deng, C. Eames, L.H.B. Nguyen, O. Pecher, K.J. Griffith, M. Courty, B. Fleutot, evolution and electrochemical mechanism of ZnMn2O4 nanoparticles anchored on
J.-N. Chotard, C.P. Grey, M.S. Islam, C. Masquelier, Crystal structures, local atomic porous carbon polyhedrons in high-performance aqueous Zn-ion batteries,
environments and ion diffusion mechanisms of scandium-substituted NASICON J. Power Sources 452 (2020), 227826.
solid electrolytes, Chem. Mater. 30 (2018) 2618–2630. [137] D.Q. Liu, Z. Shadike, R.Q. Lin, K. Qian, H. Li, K.K. Li, S.W. Wang, Q.P. Yu, M. Liu,
[126] P.A. Aparicio, J.A. Dawson, M.S. Islam, N.H. de Leeuw, Computational study of S. Ganapathy, X.Y. Qin, Q.-H. Yang, M. Wagemaker, F.Y. Kang, X.-Q. Yang, B.H. Li,
NaVOPO4 polymorphs as cathode materials for Na-Ion batteries: diffusion, Review of recent development of in situ/operando characterization techniques for
electronic properties, and cation-doping behavior, J. Phys. Chem. C 122 (2018) lithium battery research, Adv. Mater. 31 (2019), 1806620.
25829–25836. [138] Z. Shadike, E.Y. Zhao, Y.-N. Zhou, X.Q. Yu, Y. Yang, E.Y. Hu, S. Bak, L. Gu, X.-
[127] M. Chen, D. Cortie, Z. Hu, H. Jin, S. Wang, Q. Gu, W. Hua, E. Wang, W. Lai, Q. Yang, Advanced characterization techniques for sodium-ion battery studies,
L. Chen, S.-L. Chou, X.-L. Wang, S.-X. Dou, A novel graphene oxide wrapped Adv. Energy Mater. 8 (2018), 1702588.

31