Energy &Environmental Science

PERSPECTIVE

Publ

ishe

d on

20

May

201

3. D

ownl

oade

d by

EC

OL

E P

OL

YT

EC

HN

IC F

ED

DE

LA

USA

NN

E o

n 06

/11/

2015

12:

29:4

0. View Article Online

View Journal | View Issue

Charge carriers in

aSchool of Nano-Bioscience and Chemical

Science and Technology (UNIST), 100 Ba

689-798, South KoreabSchool of Chemical and Biological Engin

Gwanangno, Gwanak-gu, Seoul 151-744, SocInterdisciplinary School of Green Energy,

Technology (UNIST), 100 Banyeon-ri, Eonya

Korea. E-mail: ktlee@unist.ac.kr; Fax: +82 5

Cite this: Energy Environ. Sci., 2013, 6,2067

Received 8th March 2013Accepted 20th May 2013

DOI: 10.1039/c3ee40811f

www.rsc.org/ees

This journal is ª The Royal Society of

rechargeable batteries: Na ions vs.Li ions

Sung You Hong,a Youngjin Kim,b Yuwon Park,bc Aram Choi,c Nam-Soon Choic

and Kyu Tae Lee*c

We discuss the similarities and dissimilarities of sodium- and lithium-ion batteries in terms of negative and

positive electrodes. Compared to the comprehensive body of work on lithium-ion batteries, research on

sodium-ion batteries is still at the germination stage. Since both sodium and lithium are alkali metals,

they share similar chemical properties including ionicity, electronegativity and electrochemical reactivity.

They accordingly have comparable synthetic protocols and electrochemical performances, which

indicates that sodium-ion batteries can be successfully developed based on previously applied

approaches or methods in the lithium counterpart. The electrode materials in Li-ion batteries provide

the best library for research on Na-ion batteries because many Na-ion insertion hosts have their roots in

Li-ion insertion hosts. However, the larger size and different bonding characteristics of sodium ions

influence the thermodynamic and/or kinetic properties of sodium-ion batteries, which leads to

unexpected behaviour in electrochemical performance and reaction mechanism, compared to lithium-

ion batteries. This perspective provides a comparative overview of the major developments in the area

of positive and negative electrode materials in both Li-ion and Na-ion batteries in the past decade.

Highlighted are concepts in solid state chemistry and electrochemistry that have provided new

opportunities for tailored design that can be extended to many different electrode materials for

sodium-ion batteries.

Broader context

To meet the high demand for energy storage systems such as smart grid and electric vehicle applications, the next generation of secondary batteries is urgentlyrequired. Among the various post-rechargeable battery systems, Na-ion batteries are one of the promising choices due to the natural abundance, non-toxicity,and chemical similarity of sodium ions with lithium ions. Since sodium positions just below lithium in Group 1A of the periodic table, the heavily studiedsynthetic protocols, characterization methods, industrial techniques, and scientic equipment/facilities of Li-ion batteries can be efficiently applied for Na-ionbatteries. Sodium-ions with large ionic and orbital sizes, however, can derive different kinetic and thermodynamic properties. In this context, similarities anddifferences of Li-ion vs. Na-ion batteries are discussed regarding solid state chemistry and electrochemistry of negative and positive electrode materials.

1 Introduction

For several decades, lithium ions have been successfully utilizedas a charge carrier for secondary batteries.1,2 The outstandingelectrochemical performance of lithium ions derives from theirsmall ionic size, low atomic number, and the lowest redoxpotential.3,4 Among various metallic or semi-metallic cationslocated in s, p, d, and f blocks, lithium ions have the smallestionic radius (0.76 A). During charge and discharge, Li ions as a

Engineering, Ulsan National Institute of

nyeon-ri, Eonyang-eup, Ulju-gun, Ulsan,

eering, Seoul National University, 599

uth Korea

Ulsan National Institute of Science and

ng-eup, Ulju-gun, Ulsan, 689-798, South

2 217 2909; Tel: +82 52 217 2930

Chemistry 2013

charge carrier diffuse into electrode materials via intercalation,alloying, or conversion reactions. Considering that solid statediffusion is the major rate-determining step of electrochemicalreactions, the small ionic size permits fast kinetics by dimin-ishing the diffusion barrier.5 Furthermore, the small massnumber (6.941 amu) and low redox potential (�3.04 V vs.standard hydrogen electrode, SHE) enable high theoreticalspecic capacity and energy density of rechargeable energystorage systems.

Thanks to the excellent electrochemical performances,lithium-ion batteries (LIBs) have been successfully integratedinto mobile phones, medical and military devices, and hybridelectric vehicles.6 Lithium resources, however, are toogeographically constrained and limited to meet the globaldemands. Future energy storage systems should satisfy thefollowing requirements: cost-effectiveness, safety, sustain-ability, eco-friendliness, high capacity/power density, rate

Energy Environ. Sci., 2013, 6, 2067–2081 | 2067

Fig. 1 Voltage–capacity plots of (A) negative and (B) positive electrodematerialsfor NIBs.

Energy & Environmental Science Perspective

Publ

ishe

d on

20

May

201

3. D

ownl

oade

d by

EC

OL

E P

OL

YT

EC

HN

IC F

ED

DE

LA

USA

NN

E o

n 06

/11/

2015

12:

29:4

0.

View Article Online

capability, and large-scale manufacturing. Accordingly, sodium,magnesium, metal–air, metal–sulfur, and metal–organicbatteries are being considered for post-lithium ion batteries.4 Inparticular, sodium-ion batteries (NIBs) have been gainingincreasing attention thanks to the natural abundance and lowtoxicity of sodium resources.5,7,8 Furthermore, sodium is located

Sung You Hong graduated fromSeoul National University in2002 with BSc degree. Hongmoved to the University ofOxford to carry out his post-graduate studies (DPhil, 2005–2009) under the supervision ofProf. Ben Davis working onsynthetic organic chemistry.Following this, he joined theProf. Seeberger Group (MaxPlanck Institute) as a post-doctoral researcher continuing

to work on synthetic carbohydrate chemistry. He began indepen-dent research in 2010 at Ulsan National Institute of Science andTechnology. The Hong Group research is focused on the design andsynthesis of molecules of importance with biochemical or electro-chemical functions.

2068 | Energy Environ. Sci., 2013, 6, 2067–2081

just below lithium in the s block. Therefore, similar chemicalapproaches including a synthetic strategy, intercalation/alloy-ing/conversion chemistry, and characterization methodsutilized in electrode materials for LIBs could be applied todevelop electrode materials for NIBs.

Despite potential disadvantages, including larger size(1.02 A) of Na cations and higher redox potential (�2.71 V vs.SHE) of Na/Na+ compared to Li analogues, the different inter-actions between the guest Na cations and the host crystalstructures can inuence the kinetics as well as thermodynamicproperties of NIBs, and this may provide an avenue for abreakthrough technology to surpass LIBs.9 This perspectiveprovides a comparative overview of selected developments inboth Li-ion and Na-ion batteries in recent decades, and also theelectrochemical performances of NIB vs. LIB including redoxpotential, cycleability, and specic/volumetric capacity arecompared. Covering all of the research studies published in thistime period would be beyond the scope of the article. Thisperspective instead focuses on major developments in the areaof positive and negative electrode materials (see Fig. 1). High-lighted are concepts in solid state chemistry and electrochem-istry that have provided new insights for tailored design thatcan be extended to many different electrode materials for NIBs.

2 Negative electrode materials

Direct anodic application of elemental alkali metals (lithium orsodium) in rechargeable alkaline ion batteries causes poor cycleperformance and short-circuits because of their low meltingpoint (180.5 �C for Li and 97.7 �C for Na), high chemical reac-tivity and dendritic growth during charge and discharge. Tosolve the cycleability and safety issues of rechargeable batteries,alkali metal-insertion host materials of carbonaceouscompounds, alloy composites, metal oxides/suldes, organiccompounds containing carbonyl groups and phosphorus-basedmaterials have been extensively exploited.

Kyu Tae Lee is a Professor of theInterdisciplinary School ofGreen Energy at Ulsan NationalInstitute of Science & Tech-nology (UNIST), Korea. Hereceived his BSc (2000) and PhD(2006) in chemical engineeringfrom Seoul National University(Korea). In 2007, he joined theProf. Linda F. Nazar group at theUniversity of Waterloo (Canada)as a post-doctoral fellow. Hiscurrent research is focused on

electrode materials for Li-ion and Na-ion batteries and new elec-trochemical energy conversion/storage systems.

This journal is ª The Royal Society of Chemistry 2013

Fig. 3 Voltage profiles of (A) lithium–graphite system and (B) sodium–graphitesystem. Reproduced from ref. 11 with permission.

Perspective Energy & Environmental Science

Publ

ishe

d on

20

May

201

3. D

ownl

oade

d by

EC

OL

E P

OL

YT

EC

HN

IC F

ED

DE

LA

USA

NN

E o

n 06

/11/

2015

12:

29:4

0.

View Article Online

2.1 Carbon-based materials

Thanks to its relatively high specic capacity, stable two-dimensional (2D) structure, and low redox potential, lithiatedgraphite has been the most commonly utilized LIB anodematerial having rst been commercialized by Sony in 1991.10

Among several carbon allotropes including diamond, graphene,buckminsterfullerene, and carbon nanotubes, graphite is themost common form found in nature consisting of ABA stackingof two dimensional sheets of hexagonal sp2 carbon building upa P63/mmc space group. Up to one Li atom can reside ineach hexagonal C6 ring with a theoretical specic capacity of372 mA h g�1.

Unlike the successful anodic application of lithiatedgraphite in LIBs, the electrochemical sodium insertion intographite is proven to be not favorable.11 Lithium has variousC–Li binary phases as shown in Fig. 2,12whereas the C–Na phasediagram is not available. Also, electrochemical insertion of Naions into graphite materials has not been observed, indicatingthat sodiation of graphite might be a thermodynamicallyunfavorable process.

As shown in Fig. 3, clear differences of voltage proles wereobserved between de/lithiation and de/sodiation in graphitehosts. While reversible de/insertion of Li ions into graphite isobserved for the lithium–graphite cell, the sodium–graphitecell showed irreversible sodiation with a negligible amount ofde/sodiation. The major source of high irreversible capacity issurface passivation on graphite via irreversible electrolytedecomposition. It is noteworthy that even the small reversiblecapacity comes from an external contribution: plating/strippingof Na metals and reversible de/insertion of Na ions into carbonblack additives. DiVincenzo and Mele devised the interactionmodel based on a Thomas–Fermi theory,13 where the inter-planar alkali metal–carbon interactions are mostly determinedby the balance between coulombic attraction and hard-corerepulsion. Owing to the chemical soness of the Na–C bond, theinteraction between the guest sodium and the graphene layerbecomes very weak, explaining the lack of sodiation of graphite.

Fig. 2 Phase diagram of a C–Li binary system. Adapted from ref. 12.

This journal is ª The Royal Society of Chemistry 2013

As an effort to nd alternatives to graphite in NIBs, Doeffet al. demonstrated the reversible de/insertion of Na ions intodisordered carbons for an anode in NIBs.14 Most of the sodiumions are electrochemically inserted into nanoporous voids ofhard carbon (HC), which is built by disordered graphenestacking – the so-called ‘house of cards’ type model. The voltageproles of hard carbon are typically divided into sloping andlow-potential plateau regions.15 The insertion mechanism wasrevealed by 23Na NMRMAS. The nuclear spin quantum number(I) of 23Na is 3/2, and therefore the electrical quadrupolemoment can inuence the chemical shi and line widthdepending on the neighbouring environment. During dis-charging, two major peaks appear at +9 ppm (sharp) and �20to �30 ppm (broad). While the sharp peak corresponds to ionicsodium which is the result of the interaction with a misalignedgraphene layer, the broad chemical shi is assigned to sodiumin nanocavities.

Two independent research groups lead by Prof. Dahn15 andProf. Tirado16 demonstrated that the initial specic capacity ofhard carbon in NIBs is ca. 300 mA h g�1 close to that of graphitein LIBs. Most of the reversible capacity is related to sodiumstorage inside nanocavities. However, this hard carbon elec-trode exhibited signicant loss of capacity within 10 cycles withpoor rate performance. One of the challenging points of sodiumion batteries is the high reactivity of sodium inserted HC(Na@HC), when NaPF6 is used as a salt in electrolytes. In LIBs,the high reactivity of lithiated graphite can be reduced byadding LiPF6, since it is decomposed to form the stable ionicsalt LiF building up a passivation layer. In contrast, NaPF6 isstable even at elevated temperatures, and is not decomposed toform stable passivation layers.17,18 The reactive Na@HC therebyaccelerates the decomposition of solvents forming unstablepassivation layers, resulting in degradation of cell performance.Recently, optimization studies of cell performance using HChave been carried out.19 Komaba20 et al. improved the electro-chemical performance of cells comprised of hard carbon and

Energy Environ. Sci., 2013, 6, 2067–2081 | 2069

Fig. 5 Electrochemical performance of HCNWs: (A) CV curve, (B) voltage profile,(C) cycleability, and (D) rate capability. Reproduced from ref. 22 with permission.

Energy & Environmental Science Perspective

Publ

ishe

d on

20

May

201

3. D

ownl

oade

d by

EC

OL

E P

OL

YT

EC

HN

IC F

ED

DE

LA

USA

NN

E o

n 06

/11/

2015

12:

29:4

0.

View Article Online

NaClO4 in an EC:DEC including relatively low polarization andstable cycle performance with negligible capacity fading over100 cycles (Fig. 4).

Also, hierarchical porosity design can provide a promisingsolution to improve the solid state diffusion kinetics of Na ionsin electrode materials. Through the nanocasting method,Wenzel et al. synthesized templated carbon using porous silicaand mesophase pitch.21 Aer inltration of a carbon precursorinto the silica pores, the reaction mixture was carbonized, andthen the remaining silica architecture was removed by HFtreatment. Because of interconnections of dened macro-and meso-pores, templated carbon allows fast kinetics byreducing the diffusion lengths. The specic capacity of tem-plated carbon exceeded 100 mA h g�1 at relatively high currentrates (2 C or 5 C). However, the large surface area of the porousstructure causes a signicant amount of irreversible capacity viasolid electrolyte interface (SEI) formation, and the correspond-ing poor coulombic efficiency of the 1st cycle is detrimental inpractical applications.

Cao et al. reported high performance anode materials byapplying hollow carbon nanowires (HCNWs).22 The nano-structure was prepared by pyrolysis of a polyaniline precursorhaving a hollow nanowire morphology via self-assembly. Adisordered nanostructure was conrmed by X-ray diffraction(XRD) measurements and high resolution transmission elec-tron microscopy (HRTEM) observation. From the full width athalf-maximum (FWHM) of the (002) diffraction peak, the c-axislength was calculated to be 1.27 nm, which corresponds to 3–4graphene layers. Fig. 5 shows the electrochemical performance

Fig. 4 Electrochemical performance of a hard carbon anode. (A) Voltage profilesand (B) cycleability with (a) ethylene carbonate (EC), (b) propylene carbonate (PC),and (c) butylene carbonate (BC) solutions. Reproduced from ref. 20, withpermission.

2070 | Energy Environ. Sci., 2013, 6, 2067–2081

of HCNWs having a reversible specic capacity of 251 mA h g�1

and stable cycleability (82% retention aer 400 cycles).However, they exhibited relatively low initial coulombic effi-ciency (�50%). In research on carbon structures/architecture,various carbon precursors have been examined for theproduction of carbonaceous anode materials. For example,glucose,11,15,23 ork,17,18 pitch,24,25 carbon bers,25,26 petroleumcokes,27 and carbon black28 were used to prepare nanoporous,disordered or hard carbons.

2.2 Alloy materials

To address the issue of relatively low specic capacity ofcarbonaceous materials, alloy materials have been widelystudied in LIBs due to their high theoretical specic capacity(>500 mA h g�1). Through the electrochemical intercalationmechanism of Li ions in graphite, at least six carbons arerequired to uptake a single Li, and this corresponds to 372 mA hg�1 of specic capacity. However, alloy materials such as Sn andSi can accommodate approx. 4 Li offering high theoreticalcapacity, resulting in providing signicant improvement ofspecic capacities of 994 mA h g�1 and 4212 mA h g�1,respectively.

Alloy materials have also been examined to improve thespecic capacity for NIBs, and similar or dissimilar behavioursto LIBs were observed. Unlike the high lithium atomicconcentration in Li–Si binary alloys, the theoretical maximumsodium concentration in a Na–Si system is only 50 at% (Fig. 6a).Experimentally, a Na–Si phase diagram is not available, andonly a few Na–Si binary phases are documented.29 In addition,the theoretical redox potential of Na–Si alloy reaction is lower(<0.1 V vs. Na/Na+) than that of Li–Si (ca. 0.3 V vs. Li/Li+). Thisredox potential close to 0 V vs. Na/Na+ can cause the electro-chemical insertion of Na ions into Si being kinetically inhibitedby large polarization. Accordingly, no report has shown thereversible de/sodiation of Si-based materials yet, and moststudies on alloy anodes in NIBs have focused on Sn or Sb-basedmaterials rather than Si-based materials.

This journal is ª The Royal Society of Chemistry 2013

Fig. 6 Theoretical equilibrium redox potentials of (A) Na–Si, (B) Na–Ge, (C)Na–Sn, and (D) Na–Pb. Reproduced from ref. 30 with permission.

Perspective Energy & Environmental Science

Publ

ishe

d on

20

May

201

3. D

ownl

oade

d by

EC

OL

E P

OL

YT

EC

HN

IC F

ED

DE

LA

USA

NN

E o

n 06

/11/

2015

12:

29:4

0.

View Article Online

It was calculated based on DFT that the electrochemicalsodiation of Sn proceeds to form Na15Sn4 through several phasetransition including NaSn5, NaSn, and Na9Sn4 phases(Fig. 6).30,31 However, Wang et al. recently reported differentphase transitions via in situ de/sodiation TEM analysis of tinnanoparticles,32 inspired by the similar studies on LIBsystems.33 The sodiation of Sn nanoparticles (d ¼ 80–400 nm)proceeds in a two-phase reaction to from amorphous NaSn2

alloys (56% volume expansion), and it is further sodiated tocrystalline Na15Sn4 via a one-phase reaction (ca. 420% volumeexpansion).

The de/insertion of alkali ions from elemental Sn and Sihosts introduces detrimental mechanical strain caused by largevolume change during sodiation and desodiation (Table 1). Thisleads to cracking and pulverization of the alloy materials anddeformation of electrodes, resulting in severe capacity fadingwithin a few cycles. To keep the integrity of alloy materials,various approaches to use intermetallics and/or composites(active/inactive, active/active, active/(inactive or active)/carboncomponents) have been exploited for LIBs.34 Inactive compo-nents having a role of a buffering matrix are effective to main-tain the electrode microstructure by preventing aggregation orcoarsening of active elements. The inactive components arerequired to have several properties including immiscibility withalkali metals, high mechanical strength, small volume changeduring electrochemical reactions, and good electrical conduc-tivity. Various Sn-based active/inactive intermetallics includingSn–Cu, Sn–Mg, and Sn–Mn–C have been investigated,35–38 andthe other metallic elements (M ¼ Ni, Co, Fe, etc.) can be also

Table 1 Volume expansion in an AM system. Adapted from ref. 30 and 31

A–M systemA ¼ Li or Na

AxMy phaseat full state of

Volume expansion/%AxMy to M

Li–Sn Li15Sn4 260Li–Si Li15Si4 281Na–Sn Na15Sn4 424Na–Si NaSi 144

This journal is ª The Royal Society of Chemistry 2013

utilized as inactive buffering matrix elements. The carbona-ceous materials for the composite include graphite, porouscarbon, carbon nanotubes, and amorphous carbon.34

It is expected that a similar approach will be effective toenhance the electrochemical performance of alloy materials inNIBs, although the intermetallic or composite materials havinginactive components are still in the early stage for anodematerials of NIBs. Recently, Wang and co-workers demon-strated the electrochemical performance of a porous carbon-tincomposite.39 The composite was prepared by the carbonizationof dispersed tin oxide nanoparticles in a polymer matrix underan Ar atmosphere, and HRTEM and XRD measurements indi-cated complete conversion of tin oxide to crystalline tin metal.The electrochemical performance of the composite was exam-ined for both NIB and LIB applications (Fig. 7). The experi-mental redox potential values of de/sodiation were close to thetheoretical values based on the rst principle calculation, andfour anodic plateaus at 0.2, 0.3, 0.56, and 0.7 V were consideredto be related to the desodiation of Na15Sn4, Na9Sn4, NaSn, andNaSn5, respectively. The redox potentials of several Na–Snphases were lower by a few hundred mV than those of Li–Snbinary phases. When considering different redox potentialsbetween reference electrodes of Li/Li+ and Na/Na+ (ca. 0.3 V vs.Li/Li+), it is considered that the thermodynamic properties forthe formation of each phase in LIBs and NIBs are similar.Compared to the reversible capacity of the initial delithiation

Fig. 7 Voltage profiles of the porous C–Sn composite for (A) NIB and (B) LIBapplications. Reproduced from ref. 39 with permission.

Energy Environ. Sci., 2013, 6, 2067–2081 | 2071

Energy & Environmental Science Perspective

Publ

ishe

d on

20

May

201

3. D

ownl

oade

d by

EC

OL

E P

OL

YT

EC

HN

IC F

ED

DE

LA

USA

NN

E o

n 06

/11/

2015

12:

29:4

0.

View Article Online

(574 mA h g�1) for the LIB application, this porous carbon–tincomposite delivered smaller initial specic capacity (295 mA hg�1) and showed poor rate performance. Electrochemicalimpedance spectroscopy (EIS) revealed that Sn–C composites inNIBs experience higher mass-transfer resistance with largercharge-transfer impedance than those in LIBs, which areattributed to slower diffusivity of Na+ ions and larger SEIimpedance in NIBs, respectively. These larger impedances inNIBs result in poorer rate performance of the Sn–C compositefor NIBs.

As an active/active element approach, SnSb has gainedtremendous interests due to higher reversible capacity thanactive/inactive approaches and good reversibility to form SnSbduring delithiation. SnSb is lithiated sequentially with thefollowing distinctive conversion reactions forming Li–Sn andLi–Sb phases at each redox potential.

SnSb + 3Li++ 3e� 5 Li3Sb + Sn (1)

Sn + 3.75Li++ 3.75e� 5 0.25Li15Sn4 (2)

During delithiation, the resulting product is reversiblyrecovered into SnSb. SnSb–C composites for LIBs showedexcellent electrochemical performance with a reversiblecapacity of about 550 mA h g�1 over 300 cycles at 100 mA g�1

with 81% initial coulombic efficiency.40 The Liu group recentlydemonstrated that the SnSb–C composite for NIBs exhibitscomparable electrochemical performances to LIBs. The SnSb–Ccomposite was prepared by high-energy milling of SnSb nano-particles with carbon black at a 7 : 3 weight ratio,41 and the highreversible capacity of 544 mA h g�1 was delivered withgood cycleability (ca. 80% capacity retention aer 50 cyclesat 100 mA g�1) for NIBs.

The Sb element itself is also a promising alloy material forNIBs. Since rhombohedral Sb has a lower density than Sn due toa low atomic packing factor (�39%), Sb has more space than Snto accommodate guest alkali elements, resulting in smallervolume change during de/lithiation and de/sodiation.34 Inaddition, alkali-antimony alloys have different crystal structuresdepending on the alkali metal. While fully lithiated Li3Sb alloyshave a cubic crystal structure, the corresponding sodiatedNa3Sb alloy showed a hexagonal crystal structure.

Recently, Monconduit et al. demonstrated that Sb showspromising electrochemical performance including excellentcapacity retention up to 160 cycles with a high reversiblecapacity of about 600 mA h g�1, in spite of that micro-sized bulkSb particles were used.42 Sb surprisingly showed better cycleperformance for NIBs compared to LIBs. This is attributed tosmaller volume change during charge and discharge for NIBs.The volume of Sb, hexagonal Na3Sb, and rock salt Li3Sb is 181.1,237, and 283.8 A3, respectively. Also, the de/sodiation of Sbshowed a different reaction mechanism from the de/lithiationof Sb. Upon sodiation, Sb transforms hexagonal Na3Sb viaamorphous NaxSb, while Sb is directly changed into cubic Li3Sbon lithiation. Independently, Qian et al. also introduced a Sb–Ccomposite as an anode for NIBs, and the composite wassynthesized by simple mechanical milling of a commercial

2072 | Energy Environ. Sci., 2013, 6, 2067–2081

antimony powder with super P carbon.43 The compositereversibly delivered three sodium atoms into the hostcomposite (corresponding to a specic capacity of 640 mA hg�1) and exhibited good cycle performance (94% capacityretention over 100 cycles). Moreover, the electrochemicalperformance of the Sb–C composite for NIBs was improved dueto the addition of 5% uoroethylene carbonate (FEC) as anelectrolyte additive. Electrochemical impedance spectroscopicanalysis revealed that the FEC additive is responsible for a morestable SEI lm providing a smaller charge-transfer resistance,which results in the better electrochemical stability comparedto the FEC-free sample.

So far, only a few alloy materials have been examined as ananode for NIBs, and thus more various alloy materials arerequired to be revisited based on the library of Li-alloy mate-rials. Also, the improvement of electrochemical performance ofalloy materials for NIBs could be achievable through the use ofoptimized binders and electrolyte additives in LIBs.

2.3 Metal oxide and sulde materials

Transition metal oxides (TMOs) with various oxidation statescan be reversibly reduced and oxidized either by insertion orconversion routes with the following general eqn (3) and (4).Because of the electrochemical inactivity of bulk sized alkalimetal oxides (A2O, such as Li2O and Na2O), the conversionreaction of TMOs had been disregarded for rechargeablebatteries. In 2000, Poizot et al. reported unexpected electro-chemical performance of nanosized transition metal oxidesincluding CoO, FeO, and NiO for LIBs.44 The late 3d transitionmetal oxides showed a high specic capacity of ca. 700 mA h g�1

with excellent capacity retention (Fig. 8). The enhanced reac-tivity of Li2O is attributed to the nanosize connement effect.Aer this ground-breaking report, the conversion reaction inLIBs has been widely studied and the chemistry has beenextended to N, S, and P to replace oxides.44

(Insertion route)

MaOb + cA+ + ce� 5 AcMaOb (3)

(Conversion route)

MaOb + 2bA+ + 2be� 5 aM + bA2O (4)

M ¼ transition metal and A ¼ alkali metal.The rst transition metal oxide for NIB anodes was reported

by Alcantara et al. using spinel NiCo2O4, prepared from theprecipitation of a mixed oxalate precursor, followed by thermaldecomposition in air at 320 �C.27 The spinel oxide exhibitedpoor electrochemical performances with large polarization andsmall reversible capacity (ca. 200 mA h g�1) considering its hightheoretical specic capacity of ca. 890 mA h g�1. Early 3d tran-sition metal (Ti, V) oxides also have been utilized for anodematerials of NIBs via an insertion-type mechanism. Forexample, up to 0.5 sodium reversibly inserted into a layeredNaxVO2 host at �1.5 V vs. Na/Na+, which corresponds to 120 mAh g�1. The layered NaTiO2 can accommodate about 0.5 Naatoms per formula unit at �1 V. The relatively low redox

This journal is ª The Royal Society of Chemistry 2013

Fig. 8 Electrochemical performance of nanosized transition metal oxides: (A)voltage profiles and (B) cycle performance. Reprinted from ref. 44 withpermission.

Fig. 9 General redox processes of organic compounds including thioethers,disulfides, nitroxide radicals, and carbonyl compounds via electrochemicalmethods. Adapted from ref. 52 and 53.

Perspective Energy & Environmental Science

Publ

ishe

d on

20

May

201

3. D

ownl

oade

d by

EC

OL

E P

OL

YT

EC

HN

IC F

ED

DE

LA

USA

NN

E o

n 06

/11/

2015

12:

29:4

0.

View Article Online

potential of titanium oxides compared to other transitionmetals was ascribed to the high reducing activity of trivalenttitanium cations.45 To enhance the kinetics of sodium iontransport, amorphous TiO2 with a nanotubular structure wasexamined.46 Sodium-ion transport was dependent on theaverage diameter of nanotubes (NTs). While narrow NTsallowed poor capacity, wider NTs (>80 nm, inner diameter)showed improved capacity of 150 mA h g�1 at �1.5 V. Recently,the Palacın group reported that Na2Ti3O7 reversibly accommo-dates about two sodium atoms at about 0.3 V.47 Titanium oxidesare considered promising anode candidates owing to their lowcost, non-toxicity, and relatively low redox potential.

In the 1970s, Whittingham demonstrated that transitionmetal suldes such as TiS2 can be used as host materials ofsodium ions.48 The Ceder group explained alkali metal insertioninto sulde host materials with enthalpy values.9 The differenceof formation enthalpy values of Li2S (�466 kJ mol�1) and Na2S(�336 kJ mol�1) is relatively smaller than those of Li2O (�599 kJmol�1) and Na2O (�418 kJ mol�1). The larger orbital size ofsodium-ions compared to lithium ions may allow efficientoverlap with soer sulfur orbitals than harder oxygen. Transi-tion metal suldes including FeS2 (pyrite) and Ni3S2 (heazle-woodite) showed low potential ranges during the conversionreactions with sodium. Pyrite showed a high initial dischargecapacity of 630 mA h g�1 at �1.3 V.49 However, the specic

This journal is ª The Royal Society of Chemistry 2013

capacity dramatically decreased within 50 cycles. In the case ofheazlewoodite, the initial discharge capacity was found to be420 mA h g�1 with an average redox potential of �0.9 V. Thecapacity was retained at 81% over 15 cycles.50

2.4 Organic compound-based materials

For the past decade, major electrode materials for rechargeablebatteries have been inorganic compounds synthesized throughsolid state reaction, the sol–gel method, or simple mechanicalmilling. Despite technological advances in secondary batteriesusing inorganic compounds, high energy costs for the ceramicprocesses and the demand for electrode materials preparedfrom sustainable sources have raised tremendous concerns.3,4

Organic compounds offer alternative solutions thanks to theirlow-cost production, recyclability, and structural diversity.51–53

For example, organic sulfurs, disuldes, nitroxide radicals, andconjugated carbonyl derivatives have been utilized for thepreparation of electrode materials of LIBs (Fig. 9 and 10).52,53

However, small volumetric capacity, low cycleability, andpoor electrical conductivity have been major challenges oforganic batteries. Tarascon and coworkers provided a solutionto overcome those problems by designing conjugated carboxy-late salts.54 Dilithium terephthalate was crystallized in a spacegroup of P21/c through pi stacking and ionic interaction,thereby reducing the volume. The conjugated system betweenphenyl and carboxylate groups can help ionic and electronicconductivities of the electrode materials, while they still requirea relatively large amount of conducting carbon materials.Recently, the rst organic anode materials for NIBs wereintroduced based on a disodium terephthalate (Na2TP) struc-ture with a space group of Pbc21.55–57 Remarkably, the sodiumsalt showed excellent electrochemical performances (Fig. 11).55

Na2TP reversibly delivered about 250 mA h g�1 at about 0.4 V vs.Na/Na+. It showed high capacity retention over 90 cycles withexcellent (>98%) coulombic efficiency.

It is not easy to develop new inorganic electrodematerials forreversible de/sodiation, because of the limited number of

Energy Environ. Sci., 2013, 6, 2067–2081 | 2073

Fig. 10 Electrochemical performance of organic compounds for LIBs: (A)average discharge potential vs. discharge capacity plot and (B) power density vs.energy density plot. Reproduced from ref. 53 with permission.

Energy & Environmental Science Perspective

Publ

ishe

d on

20

May

201

3. D

ownl

oade

d by

EC

OL

E P

OL

YT

EC

HN

IC F

ED

DE

LA

USA

NN

E o

n 06

/11/

2015

12:

29:4

0.

View Article Online

crystal structures and compositions. However, the structuralrichness of organic compounds can easily provide a new line-upof electrode materials. For example, Na2TP derivatives withamino- and bromo-groups have exhibited controllable potentialchanges via inductive and resonance effects, and nitro-Na2TP

Fig. 11 Proposed reversible Na-ion insertion/deinsertion mechanism andvoltage profiles. Reprinted from ref. 55 with permission.

2074 | Energy Environ. Sci., 2013, 6, 2067–2081

showed the increase of reversible capacity through an addi-tional function of nitro group as a redox couple.

2.5 Phosphorus-based materials

Although NIBs are considered a next generation system andhave shown promising electrochemical performance includingstable cycleability and good rate capability, they still possess acritical problem in energy density: their energy density is infe-rior to that of LIBs. Most cathode and anode materials that havebeen introduced to date for NIBs show similar or slightly lowerspecic capacity and redox potential than those of LIBs, becauseit is difficult to increase the specic capacity due to limitedstorage sites and the oxidation state based on the intercalationchemistry.7,9,58 Therefore, in order to improve the energy densityof Na ion batteries, new electrode materials having a highspecic capacity should be introduced.30 Thus far, however,there have been few studies on sodium insertion materials withhigh capacity.15,20,46,47,59–62 The general strategy for storing a largenumber of ions is to use conversion chemistry such as alloyingmaterials.

Crystalline black and amorphous black phosphorus wereintroduced as promising anode materials in LIBs, because redphosphorus is known to be electrochemically inactive and blackphosphorus shows excellent reversibility relative to red phos-phorus.63–67 However, Marino et al.68 recently showed reversiblede/insertion of Li ions into crystalline red phosphorus viathe preparation of a red phosphorus/mesoporous carboncomposite by vaporization and condensation. Inspired by thereversible de/insertion of crystalline red phosphorus for Li ions,Kim et al. demonstrated excellent electrochemical performanceof amorphous red phosphorus as an anode material for NIBs.69

The amorphous red phosphorus–carbon composite powderswere synthesized through a facile and simple ball millingprocess using commercially available amorphous red phos-phorus and Super P carbon. This material exhibited a highreversible capacity of 1890 mA h g�1; this value is known to bethe highest value among reported electrode materials for Na ionbatteries (Fig. 12). Also, very stable cycle performance of negli-gible capacity fading over 30 cycles and good rate capabilitydelivering 1540 mA h g�1 at a current density of 2.86 A g�1 wereobserved. Above all, the phosphorus materials showed anappropriate redox potential of ca. 0.4 V vs.Na/Na+, which is idealfor anode applications from the view-points of safety and energydensity. In the case of LIB application, the redox potential ofphosphorus was relatively high (ca. 0.8 V vs. Li/Li+); this is notappropriate for an anode because the high redox potential ofthe anode causes loss of energy density of full cells. They alsosuggested that the electrochemical performance of amorphousred phosphorus is strongly dependent on the electricalconductivity and the volume change during sodiation anddesodiation. Accordingly, a phosphorus/carbon composite anda polyacrylic acid (PAA) binder were used to improve electricalconductivity and electrode degradation, respectively. PAA orcarboxymethyl cellulose (CMC) binders are expected to improvethe cycle performance of phosphorus-based materials due tothermally cross-linked three-dimensional interconnection of

This journal is ª The Royal Society of Chemistry 2013

Fig. 12 Voltage profiles of the red phosphorus/carbon composite. Reprintedfrom ref. 69 with permission.

Perspective Energy & Environmental Science

Publ

ishe

d on

20

May

201

3. D

ownl

oade

d by

EC

OL

E P

OL

YT

EC

HN

IC F

ED

DE

LA

USA

NN

E o

n 06

/11/

2015

12:

29:4

0.

View Article Online

those binders, as found to be effective in Si-based materialsundergoing a large volume change during lithiation anddelithiation.

Fig. 13 Electrochemical characteristics of (A) LiCrO2 and (B) NaCrO2. Repro-duced from ref. 71 with permission.

3 Positive electrode materials3.1 Layered oxide materials

Although NIBs are being considered as an alternative to replaceLIBs, research on both sodium-ion and lithium-ion intercala-tion chemistry began in the 1980s. For examples, NaxCoO2,NaxNiO2, NaCrO2, and a-NaFeO2 were examined as intercala-tion hosts of Na ions via chemical sodiation and desodiation.70

Recently, the Komaba group considered sodium chromate(NaCrO2) as a cathode for NIBs and compared it to lithiumchromate (LiCrO2).71 As shown in Fig. 13, as LiCrO2 was chargedup to 4.5 V vs. Li/Li+, it delivered a reversible capacity of only10 mA h g�1, corresponding to de/insertion of ca. 0.03 Li ions.However, it is noteworthy that NaCrO2 reversibly deliveredabout 120 mA h g�1 (0.5 mol Na ions in NaCrO2), indicating thatLiCrO2 and NaCrO2 are electrochemically inactive and active,respectively. This means that it is possible to nd new break-throughmaterials for NIBs frommaterials that might be uselessin LIBs. The electrochemical inactivity of LiCrO2 is attributed tothe irreversible migration of disproportioned Cr6+ into inter-stitial tetrahedron in LiCrO2 due to their similar size. However,such migration is inhibited in NaCrO2 because of the sizemismatch.72

Similar behavior to ACrO2 (A ¼ Li, Na) was observed inP2-Na2/3[Fe1/2Mn1/2]O2.73 Recently, the Komaba group demon-strated that partially substituted Mn for Fe in P2- and O3-typeNaxFeO2 frameworks can stabilize Fe4+ in the structure, result-ing in an increase of reversible capacity due to Fe3+/4+ redoxcouples. These Fe3+/4+ redox couples are known to be electro-chemically inactive due to the highly unstable Fe4+ in O3-type

This journal is ª The Royal Society of Chemistry 2013

LiFeO2 for LIBs.74 P2-type Na2/3[Fe1/2Mn1/2]O2 has a hexagonalstructure (space group: P63/mmc), which is isostructured withP2-NaxCoO2. O3-type Na[Fe1/2Mn1/2]O2 has a rhombohedralstructure (space group: R�3m), which is the same as O3-NaFeO2.Fig. 14 shows the voltage proles of the P2 and O3 structures.The reversible capacity of the P2 type structure is 190 mA h g�1,while the O3 structure delivers ca. 110 mA h g�1 with largepolarization. Fig. 15 shows that the energy density of P2-Na2/3[Fe1/2Mn1/2]O2 reaches 520 mW h g�1, which is similar tothat of LiFePO4 (ca. 530 mW h g�1) and higher than that of aLiMn2O4 spinel (ca. 450 mW h g�1).

NaCoO2 is another notable cathode material for NIBs. In1981, Delmas et al. compared the electrochemical characteris-tics of P2-NaxCoO2 (space group: P63/mmc) and O3-NaxCoO2

(space group: R�3m) as intercalation hosts of Na ions.75 Sodiumions reside in octahedral and prismatic sites of O3 and P2structures, respectively. The de/sodiation of the O3 structureinvolves a phase transition between P3 and O3 structures due totheir structural similarity. However, the P2 type structure ismaintained during sodiation and desodiation, which is bene-cial for structural stability during cycling. This is attributed tothe different structure of the P2 type from the P3 and O3structures, and the phase transitions between P2 and P3 or O3require Co–O bond cleavages. Fig. 16 shows the diffusivity of P2-NaxCoO2 for various compositions, which are measured usingP2-NaxCoO2 single crystals. The diffusivity of P2-NaxCoO2 was

Energy Environ. Sci., 2013, 6, 2067–2081 | 2075

Fig. 14 Galvanostatic charge/discharge curves of (A) O3-Na[Fe1/2Mn1/2]O2 and(B) P2-Na2/3[Fe1/2Mn1/2]O2. Reproduced from ref. 73 with permission.

Fig. 15 Reversible capacity vs. operating voltage ranges of the layered alkaliinsertion materials. Reproduced from ref. 73 with permission.

Fig. 16 Compositional variation of (A) inter-diffusion coefficient of Na, (B)thermodynamic factor of Na, and (C) self-diffusion coefficient of Na obtained bythe PITT technique at 300 K. Reproduced from ref. 76 with permission.

Energy & Environmental Science Perspective

Publ

ishe

d on

20

May

201

3. D

ownl

oade

d by

EC

OL

E P

OL

YT

EC

HN

IC F

ED

DE

LA

USA

NN

E o

n 06

/11/

2015

12:

29:4

0.

View Article Online

10�7 to 10�8 cm2 s�1,76 and this is not lower than that ofO3-LiCoO2 (10

�7 to 10�13 cm2 s�1) in spite of the larger size ofNa ions. This indicates that similar power density of NIBs tothat of LIBs can be obtained using layered metal oxidematerials.

O3 structure materials including LiCoO2 and Li(Ni,Co,Mn)O2 have been successfully commercialized in LIBs due totheir structural stability and excellent electrochemical perfor-mance. However, O3 structure materials are not very promisingin NIBs because of their structural instability under air.O3-NaCo1/3Ni1/3Mn1/3O2 was found to be decomposed intoNa1�xCo1/3Ni1/3Mn1/3O2 with NaOH or Na2CO3 via reaction withcarbon dioxide and water in air.77 O3-NaCo1/3Ni1/3Mn1/3O2

exhibited a relatively low reversible capacity of about120mA h g�1 (0.5 Na+) in a range of 2–3.8 V vs.Na/Na+, while O3-LiCo1/3Ni1/3Mn1/3O2 delivered about 150 mA h g�1 in a range ofca. 4 V vs. Li/Li+.78 In contrast, another solid solution analogue,O3-NaNi0.5Mn0.5O2 showed better electrochemical performance

2076 | Energy Environ. Sci., 2013, 6, 2067–2081

than O3-LiNi0.5Mn0.5O2. O3-NaNi0.5Mn0.5O2 delivered 185 mA hg�1 in a range of 2.5–4.5 V vs. Na/Na+ with stable cycle perfor-mance, and showed good rate performance, delivering 100 mAh g�1 even at 1 C (¼240 mA g�1). However, high reversiblecapacity (200 mA h g�1) of LiNi0.5Mn0.5O2 is achieved at only lowcurrent densities (0.17 mA cm�2).79 This poor rate performanceof LiNi0.5Mn0.5O2 is attributed to LiNi0.5Mn0.5O2 having 8–10%cation disorder between the Ni2+ ion in the 3a site and the Li+

ion in the 3b site due to the similar size of Ni2+ and Li+.80 Li iondiffusivity in layered transition metal oxides depends upon thecation disorder.81,82 Transition metal (M) ions in Li slabs bycation disorder disrupt the lithium ion diffusion path andincrease electrostatic repulsion between M and Li, resultingin reduced Li mobility. Also, the octahedral Li ion inLiNi0.5Mn0.5O2 is diffused through the tetrahedral site (octa-hedral–tetrahedral–octahedral), and the size of the tetrahedralsite is determined by the c-lattice parameter (Li slab space).Accordingly, cation disorder reduces the Li slab space, resultingin a higher activation barrier in lithium diffusion. However,NaNi0.5Mn0.5O2 has a negligible amount of cation disorder(0.4%) because of the larger size difference between Na+ (rNa+ ¼1.02 A) and Ni2+ (rNi2+ ¼ 0.69 A) or Mn4+ (rMn4+ ¼ 0.53 A),resulting in good rate performance.20,71 Also, the valencestate of only Ni in Li1�xNi0.5Mn0.5O2 is changed between 2+ and4+ in a range of 0 < x < 1 (ca. 4 V vs. Li/Li+) and the valence stateof Mn remains unchanged in Mn4+.83 The same behavior wasobserved in Na1�xNi0.5Mn0.5O2. The valence state of only Niis changed between 2+ and 4+ in a range of 0 < x < 1.84 Thevalence state of Mn in Na1�xNi0.5Mn0.5O2 is always 4+ regardlessof charge and discharge, and this inhibits Jahn–Teller

This journal is ª The Royal Society of Chemistry 2013

Perspective Energy & Environmental Science

Publ

ishe

d on

20

May

201

3. D

ownl

oade

d by

EC

OL

E P

OL

YT

EC

HN

IC F

ED

DE

LA

USA

NN

E o

n 06

/11/

2015

12:

29:4

0.

View Article Online

distortion of NaNi0.5Mn0.5O2 caused by unstable Mn3+. A similarreaction mechanism including the valence state and cationdisorder to O3-NaNi0.5Mn0.5O2 was also observed in P2-Na0.85[Li0.17Ni0.21Mn0.64]O2.85

3.2 Phosphate-based materials

Aer the introduction of LiFePO4 by Goodenough et al.,86

various phosphate based materials have been extensivelystudied as cathode materials in LIBs. In particular, phosphatematerials have been focused upon due to the increase of redoxpotential and good safety caused by an inductive effect and astrong oxygen–phosphorus covalent bond, respectively.However, phosphate materials have critical disadvantages ofpoor electrical and ionic conductivity. Therefore, nanosizedmaterials have been considered as a solution to overcome theseshortcomings by decreasing the solid state diffusion length of Liions, and this approach improved the electrochemical perfor-mance. LiFePO4 crystallizes in the space group of Pnma. Layersof FeO6 octahedra are corner-shared in the bc plane, and linearchains of LiO6 octahedra are edge-shared in a direction parallelto the b-axis. These chains are bridged by edge and cornershared phosphate tetrahedra, creating a stable three-dimen-sional structure. The reversible capacity and redox potential ofLiFePO4 are about 170mA h g�1 and 3.4 V vs. Li/Li+, respectively.However, NaFePO4 cannot be prepared directly as the olivinephase because the thermodynamically stable form of NaFePO4

is the mineral maricite. The M1 and M2 sites in the LiFePO4

olivine structure are occupied by Li+ and Fe2+ respectively.However, NaFePO4 maricite is exactly the reverse of LiFePO4, inwhich Na+ and Fe2+ occupy the M2 and M1 sites of the olivinestructure (Fig. 17). This gives rise to a different connectivity of

Fig. 17 Crystal structures of (A) LiFePO4 olivine and (B) NaFePO4 maricite with[010] direction projection. Reprinted from ref. 91 with permission.

This journal is ª The Royal Society of Chemistry 2013

the Fe and Na octahedra which blocks Na-ion migration path-ways, resulting in a structure that is not amenable to Na+ (de)insertion. NaFePO4 forms as an electrochemically inactivemaricite phase under conventional synthetic conditions at hightemperature, and accordingly, NaFePO4 olivine has beenobtained via electrochemical insertion of Na into FePO4.87–90

However, recently, metastable olivine phases of sodiummetal phosphates nanorods, Na[Mn1�xMx]PO4 (M ¼ Fe; Ca;Mg), were synthesized by a simple solid state reaction at lowtemperatures (#100 �C), which is attributed to a topotacticreaction that converts NH4[Mn1�xMx]PO4$H2O (M¼ Fe; Ca; Mg)(Pmn21 space group) to Na[Mn1�xMx]PO4 (Pnma space group)via direct ion exchange between NH4

+ and Na+ using moltenCH3CO2Na$3H2O.91 The connectivity of the iron and phosphatepolyhedra in the (100) plane of NaMPO4 is identical to that inthe corresponding (101) plane of NH4MPO4$H2O, and uponrapid ion exchange of NH4

+ for Na+, the adjacent sheets areknitted together by condensation of the NaO6 octahedra tocrystallize olivine NaMPO4 (Fig. 18). This synthetic method isbenecial to obtain olivine phases with negligible cationdisorder even at low synthetic temperatures, compared toconventional low temperature synthesis such as the hydro-thermal method which yields 8–10% cation disorder. Thestructure of NH4MPO4$H2O precursors is perfectly ordered dueto the bulky NH4 ions, and accordingly this order is replicated inthe structure of the resulting products. This is attributed to thesynthetic temperature being too low to induce atomic positionrearrangements such as switching of the Fe/Na sites. Also, thevolume difference of Na(Fe0.5Mn0.5)PO4 between Na-rich andNa-poor phases was 21%, which is much larger than the cor-responding volume difference for LiFePO4 (6.7%) due to thelarger size of Na ions, and this causes poor electrochemicalperformance.

Also, the different reaction mechanism of NaFePO4 wasreported from that of LiFePO4. While LiFePO4 is changed intoFePO4 via one step of a two-phase reaction, two steps of a two-phase reaction through the formation of a stable intermediatephase, Na0.7FePO4, was observed during electrochemical de/sodiation.87 Accordingly, unlike one plateau of the Li/FePO4

system, the Na/FePO4 system shows two plateaus in voltageproles. Na0.7FePO4 has a Pnma orthorhombic unit cell having

Fig. 18 Scheme of topological growth of NaMPO4. Reprinted from ref. 91 withpermission.

Energy Environ. Sci., 2013, 6, 2067–2081 | 2077

Fig. 19 Crystal structures of (A) Na3.12Fe2.44(P2O7)2 and (B) Na2FeP2O7 with[100] and [011] direction projection, respectively. Reproduced from ref. 96 and 97with permissions.

Energy & Environmental Science Perspective

Publ

ishe

d on

20

May

201

3. D

ownl

oade

d by

EC

OL

E P

OL

YT

EC

HN

IC F

ED

DE

LA

USA

NN

E o

n 06

/11/

2015

12:

29:4

0.

View Article Online

parameters intermediate between those of NaFePO4 andLiFePO4: a ¼ 10.2886(7) A, b ¼ 6.0822(4) A, c ¼ 4.9372(4) A, andV ¼ 308.95(4) A3.

In 2010, the Yamada group introduced a new pyrophosphatematerial, Li2FeP2O7, as a cathode for LIBs synthesized via solidstate reaction.92 Li2FeP2O7 has monoclinic symmetry with aspace group of P21/c. The reversible capacity of lithium ironpyrophosphate was 110 mA h g�1, corresponding to one elec-tron reaction with Fe2+/Fe3+. Unlike LiFePO4, Li2FeP2O7 showedgood rate capability without nano-sizing and carbon coatingdue to the fast lithium ion diffusion through a 2D framework inthe bc-plane.93 Also, this pyrophosphate exhibited excellentthermal stability over 500 �C in a charged state.94 Inspired by theexcellent electrochemical performance of Li2FeP2O7, sodiumiron pyrophosphates (Na2FeP2O7) were recently introduced, andshowed good electrochemical performance including a revers-ible capacity of ca. 90 mA h g�1 and good cycleability.95,96 TheBragg reections were indexed with a space group of P1 having atriclinic unit cell. In ex situ XRD patterns, Na2FeP2O7 proceededvia several phase transitions including single- and two-phasereactions at low and high redox potentials respectively. Inde-pendently and simultaneously, a new pyrophosphatecompound, Na3.12M2.44(P2O7)2 (M ¼ Fe, Fe0.5Mn0.5, Mn), wasintroduced as a cathode for NIBs.97 Na3.12Fe2.44(P2O7)2 has atriclinic structure composed of a centrosymmetrical crown ofFe2P4O22 and Fe2P4O20 connected by corner-sharing to form athree-dimensional framework, and each crown unit consists oftwo FeO6 octahedra and two P2O7 groups. The structure ofNa3.12Fe2.44(P2O7)2 is different from those of monoclinicNaFeP2O7 (ref. 98) and triclinic Na2FeP2O7,95,96 although the XRDpattern of Na3.12Fe2.44(P2O7)2 is very similar to that of Na2FeP2O7

(Fig. 19). However, the voltage proles and reversible capacity ofNa3.12Fe2.44(P2O7)2 were observed to be very similar to those ofNa2FeP2O7. Also, a Na-rich phase, Na3.32Fe2.34(P2O7)2 – amemberof the solid solution series Na4�aFe2+a/2(P2O7)2 (2/3 # a # 7/8) –was obtained via off-stoichiometric synthesis, and delivered areversible capacity of about 85 mA h g�1 at ca. 3 V vs.Na/Na+ andexhibited very stable cycle performance. Above all, it showed fastkinetics for Na ions such as Li2FeP2O7, delivering an almostconstant 72% reversible capacity at rates between C/10 and 10 Cwithout the necessity for nanosizing or carbon coating. Recently,Na2CoP2O7 having a layered structure with an orthorhombicframework was also introduced as a cathode for NIBs.Na2CoP2O7 is comprised of parallel slabs of mixed tetrahedra ofCoO4 and PO4, which forms [Co(P2O7)]22 layers parallel to [001]offering a two-dimensional Na-diffusion pathway. Na2CoP2O7

delivered a reversible capacity close to 80 mA h g�1 involving aCo3+/Co2+ redox couple at about 2–4 V vs. Na/Na+.99–101

Also, metal uorophosphates, AMPO4F (A: alkali metal, M:3d transition metal), have been focused upon as promising newcathode materials. In 2007, the Nazar group pioneered a newuorophosphate material, Na2FePO4F, as an intercalation hostfor Li ions. This compound is crystallized in the orthorhombicPbcn space group, and it is isostructured with Na2FePO4OH andNa2Co(PO4)F.102,103 Na2FePO4F showed good electrochemicalperformance including a reversible capacity of ca. 124 mA h g�1

with stable cycle performance because it has facile two-

2078 | Energy Environ. Sci., 2013, 6, 2067–2081

dimensional pathways for Li ion diffusion and small volumechange during de/lithiation (3.7%). LiNaFePO4F and Li2FePO4Fwere also obtained via chemical ion exchange of the sodiumcations for lithium within the lattice. The Tarascon groupexamined Na2FePO4F as a cathode for NIBs.104 Na ion de/intercalation was highly reversible, but Na2FePO4F showedhigher polarization than the LIB application with a reversiblecapacity of ca. 120 mA h g�1. The unit cell volume differencebetween two end members of Na2FePO4F and NaFePO4F wasonly 4%, and thus lower than that of LiFePO4 (6.7%).

In 2010, LiFePO4F tavorite was introduced independently forLIBs by the Tarascon group105 and the Nazar group,106 followingan earlier detailed report of LiVPO4F by Barker et al.107 LiFePO4Fpossesses corner-sharing one dimensional chains of [FeF2O4]octahedra in the [010] direction, with alternating tilted octa-hedra bridged by F� ligands. These chains are connected bycorner-sharing phosphate tetrahedra to create a spacious 3Dframework, which has tunnels (>3 A in diameter) along all of the[100], [010], and [001] directions as open pathways for 3D iontransport. The three-dimensional tunnel structure causes highionic conductivity, resulting in good rate performance,compared to one-dimensional LiMPO4. LiFePO4F tavoriteshowed a reversible capacity of 145 mA h g�1 (96% of theoreticalcapacity) with small polarization and good cycleability at a C/10rate and room temperature, even in a bulk particle form(�1 mm).106 Accordingly, these tavorite structured materials areconsidered promising electrode materials for NIBs due to theappropriately wide and multi-dimensional tunnels of thetavorite structure for the transport of larger Na ions. Also,

This journal is ª The Royal Society of Chemistry 2013

Perspective Energy & Environmental Science

Publ

ishe

d on

20

May

201

3. D

ownl

oade

d by

EC

OL

E P

OL

YT

EC

HN

IC F

ED

DE

LA

USA

NN

E o

n 06

/11/

2015

12:

29:4

0.

View Article Online

other phosphate- and uorophosphate-based polyanion mate-rials have been introduced for NIBs including NaVPO4F,108

Na3V2(PO4)3F3,109 Na3V2(PO4)3,110 and Na4Fe3(PO4)2(P2O7).111

4 Conclusions and perspectives

Na-ion batteries are an attractive alternative to replace Li-ionbatteries for smart grid and electric vehicle applications. Therehave been exciting developments in positive and negative elec-trode materials for Na-ion batteries in recent years. Some ofNIBs have comparable or even higher energy density comparedto the commercialized Li-ion batteries, and Na resources areinexhaustible and of lower cost. However, the electrochemicalperformance of Na-ion batteries must be improved to meet thedemands of next generation applications, although they showpromise. Moreover, Na-ion batteries are facing the same tech-nical barriers including a safety problem as Li-ion batteries,since they are operated based on similar chemistry. To solvethese issues, further in-depth understanding of the reactionmechanism and tailored design of electrode materials for Na-ion batteries must be sought. This perspective has presented acomparative overview of the major developments in the area ofpositive and negative electrode materials in both Li-ion and Na-ion batteries in the past decade, which will provide an oppor-tunity to improve the electrochemical performance of electrodematerials for Na-ion batteries.

Most carbon-based anode materials that have been intro-duced to date for Na ion batteries have shown slightly lowerspecic capacity than the commercialized graphite anode in Li-ion batteries. No matter how they can exhibit stable cycleperformance, however, these carbonaceous anode materialshaving relatively small capacity cannot meet the requirement ofnext generation rechargeable batteries in terms of high energydensity. Therefore, the research on anode materials in Na-ionbatteries is moving on the search for new anode materials withhigh specic capacity. Alloy-, metal oxide- and phosphorus-based materials are considered as promising candidatesbecause they have higher specic capacity than graphite ofLi-ion batteries. In particular, phosphorus showed excellentelectrochemical performance including high reversiblecapacity (1890 mA h g�1) and appropriate redox potential (0.4 Vvs. Na/Na+). However, phosphorus- and alloy-based materialshave large volume change during sodiation and desodiation,and accordingly, tailored design of electrode materials andimproved electrochemical cell engineering are required tomitigate these materials' shortcomings.

Various layered oxide materials including O3-NaCrO2, O3-NaNi0.5Mn0.5O2, O3-NaCo1/3Ni1/3Mn1/3O2, P2-Na2/3[Fe1/2Mn1/2]O2,and P2-Na0.85[Li0.17Ni0.21Mn0.64]O2 were developed based on thedatabase of cathode materials in Li-ion batteries, and theyshowed very promising electrochemical performance for Na-ionbatteries. However, many aspects of ion transport, reactionmechanism, ion ordering and solid-solution behavior in thesematerials are not fully understood, which requires in-depthstudies on these issues to further enhance their electrochemicalperformance. Nonetheless, safety concerns of the layered oxidessuch as Li-ion batteries may prohibit their successful

This journal is ª The Royal Society of Chemistry 2013

application in large-scale energy storage, and phosphate-basedmaterials are desirable to improve safety due to strong P–Obonding, as demonstrated for Li ion batteries. Several phos-phate-based materials including NaMPO4, Na3.12M2.44(P2O7)2,Na2FeP2O7, Na2FePO4F and Na3V2(PO4)3F3 were recently intro-duced, and exhibited promising electrochemical performance.Future research will involve methods to increase reversiblecapacity and rate capability.

Acknowledgements

This work was supported by a grant from the Energy Efficiency &Resources of the Korea Institute of Energy Technology Evalua-tion and Planning (project no. 20112010100140) funded by theKorean Ministry of Knowledge Economy, by the MSIP(Ministryof Science, ICT & Future Planning), Korea, under the C-ITRC-(Convergence Information Technology Research Center)support program (NIPA-2013-H0301-13-1009) supervised by theNIPA(National IT Industry Promotion Agency), and by theNational Research Foundation of Korea (NRF) grant funded bythe Korea Government (MEST) (no. 2011-0027950).

Notes and references

1 M. Armand and J. M. Tarascon, Nature, 2008, 451, 652–657.2 B. L. Ellis, K. T. Lee and L. F. Nazar, Chem. Mater., 2010, 22,691–714.

3 J. M. Tarascon and M. Armand, Nature, 2001, 414, 359–367.4 J. M. Tarascon, Philos. Trans. R. Soc., A, 2010, 368, 3227–3241.

5 H. K. Song, K. T. Lee, M. G. Kim, L. F. Nazar and J. Cho, Adv.Funct. Mater., 2010, 20, 3818–3834.

6 Y. S. Jung, A. S. Cavanagh, A. C. Dillon, M. D. Groner,S. M. George and S. H. Lee, J. Electrochem. Soc., 2010, 157,A75–A81.

7 V. Palomares, P. Serras, I. Villaluenga, K. B. Hueso,J. Carretero-Gonzalez and T. Rojo, Energy Environ. Sci.,2012, 5, 5884–5901.

8 B. L. Ellis and L. F. Nazar, Curr. Opin. Solid State Mater. Sci.,2012, 16, 168–177.

9 S. P. Ong, V. L. Chevrier, G. Hautier, A. Jain, C. Moore,S. Kim, X. H. Ma and G. Ceder, Energy Environ. Sci., 2011,4, 3680–3688.

10 J. O. B. M. Winter, in Handbook of Battery Materials, ed. J. O.Besenhard, Wiley-VCH, Weinheim, 1999.

11 D. A. Stevens and J. R. Dahn, J. Electrochem. Soc., 2001, 148,A803–A811.

12 T. B. Massalski and H. Okamoto, Binary Alloy PhaseDiagrams, ASM Intl., 1990.

13 D. P. DiVincenzo and E. J. Mele, Phys. Rev. B: Condens.Matter Mater. Phys., 1985, 32, 2538–2553.

14 M. M. Doeff, Y. P. Ma, S. J. Visco and L. C. Dejonghe,J. Electrochem. Soc., 1993, 140, L169–L170.

15 D. A. Stevens and J. R. Dahn, J. Electrochem. Soc., 2000, 147,1271–1273.

16 R. Alcantara, P. Lavela, G. F. Ortiz and J. L. Tirado,Electrochem. Solid-State Lett., 2005, 8, A222–A225.

Energy Environ. Sci., 2013, 6, 2067–2081 | 2079

Energy & Environmental Science Perspective

Publ

ishe

d on

20

May

201

3. D

ownl

oade

d by

EC

OL

E P

OL

YT

EC

HN

IC F

ED

DE

LA

USA

NN

E o

n 06

/11/

2015

12:

29:4

0.

View Article Online

17 X. Xia, M. N. Obrovac and J. R. Dahn, Electrochem. Solid-State Lett., 2011, 14, A130–A133.

18 X. Xia and J. R. Dahn, J. Electrochem. Soc., 2012, 159, A515–A519.

19 A. Ponrouch, E. Marchante, M. Courty, J. M. Tarascon andM. R. Palacin, Energy Environ. Sci., 2012, 5, 8572–8583.

20 S. Komaba, W. Murata, T. Ishikawa, N. Yabuuchi, T. Ozeki,T. Nakayama, A. Ogata, K. Gotoh and K. Fujiwara, Adv.Funct. Mater., 2011, 21, 3859–3867.

21 S. Wenzel, T. Hara, J. Janek and P. Adelhelm, EnergyEnviron. Sci., 2011, 4, 3342–3345.

22 Y. L. Cao, L. F. Xiao, M. L. Sushko, W. Wang, B. Schwenzer,J. Xiao, Z. M. Nie, L. V. Saraf, Z. G. Yang and J. Liu, NanoLett., 2012, 12, 3783–3787.

23 D. A. Stevens and J. R. Dahn, J. Electrochem. Soc., 2000, 147,4428–4431.

24 P. Thomas, J. Ghanbaja and D. Billaud, Electrochim. Acta,1999, 45, 423–430.

25 P. Thomas and D. Billaud, Electrochim. Acta, 2001, 46, 3359–3366.

26 P. Thomas and D. Billaud, Electrochim. Acta, 2002, 47, 3303–3307.

27 R. Alcantara, M. Jaraba, P. Lavela and J. L. Tirado, Chem.Mater., 2002, 14, 2847–2848.

28 R. Alcantara, J. M. Jimenez-Mateos, P. Lavela andJ. L. Tirado, Electrochem. Commun., 2001, 3, 639–642.

29 J. Sangster and A. D. Pelton, J. Phase Equilib., 1992, 13, 67–69.

30 V. L. Chevrier and G. Ceder, J. Electrochem. Soc., 2011, 158,A1011–A1014.

31 T. T. Tran andM. N. Obrovac, J. Electrochem. Soc., 2011, 158,A1411–A1416.

32 J. W. Wang, X. H. Liu, S. X. Mao and J. Y. Huang, Nano Lett.,2012, 12, 5897–5902.

33 J. Y. Huang, L. Zhong, C. M. Wang, J. P. Sullivan, W. Xu,L. Q. Zhang, S. X. Mao, N. S. Hudak, X. H. Liu,A. Subramanian, H. Y. Fan, L. A. Qi, A. Kushima and J. Li,Science, 2010, 330, 1515–1520.

34 C. M. Park, J. H. Kim, H. Kim and H. J. Sohn, Chem. Soc.Rev., 2010, 39, 3115–3141.

35 D. Larcher, L. Y. Beaulieu, D. D. MacNeil and J. R. Dahn,J. Electrochem. Soc., 2000, 147, 1658–1662.

36 L. Y. Beaulieu and J. R. Dahn, J. Electrochem. Soc., 2000, 147,3237–3241.

37 H. Kim, Y. J. Kim, D. G. Kim, H. J. Sohn and T. Kang, SolidState Ionics, 2001, 144, 41–49.

38 D. Larcher, A. S. Prakash, J. Saint, M. Morcrette andJ. M. Tarascon, Chem. Mater., 2004, 16, 5502–5511.

39 Y. Xu, Y. Zhu, Y. Liu and C. Wang, Adv. Energy Mater., 2013,3, 128–133.

40 C. M. Park and H. J. Sohn, Electrochim. Acta, 2009, 54, 6367–6373.

41 L. F. Xiao, Y. L. Cao, J. Xiao, W. Wang, L. Kovarik, Z. M. Nieand J. Liu, Chem. Commun., 2012, 48, 3321–3323.

42 A. Darwiche, C. Marino, M. T. Sougrati, B. Fraisse,L. Stievano and L. Monconduit, J. Am. Chem. Soc., 2012,134, 20805–20811.

2080 | Energy Environ. Sci., 2013, 6, 2067–2081

43 J. F. Qian, Y. Chen, L. Wu, Y. L. Cao, X. P. Ai and H. X. Yang,Chem. Commun., 2012, 48, 7070–7072.

44 P. Poizot, S. Laruelle, S. Grugeon, L. Dupont andJ. M. Tarascon, Nature, 2000, 407, 496–499.

45 A. Maazaz, C. Delmas and P. Hagenmuller, J. InclusionPhenom., 1983, 1, 45–51.

46 H. Xiong, M. D. Slater, M. Balasubramanian, C. S. Johnsonand T. Rajh, J. Phys. Chem. Lett., 2011, 2, 2560–2565.

47 P. Senguttuvan, G. Rousse, V. Seznec, J. M. Tarascon andM. R. Palacin, Chem. Mater., 2011, 23, 4109–4111.

48 M. S. Whittingham, Science, 1976, 192, 1126–1127.49 T. B. Kim, J. W. Choi, H. S. Ryu, G. B. Cho, K. W. Kim,

J. H. Ahn, K. K. Cho and H. J. Ahn, J. Power Sources, 2007,174, 1275–1278.

50 J. S. Kim, H. J. Ahn, H. S. Ryu, D. J. Kim, G. B. Cho,K. W. Kim, T. H. Nam and J. H. Ahn, J. Power Sources,2008, 178, 852–856.

51 H. Chen, M. Armand, G. Demailly, F. Dolhem, P. Poizot andJ. M. Tarascon, ChemSusChem, 2008, 1, 348–355.

52 P. Poizot and F. Dolhem, Energy Environ. Sci., 2011, 4, 2003–2019.

53 Y. L. Liang, Z. L. Tao and J. Chen, Adv. Energy Mater., 2012,2, 742–769.

54 M. Armand, S. Grugeon, H. Vezin, S. Laruelle, P. Ribiere,P. Poizot and J. M. Tarascon, Nat. Mater., 2009, 8, 120–125.

55 Y. Park, D. S. Shin, S. H. Woo, N. S. Choi, K. H. Shin,S. M. Oh, K. T. Lee and S. Y. Hong, Adv. Mater., 2012, 24,3562–3567.

56 L. Zhao, J. Zhao, Y.-S. Hu, H. Li, Z. Zhou, M. Armand andL. Chen, Adv. Energy Mater., 2012, 2, 962–965.

57 A. Abouimrane, W. Weng, H. Eltayeb, Y. J. Cui, J. Niklas,O. Poluektov and K. Amine, Energy Environ. Sci., 2012, 5,9632–9638.

58 S.-W. Kim, D.-H. Seo, X. Ma, G. Ceder and K. Kang, Adv.Energy Mater., 2012, 2, 710–721.

59 Y. Liang, P. Zhang, S. Yang, Z. Tao and J. Chen, Adv. EnergyMater., 2013, 3, 600–605.

60 M. D. Slater, D. Kim, E. Lee and C. S. Johnson, Adv. Funct.Mater., 2013, 23, 947–958.

61 S. I. Park, I. Gocheva, S. Okada and J.-i. Yamaki,J. Electrochem. Soc., 2011, 158, A1067–A1070.

62 K. Tang, L. Fu, R. J. White, L. Yu, M.-M. Titirici, M. Antoniettiand J. Maier, Adv. Energy Mater., 2012, 2, 873–877.

63 C. M. Park and H. J. Sohn, Adv. Mater., 2007, 19, 2465–2468.64 C. Marino, L. Boulet, P. Gaveau, B. Fraisse and

L. Monconduit, J. Mater. Chem., 2012, 22, 22713–22720.65 J. Qian, D. Qiao, X. Ai, Y. Cao and H. Yang, Chem. Commun.,

2012, 48, 8931–8933.66 L.-Q. Sun, M.-J. Li, K. Sun, S.-H. Yu, R.-S. Wang and

H.-M. Xie, J. Phys. Chem. C, 2012, 116, 14772–14779.67 L. Wang, X. He, J. Li, W. Sun, J. Gao, J. Guo and C. Jiang,

Angew. Chem., Int. Ed., 2012, 51, 9034–9037.68 C. Marino, A. Debenedetti, B. Fraisse, F. Favier and

L. Monconduit, Electrochem. Commun., 2011, 13, 346–349.69 Y. Kim, Y. Park, N.-S. Choi, J. Kim, J. Lee, J. H. Ryu, S. M. Oh

and K. T. Lee, Adv. Mater., 2013, DOI: 10.1002/adma.201204877.

This journal is ª The Royal Society of Chemistry 2013

Perspective Energy & Environmental Science

Publ

ishe

d on

20

May

201

3. D

ownl

oade

d by

EC

OL

E P

OL

YT

EC

HN

IC F

ED

DE

LA

USA

NN

E o

n 06

/11/

2015

12:

29:4

0.

View Article Online

70 J. J. Braconnier, C. Delmas, C. Fouassier andP. Hagenmuller, Mater. Res. Bull., 1980, 15, 1797–1804.

71 S. Komaba, C. Takei, T. Nakayama, A. Ogata andN. Yabuuchi, Electrochem. Commun., 2010, 12, 355–358.

72 Z. Lu and J. R. Dahn, J. Electrochem. Soc., 2003, 150, A1044–A1051.

73 N. Yabuuchi, M. Kajiyama, J. Iwatate, H. Nishikawa,S. Hitomi, R. Okuyama, R. Usui, Y. Yamada andS. Komaba, Nat. Mater., 2012, 11, 512–517.

74 K. Ado, M. Tabuchi, H. Kobayashi, H. Kageyama,O. Nakamura, Y. Inaba, R. Kanno, M. Takagi andY. Takeda, J. Electrochem. Soc., 1997, 144, L177–L180.

75 C. Delmas, J.-J. Braconnier, C. Fouassier andP. Hagenmuller, Solid State Ionics, 1981, 3–4, 165–169.

76 G. J. Shu and F. C. Chou, Phys. Rev. B: Condens. MatterMater. Phys., 2008, 78, 052101.

77 M. Sathiya, K. Hemalatha, K. Ramesha, J. M. Tarascon andA. S. Prakash, Chem. Mater., 2012, 24, 1846–1853.

78 K. S. Kang, Y. S. Meng, J. Breger, C. P. Grey and G. Ceder,Science, 2006, 311, 977–980.

79 Y. Makimura and T. Ohzuku, J. Power Sources, 2003, 119,156–160.

80 T. Ohzuku and Y. Makimura, Chem. Lett., 2001, 744–745.81 S. Venkatraman, J. Choi and A. Manthiram, Electrochem.

Commun., 2004, 6, 832–837.82 K. Kang and G. Ceder, Phys. Rev. B: Condens. Matter Mater.

Phys., 2006, 74, 094105.83 W. S. Yoon, Y. Paik, X. Q. Yang, M. Balasubramanian,

J. McBreen and C. P. Grey, Electrochem. Solid-State Lett.,2002, 5, A263–A266.

84 S. Komaba, N. Yabuuchi, T. Nakayama, A. Ogata,T. Ishikawa and I. Nakai, Inorg. Chem., 2012, 51, 6211–6220.

85 D. Kim, S. H. Kang, M. Slater, S. Rood, J. T. Vaughey,N. Karan, M. Balasubramanian and C. S. Johnson, Adv.Energy Mater., 2011, 1, 333–336.

86 A. K. Padhi, K. S. Nanjundaswamy and J. B. Goodenough,J. Electrochem. Soc., 1997, 144, 1188–1194.

87 J. N. Bridson, S. E. Quinlan and P. R. Tremaine, Chem.Mater., 1998, 10, 763–768.

88 J. Moring and E. Kostiner, J. Solid State Chem., 1986, 61,379–383.

89 S.-M. Oh, S.-T. Myung, J. Hassoun, B. Scrosati and Y.-K. Sun,Electrochem. Commun., 2012, 22, 149–152.

90 Y. Zhu, Y. Xu, Y. Liu, C. Luo and C. Wang, Nanoscale, 2013,5, 780–787.

91 K. T. Lee, T. N. Ramesh, F. Nan, G. Botton and L. F. Nazar,Chem. Mater., 2011, 23, 3593–3600.

This journal is ª The Royal Society of Chemistry 2013

92 S.-i. Nishimura, M. Nakamura, R. Natsui and A. Yamada,J. Am. Chem. Soc., 2010, 132, 13596–13597.

93 J. M. Clark, S.-i. Nishimura, A. Yamada and M. S. Islam,Angew. Chem., Int. Ed., 2012, 51, 13149–13153.

94 H. Kim, S. Lee, Y.-U. Park, H. Kim, J. Kim, S. Jeon andK. Kang, Chem. Mater., 2011, 23, 3930–3937.

95 P. Barpanda, T. Ye, S.-i. Nishimura, S.-C. Chung, Y. Yamada,M. Okubo, H. Zhou and A. Yamada, Electrochem. Commun.,2012, 24, 116–119.

96 H. Kim, R. A. Shakoor, C. Park, S. Y. Lim, J.-S. Kim, Y. N. Jo,W. Cho, K. Miyasaka, R. Kahraman, Y. Jung and J. W. Choi,Adv. Funct. Mater., 2012, 1147–1155.

97 K.-H. Ha, S. H. Woo, D. Mok, N.-S. Choi, Y. Park, S. M. Oh,Y. Kim, J. Kim, J. Lee, L. F. Nazar and K. T. Lee, Adv. EnergyMater., 2013, 3, 770–776.

98 P. Barpanda, S. Nishimura and A. Yamada, Adv. EnergyMater., 2012, 2, 841–859.

99 P. Barpanda, J. C. Lu, T. Ye, M. Kajiyama, S. C. Chung,N. Yabuuchi, S. Komaba and A. Yamada, RSC Adv., 2013,3, 3857–3860.

100 P. Barpanda, M. Avdeev, C. D. Ling, J. C. Lu and A. Yamada,Inorg. Chem., 2013, 52, 395–401.

101 F. Sanz, C. Parada, J. M. Rojo, C. Ruiz-Valero and R. Saez-Puche, J. Solid State Chem., 1999, 145, 604–611.

102 B. L. Ellis, W. R. M. Makahnouk, Y. Makimura, K. Toghilland L. F. Nazar, Nat. Mater., 2007, 6, 749–753.

103 B. L. Ellis, W. R. M. Makahnouk, W. N. Rowan-Weetaluktuk, D. H. Ryan and L. F. Nazar, Chem. Mater.,2010, 22, 1059–1070.

104 N. Recham, J.-N. Chotard, L. Dupont, K. Djellab,M. Armand and J.-M. Tarascon, J. Electrochem. Soc., 2009,156, A993–A999.

105 N. Recham, J. N. Chotard, J. C. Jumas, L. Laffont,M. Armand and J. M. Tarascon, Chem. Mater., 2010, 22,1142–1148.

106 T. N. Ramesh, K. T. Lee, B. Ellis and L. F. Nazar,Electrochem. Solid-State Lett., 2010, 13, A43–A47.

107 J. Barker, M. Y. Saidi and J. L. Swoyer, J. Electrochem. Soc.,2003, 150, A1394–A1398.

108 J. Barker, M. Y. Saidi and J. L. Swoyer, Electrochem. Solid-State Lett., 2003, 6, A1–A4.

109 R. K. B. Gover, A. Bryan, P. Burns and J. Barker, Solid StateIonics, 2006, 177, 1495–1500.

110 Z. L. Jian, L. Zhao, H. L. Pan, Y. S. Hu, H. Li, W. Chen andL. Q. Chen, Electrochem. Commun., 2012, 14, 86–89.

111 H. Kim, I. Park, D.-H. Seo, S. Lee, S.-W. Kim, W. J. Kwon,Y.-U. Park, C. S. Kim, S. Jeon and K. Kang, J. Am. Chem.Soc., 2012, 134, 10369–10372.

Energy Environ. Sci., 2013, 6, 2067–2081 | 2081