Nanostructured positive electrode materials for post

0 downloads 0 Views 4MB Size Report
Li2S4, and Li2S3) and insoluble sulfides (Li2S2/Li2S) when using carbonate and/or ether liquid electrolytes. The overall redox couple of a Li–S battery can be ...
Energy & Environmental Science View Article Online

Published on 26 September 2016. Downloaded by SLUB DRESDEN on 29/11/2016 20:29:33.

REVIEW

Cite this: Energy Environ. Sci., 2016, 9, 3570

View Journal | View Issue

Nanostructured positive electrode materials for post-lithium ion batteries Faxing Wang,†a Xiongwei Wu,†b Chunyang Li,a Yusong Zhu,*a Lijun Fu,*a Yuping Wu*abc and Xiang Liua Nanotechnology has opened up new frontiers in materials science and engineering in the past several decades. Considerable efforts on nanostructured electrode materials have been made in recent years to fulfill the future requirements of electrochemical energy storage. Compared to bulk materials, most of these nanostructured electrode materials improve the thermodynamic and kinetic properties of electrochemical reactions for achieving high energy and power densities. Here we briefly review the state-of-the-art research activities in the area of nanostructured positive electrode materials for post-lithium ion batteries, including Li–S batteries, Li–Se batteries, aqueous rechargeable lithium batteries, Li–O2 batteries, Na-ion batteries, Mg-ion batteries and Al-ion batteries. These future rechargeable battery systems may offer increased energy densities, reduced cost, and more environmental benignity. A particular focus is directed to the design principles of these nanostructured positive electrode materials and how nanostructuring

Received 18th July 2016, Accepted 26th September 2016

influences electrochemical performance. Moreover, the recent achievements in nanostructured positive

DOI: 10.1039/c6ee02070d

Zn-ion batteries, F- and Cl-ion batteries, Na–, K– and Al–S batteries, Na– and K–O2 batteries, Li–CO2

electrode materials for some of the latest emerging rechargeable batteries are also summarized, such as batteries, novel Zn–air batteries, and hybrid redox flow batteries. To facilitate further research and

www.rsc.org/ees

development, some future research trends and directions are finally discussed.

Broader context The application of rechargeable battery systems has decreased our consumption of fossil fuels, thereby reducing CO2 emissions and the consequent effects on climate change. Over the past nearly three decades, lithium ion batteries (LIBs) have been identified as some of the most promising energy storage devices to power portable electronic devices as well as automobiles. However, the need for higher energy and power densities, a longer cycle life and improved safety along with a lower cost cannot be fulfilled by the current state-of-the-art LIBs. Therefore, alternative rechargeable battery systems going beyond LIBs are being pursued. Now research on post Li-ion batteries is advancing fast towards lowering the costs, improving the specific energy density and reducing environmental and safety problems. From the point of view of materials, here we present a critical overview of fundamental research activities on positive electrode materials for various battery systems, including Li–S batteries, Li–Se batteries, aqueous rechargeable lithium batteries, Li–O2 batteries, Na-ion batteries, Mg-ion batteries, Al-ion batteries and some of the latest emerging rechargeable batteries (Zn-ion batteries, F- and Cl-ion batteries, Na–, K– and Al–S batteries, Na– and K–O2 batteries, Li–CO2 batteries and hybrid redox flow batteries).

1. Introduction Interest in electrical energy storage has increased significantly over the past few decades due to growing global energy demands coupled with environmental concerns. Electrical energy storage in the form of rechargeable batteries can be used not only as the a

College of Energy and Institute for Advanced Materials, Nanjing Tech University, Nanjing 211816, Jiangsu Province, China. E-mail: [email protected] b College of Science, Hunan Agriculture University, Changsha, China c Department of Chemistry and Shanghai Key Laboratory of Molecular Catalysis and Innovative Materials, Fudan University, Shanghai 200433, China † Equal contribution.

3570 | Energy Environ. Sci., 2016, 9, 3570--3611

power sources for small devices such as laptops and cellphones, but also as a back-up energy supply for stationary energy storage systems such as national electric grids and transportation applications like electric vehicles (EVs) and spacecrafts. In addition, military requirements are another strong driving force behind the research and development of batteries. All of the electronic gizmos of a modern soldier (such as night vision goggles, flashlights, radios, and GPSs) are powered by batteries. Therefore, rechargeable batteries are now a competitive commodity of national and strategic significance.1 About 235 years ago, Italian anatomists found the bioelectricity phenomenon when dissecting a frog. Inspired by

This journal is © The Royal Society of Chemistry 2016

View Article Online

Published on 26 September 2016. Downloaded by SLUB DRESDEN on 29/11/2016 20:29:33.

Review

Fig. 1 (a) The roadmap of battery development. (b) An overview of nanostructured positive electrode materials for post-lithium ion batteries.

this, in 1800, Alessandro Volta from the University of Pavia (Italy) constructed the first battery (Fig. 1a). He assembled a pile of alternate silver (or brass or copper) and zinc (or tin) discs saturated with brine, which was separated by a piece of cloth. When the two electrodes were connected by a wire conductor, a

Dr Yuping Wu is a distinguished professor at Nanjing Tech University, Nanjing, China. Since 1994, he has published over 230 peerreviewed papers, written 6 monographs whose sales are above 30 000 copies and 8 chapters, translated one book, and got 22 issued patents. His present Hindex is 53. Within 10 years, 32 papers are listed as ESI highly cited papers, and he is one of the highly cited researchers in Yuping Wu the world. His research has led to the creation of four companies on anode materials, cathode materials, separators and power lithium ion batteries, respectively. His research is mainly focused on lithium ion batteries, aqueous rechargeable lithium batteries, supercapacitors, other new energy storage and conversion systems and solar hydrogen.

This journal is © The Royal Society of Chemistry 2016

Energy & Environmental Science

continuous current of electricity was produced.2a Because of this famous experiment, his name is commemorated for all time by the unit of electrical potential (the volt). In 1836, Prof. John Daniell from King’s College (London, UK) introduced the first practical galvanic battery (Zn/ZnSO4 + CuSO4/Cu, also called a Daniell cell), which presented a voltage of 1.1 V. The next major advance in the development of batteries was the invention of the Zn/NH4Cl/C primary battery in 1866. Actually, the first effective rechargeable battery (the famous lead–acid battery) was demonstrated by a French chemist in 1859. Then it achieved commercialization in 1882. In Sweden, Waldemar Jungner took out a patent for a ‘nickel–cadmium battery’ in 1895. Similar to nickel//cadmium batteries, Mr Thomas Edison developed and patented the ‘nickel–iron battery’. Interestingly, when these two inventors learnt of each other’s work and patent, there was considerable controversy and rivalry. After that, Mr Edison and Henry Ford (the founder of the Ford Motor Company) worked for some years on an electric automobile, but the so-called ‘‘electric automobile’’ soon gave way to gasolinepowered automobiles.1 In the 1920s and 1930s, zinc–nickel oxyhydroxide (Zn–NiOOH) batteries were also commercially available, but they had short cycle-lives and failed to gain market acceptance. It is only in the last couple of decades that the nickel–metal-hydride has become a commercially popular product. In the 1970s, the first oil crisis forced researchers to look for new alternative energy sources. Research carried out at the Exxon Laboratories (USA) in the 1970s showed that Li+ ions are electrochemically ‘intercalated’ into the crystal lattice of titanium disulfide (TiS2).2b At the AT&T Bell Laboratories (USA), researchers focused on niobium triselenide (NbSe3) as a positive electrode with up to three Li+ ions incorporated per formula unit.2c A significant advance was made in 1980 with the discovery at Oxford University (UK) that Li+ ions could be electrochemically withdrawn from the LiCoO2 structure and inserted reversibly.2d Throughout the 1980s both in the USA and Japan, various attempts were made to manufacture and market rechargeable lithium batteries that applied lithium metal as the negative electrode, but eventually these batteries were withdrawn as a result of safety related concerns on dendrite growth. Lithium is not normally electrodeposited as a smooth layer, but as a ‘mossy’ deposit. In the early 1990s, the Sony Corporation (Japan) first commercialized lithium ion batteries employing two intercalation electrodes with Li+ ions shuttling back and forth between the positive electrode (LiCoO2) and the negative electrode (graphite). This kind of battery contains no metallic lithium and is therefore much safer than the earlier one (lithium–metal battery). Other Japanese manufacturers soon entered the market, followed closely by American, European and Chinese companies. The subsequent growth in sales of the Li ion batteries was truly phenomenal and the ‘Li era’ arrived in the early 2000s.2c So far, the current lithium ion batteries, which are based on intercalation positive electrode materials as some of the most successful energy storage systems seen during the past 30 years, are encountering different kinds of challenges and problems.2e Although lithium ion batteries are now dominating the market

Energy Environ. Sci., 2016, 9, 3570--3611 | 3571

View Article Online

Published on 26 September 2016. Downloaded by SLUB DRESDEN on 29/11/2016 20:29:33.

Energy & Environmental Science

of portable electronic devices, it has been difficult to create batteries with the power and energy densities, safety, cycleability and cost to meet the requirements of large-scale application for stationary energy storage systems and transportation applications. These problems can be outlined more fully as follows. (1) One of the major issues is safety. Overcharging or heating usually results in the decomposition of the charged positive electrodes and electrolytes. The electrolyte decomposition ruins the battery while the decomposition of the positive electrode leads to the liberation of oxygen gas and dangerous fire particularly if the battery vent failed to operate correctly. Particular care is needed when Li ion batteries are used in series or parallel combinations in a battery pack. (2) The energy densities of current Li ion batteries remain insufficient for many emerging applications such as vehicle electrification since the speed of increase in specific energy technologies is slowing down. After all, the capacities of the positive electrode (insertion-oxide) approach an intrinsic specific capacity limit of less than 335 mA h g1. (3) High cost remains another critical barrier to the widespread scale-up of Li ion batteries due to the special cell assembly technology required, the requirement of a strictly dry environment during manufacturing processes, and the high price of raw metals, organic electrolytes, and lithium salts. Li ion batteries were originally developed as a high-energy source for portable electronic devices. However, nowadays Li ion batteries with a megawatt hour (MW h) scale have already been developed and planned to test for the storage of electricity generated from solar cells and wind turbines.2f The safety and cost are relatively more important in large-scale stationary energy storage systems. (4) The limited ion conductivities of the organic electrolytes in Li ion batteries lead to relatively low power densities, which need a long charging time. Following the battery development roadmap in Fig. 1a, what age will the rechargeable batteries be in the future for many years? As a result, batteries instead of lithium ion ones or post-lithium ion batteries should be explored. With a goal of increasing energy density, researchers are pursuing alternative positive electrode materials such as S, Se and O2 that can offer higher capacities than conventional intercalation positive electrode materials.3 In addition, the use of flammable organic electrolytes might lead to thermal run-away and safety accidents, and thus aqueous rechargeable lithium batteries (ARLBs) were suggested.4 The huge exploitation of Li resources will eventually one day lead to their depletion since the lithium element reserves are limited and are mainly located in relatively few geographical areas. In this context, Na-/Mg-/Al-ion batteries were proposed.5,6 Compared to Li resources, resources of other elements (Na, Mg and Al) are abundant across the world with a low price and these other metals are also easy to recycle. Besides, the recent emerging rechargeable batteries such as Zn-ion batteries, F- and Cl-ion batteries, Li–CO2 batteries, and hybrid redox flow batteries are also expected to be highly promising candidates for sustainable energy storage systems. At the same time, it is now well known that engineering electrode materials in the nanoscale range plays a significant role in the field of electrochemical energy storage. Currently, scientists have synthesized positive electrode materials in a variety of

3572 | Energy Environ. Sci., 2016, 9, 3570--3611

Review

nanostructures including zero-dimensional (0D) (nanoparticles, and quantum dots), one-dimensional (1D) (nanowires, nanorods, and nanotubes), two-dimensional (2D) (nanowalls, and nanosheets), and three-dimensional (3D) hierarchical ones.7 In most cases, the specific capacity and rate capability of the positive electrode materials are improved when they are nanostructured. Chemists have explained these enhancements in terms of shorter ion diffusion paths through the solid state, since solid state diffusion is relatively sluggish compared to the diffusion through a liquid electrolyte.8 Moreover, small domain sizes can also reduce intercalation stress.9a Besides, some electrochemical reactions cannot occur in bulk counterparts, unless by turning the active materials into nanosized or nanoporous structures.9b In general, the size effects of nanostructured materials can improve the thermodynamics and kinetics of electrochemical reactions.7a,9c On one hand, thermodynamics describe the maximum energy released or stored for an electro-chemical reaction. High energy density (E) requires a large specific capacity (Q) and high voltage (V) between positive and negative electrodes. Nanostructures can increase the utilization of the active materials to improve the capacity (Q) and reduce polarization to enhance the working voltage (V). On the other hand, kinetics usually decide the reaction rate which is associated with charge transport (in the electrolyte and the active materials) and charge transfer (in the two-phase interface).7a High power requires both electrons and ions to be highly mobile throughout the electrode material and electrolytes. Nanostructured materials can shorten the electronic transport and ionic diffusion distance to improve the power density. However, the small size and the large specific surface area may have adverse impacts on some applications of nanomaterials. A large electrolyte/electrode surface area may lead to more significant side reactions with the electrolyte. Another potential disadvantage is the less dense packing density leading to lower volumetric energy density.10 A hierarchical micro–nano structure becomes an effective strategy to solve the problems. The tradeoff between using nanostructured electrodes and designing dense architectures may give rise to new scientific questions at the mesoscale. Therefore, nanostructured electrode materials are creating both opportunities and challenges for enhanced energy storage. The practical question is, which properties of these altered nanomaterials can be used to improve the performance of batteries? Here, the major developments in the area of nanostructured positive electrode materials for post-lithium ion batteries are summarized. Considering the recent interest in this field and our own expertise, special focus is put on the design principles of these nanostructured positive electrode materials and how nanostructuring influences electrochemical performance. As shown in Fig. 1b, the topics that will be discussed in this review include nanostructured positive electrode materials for (1) Li–S, (2) Li–Se, (3) ARLBs, (4) Li–O2, (5) Na-ion, (6) Mg-ion, (7) Al-ion batteries and (8) other rechargeable batteries such as Zn-ion batteries, F- and Cl-ion batteries, Na–, K– and Al–S batteries, Na– and K–O2 batteries, Li–CO2 batteries, Zn–air batteries, and hybrid redox flow batteries. At the end of this review, an outlook

This journal is © The Royal Society of Chemistry 2016

View Article Online

Review

Energy & Environmental Science

Published on 26 September 2016. Downloaded by SLUB DRESDEN on 29/11/2016 20:29:33.

is provided to highlight latest research trends in designing nanostructured positive electrode materials. It is aimed at providing some rational understanding of the effects of nanostructure engineering on the electrochemical performance of the positive electrode materials. It is anticipated that this review can shed some light on the development of advanced positive electrode materials for post-lithium ion batteries.

2. Nanostructured positive electrode materials for rechargeable Li–S batteries Sulfur, one of the most abundant elements on the Earth, is a positive electrode material that can accept up to two electrons per atom. During the charge/discharge process, S8 undergoes a series of structural and morphological changes involving the formation of soluble lithium polysulfides Li2Sx (Li2S8, Li2S6, Li2S4, and Li2S3) and insoluble sulfides (Li2S2/Li2S) when using carbonate and/or ether liquid electrolytes. The overall redox couple of a Li–S battery can be described using reaction (1): S8 + 16Li+ + 16e 2 8Li2S

(1)

improves the safety. However, there are still several challenges that impede their practical applications. The first main issue is that sulfur is both ionically and electrically insulating (its electronic conductivity is 5  1030 S cm1 at 25 1C). The second is related to the volume change of S particles during charge and discharge. When S is fully converted into Li2S, the volume expansion is as large as 80% because the densities of S and Li2S are 2.03 and 1.66 g cm3, respectively.11 Furthermore, another problem arises because of the intermediate discharge products Li2Sx (Li2S8, Li2S6, Li2S4, and Li2S3). These soluble S-based species can diffuse out of the electrode, through the separator, to the lithium negative electrode. This process (the so-called shuttle effect) results in a permanent loss of active materials, which accounts for the dramatic loss of discharge capacity during cycling. Tremendous progress has been made in understanding and improving the performance of S positive electrode materials. Generally, this progress in nanostructured S can be divided into several categories: (1) S–carbon nanohybrids (such as S–CNTs or CNFs, S–graphene and S–porous carbon), (2) core–shell structured S-based nanocomposites (such as S-conductive polymers), and (3) nanostructured Li2S.

+

and it occurs at a potential of 2.15 V (vs. Li /Li). As a result, sulfur as a positive electrode has a high theoretical capacity of 1675 mA h g1. An order of magnitude higher than those of the conventional intercalation positive electrode materials can be achieved, and the packaged Li–S batteries can arrive at an energy density of 400–600 W h kg1.11–13 The first developments of the Li–S battery were aimed at improving the energy density of primary batteries through the complete utilization of a two-electron reduction of sulfur in the 1960s.14,15 Moving from the traditional intercalation positive electrode to S has many benefits besides the higher capacity, such as the low-cost S replacing expensive transition metals like Co, and using environmentally friendly sulfur in comparison with certain toxic transition-metal compounds.16 The lower potential (2.15 V vs. Li+/Li) is not detrimental for practical applications because the gravimetric capacity of sulfur is the highest of any other solid positive electrode material. Furthermore, the low operating voltage without decreasing energy density

Table 1

2.1

Sulfur–carbon nanohybrids

2.1.1 S–CNTs and S–CNFs (1D). Carbon nanotubes (CNTs) and carbon nanofibers (CNFs) in Li–S batteries have long been investigated as highly conductive agents. They present not only a superior electrically conductive network, but also a high tensile strength that stabilizes the S positive electrode structure. This stability is especially critical to accommodate the severe volume change that occurs in the S electrode during the discharge/ charge process. Some characteristics of them are summarized in Table 1. Both chemical and physical routes have been demonstrated for the fabrication of S–CNT based 1D hybrid nanostructures. In the case of physical routes, most of the sulfur incorporated into the CNTs was obtained by heating the sulfur/carbon mixture at around 155 1C under the protection of inert gas. This is because elemental S (S8) turns into liquid and presents the lowest viscosity at 155 1C, so the liquid S can be infused into

Some characteristics of 1D S–carbon nanohybrids for Li–S batteries

Positive electrode

Sulfur content

Preparation method

S-disordered CNTs Small sulfur allotropes S2–4–MPC–CNTs S–CNTs

Template method + physical evaporation process Hydrothermal method + physical melt-diffusion process Template method + chemical reduction process

S-vertical aligned CNTs S–CNF S–CNTs S–porous CNF

Chemical vapor deposition + physical infiltration Electrospinning method + physical evaporation process Chemical vapor deposition + chemical precipitation process Electrospinning method + chemical deposition process

S–porous CNF–CNTs

Electrospinning method + template method + physical evaporation process Ball-milling method + solvent exchange process Ball-milling method + physical melt-diffusion process

S–CNTs S–porous C–CNTs

This journal is © The Royal Society of Chemistry 2016

40 40 23 50 70 65 80 42 60 55

wt% wt% wt% + wt% wt% wt% wt% wt% wt% wt%

58 wt% 82 wt%

Rate capability

Cycling behavior

Ref.

2.4 A g 8.35 A g1 6 A g1

100 200 100

17a 17b 18a

C/13 0.1 A g1 0.1 A g1 0.34 A g1

40 70 30 30

18b 19a 19b 20

5C

200

21

1.6 A g1 3.2 A g1

200 150

22 23

1

Energy Environ. Sci., 2016, 9, 3570--3611 | 3573

View Article Online

Published on 26 September 2016. Downloaded by SLUB DRESDEN on 29/11/2016 20:29:33.

Energy & Environmental Science

Review

the CNT host structure at this temperature. The strategy using CNT–microporous carbon (MPC)–small sulfur S2–4 (S2, S3 and S4) nanohybrids for Li–S batteries has been previously demonstrated.17b The small S2–4 molecules avoid the unfavorable transition between S8 and S42 during the discharging/charging process. Besides, when S is vaporized at an elevated temperature in a vacuum, its vapor can go into the carbon voids and even into the graphite layers of graphitic clusters, that liquid S cannot reach directly. For example, S impregnation into the CNTs was carried out at high temperatures (500 1C).17a Interestingly, the heat treatment at 500 1C could break down the S8 molecule into S6 or S2, and the conventional Li–S reactions with dissolvable polysulfide intermediate products are altered. The use of templates and chemical vapor deposition (CVD) are also two common methods to synthesise S–CNT nanohybrids. By using a sulphate-containing AAO template, S–CNTs could be assembled into a membrane as a binder-free, flexible positive electrode.18a The incorporation of S into the AAO-templated CNTs is due to the carbothermal reduction (2): SO42 + C - S + CO + CO2

(2)

Another binder-free electrode containing vertically aligned-CNTs, which is subsequently infiltrated with S, was also introduced as a promising choice for Li–S batteries.18b The vertically alignedCNTs were directly synthesized on a Ni current collector. This nanocomposite electrode could enable high sulfur loadings up to 70 wt%. It is worth mentioning that the preparation of S–CNTs in a large quantity and at a low cost remains a challenge, which partly countervails the merits of low price Li–S batteries. Recently, a novel S vapor deposition strategy was demonstrated to synthesize S–C nanohybrids on a large-scale.19a This S vapor deposition provides a simple, continuous, and economical strategy to produce positive electrodes at a high speed and low cost. Precipitation methods with a low cost also show promising industrial potential in preparing S–CNT nanohybrids.19b First, acid treated CNTs were dispersed in a Na2S2O3 solution. Then, H2SO4 was added dropwise to the solution. The reaction for the sulfur preparation is as follows: Na2S2O3 + H2SO4 - Na2SO4 + Sk + H2SO3

(3)

In addition, 1D electrospun carbon nanofibers (CNFs) are attractive for the construction of S–CNF nanohybrids.20,21 Electrospinning is another simple and effective technology with scale-up potential for a wide range of nanomaterials aimed at industrial production. Through electrospinning, carbonization and solution-based chemical reaction–deposition methods, S encapsulated into porous carbon CNFs was recently synthesized.20 Another choice for the fabrication of 1D S-based nanohybrids on a large scale is the ball-milling method.22,23 S was firstly uniformly coated on the surface of functional CNTs with an S layer of thickness 10–20 nm. Then the nanoscale S–CNT intermediate composite was ball-milled to form an interwoven and porous sphere architecture with large pores (around 1 mm to 5 mm).22

3574 | Energy Environ. Sci., 2016, 9, 3570--3611

The flexible S–CNT membrane could sustain a stress of 10 MPa with 9% strain and shows a high electrical conductivity of 800 S m1, which remains unchanged after 12 000 bend cycles.18a At a high current density (6 A g1), the discharge capacity of the S–CNT membrane is attained at 712 mA h g1 (23 wt% S) and 520 mA h g1 (50 wt% S). The overall capacity of the flexible electrode correspondingly reaches 163 mA h g1 (23 wt% S) and 260 mA h g1 (50 wt% S). After a 100 cycle test, the overall morphology and structure of the S–CNT membrane are well preserved, which indicates that the stability of S confined in the walls of the CNTs during lithiation/de-lithiation is very good. Like the S–CNT membrane, S–CNF also exhibits a stable cycling behavior with a large capacity retention of around 95% at the 50th cycle.19a,20 The porous S–CNF nanocomposite with 42 wt% S maintains a stable discharge capacity of about 1400 mA h g1 at 0.05C, 1100 mA h g1 at 0.1C and 900 mA h g1 at 0.2C.20 This porous S–CNF nanocomposite electrode also has a high reversible capacity after at least 30 cycles at higher rates. The high performance is attributed to the unique structures. First, the S can be effectively encapsulated into the small-sized pores and therefore the small size of sulfur in the CNT@CNFs shortens the transport pathways for both electrons and Li ions. Second, the high surface area and the tubular structure of the CNTs can trap dissolved polysulfides and accommodate the large sulfur volumetric expansion during lithiation.21 Different from most S–C materials with a nanostructure, the micrometer scale S–CNT nanomicrosphere with a large pore structure could exhibit high sulfur utilization.22 Its hollow macro-sphere architecture could constrain more electrolytes within the electrode, thus providing a high rate capability. It remains 650 mA h g1 after 200 cycles at 1 A g1. Meanwhile, the functional groups on the CNT surface further provide a chemical gradient which could retard the diffusion of polysulfides out of the electrode. Interestingly, a recent study demonstrates that complex synthesis and surface modification are not necessary to achieve high-performance Li–S batteries. They use conductive multiwalled CNT paper acting as a pseudo-upper current collector.24 Compared to the Li–S batteries without CNT films, the intercalation of the CNT films into Li–S batteries has an additional advantage of localizing the polysulfides in the electrolyte, thus avoiding the unwanted migration and diffusion of the dissolved species to the negative electrode. Moreover, this novel configuration plays a key role in reducing the interfacial resistance for the sulfur electrodes. 2.1.2 S–graphene (2D). The Nobel Physical Prize-winning discovery of graphene, a one-atom-thick 2D single layer of sp2-bonded carbon, has created an entirely new branch of carbon chemistry and led to an explosion of interest in graphene.25 In light of its unique structural properties, graphene exhibits many advantages such as extraordinarily high electrical and thermal conductivity, great mechanical strength and high surface area, making it a potential candidate for applications in energy storage fields.26 The most produced graphene is usually decorated with epoxy and hydroxyl groups, which can enhance the

This journal is © The Royal Society of Chemistry 2016

View Article Online

Published on 26 September 2016. Downloaded by SLUB DRESDEN on 29/11/2016 20:29:33.

Review

binding of S with the C–C bonds due to the induced ripples caused by the functional groups. There have been a lot of attempts to fabricate S–graphene nanohybrids using different techniques such as solvothermal and sulfur vapor methods,27 thermal expansion and melt-diffusion strategies,28 chemical reduction graphene oxide (GO) and meltdiffusion strategies,29,30 sulfur-assisted exfoliation of graphite,31 and an electrochemical assembly strategy.32 Among them, chemically reduced GO is not only scalable but can also provide graphene with processibility and new functions. However, the graphene obtained using this method suffers from two main drawbacks: (i) the preparation produces a lot of pollution and the cost is high due to the high usage of acid and strong oxidants; and (ii) it also leads to severe structural damage to the graphene layer, resulting in relatively poor conductivity and stability. Thus, scotch-tape-like sulfur-assisted exfoliation of graphite was developed to produce S–graphene nanocomposites with graphene sheets of low defects.31 The intimate interaction between sulfur and graphene, attributed to the similar electronegativities of the two elements, is stronger than the van der Waals forces between adjacent p–p stacked graphene layers. This causes cleavage of the graphene layers when the sulfur molecules stick to the surface and edges of the graphite, similar to Scotch tape in micromechanical exfoliation processes. Metal organic frameworks (MOFs), an intriguing family of hybrid porous materials, have attracted increasing attention as the host for S impregnation owing to their extra large surface area, pore volume and tunable opening.33 Specifically, a graphene/ chromium-MOF (MIL-101) composite was investigated to serve as a host for S immobilization in Li–S batteries.33b The S was firstly infiltrated into the MIL-101 using a melting-diffusion method. Then graphene was wrapped on the surface of the MIL-101/S via strong electrostatic attractions. The unique structure with a large specific area and a conductive shell ensures a high dispersion of sulfur in the composite. The early synthesis of a S–graphene nanocomposite utilizes large sulfur particles enveloped by graphene/graphene oxide sheets with a polymer layer buffering the sulfur.34 The resulting nanocomposite shows stable specific capacities up to 600 mA h g1 over more than 100 cycles. The graphene and polymer coating layers are important for accommodating the volume expansion

Energy & Environmental Science

of the coated sulfur particles during discharge, trapping soluble polysulfide intermediates, and rendering the sulfur particles with electrical conduction. However, the stacking of the graphene nanosheets during electrode formation leads to a significant reduction in the active surface area of the 2D structures. Thus, a nanocomposite with S entrapped into hierarchical porous graphene is obtained and high discharging capacities of 1068 and 543 mA h g1 at 0.5 and 10C, respectively, are achieved.28a Furthermore, a discharging capacity of 386 mA h g1 can be retained at an ultra-low temperature of 40 1C, which far exceeds the operating range of conventional lithium-ion batteries. An excellent rate capability and endurance to over-discharge were achieved in a unique interleaved expanded graphiteembedded sulphur nanocomposite.29 Its discharge capacity is 337 mA h g1 at the highest current density of 28 A g1, which corresponds to a time of 35 s to fully discharge the total capacity. In this nanocomposite, graphene layers can act as mini-current collectors which facilitate the fast transportation of electrons during the charge/discharge process. A similar excellent rate capability was also achieved by vertically aligning S–graphene nanowalls onto an electrically conductive substrate.32 A high reversible capacity of over 400 mA h g1 at 13.36 A g1 was achieved. In each individual S–graphene nanowalls, S nanoparticles are homogeneously anchored in between the graphene layers and ordered graphene arrays are arranged perpendicularly to the substrates, which are favorable for the fast diffusions of both lithium and electrons. The graphene sheets with outstanding electron conductivity (1820 S cm1) prepared from the sulfur-assisted exfoliation of graphite also provide an ideal model to homogenously anchor S and remedy the insulating properties of S.31 Compared to the above S–graphene, the nanocomposite of S and graphene–MOFs delivers a much higher discharge capacity of 1192 mA h g1 (Fig. 2a), which is also higher than that of the S–MOF nanocomposite (721 mA h g1). The MOF possesses a large surface area and micro-porous aperture windows, which can supply abundant adsorption sites for sulfur as well as the polysulfides. Moreover, the larger cages associated with MOFs help to alleviate the deposition of the reduced lithium sulfides and to improve the penetration of electrolytes into the composite. But the MOF host has much lower electronic

Fig. 2 (a) Galvanostatic discharge/charge profiles of S–MOF–graphene and S–MOF nanocomposites at a rate of 0.1C (modified from ref. 33b, copyright permission from the Royal Society of Chemistry). (b) The XAS spectra of GO and S–GO nanocomposites and (c) representative pattern of GO immobilizing S. (Yellow, red, and white balls denote S, O, and H atoms, respectively, while the others are C atoms) (modified from ref. 37, copyright permission from the American Chemical Society).

This journal is © The Royal Society of Chemistry 2016

Energy Environ. Sci., 2016, 9, 3570--3611 | 3575

View Article Online

Published on 26 September 2016. Downloaded by SLUB DRESDEN on 29/11/2016 20:29:33.

Energy & Environmental Science

conductivity than carbon materials, which leads to a relatively low capacity compared to other S–carbon nanohybrids as discussed in the above section. After graphene wrapping on the surface of the S–MOF nanocomposite, the capacity is greatly increased.33 Besides graphene, graphene oxide (GO) is also effective as a network to trap soluble polysulfide intermediates.35–37 A hierarchically micro/mesoporous microwave exfoliated graphene oxide (MEGO) with a high surface area was utilized as a superior carbon host material for high S loading.36 Due to the extremely high surface area (up to 3000 m2 g1) and large pore volume (up to 2.14 cm3 g1), the MEGO could achieve a high sulfur loading of 75 wt%, leading to a high overall energy density. Likely, a uniform and thin S coating on the graphene oxide sheets was created to immobilize sulfur and lithium polysulfides via the reactive functional groups on graphene oxide.37 A soft X-ray absorption spectroscopy (XAS) measurement probes the chemical bonding on the S–GO surface (Fig. 2b). Compared with GO, the increase in the sharpness of the p* (line A) and excitonic state (line D) for the S–GO nanocomposites suggests that the ordering of the sp2-hybridized carbon structure is better formatted after S is incorporated (Fig. 2b). Line C originating from a different functional group (possibly the C–O bond) on the GO is significantly weakened when incorporated with S, which means that strong chemical interaction between S and the functional groups of GO occurs and S can partially reduce the GO. Both epoxy and hydroxyl groups can enhance the binding of S to the C–C bonds due to the induced ripples caused by the functional groups (Fig. 2c). This S–GO nanocomposite displays an excellent capacity stability of about 954 mA h g1 after 50 cycles. The GO provides highly reactive functional groups on its surface that can serve as immobilizers to hold the S. Interestingly, Nafion coating on S–graphene sheet nanocomposites could improve the cycling performance of Li–S batteries.38 A Nafion polymer is a cation exchange material with a high density of negatively charged sulfonate groups. This allows the penetration of Li+ cations into the nanocomposite electrode while at the same time suppressing polysulfide anions from diffusing across the Nafion barrier because of static–electric repulsion. This concept was also demonstrated using Nafion-based membranes for Li–S batteries.39 However, the ion exchange membrane (Nafion) may become another bottleneck in the commercialization of Li–S batteries because of its high cost. Considering the trade-off between cost and electrochemical performance, some polymers with a low cost can be used to modify Nafion. 2.1.3 S–nanoporous carbon (3D). Introducing pores into carbon and S can increase both the electrical and ion conductivity of the sulfur-based electrode while at the same time retard the polysulfide shuttle phenomenon through high surface area carbon adsorbents. Studies on S–nanoporous carbon have increased rapidly since the report in 2009 of high-capacity Li–S batteries with highly ordered mesoporous carbon and S as the positive electrode.40a According to the definition by the IUPAC, the pore can be classified into three

3576 | Energy Environ. Sci., 2016, 9, 3570--3611

Review

types: microporous (o2 nm), mesoporous (2–50 nm) and macroporous (450 nm). Various nanoporous carbon materials have been utilized to improve the electrochemical properties of S positive electrodes.40 Sucrose as a carbon precursor was applied to form microporous carbon spheres with a very narrow pore size distribution of less than 1 nm.41 S was encapsulated into the micropores of carbon spheres via thermal treatment of a mixture of sublimed sulfur and carbon spheres. The S content in the nanocomposite is 42 wt%. Most of the recently reported research studies have focused on S–mesoporous carbon nanocomposites. The templated method is a powerful approach to create mesoporous carbon materials with a tunable pore size, large specific surface area and interconnected pore network. The templates include silicas (like SBA-15, silica sphere),40 triblock copolymer Pluronic42–45 and MOFs.46–48 For example, a porous carbon with a uniform distribution of mesopores (7.3 nm) and micropores (o2 nm) was synthesized through a soft-template synthesis method, followed by KOH activation.43 Using a multiple template approach also created bimodal carbons with various pore sizes.42,44,45 In recent years, the synthesis of porous carbon materials from MOFs has attracted increasing attention due to the facile preparation procedures, high carbon yield, and unique porous structures.49 MOFs are well known to serve as self-sacrificial templates and the decomposition of MOFs under inert conditions affords porous carbon materials due to the presence of organic ligands in their framework structures.46–48 When S embedded in MOFderived hierarchically porous carbon nanoplates is evaluated as a positive electrode for Li–S batteries, it also demonstrates high specific capacity and an excellent cycling performance.47 Moreover, a MOF as a precursor has the advantages of low solubility and non-volatility, which increases the yield of carbon products from the starting chemicals. In addition, by using a biomineralization-induced self-assembly reaction of polyamine–silicon-resol on a collagen sponge, a carbon capsule monolith possessing ultra-small hollow nanocores and ultrathin nanoshells with a high surface area and porosity was synthesized very recently.50 The carbon capsule monolith can serve as an excellent solvent-restricted, ionic–electronic conductive ‘nanoreactor’ for the electrochemical redox reaction between S and Li. The S–microporous carbon sphere composite presents an excellent cycling performance.41 After 500 cycles, the remained capacity is still above 600 mA h g1. However, it exhibits an unusual discharge profile that does not show the characteristic two voltage plateaus of most reported Li–S batteries. This may be due to the reaction of the carbon with S to form a bonded C–S composite during the heat treatment.51 The design of S–mesoporous carbon nanocomposites was investigated in detail.42–45 The mesoporous carbon prepared using a co-assembly method has a surface area of 2102 m2 g1, a pore volume of 2.0 cm3 g1 and a unique bimodal mesoporous (5.6/2.3 nm) structure.42 The resulting nanocomposite with 60 wt% sulfur shows an initial discharge capacity of 1138 mA h g1 and an excellent rate capability of 6C.

This journal is © The Royal Society of Chemistry 2016

View Article Online

Published on 26 September 2016. Downloaded by SLUB DRESDEN on 29/11/2016 20:29:33.

Review

Fig. 3 (a) The pore size distribution of bimodal mesoporous carbon loaded S with various C/S ratios (modified from ref. 45, copyright permission from the Royal Society of Chemistry). (b) The cyclic voltammograms of S–porous hollow carbon nanocomposites at a sweep rate of 0.2 mV s1 (modified from ref. 50, copyright permission from the Royal Society of Chemistry).

The large internal porosity and surface area of the hierarchical porous carbon are essential for achieving the high utilization of S. In such a bimodal pore structure, (1) the small pores contain the majority of the S mass and aid in suppressing the diffusion of polysulfide species into the electrolyte; and (2) the larger pores facilitate electrolyte ingress throughout the structure. As shown in Fig. 3a, the small-pore volume decreases much faster upon S loading than that of the large pores. This suggests that the smaller pores are preferentially filled at first when S is impregnated into such a carbon host.45 The partially graphitic structure of the carbon framework can provide mechanical stability to the deposited sulfur. No changes in the CV peak positions or peak currents are observed, even after 60 scans, confirming the electrochemical stability of the S–hollow carbon nanocomposites. They retain 91% of their initial capacity after 100 cycles at a moderate C/5 rate.40b Differently, in the case of the porous hollow carbon nanocapsule– sulfur composite (Fig. 3b), the reduction peak current at 2.0 V vs. Li+/Li gradually increases with the cycle number during the initial several CV scans, which could be due to the progressive penetration of electrolytes into all the pores in the electrode.50 To attain a good supply of Li+ ions and enough empty porous space to accommodate the polysulfide anions, partial sulfur loading is also used instead of full S loading.52 2.2

Core–shell structured S based nanocomposites

Conducting polymers are considered as a category of promising supplementary materials to couple with S for Li–S batteries. Polymer-based processes are feasible below 100 1C since polymers are soluble or dispersible in various solvents, which are unlike carbon nanostructures that usually require carbonization processes at a high temperature during the fabrication.11 In addition, the advantages of conducting polymers also include relatively high electronic/ionic conductivity and the feasibility of forming uniform shells. Many experimental approaches have been reported for the synthesis of core–shell structured S–polymer composites. In most cases, these hybrid nanostructures are fabricated using multi-step approaches. For example, a 3D, cross-linked, structurally stable S–polyaniline (PANi) nanotube was reported.53

This journal is © The Royal Society of Chemistry 2016

Energy & Environmental Science

The polyaniline nanotube was formed at first through a onestep self-assembly process in ice water. Next, S and the polyaniline tubes were mixed together and heated to 280 1C. The S was both physically and chemically confined in the nanotubes at the molecular level. Another two similar examples of S–PANi being fabricated using a multi-step method are reported in the literature.54,55 In the former work, a more complicated polyaniline–S yolk–shell nanocomposite was obtained through heating vulcanization of a polyaniline–S core–shell structure.54 In the latter, the C–PANi–S particle congeries were uniformly coated with conductive polyaniline, thus forming the C–PANi–S@PANi composite with a multi-core–shell structure and a conductive polymer network.55 Besides PANi, core–shell structured S composites with other polymers have also been synthesized including polythiophene (PTh),56 polypyrrole (PPy),57 poly-N-vinylcarbazole (PVK),58 polyvinylpyrrolidone (PVP),59 and poly-3,4-ethylenedioxythiophene (PEDOT).60 Among them, the most intriguing example is the fabrication of monodisperse core–shell structured S–PVP by a scalable, room temperature, one-step, and bottom-up approach.59 The PVP molecules can self-assemble into a hollow spherical micelle with their hydrophobic alkyl back-bones pointed toward the interior of the micelle wall and the hydrophilic amide group outward into the water. Its hydrophobic nature made S grow preferentially onto the hydrophobic portion of the PVP micelles. Many PVP molecules can further absorb on the S nanoparticle surface if S is exposed to water during growth owing to the interaction between S and PVP. The PVP shell coating on the monodispersity of the S nanospheres can minimize the polysulfide dissolution. Moreover, empty space has also been engineered into the particles for S to expand inward instead of outward upon lithiation. Hollow carbon can also encapsulate S for effective trapping of polysulfides.61a The hollow carbon arrays were fabricated using anodic aluminum oxide (AAO) templates, through thermal carbonization of polystyrene. The AAO template also facilitates sulfur infusion into the hollow fibers and prevents sulfur from coating onto the exterior carbon wall. Another choice for the fabrication of the core–shell structured S–carbon nanocomposite is to mill commercial sulfur powder with carbon black in N-methyl-2-pyrrolidone (NMP).61b Furthermore, a novel homogeneous precipitation method has been used to synthesize core–shell structured S-activated carbon aerogels.61c This method is based on the following reactions: (x  1)S + (NH4)2S - (NH4)2Sx,

(4)

(NH4)2Sx - 2NH3 + H2S + (x  1)S (x = 2–6)

(5)

(NH4)2S has been widely used to extract sulfur from the sulfur residue in the hydrometallurgy industry. Another emerging avenue for synthesizing core–shell structured S–C is the molecular layer deposition (MLD) technique.62 MLD, an analogy of atomic layer deposition (ALD), can provide precise control of film thickness and allows for conformal film growth over nanostructure substrates.

Energy Environ. Sci., 2016, 9, 3570--3611 | 3577

View Article Online

Published on 26 September 2016. Downloaded by SLUB DRESDEN on 29/11/2016 20:29:33.

Energy & Environmental Science

The soft polymer matrix can accommodate the reversible electrochemical conversion reaction between S and Li2S as well as the volumetric change. The 3D, cross-linked, structurally stable S–polyaniline (PANi) nanotube retains a capacity of 568 mA h g1 after 100 cycles and 432 mA h g1 after 500 cycles at a rate of 1C.53 It is believed that the polymer matrix, PANi, functions as a self-breathing, flexible framework during charge/discharge, and reduces stress and structural degradation. Following this concept, another two excellent examples of S–PANi have been reported recently.54,55 In both cases, diffusion of lithium polysulfides is effectively prevented. Other core–shell structured composites like PThs, PVK, PPy and PEDOT with S also show good electrochemical performances. For example, the specific capacities of S and S–PThs are comparable at a relatively low current density. However, as the current increases, the specific capacity of the core– shell structured S–PTh positive electrode is found to be overwhelmingly better than that of the pure S one.56 This demonstrates that the kinetics of the Li–S redox reaction in the S–PTh electrode is faster than that of the pure S electrodes. S–PTh as a core–shell structure material can improve the conductivities of the composites. Quantum dots (QDs) have been widely studied because of their interesting optical and electrical properties. A novel core– shell S quantum dot/PVK (SQD/PVK) nanocomposite maintains a specific capacity of 687 mA h g1 after 200 cycles at 0.5C, corresponding to an 89.7% capacity retention with only 0.05% capacity degradation per cycle.58 The uniformly dispersed SQDs in the PVK have a large inner void space, which help to accommodate the huge volume expansion during cycling. A better cycling behavior was also achieved using monodisperse PVPencapsulated hollow S nanospheres.59 Over long-term cycling of 1000 cycles at 0.5C, the capacity decay is as low as 0.046% per cycle and the average coulombic efficiency is 98.5%.59 Only 28% of the total S mass was found to be dissolved in the electrolyte after 500 cycles. In contrast, for the electrodes made from virginal S particles, ICP analysis shows a loss of 81% of the total S mass into the electrolyte after the first discharge. This indicates effective trapping of polysulfides by the PVP shells. For all the core–shell structured composites discussed above, the specific capacities are much higher than those of the individual components, especially at high current densities. Additionally, the cycling stability as well as rate capability is dramatically improved. It is concluded that the shell (polymer) could minimize the large volume changes, increase the conductivities and trap polysulfides effectively. However, there are still some differences when different polymers are used. The roles of different conductive polymers (PANi, PPy and PEDOT) on the electrochemical performances of the S electrode were systematically investigated.60 PEDOT was found to be the best among the three conductive polymers to stabilize the cycling performance of the sulfur electrode. To elucidate the interaction between the LixS (0 o x r 2) species and conductive polymers (PEDOT, PPy and PANi), ab initio simulations in the framework of density functional theory were performed. In the case of PEDOT, both the oxygen and sulfur atoms strongly bind

3578 | Energy Environ. Sci., 2016, 9, 3570--3611

Review

with the lithium atom in Li2S to form a chelated coordination structure, which gives a strong binding energy of 1.22 eV. As for PANi and PPy, only a weaker interaction with Li2S was found because there is only separate p–s coordination between the heteroatoms and the lithium atom. The respective binding energies are much lower (0.67 eV for PANi and 0.64 eV for PPy). The significance of this work is that it unraveled different structural configurations of conductive polymer–S composites. This strong binding affinity of PEDOT with LixS (0 o x r 2) can effectively reduce polysulfide diffusion into the electrolyte and thus contribute to a more stable cycling performance compared to PPy and PANi. S–CNT nanohyrids have previously been discussed in Section 2.1.1 but S was mainly coated onto the outer surface. Consequently sulfur is exposed to the electrolyte without any capping, and the dissolution issue is not fully solved. Recently, CNT-encapsulated S, in which S is coated only onto the inner surface of hollow carbon nanofibers instead of the exterior surface, has delivered an initial discharge capacity of around 1560 mA h g1, approaching the theoretical capacity of S.61a S is effectively contained in the high-aspect ratio hollow carbon nanofibers, and its contact with the electrolyte is limited to only the two open ends of the CNTs. 2.3

Nanostructured Li2S electrodes for Li–S batteries

Li2S has a high specific capacity of 1166 mA h g1 and its voltage profile is similar to that of S. In Li–S full batteries, the negative electrode uses lithium metal balanced with the amount of S in the positive electrode. In lab-scale half-cells, lithium is always present in significant excess. The lithium negative electrode presents significant challenges. Since lithium is stored in the Li2S positive electrode, the metallic lithium can be replaced by a high-capacity Li-free negative electrode (such as tin or silicon) when Li2S is used as the positive electrode. Li2S has a very high melting point (4900 1C), allowing it to perform even at higher temperatures. However, the high melting point makes it difficult to synthesize Li2S–C nanocomposites to some extent. Another main hindrance to utilize Li2S is that it is both electronically and ionically insulating. Even worse, it is very sensitive to moisture and needs to be activated when coupled with negative electrodes. In the case of a micron-sized Li2S–C composite, their discharge capacities are below 500 mA h g1 at room temperature.63a,b The Li+ ion diffusivity in Li2S is as low as 1015 cm2 s1.63c There is a potential barrier during the initial charge to fully convert Li2S into sulfur, which can be overcome by applying a high cutoff voltage or a low charge current density.63c Of course, nanostructuring Li2S is an effective approach to decrease the potential barrier and improve its lithium storage capability. A Li2S–mesoporous carbon nanocomposite was synthesized by chemically lithiating S–CMK-3 nanocomposites with n-butyllithium.63d The strong reductant reduces S to Li2S through reaction (6): 2C4H9–Li + 2S 2 C4H9–S–C4H9 + 2Li2S

(6)

The thioether byproduct was evaporated afterwards by heating and Li2S was trapped in the several nanometer–diameter pores

This journal is © The Royal Society of Chemistry 2016

View Article Online

Published on 26 September 2016. Downloaded by SLUB DRESDEN on 29/11/2016 20:29:33.

Review

of the CMK-3. Li2S–C nanocomposites were also prepared via hand grinding and mechanical milling.63e A new bottom-up and hard template technique was proposed to tailor core–shell Li2S–C composites.64a The size of the as-prepared Li2S is around 100 nm and the outside carbon shell has a thickness of 20–50 nm. Interestingly, it was reported that the lithiated graphite electrode could chemically reduce in situ the polysulfide Li2S6 in liquid electrolytes to insoluble Li2S.64b This chemical reduction reaction process is evident from the reduction of the d002 spacing of graphite from 3.56 to 3.37 Å. About 48% of the sulfur in the polysulfide was converted to Li2S. Carbon coating on Li2S nanocomposites can also be achieved through the carbonization or pyrolysis of polymers (such as polystyrene64c,d and polyvinylpyrrolidone64e, f ) and ionic liquid precursors.64g A one-dimensional Li2S–C composite has been reported by combining electrospinning and subsequent pyrolysis technology.65a,b However, chemical synthesis of Li2S–C nanocomposites requires high temperature carbonization in an inert atmosphere to protect Li2S from moisture.64,65a,b In this regard, a solution-based method is developed to prepare the 1D Li2S–C nanocomposites.65c,d Usually, Li2S is first dissolved in ethanol, and then the addition of CNTs into the solution prior to evaporation leads to nucleation and growth of Li2S heterogeneously on the CNT surface sites. Such a process is carried out at low temperature, achieving high Li2S loading, which makes it a low cost and scalable approach. Recently, 2D thermally exfoliated graphene has been utilized to load S via melt-diffusion. The S–graphene nanocomposite is then chemically lithiated with lithium triethylborohydride (LiEt3BH) to form uniformly dispersed Li2S nanoparticles with a size of 8.5 nm.66a Another in situ formed Li2S–graphene composite was prepared through one-pot pyrolysis of a mixture of graphene nanoplatelet aggregates and low-cost lithium sulfate (Li2SO4).66b Part of the graphene nanoplatelet aggregates serve as a reductant to reduce Li2SO4 and create Li2S at an elevated temperature. Meanwhile, the remaining graphene nanoplatelet aggregates are etched in situ into thin graphene sheets via this chemical reaction. A simple drop coating method was also developed to synthesize a free-standing and binderfree Li2S/RGO paper without the use of viscous slurry and metal substrates.66c Nanosized Li2S particles with diameters of 25–50 nm were distributed uniformly inside the long-range interconnected RGO paper. The roles of nitrogen doping in improving the electrochemical performance of graphene was also demonstrated in Li2S electrodes.66d A new-generation highly nitridated graphene (HNG) was synthesized via thermal nitridation of a graphene oxide/polyol mixture. Next, amorphous Li2S3 was coated on both sides of HNG to form sandwiched structures via solvent evaporation. Finally, Li2S3 can be transformed into cubic Li2S by heat treatment at 300 1C for 1 h.66d Besides 2D graphene nanosheets, 3D N- or B-doped graphene aerogel is a good carbon matrix to disperse nanosized Li2S.66e The Li2S–graphene can be further coated with a conformal protective carbon layer via CVD deposition.66f An initial discharge capacity of 950 mA h g1 is achieved in the Li2S–CMK-3 nanocomposite, which is about 80% of the

This journal is © The Royal Society of Chemistry 2016

Energy & Environmental Science

theoretical value and thirty times that of 10 mm-sized commercial Li2S particles.63d The ball-milled Li2S–C nanocomposite shows a high specific capacity close to the theoretical value of Li2S, but large voltage hysteresis appears.63e The core–shell Li2S–C composites demonstrate a good capacity retention of 60% after prolonged 300 cycles at 0.5C.64a The Li2S–graphite presents a stable discharge capacity of B800 mA h g1 over 50 cycles.64b It is believed that the utilization of sulfur can be further improved with smart designed graphite and electrode structures. The enhanced electrochemical properties of Li2S are usually achieved after the carbon coating,64c–g which are attributed to the increased electronic and ionic conductivities and the alleviation of the shuttle effects of intermediate lithium polysulfides during the discharge/charge process. The mesopores in the 1D Li2S–C nanofibers facilitate Li+ ion transport in the active materials, thus favoring the formation of polysulfides and utilization of the active materials, enabling a low activation potential of B2.6 V (vs. Li+/Li) and an extraordinarily stable initial charge plateau of B2.5 V (vs. Li+/Li).65a Similarly, Li2S nanocrystals with a size less than 10 nm formed uniformly in the pores of the CNT networks also show an unprecedented low potential difference (0.1 V) during charge/ discharge.65c Without any electrolyte additives, another Li2S–CNT nanocomposite offers a stable electrochemical performance with up to 90% capacity retention after 100 cycles.65d It was found that the formation of amorphous Li2S could improve the cycling stability when using Li2S–CNFs as the positive electrode.65b The amorphous Li2S is more easily oxidized to soluble polysulfides during charging compared to the crystalline Li2S. The amorphous Li2S discharge product is responsible for the small overvoltage and high reversibility at the high current density.65b The Li2S-thermally exfoliated graphene nanocomposite exhibits an initial capacity of 1119 mA h g1 with a negligible charged potential barrier in the first cycle.66a When coupled with a Si negative electrode, the nanocomposite delivers a high specific capacity of 900 mA h g1. In another report, an in situ formed Li2S–graphene composite shows an initial discharge capacity of 693 mA h g1 and a 40th discharge capacity of 508 mA h g1.66b The flexible Li2S–RGO paper electrode shows an excellent cycle life with a reversible discharge capacity of 816 mA h g1 after 150 cycles at 0.1C.66c A Li2S–N doped graphene composite can increase its ultralong cycle life to 3000 cycles, which is the longest cycle life demonstrated so far.66d Likely, a graphene–Li2S–carbon nanocomposite retains B97% of the initial capacity of B1040 mA h g1 at 0.5C after over 700 cycles.66f The 2D layer structure of graphene plays an important role in transporting electrons and ions, maximizing the utilization of the active Li2S, and accommodating the stress generated from the large volume change during the charge and discharge process. In addition, heteroatom nitrogen or boron doping is beneficial for fast charge transference, and improves the affinity between the nonpolar graphene framework and polar polysulfide species.66e The N atoms in PPy are found to possess a favorable Li–N interaction with Li2S, as shown via ab initio simulations, and

Energy Environ. Sci., 2016, 9, 3570--3611 | 3579

View Article Online

Published on 26 September 2016. Downloaded by SLUB DRESDEN on 29/11/2016 20:29:33.

Energy & Environmental Science

was also proven from the XPS and Raman spectra.67a This enables PPy to bind strongly onto and cover the surface of Li2S to constrain intermediate Li2Sn species during cycling. Thus Li2S–PPy composites demonstrate a high discharge capacity of 785 mA h g1 with stable cycling over prolonged 400 charge/ discharge cycles. Polyvinylpyrrolidone (PVP) is also used with polyethylene oxide (PEO) as a mixed binder for mechanically milled Li2S.67b The PVP is effective in anchoring polysulfide and preventing its dissolution into the electrolyte while the PEO binder is expected to build a Li+ ion diffusion path in the dense electrode. However, the insulating PVP may also act as a barrier for electron transport when it covers the surface of Li2S. From the perspective of battery architecture, some new strategies to efficiently utilize Li2S have been reported very recently.67c,d Specifically, an N-doped CNT film is applied on top of a Li2S–C electrode, which serves not only as a top current collector but also as a barrier layer to effectively impede the polysulfide diffusion and enhance the utilization of active materials (Fig. 4a and b). In contrast, the solution color of the right side changed from colorless to light yellow after 20 h for the normal cell configuration with a bottom current collector (Fig. 4c and d), which means that Li2S8 can diffuse through the virginal separator.67c Another advanced Li–Li2S battery with a new cell architecture is achieved with a dualphase electrolyte separated by a lithium super ionic conductor (LISICON).67d The electrolyte for the Li2S positive electrode and

Review

the electrolyte for the negative electrode are separated by a LISICON film of Li1+x+yAlxTi2xSiyP3yO12 (LATP), as shown in Fig. 4e. The LATP is permeable to Li+ ions while it is impermeable to polysulfide-anions. Therefore, parasitic reactions associated with the polysulfide shuttle process are totally eliminated. Furthermore, this approach provides an effective way to efficiently utilize Li2S with the elimination of self-discharge. Polysulfides constrained in the electrolyte of the positive compartment can serve as redox mediators to chemically react with Li2S. Accordingly, a small amount of carbon additive is enough to activate a large amount of Li2S. However, the LISICON film also presents some disadvantages for the cell performance, which will be discussed in Section 4 in this review.

3. Nanostructured positive electrode materials for rechargeable Li–Se batteries Se, [Ar]3d104s24p4, is chemically similar to S. The initial work of developing Se as a positive electrode for rechargeable batteries was conducted in the Argonne National Laboratory in 2012.68 The major allotropes of Se include trigonal Se (t-Se, thermodynamic stable phase constructed from Se chains) and various monoclinic Se (constructed from Se8 rings).69 The overall redox reaction of Se is shown in eqn (7): Se + 2Li+ + 2e 2 Li2Se

Fig. 4 The illustration of the configuration of a Li–Li2S@C battery (a) with a top current collector and (b) with a bottom current collector. Photos showing the different diffusion situations of polysulfides in H-type glass containers through (c) a virginal separator and (d) a separator with the nitrogen-doped carbon nanotube (N–CNT) film after 20 h (modified from ref. 67c, copyright permission from the Royal Society of Chemistry), and (e) schematic architecture for the Li–Li2S battery composed of a dual-phase electrolyte using a ceramic separator (modified from ref. 67d, copyright permission from the Royal Society of Chemistry).

3580 | Energy Environ. Sci., 2016, 9, 3570--3611

(7)

Although Se has a lower theoretical gravimetric capacity (675 mA h g1) than S, its higher tap density (ca. 2.5 times that of S) could offset the deficiency and exhibit a high theoretical volumetric capacity density (3253 mA h cm3), comparable to that of S (3467 mA h cm3).70 Moreover, the electronic conductivity of Se (1  103 S m1) is considerably higher than that of S (5  1030 S m1). Unlike the Li–S system discussed in Section 2, Li–Se batteries can be cycled to high voltages (up to 4.6 V) without failure.68 It has been proven that organic electrolytes have a great impact on the electrochemistry of Li–Se batteries. Currently, two types of electrolytes are generally applied in research on Li–Se batteries. The first type is a carbonate-based electrolyte, such as LiPF6 in ethylene carbonate (EC), dimethyl carbonate (DMC) and so on. The second type is an ether-based electrolyte that is a Li salt (typically LiTFSI) in ether mixtures, like 1,3-dioxolane (DOL) and 1,2-dimethoxyethane (DME). A study on the lithiation/delithiation mechanism of Li–Se batteries in an ether-based electrolyte shows that its electrochemical reaction process is almost the same as the well-known stepwise reaction in Li–S batteries.71 Polyselenides are also soluble in ether. In contrast, Se is directly reduced to Li2Se during discharge in carbonate-based electrolytes without an intermediate phase being detected.72 The reason for this is that the redox products Se and Li2Se, as well as lithium polyselenides, are insoluble in these electrolytes. Currently, various nanostructured Se nanocomposite materials have been developed as the positive electrode for Li–Se batteries,

This journal is © The Royal Society of Chemistry 2016

View Article Online

Review

Energy & Environmental Science

including 1D nanocomposites (Se–CNTs, CNFs, and Se–PANi) and 3D nanocomposites (Se–porous C and Se–C spheres).

Published on 26 September 2016. Downloaded by SLUB DRESDEN on 29/11/2016 20:29:33.

3.1

1D nanocomposites

As one of the most common and effective methods, the electrospinning technique is widely applied to fabricate various Se-based 1D nanocomposites.73–75 For example, a 1D fiber composed of Se-carbonized polyacrylonitrile was synthesized by heating selenium–polyacrylonitrile fibers via an electrospinning technique at 600 1C.73 During the thermal treatment, the reactions of PAN with selenium are as follows: (8) The Se molecules can be confined by N-containing carbon ring structures in the form of energy-storing selenium side chains in the carbonized PAN matrix. The use of a hydrothermal synthesis method has also been successfully utilized to synthesize 1D Se–C nanocomposites.76,77 First, sodium selenite is treated with ascorbic acid to obtain Se as shown in eqn (9): Na2SeO3 + 2C6H8O6 + 4H+ - C6H7NaO6 + Se + 3H2O

(9)

Then the mixture is transferred into a Teflon-lined autoclave followed by hydrothermal treatment at 200 1C for 24 h. Ascorbic acid is utilized as both a reductant and a carbon source to form a carbon shell.77 Fig. 5a and b present the SEM and TEM micrographs of the as-synthesized Se–C core–shell nanocomposites. They show exclusively a 1D structure with a micrometer length distribution and diameters of 500–700 nm. These SEM and TEM micrographs also demonstrate the structural characteristics of the core–shell structures with an inner core and an outer shell. The outer carbon shell can enhance the conductivity of the core–shell composites, and a thicker selenium core indicates

that these 1D nanocomposites can load enough active materials. Other 1D core–shell nanocomposites of graphene encapsulated Se–polyaniline (G@Se–PANi) have been designed and synthesized under the conditions of low temperature without heating.78 The conductive PANi is at first closely covered onto the surface of the Se nanowires via an in situ chemical oxidative-polymerization method. In the acidic solution, the surface of the PANi shell is positively charged because of the existence of amino functional groups. Thus electrostatic interactions do exist between negatively charged hydrophilic GO nanosheets and positively charged PANi shells. The GO is finally chemically reduced to reduced graphene oxide (RGO). In this nanocomposite, Se nanowires are well-sealed in the PANi layer with a thickness of 25 nm forming a core–shell structure and the Se–PANI core–shell nanowires are uniformly encapsulated in the graphene nanosheets (Fig. 5c). The uniform distribution of Se, N and C elements is confirmed by the elemental mapping image. The Se-carbonized polyacrylonitrile fiber electrode shows a discharge plateau at a low voltage in the first cycle.73 This is because additional energy is needed to dissociate Se from the complex bond between selenium and the heterocyclic compound during the reaction of Se and Li+ ions, which leads to a more negative potential. The initial discharge and charge capacities of the Se–CNFs are 1159 mA h g1 and 659 mA h g1, respectively, showing a 56.9% initial coulombic efficiency.74 A similar phenomenon is also found in 1D Se–CNF–CNT nanocomposites.75 It is believed that the formation of an irreversible solid electrolyte interphase is one reason for the low coulombic efficiency. However, this 1D nanocomposite delivers an excellent cycling behavior and still shows an exceptionally high capacity of 516 mA h g1 after 900 cycles. The original textural properties (shape, size, and structural integrity) could be well retained after 900 cycles, which indicates high structural stability. Of course, the surface of the fibers becomes rough after extended cycling due to SEI formation and some residual electrolytes.

Fig. 5 (a) SEM and (b) TEM micrographs of 1D Se–C core–shell nanocomposites, (c) TEM micrograph and elemental mapping images of graphene encapsulated Se–PANi core–shell nanowires (G@Se–PANi), and cyclic voltammogram profiles of the (d) 1D Se–C core–shell nanocomposites in etherbased electrolyte and (e) G@Se/PANi in carbonate-based electrolytes (modified from ref. 77 and 78, copyright permission from Elsevier).

This journal is © The Royal Society of Chemistry 2016

Energy Environ. Sci., 2016, 9, 3570--3611 | 3581

View Article Online

Published on 26 September 2016. Downloaded by SLUB DRESDEN on 29/11/2016 20:29:33.

Energy & Environmental Science

Review

Fig. 5d shows the CV curves of the 1D selenium/carbon core– shell nanocomposites at a scan rate of 0.2 mV s1 using Li salt in ether mixtures as the electrolyte. It has two obvious cathodic peaks and one anodic peak. The cathodic peaks at around 2.2 and 2.0 V (vs. Li+/Li) during discharge correspond to the reduction of Se to high-order polyselenides Li2Sen (n Z 4) and further reduction to Li2Se2 and Li2Se. This is different from the single-plateau mechanism of the Li–Se battery in the carbonatebased electrolytes.72,78 Fig. 5e exhibits the CVs of the G@Se–PANi electrode for the first 3 cycles in 1.0 M LiPF6 in ethylene carbonate (EC)/diethyl carbonate (DEC). There is only one pair of reversible redox peaks for G@Se–PANi, suggesting that a single phase change reaction takes place during the lithium uptake/ release processes in the carbonate-based electrolyte. Due to the insolubility of Se, Li2Se and lithium polyselenides in the EC/DMC electrolyte, no shuttle effect was observed. Thus, the cycling behavior of G@Se–PANi in carbonate-based electrolytes is much better than that of 1D Se/carbon core–shell nanocomposites in ether-based electrolytes.77,78 3.2

3D nanocomposites

Porous carbon spheres have been shown to be promising matrix materials which can encapsulate Se to form a 3D hierarchical architecture.79–82 Traditionally, the hard template method is one of the most popular methods for the synthesis of porous carbon spheres.80,81 The hard template is usually a silica microsphere. Besides the hard template method, a hydrothermal process combined with a KOH activation technique has also been used to prepare porous carbon nanospheres.79,82 Then Se is loaded into porous carbon nanospheres using a melting-diffusion method similar to a S–C system. In addition, nanoporous Se could be prepared using a simple mechanical method adopting nano-CaCO3 as a template, as reported by our group.83 The prepared nanoporous structure consists of nanopores and nanowalls. Such a hierarchical architecture can avoid the aggregation and retain the large surface area. Like the S positive electrode, combining Se with carbon materials is an effective means to improve the specific capacity and rate capability. The initial discharge capacity of the Se-porous hollow carbon sphere (PHCS) composite is 590 mA h g1, which is 87.4% of the theoretical capacity.80 After 50 cycles, the remaining capacity is 338 mA h g1, much higher than that of the pristine Se electrode (only 75 mA h g1 after 50 cycles). The porous hollow framework of the PHCS not only confines selenium and minimizes the loss of lithium polyselenide in the ether-based electrolyte, but also facilitates the transport of Li+ ions. Mesoporous carbon microspheres (MCMs) with tunable pore sizes (3.8 nm, 5 nm, 6.5 nm and 9.5 nm) were prepared and further utilized as the conducting host to incorporate selenium for Li–Se batteries.81 Among the Se–MCM series, MCM–Se with a pore size of 3.8 nm presents an initial discharge capacity of 513 mA h g1 at 0.5C with the best cycling performance. The smaller pores could constrain the diffusion of polyselenide species via a stronger confinement effect in the ether-based electrolyte, thus improving the cycling performance. Interestingly, a molecule transformation from ring-like Se8 to

3582 | Energy Environ. Sci., 2016, 9, 3570--3611

Fig. 6 XPS spectra of Se3d and C1s: (a) and (b) pristine Se confined porous carbon nanosphere (Se/PCN) electrode, and (c) and (d) the Se/PCN electrode after the 1st discharge/charge cycle at 3.0 V vs. Li+/Li. (e) Rate capability of Se/PCNs. The electrolyte was 1 mol l1 LiPF6 in a mixed solvent of EC and DMC (1/1, v/v) (modified from ref. 82, copyright permission from Elsevier). (f) Cycling behaviors of nanoporous Se (NP–Se) and commercial Se particles (CP–Se). The electrolyte was a standard 1 mol l1 LiPF6 solution in a 1 : 1 : 1 mixture of EC, DMC, and DEC (modified from ref. 83, copyright permission from the Royal Society of Chemistry).

chain-like Sen in porous carbon nanosphere (Se/PCN) nanocomposites was found using X-ray photoelectron spectroscopy (XPS) (Fig. 6a–d).82 In addition to the normal two peaks at 56.0 and 55.1 eV for the fresh electrode (Fig. 6a), two more peaks at 59.0 and 58.1 eV were observed after the 1st discharge/charge cycle (Fig. 6c). The authors thought that these two new peaks together with an additional peak at 290.2 eV in Fig. 6d may come from the Se–Se bonds located at the chain end of the chain-like Sen molecules or from the reactions between Se and the carbonyl groups producing some products containing Se–C and/or Se–O bonds. Similar to Se–MCMs, a high rate behavior is also attained for the Se/PCN nanocomposite, as shown in Fig. 6e. It demonstrates an extraordinarily fast charging/ discharging ability, which can be charged (discharged) with 67% initial capacity (at 0.2C) in 4 min (15C), respectively. Remarkably, it also exhibits an impressive cycling stability over 1200 cycles with a capacity decay as low as 0.03% per cycle. The superior overall battery performance is ascribed to the strong adsorbing ability of Se/Li2Sex provided by the microporerich carbon structure.82 Recently, without adding any carbon, satisfactory electrochemical performances have been obtained for nanoporous selenium (NP–Se) electrodes.83 Due to the special porous structure of the NP–Se, and the large contact area between the building blocks and the electrolyte, it exhibits a reversible capacity of 338 mA h g1, which is much higher

This journal is © The Royal Society of Chemistry 2016

View Article Online

Review

Energy & Environmental Science

Published on 26 September 2016. Downloaded by SLUB DRESDEN on 29/11/2016 20:29:33.

than that of CP–Se (Fig. 6f). Furthermore, the NP–Se shows a capacity of 206 mA h g1, while the CP–Se only retains 60 mA h g1, which is ascribed to those nanopores and nanowalls avoiding the aggregation and alleviating the structural strain associated with Li+ ions during charging and discharging.

4. Nanostructured positive electrode materials for aqueous rechargeable lithium batteries The experience from Li-ion battery technology shows that the use of flammable organic electrolytes might lead to thermal runaway and safety accidents. In this regard, aqueous rechargeable lithium batteries (ARLBs) have been investigated and demonstrated in the past two decades. ARLBs were first proposed in 1994.84 However, the inferior capacity retention and low energy density seriously limit their practical applications. A decade later, Wu and his coworkers demonstrated an ARLB with a promising electrochemical performance by using Li2SO4 as the electrolyte, and LiV3O8 and LiCoO2 as the negative and positive electrodes, respectively.4 After that, research on ARLBs quickly became a hot topic.85–87 ARLBs can overcome some disadvantages of Li-ion batteries. It is inherently safe by avoiding the use of flammable organic electrolytes. Moreover, the cost of the aqueous electrolyte is low because the expensive salts can be replaced with cheap ones. In addition, the ionic conductivity of aqueous electrolytes is high, about two orders of magnitude higher than that of organic electrolytes, which ensures high rate capability and thus high specific power.87 Compared to the key factors (power and energy densities, cycleability and safety) when a battery is used in electric vehicles, the cost together with safety is more emphasized in large-scale stationary energy storage systems, such as smart grids. With this in mind, the aqueous rechargeable batteries have potential applications in stationary energy storage systems in the future. So far, some reviews were published with the scope on ARLBs.86a,88–90 The discussions on ARLBs, here, are summarized in Table 2. In generation-I ARLB systems, intercalation compounds

Table 2

with an intercalation potential of 3–4 V (vs. Li+/Li) are usually used as the positive electrode materials.85,91–95 To achieve high power ARLBs, nano-structuring electrode materials were designed. For example, a LiMn2O4 nanotube with a preferred orientation of (400) planes was prepared by using CNTs as the sacrificial templates by Wu’s group.94 Its charge capacity reaches 60 mA h g1 (54% of the capacity) at a charge rate of 600 C as a positive electrode material for ARLBs.94b However, the energy density of generation-I ARLB systems is still much lower than that of conventional lithium ion batteries due to the narrow electrochemical window of water. In generation-II ARLB systems, Li or Mg metal coated with LISICON films and gel polymer electrolytes is used as a negative electrode.96–99 Meanwhile, the positive electrode materials are still intercalation compounds with various nanostructures (like LiMn2O4,96 LiCoO2 nanoparticles97 and macroporous LiFePO498), and Li2SO4 aqueous solution is used as the electrolyte. These battery systems deliver an output voltage of about 3–4 V, about 3 times higher than that of traditional ARLBs, and their energy density was calculated to be around 300 W h kg1. An important element for generation-II ARLB systems is the LISICON plate (Li1+x+yAlxTi2xSiyP3yO12) that separates the negative electrode (Li or Mg) and the aqueous solution on the positive electrode side. Since protons are smaller than Li+ ions, there is a concern that protons may transfer through the LISICON plate and be reduced by lithium metal to form hydrogen, which is detrimental for long term stability and presents safety issues. However, all the recent investigations indicate that the proton cannot go through the LISICON glass, and further confirm its stability in the neutral electrolytes since the content of protons in neutral solution is 107 mol1 or 0.1 ppm, much less than the content of HF in the organic electrolytes (usually about 10 ppm).100a,b However, the LISICON is not stable when brought into contact with lithium metal. To solve this problem, our group designed a sandwiched gel polymer electrolyte which was coated between lithium metal and LISICON.96–99 Generation-III ARLB systems were also suggested and were fabricated using Br3/Br (or I3/I) couples as the positive electrode and metallic Li as the negative electrode as shown in eqn (10): Br3 2 Br + Br2 + 2e 2 3Br

(10)

Some characteristics of several types of aqueous rechargeable lithium batteries (ARLBs)

Type

Electrode material or composition

Preparation method for the positive electrode material

Voltage (V) Rate behavior

Generation I

LiCoO2 nanoparticles Porous LiMn2O4 V2O5//LiMn2O4 MoO3//LiMn2O4 LiMn2O4 nanotube Zn//LiNi1/3Co1/3Mn1/3O2–CNTs–RGO PbSO4//LiMn2O4 nanocubes

Sol–gel Template Sol–gel Sol–gel Template Solid state reaction + hydrothermal Hydrothermal + solid state reaction

0–1.8 0–1.8 0–1.8 0–1.95 0–1.8 1–1.9 0–1.8

10 A g1 10 A g1 0.2 A g1 — 66 A g1 0.5 A g1 3 A g1

Generation II

Li–LISCON–GPE//LiMn2O4 nanoparticles Li–LISCON–GPE//LiCoO2 nanoparticles Li–LISCON–GPE//macroporous LiFePO4 Mg//LiFePO4

Sol–gel Sol–gel Template Template

3.7–4.25 3.5–4.3 2.5–4.0 1.75–3.5

0.1 A g1 0.15 A g1 1.5 A g1 0.05 A g1

— —

3.7–4.25 2.75–4.25

12.32 mA cm2 12 mA cm2

Generation III Li//Br2 Li//I2

This journal is © The Royal Society of Chemistry 2016

Cycling performance Ref. 40 10 000 500 150 1200 50 110 30 20 50 20 100 100

91 92 93a 94a 94b 95a 95b 96 97 98 99 100c 100d

Energy Environ. Sci., 2016, 9, 3570--3611 | 3583

View Article Online

Published on 26 September 2016. Downloaded by SLUB DRESDEN on 29/11/2016 20:29:33.

Energy & Environmental Science

Review

The tribromide/bromide (Br2/Br) couple has a stable redox potential of 4.1 V (vs. Li+/Li), and a specific capacity of 335 mA h g1.100b,c The energy density of the Generation-III ARLB system, based on the mass of Li and Br2, is about 1220 W h kg1, much higher than those of the 2G ARLBs and lithium ion batteries. It should be noted that using carbon materials (such as AC and acetylene black) as the catalyst can promote the redox reaction of Br2/Br. The commercially available lithium superionic conductor (LISICON) has a Li-ion conductivity in the order of 104 S cm1 which is lower than that of Li+ or Br in the aqueous phase (102 S cm1), and comparable to that of Li+ ions in the gel polymer electrolyte (104 S cm1). To enhance the ionic conductivity of LISICON films and gel polymer electrolytes, the development of a suitable solid electrolyte is still ongoing.

5. Nanostructured positive electrode materials for Li–O2 batteries The Li–O2 battery was first introduced in 1996, and it was composed of lithium, an organic-impregnated polyacrylonitrile electrolyte, and a carbon electrode.101 The Li–O2 battery with organic carbonates as the electrolyte was demonstrated, which results in decomposition of electrolytes irreversibly at the cathode upon discharge.102 Later, ether electrolytes were found to be more stable with the reduced O2 species during the electrochemical process than organic carbonate. After that, research on Li–O2 batteries quickly became a hot topic. A typical rechargeable nonaqueous Li–O2 battery is comprised of a Li metal as the negative electrode, a Li conducting organic electrolyte and a catalyst with oxygen as a positive electrode. According to its electrochemical reaction (11), it shows an extremely large theoretical specific energy of above 3500 W h kg1 including the mass of lithium and oxygen.3 This has attracted scientists worldwide to pursue this fascinating energy storage system. 2Li+ + O2 + 2e 2 Li2O2

(11)

At the positive electrode, O2 from the atmosphere enters the porous catalyst and is reduced to Li2O2 at the electrode surface during the discharging process (oxygen reduction reaction, ORR). The peroxide is then decomposed upon charging (oxygen evolution reaction, OER).3 However, the real mechanism for nonaqueous Li–O2 batteries is very complicated and the electrolyte plays a crucial role. For example, it was reported that carbonate solvents could be attacked by oxygen radicals, which leads to the formation of lithium carbonate (Li2CO3), instead of the previously claimed discharge product Li2O2.103,104 Similarly, the decomposition of organic carbonates was also observed.105,106 Besides the electrolyte, critical challenges include low round-trip efficiency (power efficiency), high overpotentials, poor cycle life and low power density or rate capability. Recently, there have been several high quality reviews on Li–O2 (or air) batteries.3,107,108 Now it is well realized that one of the bottlenecks for a non-aqueous Li–O2 battery is the positive electrode (also named as the air electrode), in which the sluggish ORR/OER kinetics would increase the overpotential.

3584 | Energy Environ. Sci., 2016, 9, 3570--3611

Both the materials and architecture of the air electrodes influence their performance significantly. The discussion, here, will be concentrated on designing a nanostructured positive electrode to develop efficient electrocatalysts to promote the ORR and OER for high power density, good cycling stability and high energy efficiency. It should be pointed out that most reported Li–air batteries are tested in pure oxygen in order to avoid contaminants from air (especially water). Therefore, ‘‘Li–air’’ and ‘‘Li–O2’’, and ‘‘air electrode’’ materials are usually used as the same in this review without differentiation. 5.1

0D nanocatalysts

Noble metals, like Pt and Au, have been demonstrated to show excellent catalytic performances in Li–O2 batteries. Pt–Au nanoparticles were synthesized by reducing HAuCl4 and H2PtCl6 in oleylamine and then loaded onto Vulcan carbon (VC) to yield 40 wt% PtAu/C.109 To remove the surfactant, the catalyst was then thermally treated at 250 1C in dry air. Transmission electron micrographs (TEMs) show that Pt–Au nanoparticles are uniformly distributed on carbon (Fig. 7a) with a volumeaveraged diameter of 7.3 nm. In addition, a green synthetic process was reported to prepare Au–Pt core–shell nanoparticle chains by the integration of electrical dipole-induced selfassembly and the simultaneous epitaxial growth of the Pt shell.110 In this work, compressed hydrogen was used as the driving force for producing the assembled nanochain networks by unbalancing the attractive van der Waals force and residual electrostatic repulsions, and increasing the chance of Brownian collisions. As shown in the TEM micrograph (Fig. 7b), the nanoparticles are connected to one another and form a linear assembly with observable sub-branches. In addition, the shell thickness/coverage could be easily controlled by simply adjusting the dosage of second metal ions (such as Pt) in solution. Besides noble metals, another recent example of 0D nanoparticles as the positive electrode for Li–O2 batteries, is a spinel CuCo2O4 nanocrystal fabricated using a simple and low-cost urea combustion method.111 Pt–Au/C nanoparticles exhibit bifunctional catalytic activity, where Au and Pt atoms are primarily responsible for the ORR and OER kinetics in Li–O2 batteries, respectively.109 The Au–Pt core–shell chain electrodes show lower charge and discharge overpotentials in comparison with the monodispersed Au–Pt core–shell nanoparticles and the mixture of Au and Pt nanoparticles.110 Furthermore, the Au–Pt core–shell network electrode exhibits a good cycling stability due to the interconnected bimetal nanostructures, which greatly increase the electron transport in the electrocatalysts compared to the separated noble metal nanoparticles. These nanometer-scale catalysts based on precious metals109,110 were reported to have higher discharge voltages than metal oxides.112,113 However, binary metal oxide (CuCo2O4) catalysts could effectively reduce the charge/discharge polarization of Li–O2 batteries.111 The air electrode with CuCo2O4/KB shows a high capacity of 7962 mA h g1 with a discharge/recharge voltage gap of 0.95 V at 0.05 A g1. With a cut-off capacity of 1000 mA h g1 at 0.4 A g1, this battery could be discharged and charged for

This journal is © The Royal Society of Chemistry 2016

View Article Online

Published on 26 September 2016. Downloaded by SLUB DRESDEN on 29/11/2016 20:29:33.

Review

Fig. 7 TEM micrographs of (a) Pt–Au/C (modified from ref. 109, copyright permission from the American Chemical Society) and (b) Au–Pt core–shell nanoparticle chains (modified from ref. 110, copyright permission from the Royal Society of Chemistry), (c) CV curves of Li–O2 batteries with Vulcan XC-72 (VC), m-TiN/VC, and n-TiN/VC as positive electrodes (catalysts) under an O2 atmosphere at 0.05 mV s1 and (d) FT-IR spectra of the positive electrode at the 5th cycles (modified from ref. 116, copyright permission from the Royal Society of Chemistry).

more than 20 cycles, which indicates the superior ORR and OER catalytic activities of CuCo2O4. Transition metal nitrides (or composites) and nitrogen doped carbon materials have been demonstrated to be efficient catalysts in replacing noble metals for the ORR in aqueous fuel cells.114,115 As one of the metal nitride families, TiN possesses high electronic conductivity and good electrochemical activity, which enables it to be widely applied in electrochemistry. The applicability of TiN in non-aqueous Li–O2 batteries was recently reported.116 Fig. 7c shows CVs of Li–O2 batteries with Vulcan XC-72 (VC), micro-sized TiN and VC (m-TiN/VC) and TiN nanoparticles supported on VC (n-TiN/VC). In the cathodic scan, n-TiN/VC presents a higher ORR current than the other two samples (m-TiN/VC and VC), which suggests that n-TiN/VC has better ORR catalytic activities. When discharged to 2.0 V (vs. Li+/Li), n-TiN/VC presents a larger capacity of 6407 mA h g1 in comparison with m-TiN/VC and VC. This enhanced performance can be ascribed to the high catalytic activity of the 0D TiN nanoparticles and the intrinsic contact between them and VC. The superior catalytic activity and high conductivity of TiN make it applicable as an alternative support to carbon in nonaqueous Li–O2 batteries. However, electrolyte decomposition is observed when it is used as a catalyst (Fig. 7d). The sign of Li2CO3 can be found in the discharged electrodes (both n-TiN/VC and m-TiN/VC), but not in the recharged electrodes. The catalytic mechanisms of TiN and many other catalysts (such as metal oxides) in the ORR or OER in non-aqueous Li–O2 batteries are still unknown. A further fundamental understanding of the reaction mechanisms of Li–O2 batteries is required. 5.2

1D nanocatalysts

1D carbon (like CNFs and CNTs) and its supported metal or metal oxide have received attention as catalysts in Li–O2 batteries.

This journal is © The Royal Society of Chemistry 2016

Energy & Environmental Science

For instance, a 1D nanostructured air electrode consisting of nitrogen doped CNT arrays (CNTAs) and a backing carbon paper was demonstrated, in which the sparsely populated, vertically aligned and straight CNTAs were directly grown on carbon fiber papers using electrodeposited Ni catalysts followed by plasma enhanced chemical vapor deposition (PECVD).117 In the case of the 1D carbon supported noble metal, Pt–CNTs have been actively studied in non-aqueous Li–O2 batteries.118,119 The CNT electrode was coated with Pt nanoparticles by sputtering and the percentage of Pt on the CNTs is approximately 66 wt%.118 After the deposition of the Pt nanoparticles, the cross-linked CNT sheet with the open framework was maintained, which ensured that all the CNT surfaces could be used as active reaction sites. The CNTs are approximately 15 nm thick, and the size of the Pt nanoparticles is about 3–4 nm. A CNT supported with Ru nanoparticles was designed using a wet chemical method.120 The supported Ru nanoparticles can function as both ORR and OER catalysts while the CNTs are interpenetrated together to form a porous network, which can facilitate the transport of both electrons and oxygen. There is a significant disadvantage to using noble metal catalysts (such as Pt and Ru) due to their high cost. The development of low cost active catalysts is critical for the effective operation of Li–O2 batteries. Thus, porous nonprecious Fe–CNF catalysts have also been developed using the electrospinning method.121 The acid leaching and repyrolysis of the Fe based nanofibrous catalyst introduced porosity into the carbon nanofibers while maintaining the active sites. Another 1D nanostructured positive electrode material for low-cost Li–O2 batteries is CNTs supported with metal oxides. For example, a RuO2 shell was uniformly coated on the surface of the CNT core using a simple sol–gel method.122 This 1D nanocomposite consists of a CNT core with a diameter of about 15 nm and a RuO2 shell with a thickness of about 3.5 nm. Additionally, in situ deposition of MnO2 nanorods onto carbon was carried out at room temperature through a simple reaction (12) between manganese sulfate and potassium permanganate.123 3MnSO4 + 2KMnO4 + 2H2O - 5MnO2 + K2SO4 + 2H2SO4 (12) Interestingly, a 1D a-MnO2/ramsdellite-MnO2 nanocomposite was obtained via acid treatment of the Li2MnO3 nanorods.124 Li2MnO3 (‘Li2OMnO2’) is usually used as the positive electrode material for Li-ion batteries with a high specific capacity.125 The material H[H0.18Li0.15Mn0.67]O1.9, obtained by the delithiation of layered Li2MnO3 using acid, shows a 2-fold increase in oxygen reduction ability versus Li2MnO3.126 Manganite (MnOOH) materials are also active electrocatalysts for the Li–O2 battery. The high aspect ratio a-MnOOH nanowires have been successfully synthesized through a simple one-step hydrothermal process using only potassium permanganate (KMnO4) and polyvinylpyrrolidone (PVP).127 Aiming at searching for new ORR catalysts, a series of perovskite oxides were developed as a promising electrocatalyst for Li–O2 batteries. These oxides have a high electronic/ionic conductivity and catalytic activity. Recently, porous La0.75Sr0.25MnO3 nanotubes were prepared by combining the electrospinning technique with a heating method.128

Energy Environ. Sci., 2016, 9, 3570--3611 | 3585

View Article Online

Published on 26 September 2016. Downloaded by SLUB DRESDEN on 29/11/2016 20:29:33.

Energy & Environmental Science

The Pt@CNT air electrode could successfully retain the optimal air pathways and provide the conditions for effective catalytic activity.118 It delivers a cycling performance over 100 cycles, with 1000 mA h g1 at a high current rate of 2 A g1. It was found that the OER-enhancing Pt catalyst could control the morphology of the discharge products, thereby contributing to the enhanced cycling stability. Ru@CNTs and RuO2@CNTs have also been reported to be good candidates as both the ORR and OER catalysts in non-aqueous Li–O2 batteries.120,122 The resultant Ru@CNTs were found to efficiently reduce the charge overpotential below 4.0 V at 0.5 A g1.120 Carbon materials exhibit sufficient catalytic activity for the ORR but low catalytic activity for the OER. More seriously, a carbon electrode can react with Li2O2 to form Li2CO3. The core–shell structured RuO2@CNTs could effectively prevent direct contact between the CNTs and the discharge product Li2O2, thus avoiding or reducing the formation of Li2CO3. As discussed above, the corrosion of carbon support materials has been identified as the major factor for cell failure in most carbon-supported Li–O2 batteries. One of the alternative routes is to use an air electrode based on carbon free materials (such as MnO2, perovskite oxides, and polymers) which act as both a catalyst and support. Although porous a-MnO2 delivers a low capacity (1400 mA h g1 at 0.1 A g1), the charge potential has been lowered to 3.5 V which is one of the lowest charge potentials reported for Li–O2 batteries.123 Different from a-MnO2, the acid-treated Li2MnO3 electrode consists of a composite of a-MnO2/ramsdellite-MnO2 that delivers a significantly high capacity up to 5000 mA h g1 during the early cycles.124 The Li2O-stabilized and partially lithiated electrode component, 0.15Li2Oa-LixMnO2, that has Mn4+/3+ characteristics may facilitate the Li2O2/Li2O discharge/charge chemistries providing dual electrode/electrocatalyst functionality. Besides MnO2, the MnOOH nanowires show high catalytic activity for the ORR and OER in nonaqueous electrolytes.127 Recently, perovskite-type compounds as ORR catalysts have also attracted significant attention in the field of Li–O2 batteries. The porous La0.75Sr0.25MnO3 nanotubes significantly suppressed the ORR and especially OER overpotentials and thus improved the round-trip efficiency.128 The PPy nanotube supported air electrode exhibits a better cycling stability and rate capability than the conventional carbon supported electrode.129 The PPy also has many advantages, such as high electronic conductivity, high chemical and electrochemical stability, easy synthesis, and especially higher polarity than the carbon materials. 5.3

2D nanocatalysts

Graphene nanosheets have been identified as excellent catalysts for ORR and OER due to their extraordinary electronic conductivity (103 to 104 S m1), large surface area (theoretical value of 2630 m2 g1) and controllable surface defects. Most of the graphene nanosheets were prepared via the chemical reduction of exfoliated graphite oxide nanosheets.130–132 Then some catalysts (like MnO2131 and Ru132) can nucleate and grow on the surface of graphene. For example, a 2D a-MnO2/graphene hybrid was synthesized via the redox reaction of graphene and

3586 | Energy Environ. Sci., 2016, 9, 3570--3611

Review

KMnO4 under acidic conditions.131 First, oxygen-containing groups (–COOH, –CO and –COH) were firstly introduced on the surfaces of graphene via a moderate acid treatment. Then, a-MnO2 nanorods were generated by the reaction of MnO4 in the solution and Mn2+ resulted from the reaction of MnO4 with carbon. The graphene nanosheets acted as both support and reductants. Besides the chemical reduction of GO, graphene can also be synthesized by electrochemical exfoliation of the graphite papers in aqueous solution of Na2SO4, followed by annealing in an inert gas atmosphere to control the amount of structural defects.133 During the electrochemical exfoliation process, ions intercalated into graphene interlayers result in the expansion of the interlayer spacing. Such a process is advantageous with high yield, environmentally friendliness, and low cost. Both sides of other 2D carbon or carbon-like structures (such as carbon nanosheets and MoS2 nanosheets) can also be accessed by oxygen and used as the air electrode for Li–O2 batteries.134 Novel graphene oxide-derived carbon and carbon nanosheet supported Co nanoparticles were prepared using a sol–gel method.134a,b In the case of the Co–C nanocomposite, it shows an interesting structure where hundreds of ‘carpenterworms’, carbon nanotubes overspread on the 2D carbon nanosheets.134b Molybdenum disulfide nanosheets decorated with gold nanoparticles (MoS2/AuNPs) were prepared using hydrazine to reduce (NH4)2MoS4 and HAuCl4 synchronously in an aqueous solution at 200 1C.134c The AuNP, as a novel inhibitor material, could mediate the growth of layered MoS2 materials. The pure MoS2 synthesized in the same hydrothermal system without adding the HAuCl4 precursor consists of uniform nanoflowers. However, the morphology of MoS2 was transformed from nanoflowers to nanosheets after adding the HAuCl4 precursor to the reaction system. Ternary metal oxides with 2D nanosheet architecture, like NiCo2O4135,136 and ZnCo2O4137 have also been successfully obtained as catalysts for Li–O2 batteries. Both NiCo2O4 and ZnCo2O4 nanosheets were synthesized via a facial hydrothermal method followed by low temperature calcination.136 Pure graphene nanosheets have been investigated as air electrodes for Li–O2 batteries with alkyl carbonate electrolytes.130 The graphene nanosheet electrodes exhibit a much better cycling stability and lower overpotential than those containing VC (vinyl carbonate). The pores between the carbon nanoparticles in the VC air electrode were filled by the reduction products after discharging. This could reduce further oxygen diffusion, leading to an increase of the polarization. However, the graphene air electrode presents sufficient pores between the graphene agglomerates, which could facilitate the oxygen diffusion and electrolyte impregnation. Furthermore, the pores inside the air electrode are only partially filled by the insoluble reduction products after cycling. This suggests that the capacities of the air electrodes are associated with the pore volume and the special porous structure. This hypothesis has been confirmed in another report.131 The a-MnO2/graphene hybrid and the mixture of a-MnO2/graphene have a similar specific surface area but different pore volumes and average pore sizes. The larger pore volume of the hybrid is due to the longer

This journal is © The Royal Society of Chemistry 2016

View Article Online

Published on 26 September 2016. Downloaded by SLUB DRESDEN on 29/11/2016 20:29:33.

Review

distance between the MnO2 particles, which is enlarged by graphene. The larger pore volume of the hybrid guarantees more space to accommodate the discharging products, thus providing a higher reversible capacity than the mixture. In another bifunctional catalyst, reduced graphene oxide co-doped with N in combination with Fe and Co acts as the ORR catalyst, and ruthenium nanoparticles act as the OER catalyst as well as the ORR catalyst.132 A Li–O2 battery with this catalyst as the positive electrode demonstrates an ultra-high reversible capacity of 23 905 mA h g1 at 0.2 A g1. Even at a high discharge rate (1 and 2 A g1), it also shows very high capacities of 14 560 and 6420 mA h g1, respectively. This bifunctional design guarantees rapid ORR and OER kinetics, leading to low charge/ discharge overpotentials and consequently an excellent electrochemical performance. Moreover, the 2D nanosheet structure of Ru–FeCoN/rGO is very conducive to the formation of Li2O2 nanocrystals with a regular architecture. The round-trip (power) efficiencies of Li–O2 batteries based on graphene are affected by the structural defects of graphene. The performance of graphene nanosheets with different ID/IG ratios was investigated.133 The D-band corresponds to an A1g vibration mode of carbon atoms with dangling bonds in plane terminations of disordered graphite. The G-band is attributed to the E2g vibration mode of sp2-bonded carbon atoms in a twodimensional hexagonal lattice. A higher ID/IG means a lower graphitization degree. As a result, the ID/IG is related to the defect densities in the graphene sample. It is found that the graphene exhibits lower round-trip efficiencies and generates more side products for samples with higher ID/IG values. Annealing in Ar can remove some structural defects in the graphene sheets. A graphene foam electrode annealed at 800 1C with a lower ID/IG delivers a round-trip efficiency of up to 80% with a stable discharge voltage at 2.8 V and a stable charge voltage below 3.8 V for 20 cycles. The electrolyte could also influence the catalytic activity. A graphene oxide-derived 2D carbon electrode for the Li–O2 battery in DMSO based electrolytes shows a higher discharge capacity of 10 600 mA h g1.134a This work confirms the feasibility and efficiency of DMSO for high performance Li–O2 batteries. Improved cycling performance is also achieved by replacing commercial KB carbon electrodes with this graphene oxide-derived 2D carbon positive electrode. NiCo2O4, a well-known cobalt–nickel spinel oxide, has two solid-state redox couples (Co3+/Co2+ and Ni3+/Ni2+) in its structure. Recently, it has been considered as a catalyst for Li–O2 batteries.135,136 It exhibits a higher reversible capacity, lower charge/discharge overpotential, and better cycling stability than pristine carbon black. Following a similar idea, a ZnCo2O4 nanosheet-based Li–O2 battery also delivers a much higher discharging capacity than the super P-based one.137 These results indicate that both NiCo2O4 and ZnCo2O4 nanosheets show high catalytic activity towards ORR and OER. 5.4

3D nanocatalysts

The 2D graphene nanosheets tend to aggregate or restack owing to the strong van der Waals interactions between the graphene planes. Consequently, some unique properties that

This journal is © The Royal Society of Chemistry 2016

Energy & Environmental Science

individual sheets possess such as high specific surface area and peculiar electron transport are significantly decreased or even lost.138 In this regard, 3D graphene air electrodes consisting of interconnected pore channels were developed to prevent graphene from restacking and even reinforce their electronic conductivity.139,140 For example, the construction of hierarchically porous air electrodes with 3D graphene sheets has been achieved using a colloidal microemulsion approach.139b A binderfree air electrode consisting of a 3D graphene/Co3O4 structure was prepared through a two-step method. Firstly, 3D graphene was deposited on a Ni foam via CVD. Subsequently, Co3O4 nanosheets were grown on the graphene through a hydrothermal reaction.139a Cross-linked Co3O4 nanosheets with an open pore structure were fully and vertically distributed throughout the 3D graphene skeleton. Another macroporous graphene@graphitic carbon nitride (g-C3N4) nanocomposite was also obtained from the self-assembly of graphene and g-C3N4 nanosheets via a hydrothermal method.140 The template method is well used for the synthesis of 3D catalysts. The ordered mesoporous/ macroporous carbon sphere arrays were successfully obtained using macroporous silica and F127 as the templates.141 This macroporous silica skeleton was obtained in a one-step reaction from silica alkoxide precursors that were templated around the polystyrene spheres. Besides 3D carbon materials, 3D transition metal oxides (such as MnO2,142–145 NiCo2O4146 and perovskite-based ones147) have also been designed as catalysts for Li–O2 batteries. A simple one-step precipitation synthesis of titanium containing g-MnO2 hollow spheres with enhanced catalytic activity was reported without adding surfactants or templates.142 A Ti ion was used as the morphology directing agent during the precipitation process. A self-transformation process by localized Ostwald ripening was proposed to explain the formation mechanism of the titanium containing g-MnO2 hollow spheres. A sponge-like e-MnO2 nanostructure was synthesized via direct growth of e-MnO2 on Ni foam through an electrodeposition route.143 Recently, 3D macroscopic assemblies, such as aerogels and hydrogels, represented a category of materials that feature low density, large open pores and a high inner surface area. A bottom-up assembly of 2D MnO2 nanosheets into 3D aerogels was reported.144 The dissolution–crystallization and ice-assisted freeze-drying processes played a critical role in the rational assembly of the 2D nanosheets. The as-obtained 3D aerogels combine a number of excellent structural properties such as highly exposed active sites, rich porosity and a 3D continuous network. Like the 3D ordered mesoporous/macroporous carbon sphere,141 mesoporous MnO2, foam-like NiCo2O4 and macroporous LaFeO3 were also synthesized through a template (mesoporous silica,145 starch146 or polystyrene sphere147) replication approach. For example, soluble starch was used as the template for the fabrication of 3D foam-like NiCo2O4 by Wu’s group very recently, as shown in Fig. 8a.146 Starch is a natural polysaccharide polymer with abundant hydroxyl groups, in which cations (Ni2+ and Co2+) can favorably bind with the –OH containing groups on the starch molecules. Moreover, when the starch template was removed at a high temperature in air, the release

Energy Environ. Sci., 2016, 9, 3570--3611 | 3587

View Article Online

Published on 26 September 2016. Downloaded by SLUB DRESDEN on 29/11/2016 20:29:33.

Energy & Environmental Science

Review

Fig. 8 (a) Schematic illustration of the preparation of the 3D foam-like NiCo2O4. (b) Linear sweep voltammetry (LSV) measurements at a rotation speed of 1600 rpm for 3D foam-like NiCo2O4 and super P. (c) Initial discharge–charge curves of the 3D foam-like NiCo2O4 and Super P at 0.2 A g1. (d) XRD patterns of the 3D foam-like NiCo2O4 electrode (before discharging, after discharging, and after recharging) (modified from ref. 146, copyright permission from Wiley).

of CO2 could lead to a large amount of mesopores in the NiCo2O4. In another report, the close packing order of the original polystyrene template was successfully preserved after the calcination process.147 The interconnected inorganic walls created a ‘‘honeycomb’’ pore structure in the 3D structure. The free-standing and binder-free 3D graphene–Co3O4 electrode was used directly as a positive electrode in Li–O2 batteries.139a Compared with conventional paste coating methods with the need for a binder, the direct contact of Co3O4 nanosheets with the highly conductive graphene could facilitate continuous and high electron transfer flux throughout the electrode. Recently, it was found that graphitic–carbon nitride (referred to as g-C3N4) could provide a large amount of active reaction sites. To improve its electron transfer capability, a macroporous graphene@ graphitic carbon nitride (g-C3N4) composite was reported as an air electrode.140 In this nanocomposite, the g-C3N4 nanosheets act as efficient electrocatalysts, and the macroporous graphene nanosheets provide space for Li2O2 to deposit and also promote the electron transfer. Excellent cycling performance, with a terminal voltage higher than 2.4 V after 105 cycles based on 1000 mA h g1, is achieved. Ordered hierarchical mesoporous/ macroporous carbon is also a promising catalyst.141 The ordered mesoporous channels in these materials can effectively facilitate Li+ diffusion and electron transfer while those macropores can provide a space for O2 diffusion and O2/Li2O2 conversion. MnO2 is one of the widely studied air electrode materials for Li–O2 batteries. 3D MnO2 with different nanostructures (like hollow spheres,142 sponge-like nanostructures,143 aerogels144 and mesoporous nanostructures145) has been employed as a positive electrode material for Li–O2 batteries. The e-MnO2 has abundant structural defects, which are considered as favorable catalytic sites. The 3D e-MnO2/Ni electrode exhibits a high-rate capability (discharge capacity of 6300 mA h g1 at a current density of 0.5 A g1) and considerable cyclability (exceeding 120 cycles) without controlling the discharge depth.143 The 3D MnO2 aerogel has a significantly increased discharge capacity in comparison to its powder-like counterpart (4581 vs. 3903 mA h g1).144

3588 | Energy Environ. Sci., 2016, 9, 3570--3611

Interestingly, decayed Li–MnO2 batteries can be further utilized as rechargeable Li–O2 batteries.145 Modestly lithiated Li0.5MnO2 exhibits the best performance with an enhanced round-trip efficiency (ca. 76%), high cycling ability (190 cycles), and high discharge capacity (10 823 mA h g1), which outperforms many MnO2 and other transition metal-based electrocatalysts.142–144 Since a good oxygen electrocatalyst should not interact with O2 too weakly or too strongly, Li0.5MnO2 with a modest oxygenbinding energy shows very high ORR activity. NiCo2O4 is a typical ternary spinel nickel cobalt oxide with the advantages of higher electronic conductivity and better bifunctional catalytic activity toward the ORR and OER than Co3O4 and NiO. In the case of our prepared 3D foam-like NiCo2O4 catalyst, its positive shifts of the onset potential and the half-wave potential (E1/2) in O2-saturated 0.1 M KOH solution are 0.15 and 0.34 V (vs. AgCl/Ag), respectively, which are better than those of the Super P (Fig. 8b).146 In the electrolyte containing O2-saturated 1.0 M LiCF3SO3/TEGDME, the Li–O2 battery with this 3D foam-like NiCo2O4 exhibits a discharge and charge overpotential of 0.2 and 0.97 V, respectively. These values are lower than those of the Li–O2 batteries with the pure Super P at the same current density (Fig. 8c). The initial discharge capacity of the 3D foam-like NiCo2O4 electrode is 10 138 mA h g1. Such a 3D foam-like structure not only provides more contact sites and a larger space for Li2O2 deposition, but also simultaneously improves the transport of oxygen and electrolytes. Some new diffraction peaks, which could be assigned to the (100), (101), and (110) peaks of Li2O2, could be observed for the 3D foam-like NiCo2O4 electrode after the first discharge, and no Li2CO3 phase was found. Some Li2O2 diffraction peaks disappeared when the battery was recharged (Fig. 8d). These results indicate that Li2O2 is a major crystalline discharge product. Similar effects have been demonstrated in the perovskite-based catalysts. The electrocatalytic activity of a 3D ordered macroporous LaFeO3 (3DOM-LFO) catalyst supported on commercial Super P carbon (SP) for ORR and OER was examined in Li–O2 batteries.147 There was almost no

This journal is © The Royal Society of Chemistry 2016

View Article Online

Published on 26 September 2016. Downloaded by SLUB DRESDEN on 29/11/2016 20:29:33.

Review

degradation of the Li–O2 battery with the 3DOM-LFO/SP electrode even after 124 cycles. In contrast, the operation of the Li–O2 battery without the 3DOM-LFO/SP catalyst was limited to only 43 cycles. The unique ‘honeycomb’ porous structure unit of 3DOM-LFO leads to a favorable 3D frame electrode structure which can thus provide a sufficient void volume for Li2O2 deposition. In addition, this 3D porous structure would offer more abundant oxygen and electrolyte transportation paths, helping to get uniform O2 and electrolyte distribution inside the electrode.

6. Nanostructured positive electrode materials for Na-ion batteries Historically, progress in room temperature Na-ion batteries (NIBs) is parallel to that of Li-ion batteries, which was initiated in the early 1980s.148,149 Unfortunately, Na-based systems rapidly lost color because Na+ ions are much larger in radius than Li+ ions and it is more difficult to find a suitable host material to accommodate the Na+ ions. However, for sustainability reasons, the Na-ion batteries have recaptured the scientific community’s attention in recent years.150–152 It is hard to reach the energy density seen in Li-ion batteries when using Na-ion batteries because the Na atom is three times heavier than the lithium atom. Moreover, the standard electrochemical potential of Na (2.71 V vs. SHE) is higher than that of Li (3.05 V vs. SHE). However, it shows great potential applications in large-scale stationary energy storage systems because its cost would be relatively lower compared to that of the lithium ion battery and the source of sodium is more bountiful than that of lithium. In this context, many efforts have been made on the electrode materials of Na-ion batteries for higher capacity, better rate capability and a longer cycle life. In this section, some representative nanostructured positive electrode materials for Na-ion batteries in the past several years are discussed. 6.1

0D nanomaterials

NaFePO4 nanoparticles with a size of 50 nm were synthesized through a simple solid-state method followed by ball-milling with conductive carbon.153 The amount of carbon in the carbon-mixed NaFePO4 is approximately 20 wt%. It could also be prepared via chemical delithiation of LiFePO4 powder followed by electrochemical sodiation of FePO4.154 However, this process is not economic. Layered Na0.66Fe0.5Mn0.5O2155 and Na4Co2.4Mn0.3Ni0.3(PO4)2P2O7156 were successfully prepared by using the sol–gel method. The sol–gel method is another general method that can be used for the synthesis of nanocrystallites. Compared to solid-state reactions, the sol–gel method has the advantage of producing a homogenous mixture of the starting ingredients at the atomic or molecular level, leading to the formation of a smaller particle size with uniform particle size distribution. Another traditional route to synthesize nanocrystallites is the hydrothermal method. As one good example, a vanadyl(IV) compound, Na3(VO)2(PO4)2F, has been successfully synthesized

This journal is © The Royal Society of Chemistry 2016

Energy & Environmental Science

using a novel single-step hydrothermal method.157 This material was then coated with carbon via an impregnation method followed by either a long thermal treatment or a flash thermal treatment. Very recently, an energy-efficient biotemplated route was applied to synthesize FePO4 nanoparticles for Na-ion batteries.158 Self-assembled M13 viruses and single wall carbon nanotubes (CNTs) were used as templates. The M13 virus is a filamentous bacteriophage with a length of 880 nm and a diameter of 6.5 nm. With the aid of the M13 virus, amorphous FePO4, with particles as small as 20 nm, was obtained at room temperature. The most stable polymorph of NaFePO4 is maricite, which is structurally analogous to LiFePO4. Its theoretical specific capacity is 154 mA h g1. Nano-sized maricite NaFePO4 nanoparticles in the NaPF6 (1 : 1 EC/PC) electrolyte deliver a capacity of 142 mA h g1 and outstanding cyclability with a negligible capacity fade after 200 cycles (95% retention of the initial cycle).153 Compared to the NaPF6 (1 : 1 EC/PC) electrolyte, NaFePO4 in a NaTFSI-incorporated ionic liquid electrolyte shows a lower discharge capacity (120 mA h g1) and worse cycling behavior.154 However, this ionic liquid electrolyte has a high thermal stability (4400 1C) and non-flammability, and is thus ideal for high-safety applications. More new positive electrode materials for Na ion batteries were reported with higher specific capacity and working voltage compared to NaFePO4.155–157 A discharge capacity of 158 mA h g1 was produced using Na0.66Fe0.5Mn0.5O2 in the range of 1.5–4.3 V (vs. Na+/Na) during the first cycle.155 It has two peaks at around 2.15 and 3.85 V (vs. Na+/Na) in the cathodic sweep, and two peaks are observed (ca. 2.07 and 3.80 V) in the anodic sweep. Na4Co2.4Mn0.3Ni0.3(PO4)2P2O7 has two redox couples around 4.2 V and 4.6 V (vs. Na+/Na).156 Charge compensation using Co, Mn and Ni ions proceeds simultaneously, which provides a high mixed potential in each redox reaction. The vanadyl(IV) compound, Na3(VO)2(PO4)2F, was first proposed in 2006.159 The effects of carbon coating on the electrochemical performance of Na3(VO)2(PO4)2F nanoparticles were investigated.157 At C/20 the uncoated material exhibits a reversible capacity of 91 mA h g1 whereas the material coated with carbon via a flash thermal treatment has a discharging capacity of 102 mA h g1. The difference can be related to an enhancement of the electronic conductivity and the Na+ ion diffusion ability of the electrode materials when carbon coating is applied. Different from NaFePO4, FePO4 nanoparticles deliver 152 mA h g1 in the first discharge, but the first charge capacity is only 104 mA h g1. The second discharge capacity drops further to 97 mA h g1.158 The irreversible capacity loss in the first cycle indicates that the Na-ion is trapped in the amorphous FePO4. The cause is unclear at present. Upon dispersing 1 wt% CNTs with a high surface contact with the grown FePO4, the charge transfer is improved, leading to both a higher capacity and improved capacity retention. Fluoride-based positive electrode materials were extensively studied for Li-ion batteries and were also proposed for Na-ion batteries with the aim of overcoming the theoretical specific capacity limit of the intercalation compounds. A capacity

Energy Environ. Sci., 2016, 9, 3570--3611 | 3589

View Article Online

Energy & Environmental Science

Review

higher than 220 mA h g1 and a lifetime of at least 50 cycles were demonstrated for FeF30.5H2O nanoparticles (10 nm) embedded in a fragmentated CNT matrix.160 The CNTs do not uniformly wire the grains but rather are locally entangled in the regions where a lot of very small nanoparticles are assembled.

Published on 26 September 2016. Downloaded by SLUB DRESDEN on 29/11/2016 20:29:33.

6.2

1D nanomaterials

The hydrothermal or solvothermal method has been widely used for the synthesis of various 1D nanostructured positive electrode materials for Na-ion batteries, including NaV3O8 nanowires,161 V2O5 nanotubes,162,163 V2O5 nanobelts164 and Na0.44MnO2 nanowires.165a The main advantages of the hydrothermal or solvothermal processes include fast reaction kinetics, a short processing time, good phase purity, high crystallinity, low cost and so on. In addition, Na2.7Ru4O9 nanorods were synthesized at 700–850 1C via a solid-state reaction under Ar flow.166 A 5 wt% excess of Na2CO3 was added to compensate for the evaporation of sodium during the synthesis. Tunnel-structured Na0.54Mn0.50Ti0.51O2 nanorods were prepared using a facile molten salt method, in which CH3COONa, Mn(CH3COO)2 and TiO2 reacted at 800 1C in a mixture of molten NaCl–KCl.167a In comparison with LiV3O8, a larger distance between the [V3O8] layers of NaV3O8 suggests its possible use as a Na-ion intercalation host. NaV3O8xH2O nanowires were proposed as high-capacity positive electrode materials for Na-ion batteries.161 They exhibit two pairs of redox peaks, with the oxidation peaks located at about 2.73 V (vs. Na+/Na) and 3.59 V (vs. Na+/Na) and the corresponding reduction peaks at about 1.87 V (vs. Na+/Na) and 3.08 V (vs. Na+/Na), respectively. They possess a high sodium intercalation/deintercalation capacity, giving an initial specific discharge capacity of 170 mA h g1 with a capacity retention of 51.9% after 50 cycles. The crystal water in the NaV3O8xH2O host could be almost completely removed after the appropriate heat-treatment. The annealed NaV3O8 exhibits a much better cycling stability, with a capacity retention of 91.1% after 50 cycles. This improvement resulted from the contracted crystal volume and the increased crystallinity. Vanadium oxides have been the focus of many recent studies regarding high capacity positive electrode materials for Na-ion batteries.162–164 Single-crystalline bilayered V2O5 nanobelts with a large interlayer spacing of (001) crystal planes exhibit a high capacity of 231 mA h g1.164 The as-prepared V2O5 presents a noticeable distance between the closest approach bilayer stacks (11.53 Å) along the c-axis (Fig. 9a), which was confirmed using XRD, Raman, FT-IR, and TEM. It delivers a high discharge capacity of 170 mA h g1 after 100 cycles (Fig. 9b). The superior electrochemical performance is ascribed to the unique bilayered vanadium oxide nanobelts with dominantly exposed {100} crystal planes, which provide large interlayer spacing for facile Na-ion intercalation/deintercalation. The voltage plateau is located at about 2.4 V (vs. Na+/Na). It still delivers a discharge capacity of 170 mA h g1 after 100 cycles (Fig. 8b). In contrast, VOx nanotubes exhibit faster capacity fading.162,163 This is probably because of the lower valence of vanadium in the material and the existence of amine templates, which were used to fabricate the nanotubes in a large amount. In addition, the

3590 | Energy Environ. Sci., 2016, 9, 3570--3611

Fig. 9 (a) Refined structural model of the bipyramidal layered structure of V2O5 viewed along the [010] and [100], and (b) cycling behavior of the V2O5 nanobelt and the high-magnification TEM image is shown in the inset (modified from ref. 164, copyright permission from the American Chemical Society).

exposed crystal planes also mean a lot for electrode materials.164 Electrochemical energy storage in rechargeable batteries involves chemical reactions at the surface or interface. The discharge/ charge processes are accompanied by the transport of active ions across the surface of the crystals. So the interaction between the active ions in the electrolyte and on the surface of the crystals is essential. In fact, active ions can quickly shuttle back and forth in the crystals along some directions while they are hard to diffuse in the crystal along other directions because a mass of other atoms obstructs their passage. In addition, some facets have a relatively high surface energy and these crystal facets provide reactive sites for fast redox reactions during the charge and discharge processes. More detailed characteristics of electrode materials with tailored facets for electrochemical energy storage can be found in another recent review.167b Among all the positive materials identified as possible candidates for use in Na-ion battery applications, Na0.44MnO2 is particularly attractive because its crystal structure has suitable large-size tunnels for sodium incorporation.165 The singlecrystalline Na0.44MnO2 nanowire shows an initial charge capacity of 50 mA h g1 corresponding to 0.18 Na-ion deintercalation from Na0.44 MnO2, in which the potential plateaus were observed at 3.2 and 3.45 V (vs. Na+/Na).165a The following discharge/charge cycle displays a reversible capacity of 120 mA h g1, corresponding to 0.43 Na-ion intercalation/deintercalation in NaxMnO2 (0.26 o x o 0.69). In addition to the two above mentioned potential plateaus at 3.45 and 3.2 V, four potential plateaus at 3.0, 2.6, 2.5 and 2.2 V (vs. Na+/Na) were also observed during both the charging and discharging processes. These potential plateaus can be explained due to the Na-ion intercalation/deintercalation via the multiple-phase process.165c Motivated by the search for other structures similar to Na0.44MnO2, some new Na ion intercalation materials, Na2.7Ru4O9 and Na0.54Mn0.5Ti0.51O2, were reported.166,167 The unique network structure of [RuO6] octahedra provides 1D chains that are large enough to accommodate three Na sites. Na2.7Ru4O9 shows three redox peaks between 2.8 and 3.9 V (vs. Na+/Na). Electrochemical data and ex situ XRD results indicate that once the Na ions are removed from the Na2.7Ru4O9 structure, only a portion (o60%) could be re-intercalated.166 The Na0.54Mn0.5Ti0.51O2 nanorods deliver a discharge capacity of 100 mA h g1 after 150 cycles at 0.028 A g1, and 67 mA h g1 after 400 cycles at 0.140 A g1.167a

This journal is © The Royal Society of Chemistry 2016

View Article Online

Review

Energy & Environmental Science

These nanorods are grown in a direction perpendicular to the sodium ion tunnels, greatly shortening the diffusion distance of the Na ions and benefiting the transfer kinetics. This result once again proves the importance of the tailored facets in the electrode materials.

Published on 26 September 2016. Downloaded by SLUB DRESDEN on 29/11/2016 20:29:33.

6.3

2D nanomaterials

Similar to the 1D nanostructured positive electrode materials for Na-ion batteries, the hydrothermal method is an effective way to synthesize 2D nanostructured positive electrode materials. Single crystalline Na0.7MnO2 nanoplates have been synthesized using a hydrothermal method.168 The as-prepared material consists of uniform rhombus-shaped nanoplates without any other irregular shapes. The sol–gel process is another popular method for the synthesis of 2D nanostructured positive electrode materials for Na-ion batteries, such as NaV3O8 nanosheets,169 Na3V2(PO4)3–RGO nanocomposites170 and sheetlike structure FeF3–graphene composites.171 Additionally, NaFeSiO4 was obtained through electrochemical Li–Na ion-exchange.172 First, the carbon-coated Li2FeSiO4 nanoparticles were synthesized using a modified sol–gel method. Then carbon-coated Li2FeSiO4 nanoparticles with graphene nanosheets were prepared through pyrolysis at high temperature. Finally, carboncoated Li2FeSiO4 nanoparticles and graphene nanosheets were delithiated in Li ion batteries and then sodiated. As to its uneconomical part of this synthesis method, we mentioned this in the previous section. NaxMnO2 can be categorized into two different types of structures: tunnel structures (x = 0.4 and 0.44) and layered structures (x = 1.0 and 0.7). The tunnel structure was discussed in the above section. For P2-phase Na0.7MnO2, the Mn3+/Mn4+ is responsible for the conduction in NaxMnO2. Recently, the Na0.7MnO2 nanoplates have demonstrated a near theoretical capacity of 164 mA h g1 (its theoretical capacity is 170 mA h g1).168 The charge/discharge profiles of the Na0.7MnO2 nanoplates are very similar to those of the reported Na0.44MnO2 materials. Moreover, they exhibit a good performance at increased current densities. The improvement is ascribed to the predominantly exposed active (100) facet, which could facilitate fast Na+ ion intercalation/deintercalation during the discharging and charging processes. However, its capacity decay is still obvious due to the Jahn–Teller effect of Mn3+/Mn4+. As a member of the vanadium oxide family, NaV3O8 has been proved to be a promising positive electrode material for Na-ion storage. The PPy@NaV3O8 nanosheet core–shell composites show two distinct pairs of redox peaks located at 3.35 and

Table 3

2.35 V (vs. Na+/Na) (cathodic, reduction), and at 3.47 and 2.61 V (anodic, oxidation), respectively, and all the redox peaks almost overlap in the first three cycles.169 NASICON (Na+ superionic conductor)-related compounds have been shown to be another promising positive electrode material for Na-ion batteries. The rhombohedral NASICON form of Na3V2(PO4)3 was firstly tested as a positive electrode material for Na-ion batteries in 2002. Very recently, flexible and binder-free electrodes of Na3V2(PO4)3/reduced graphene oxide (NVP/RGO) were reported.170 The NVP/RGO paper-like electrode delivers a reversible capacity of 113 mA h g1 at 0.1 A g1 and a good capacity retention of 96.6% after 120 cycles. The ultrafine NVP nanocrystals are perfectly incorporated into the interconnected graphene framework, resulting in the integrity of the electronic connectivity being maintained and relieving the volume effect during the Na+ ion inserting/removing processes. In a similar way, FeF3 nanosheets were loaded on the surface of the graphene sheets to form a hybrid of sheet-like structures.171 The FeF3 nanosheets and graphene sheets stack with each other to form a hierarchical electron/ion conducting network, which contributes positively to the large reversible Na-storage capacity. In an attempt to increase the specific capacity, carboncoated Na2FeSO4 on graphene nanosheets was obtained.172 Its discharge capacities reach 250 and 175 mA h g1 at the current densities of 0.02 and 0.05 A g1, respectively. This indicates that a two-electron reaction is possible for the positive materials of Na-ion batteries. 6.4

3D nanomaterials

Many 3D positive electrode materials for Na-ion batteries have been reported and some characteristics of them are summarized in Table 3. As discussed in the above section, Na3V2(PO4)3 (NVP) has a typical NASICON structure, which provides an open and 3D framework for Na+ ion immigration. Many experimental approaches were reported for the synthesis of 3D Na3V2(PO4)3 (such as porous hollow spheres and carbon-coated Na3V2(PO4)3 in mesoporous carbon), including the sol–gel method combined with a freeze-drying process,173a the ultrasonic spray pyrolysis process174 and the nanocasting technique.175 Solvothermal methods are other commonly used methods to fabricate 3D nanostructured positive electrode materials. A fluffy Na0.67FePO4–CNT nanocactus and Na2Fe3xMnx(PO4)3 micro/nanocompound were successfully prepared through a solvothermal method.176,177 The porous Na0.67FePO4 with a nanocactus-like morphology is composed of nanorods with an open 3D structure.176 The Na2Fe3xMnx(PO4)3 compounds are formed into complex 3D hierarchical structures

Some characteristics of 3D positive electrode materials for Na-ion batteries

Positive electrode

Preparation method

Specific capacity Rate capability Cycling behavior Ref.

Porous Na3V2(PO4)3/C Na3V2(PO4)3/C hollow microspheres Na3V2(PO4)3/C mesoporous C Na0.67FePO4/CNTs nanocactus Na2Fe3xMnx(PO4)3 micro/nanocompounds V6O13 micro-flowers Na4Fe(CN)6 microcubes

Sol–gel method + freeze-drying process Ultrasonic spray pyrolysis process Nanocasting technique Solvothermal method Solvothermal method Hydrothermal method Precipitation process

118.9 mA h g1 100 mA h g1 115 mA h g1 138 mA h g1 120 mA h g1 225 mA h g1 170 mA h g1

This journal is © The Royal Society of Chemistry 2016

0.59 A g1 1 A g1 5C 0.2 A g1 0.005 A g1 0.16 A g1 0.6 A g1

100 300 1000 50 50 30 150

173 174 175 176 177 178a 179e

Energy Environ. Sci., 2016, 9, 3570--3611 | 3591

View Article Online

Published on 26 September 2016. Downloaded by SLUB DRESDEN on 29/11/2016 20:29:33.

Energy & Environmental Science

assembled from well-defined, low-dimensional nanorods.177 A simple and versatile method for the preparation of V6O13 microflowers was also developed via a simple hydrothermal route with the aid of an alkali metal nitrate salt.178a The synthesis was performed in a water solvent using ammonium metavanadate, oxalic acid and nitrate as reactants. Prussian blue (PB) and its analogues (PBA) were also explored as Na-storage positive electrodes.179 High-quality Na0.61Fe[Fe(CN)6]0.94 microcubes were obtained by utilizing Na4Fe(CN)6 as the single iron-source precursor. It exhibits a low zeolite water content and a small number of [Fe(CN)6] vacancies in the crystal framework.179e Both the carbon coated porous Na3V2(PO4)3 and hollow Na3V2(PO4)3 spheres show high rate capabilities.173a,174 It is generally accepted that for a porous or hollow Na3V2(PO4)3 electrode material, various pores provide good access of the electrolyte to the electrode surface. In addition, the surface area of the porous material is relatively large, thus facilitating charge transfer across the electrode/electrolyte interface. Moreover, the hollow interior usually provides extra free space for alleviating the structural strain associated with repeated Na+ ion insertion/ extraction processes, giving rise to improved cycling stability.173b The effective diffusion of Na+ and e into and out of the NVP can increase the specific capacities particularly at higher charge/ discharge rates. NVP@C@CMK-3 delivers the highest capacity of 115 mA h g1, corresponding to 97% theoretical capacity.175 The pure NVP without carbon coating shows the worst performance, only delivering an initial capacity of 83 mA h g1. After three cycles, the charge transfer resistance (Rct) of the NVP@C@CMK-3 is 207 O, which is lower than that of NVP@CMK-3 (270 O). This further demonstrates that the large surface area of the CMK-3 allows sufficient infiltration of electrolytes. The fluffy Na0.67FePO4–CNT nanocactus delivers no obvious capacity fading over 50 cycles. During the sodiation/desodiation process, the route of the Na ions was confirmed using X-ray absorption near edge spectroscopy (XANES) and extended X-ray absorption fine structure spectroscopy (EXAFS).176 In the case of alluaudite Na2Fe3xMnx(PO4)3, the relationship between the structure and physicochemical properties was systematically investigated.177 After the introduction of Mn ions on the Fe sites, the long-range order maintains the isostructural framework. The long-range structure, which provides a unique micro/ nanomorphology, plays an important role in improving electrochemical performance by offering a short transfer path for electrons and ions. V6O13 has been extensively investigated as a positive electrode material for lithium-ion batteries since 1979.178b However, its electronic conductivity decreases upon lithium intercalation.178c However, V6O13 microflowers with a small amount of impurity phase display a high capacity up to 225 mA h g1 for Na ion batteries.178a Moreover, the impurity phase in the V6O13 microflowers may lead to fewer phase transitions upon sodium intercalation. Prussian blue analogues (PBAs), one class of MOFs, have received attention from the community owing to their impressive electrochemical performance as well as easy synthesis procedures.179 The high-quality Na0.61Fe[Fe(CN)6]0.94 microcubes,

3592 | Energy Environ. Sci., 2016, 9, 3570--3611

Review

with a low crystal water content and a small number of [Fe(CN)6] vacancies, exhibit a high specific capacity of 170 mA h g1 by realizing a two-electron reaction.179e However, the occupancy of interstitial sites by water molecules could inhibit Na+ ion intercalation and transport. Even worse is that the decomposition of water deteriorates the coulombic efficiency and cycling performance. In addition, the [Fe(CN)6] vacancies can decrease the electronic conductivity of PB and lead to framework collapse and a disordered lattice during cycling. 6.5

Aqueous rechargeable sodium batteries (ARSBs)

ARSBs appear to be an attractive alternative to their lithium counterparts (discussed in Section 4) for electric energy storage, because of the widespread availability and low cost of sodium resources. Currently, the positive electrode materials for ARSBs mainly include NaxMnO2,180,181 MnO2,182–184 Na3V2(PO4)3185,186 and Prussian blue (PB).187,188 Among various Na intercalation compounds, Na0.44MnO2 has been the most widely investigated. The Na+ ion deintercalation/intercalation out of/into the Na0.44xMnO2 in an aqueous electrolyte was first investigated in 2007.189 Recently, an ARSB based on the Na0.44MnO2 nanorod as the positive electrode was presented, which demonstrated its ultrafast rate performance comparable to that of a supercapacitor.180 At a current rate as high as 90C, the retained capacity is more than 70% of that at 1C. Such a high rate performance is hard to achieve in organic electrolytes. The electrochemical activity of Na0.95MnO2 nanorods in the aqueous electrolyte was reported by our group.181 There are a couple of redox peaks at 1.4 and 1.7 V (vs. Zn2+/Zn), respectively, which correspond to the deintercalation and intercalation of Na+ ions in the aqueous electrolyte. A stable cycling performance is obtained up to 1000 cycles at 4C with metallic zinc as the negative electrode. Though the energy density of this ARSB could not be comparable with those of Li-ion batteries, it is very safe and the cost is also very low. Up to now, several crystallographic phases of MnO2, i.e., a-, b-, g-, d- and l-phases, have been reported. Among them, spinel-type l-MnO2,183 birnessitetype d-MnO2184 and g-MnO2185 show Na+ ion deintercalation/ intercalation behaviors. A specific capacity of approximately 225 mA h g1 was obtained from g-MnO2 at a current rate of 8 mA g1 in a 7 M NaOH aqueous electrolyte.185 However, g-MnO2 underwent a structural transition during the initial discharge, resulting in 24% capacity decay after 25 cycles. Besides NaxMnO2 and MnO2, the carbon-coated Na3V2(PO4)3 and PB were applied as the positive electrode materials for ARSBs. The diffusion-controlled Na ion deintercalation/ intercalation of the Na3V2(PO4)3 in the aqueous electrolyte was revealed via CV experiments with various scan rates.186 The redox potential was determined to be 0.4 V (vs. SCE). PB and its analogues were investigated as hosts for alkali ions several years ago.190 It has been reported that there is a stable Na+ ion cyclability into potassium copper hexacyanoferrate in an aqueous electrolyte.187a NiCHF (described as K0.6Ni1.2Fe(CN)63.6H2O) delivers a redox potential of 0.59 V (vs. SHE). To improve the redox potential, the solid-solution phase of CuxNi1xHCF was proposed.187b The redox potential increases from 0.6 to 1.0 V

This journal is © The Royal Society of Chemistry 2016

View Article Online

Published on 26 September 2016. Downloaded by SLUB DRESDEN on 29/11/2016 20:29:33.

Review

(vs. SCE) as the Cu content increases without a significant capacity loss. It should be noted that the series of Ni- and Cu-based PB positive electrodes cannot be paired with conventional Na-free negative electrodes because they initially exist in a charged state. Prior to the assembly of a full cell, this Na free PB needs to be pre-conditioned in half-cells. To address this, sodiated PB, Na2NiHCF, was introduced.188 The ARSB, based on the Na-rich Na2–NiFe(CN)6 positive electrode and the Na-deficient NaTi2(PO4)3 negative electrode in the aqueous Na2SO4 electrolyte, gives an average output voltage of 1.27 V and could retain 88% of its initial capacity after 250 cycles at the 5C rate.

7. Nanostructured positive electrode materials for Mg-ion batteries The first work on Mg rechargeable battery systems was presented in 1990.191 A large number of transition metal oxides (Co3O4, Mn2O3, Mn3O4, MoO3, PbO2, Pb3O4, RuO2, V2O5 and WO3), sulfides (TiS2, VS2 and ZrS2), and borides (MoB2, TiB2 and ZrB2) were tested for Mg2+ ion intercalation abilities. Among them, TiB2 shows the highest capacity (324 mA h g1). Mg is a benign and the 5th most abundant element in the earth’s crust. The divalent nature of Mg-ions has been a hurdle for fast intercalation kinetics, but can be an opportunity for realizing an utmost capacity.192–194 At present there are a few companies trying to develop rechargeable Mg batteries, including Sony, LG, Toyota and Pellion Technologies (a company spin out of Massachusetts Institute of Technology). The last decade witnessed a tremendous number of works dedicated to Mg-ion batteries. 7.1

Chevrel phase (CP): Mo6T8 (T = S or Se)

The second breakthrough in Mg-ion batteries was attained in 2000 by demonstrating a high-rate positive electrode material based on Chevrel phases (CPs) of the Mo6T8 family (T = S, Se or their combination).195 They were cycled over 2000 times at 100% depth of discharge with a capacity fading of less than 15%. The impressive cycle life spurred a surge of interest in this topic. Now CPs have been known to be capable of intercalating monovalent and multivalent cations such as Li+, Na+, Zn2+, Cd2+, Ni2+, Mn2+, Co2+ and Fe2+.193 To synthesize CP, Cu2Mo6S8 composition is the first reported one prepared at a high temperature. The overall reaction requires 4 days, and also includes heating to 1250 1C for 1–2 days. Then Cu is extracted from the synthetic product Cu2Mo6S8 in a 6 mol L1 HCl/H2O (1 : 1) solution under an air atmosphere.196a Recently, a two-step solution chemistry route was used to synthesize 3D cuboidal shaped Cu2Mo6S8 with 5 h annealing at 1000 1C under a reducing atmosphere.196b The Cu(NH4)xMo3S9 compound upon heating under a 6.5% H2 + Ar atmosphere directly yields a Cu2Mo6S8 phase according to the following reaction (13): 2Cu(NH4)xMo3S9 + 10H2 - Cu2Mo6S8 + 10H2S + 2xNH3 + xH2 (13)

This journal is © The Royal Society of Chemistry 2016

Energy & Environmental Science

This approach marks a new synthesis route different from hitherto described conventional solid-state methods. This material is able to theoretically insert two Mg2+ ions per Mo6S8 unit upon the first discharge of the electrode. However, not all the intercalated Mg2+ ions can be deintercalated at room temperature, usually resulting in a 20–25% capacity loss. The full capacity of Mo6S8 can only be realized at elevated temperatures. One effective way of improving the intercalation/ deintercalation kinetics of CP electrodes is by nanosizing the CP powder.196a A decrease in the average size of the particles in contact with the solution means a shorter diffusion length for Mg2+ ions in the host and a higher surface area, resulting in a decrease of the diffusion resistance. The Mg trapping in sulfide CP during the initial cycle is caused by an unusual arrangement of the inner sites for cation intercalation. Therefore, the inherent kinetic limit cannot be completely resolved through nanosizing the CP powder. Compared to Mo6S8, Mo6Se8 possesses increased cell parameters. The increased cell parameters lead to greatly increased Mg ion mobility. As a result, the intercalation/ deintercalation of Mg into Mo6Se8 is fully reversible.197 Similarly, partial substitution of S with Se in the CP also has a positive effect. The Se is a larger anion than sulfur (Se2: 1.84 Å and S2: 1.7 Å), and the Se–Se and Mo–Se interatomic distances are larger than those of S–S and Mo–S in the CPs. When Se is incorporated into Mo6S8ySey up to y = 2, the room-temperature obtainable specific capacity is increased largely. Above 80% capacity retention at 1C is observed for Se-substituted CPs at 15 1C, while the capacity retention is about 50% for pure Mo6S8.198 7.2

Transition metal oxides

Reversible intercalation of Mg ions has been reported in transition metal oxides including V2O5, MnO2, MoO3 and TiO2. The combustion flame-chemical vapor condensation process (CF-CVC) was used to produce nanocrystalline V2O5 of small primary particles (25 nm) with minimal aggregation in the vapor phase.199 V2O5 nanotubes are obtained as the main product in a sol–gel reaction followed by hydrothermal treatment from V2O5 precursors and primary amines.200 This V2O5 has a considerably large distance between the layers (ca. 3.5 nm), and wide inner (15–40 nm) and outer (60–100 nm) diameters. Hydrated V2O5 xerogels were prepared via the reaction of hydrogen peroxide and metallic vanadium powder.201 To obtain V2O5 aerogels, V2O5 hydrogels are first synthesized through ion-exchange processing of sodium metavanadate.202 V2O5 aerogels are then obtained by exchanging the water with acetone and liquid CO2, and then carrying out supercritical drying with liquid CO2. The various crystal phases of manganese oxide (a-MnO2,203 d-MnO2,204 and l-MnO2205,206) have also been investigated as positive electrodes for Mg ion batteries. As mentioned in the earlier sections, the preparation of MnO2 mainly includes a solid state reaction and a hydrothermal method. V2O5 is a very important metal oxide with intriguing properties and widespread applications. The electrochemical reactivity of cations such as Ca2+, Mg2+, and Y3+ in crystalline V2O5 materials was investigated.199 Compared to Y3+, the chemical/electrochemical

Energy Environ. Sci., 2016, 9, 3570--3611 | 3593

View Article Online

Published on 26 September 2016. Downloaded by SLUB DRESDEN on 29/11/2016 20:29:33.

Energy & Environmental Science

intercalation of Mg2+ into V2O5 is slow due to the extremely slow diffusion rate of the Mg2+ ion in the crystal. The V2O5 nanotube has an advantageous shape for Mg ion intercalation,200 including the considerably large distance between the layers, wide inner and outer diameters and the open tube ends. AC (alternative current) impedance demonstrates that Mg2+ ion diffusion into the nanotubes is faster than that into polycrystalline V2O5. In some studies, V2O5 xerogels201 and aerogels202 were also investigated to insert and extract Mg ions. Gels are porous structures with large surface areas which allow for more surface diffusion. V2O5 xerogels exhibit electrochemical intercalation up to 1.84 moles of Mg2+ ions per mole of V2O5.201 V2O5 aerogels usually have surface areas several times larger than V2O5 xerogels. However, V2O5 aerogels have only been observed to electrochemically insert 0.6 moles of Mg2+ ions per mole of V2O5.202 The reason for this is unclear and more work needs to be completed to compare their electrochemistry. MnO2 nanomaterials with different crystal structures display various performances for Mg ion batteries. The magnesium electrochemistry for nanosized a-MnO2 structures (hollandite with 2  2 tunnels) was recently investigated.203 It offers a specific capacity of 280 mA h g1 in the first discharge using the Mg salt/THF electrolyte. However, the discharge/charge capacity fades during cycling due to the partial collapse of the a-MnO2 tunnel structure associated with Mg2+ ion intercalation/deintercalation, which has been observed using EXAFS (extended X-ray absorption fine structure) analysis. The Mg2+ ion deintercalation/intercalation out/into d-MnO2 shows a lower discharge capacity (70 mA h g1) than the a-phase, but the d-MnO2 is able to retain more of its original capacity over longer cycles.204 Spinel l-MnO2 has an open circuit potential of 1.2 V (vs. Mg2+/Mg) and delivers a specific capacity up to 80 mA h g1 in the 0.1 mol l1 Mg(ClO4)2–0.5 mol l1 H2O/THF electrolyte.205 However, H2O should be discarded in systems containing Mg metal as they are not compatible with Mg metal. Very recently, l-MnO2 has shown a high initial discharge capacity of 545.6 mA h g1 in a 0.5 mol l1 MgCl2 aqueous electrolyte in a three-electrode test system (graphite nanosheets and SCE were used as the counter and reference electrodes, respectively).206 After the intercalation of Mg2+ ions, MgMn2O4 still retains the spinel structure. However, l-MnO2 is usually prepared by etching Li out of the LiMn2O4 spinel, while LiMn2O4 is a positive electrode of LIBs. This preparation method is uneconomical. Alpha-MoO3 has a distinctive layer structure, and each layer is composed of [MoO6]6 octahedra in which each Mo atom is surrounded by a distorted oxygen octahedron. Interaction between adjacent layers is through weak van der Waals forces. Electrochemical reversible insertion/extraction of Mg2+ in MoO3 was first demonstrated in the MgCl2/EMIC/AlC13 molten salt electrolyte at 80 1C.207 In another study, MoO3 films from electrodeposition were reversibly cycled at a relatively high capacity (220 mA h g1) at a potential of around 1.8 V (vs. Mg2+/Mg).208 XRD and Raman spectra clearly show that a two-phase reaction occurs during the magnesiation of MoO3. When TiO2 is used as a positive electrode material for

3594 | Energy Environ. Sci., 2016, 9, 3570--3611

Review

rechargeable Mg ion batteries, the electrolyte is the key factor in achieving electrochemical activity. It exhibits a poor Mg2+ ion intercalation/deintercalation performance in the organic magnesium organohaloaluminate electrolyte, i.e. (PhMgCl)2AlCl3/THF.209 However, when a highly compatible electrolyte of Mg(BH4)2/LiBH4/tetraglyme is used, the commercial TiO2 (25–30 nm in size) delivers a high specific capacity (155 mA h g1) and excellent cycling stability (90 cycles). It has well-defined charging and discharging plateaus at approximately 0.87 and 1.16 V (vs. Mg2+/Mg), respectively. The co-intercalation of Li+ and Mg2+ may contribute to this exceptional performance of the batteries. 7.3

Transition metal sulfides

Molybdenum disulfide has been the subject of studies in part due to its layered MoS2 structure. Several solution-based chemical reaction routes at low temperatures have been designed to prepare various MoS2 nanostructures, including hollow-cage fullerene-like nanoparticles, fibrous floccus, and spherical nanovesicles.210 Graphene-like MoS2 was synthesized via a solvothermal route from the reaction of MoO3 and thioacetamide with pyridine as the solvent.211 Similarly, RGO-supported MoS2 hybrids212 and sandwich-structured MoS2/C microspheres213 were fabricated via a hydrothermal method followed by a heat-treatment. Layered MoS2 and graphene nanosheets in the hybrids interlace with each other to form novel sandwich-structured microspheres. The weak van der Waals interaction between layers and the large interlayer distance makes MoS2 an attractive host for Mg2+ ion intercalation. MoS2 with hollow-cage fullerene-like particles, fibrous floccus, and spherical nanovesicle structures has a capacity of 2–25 mA h g1 and a discharge voltage of around 1.5 V (vs. Mg2+/Mg).210 DFT calculations based on first principles show that in a zigzag MoS2 nanoribbon the theoretical capacity of 232.2 mA h g1 could be reached.214 Further work with an exfoliated graphene-like MoS2 achieves a first discharge capacity of 170 mA h g1 and a high operating voltage of 1.8 V (vs. Mg2+/Mg).211 Moreover, 95% of the initial capacity is kept after 50 cycles. The lattice fringe spacings of graphenelike MoS2 are measured to be in the range of 0.65–0.70 nm, which are larger than those of bulk MoS2 (0.63 nm). This expanded interlayer distance and graphene-like morphology lead to this high reversible capacity. The MoS2/RGO hybrids also exhibit higher discharge capacities as well as better cycling performances than pristine MoS2.212 Recently, sandwichstructured graphene-like MoS2/C microspheres presented a good cycling stability and initial discharge capacities that could reach 213 mA h g1.213 These results exceed those typically reported for CPs in Mg ion batteries.195–198 Besides MoS2, TiS2 nanotubes could be reversibly intercalated and deintercalated by large amounts of Mg ions in the channels held by van der Waals interactions.215 CV shows that a peak appears at about 0.98 V (vs. Mg2+/Mg), attributing to Mg2+ ion intercalation into the TiS2. During the following Mg2+ deintercalation, a peak is observed at 1.2 V (vs. Mg2+/Mg). The discharge capacity is 180 mA h g1 after 80 cycles at a low current rate. In the TiS2 nanotubes, the van der Waals gap between the S–Ti–S and S–Ti–S layers could

This journal is © The Royal Society of Chemistry 2016

View Article Online

Review

be fully utilized during the intercalation process. Other sulfide based compounds including copper sulfide,216 nickel sulfide,216 and NbS3217 have also been investigated as positive electrode materials for Mg-ion batteries, but in general exhibit poor electrochemical performances. Further work on their materials and intercalation mechanism is in progress.

Published on 26 September 2016. Downloaded by SLUB DRESDEN on 29/11/2016 20:29:33.

7.4

Olivine compound MgMSiO4 (M = Mn, Co, Fe)

The first report of Mg-ion batteries utilizing a silicate based electrode (Mg1.03Mn0.97SiO4 nanoparticles) outlined their synthesis using a modified sol–gel approach and in situ carbon coating.218,219 Then a hierarchical CNT/C/Mg1.03Mn0.97SiO4 nanocomposite was prepared via a molten salt method followed by a chemical vapor deposition (CVD) method.220 The Mg1.03Mn0.97SiO4 nanoparticle with a carbon layer adhered to a single CNT. Mesoporous MgMnSiO4 could be prepared using mesoporous silica SBA-15 or KIT-6 as both a template and a silicon source.221 The hierarchically porous MgCoSiO4 with both macropores and mesopores was synthesized using a solution-based templating method with PS colloidal crystals as the templates.222 The 3D hierarchically porous architectures consist of a network of macropores with several nanometer walls and mesopores at the boundaries of the grains or between grains. Magnesium iron silicate (MgFeSiO4) could be prepared using a molten salt method223 and an ion-exchange method.224 KCl with a melting point of 780 1C was used as a flux. The ionexchange method involves the electrochemical deintercalation of 2Li+ ions from Li2FeSiO4 and electrochemically replacing them with Mg2+ ions. The Mg-storage capacity of the silicate is strongly dependent on the particle size and the electronic conduction of the material. Carbon coating could enhance the electronic conductivity and suppress the particle growth.218 Carbon coated Mg1.03Mn0.97SiO4 nanoparticles deliver a discharge capacity of 244 mA h g1 with an operating voltage of around 1.6 V (vs. Mg2+/Mg).219 However, the specific capacity of the microsized Mg1.03Mn0.97SiO4 is only 43.2 mA h g1 and the discharge voltage plateau is blurred. A recent work involving manganese silicates adding CNTs approaches the theoretical capacity of 314 mA h g1.220 The CNTs provide a conductive network to enhance electronic conductivity, but they also increase structural strength. CNTs have an extremely high elastic modulus, which aids in holding the electrode together during cycling. Electrochemical data also demonstrate that mesoporous Mg1.03Mn0.97SiO4 has larger discharge capacities and a higher flat plateau compared with the bulk form.221 When used as a positive electrode for Mg ion batteries, MgCoSiO4 has an operating voltage plateau of around 1.6 V (vs. Mg2+/Mg). The voltage does not increase compared to Mg1.03Mn0.97SiO4.222 MgFeSiO4 is another good host for Mg2+ ion intercalation. MgFeSiO4 synthesized at 900 1C could deliver 125 mA h g1 for its initial discharge capacity and a 91.4% capacity retention in the 20th cycle at 0.1C.223 It is noteworthy that nearly 25% of the capacity is lost from a small increase of discharge current from 0.1C to 0.3C. Ion-exchanged MgFeSiO4 is also demonstrated as a feasible positive material for Mg ion batteries.224 It provides a

This journal is © The Royal Society of Chemistry 2016

Energy & Environmental Science

capacity of more than 300 mA h g1 at an average potential of 2.4 V (vs. Mg2+/Mg), with good retention upon cycling in Mg(TFSI)2/acetonitrile electrolytes using a three electrode test system. However, magnesium deposition/dissolution cannot be achieved using Mg(TFSI)2 in acetonitrile as a solvent. When Mg(TFSI)2/triglyme is used as the electrolyte, only half of the theoretical capacity of the positive electrode material could be obtained (166 mA h g1) in a prototype full battery. This presumably arises from the low Mg2+ cation flux in the triglyme solvent. Further improvements in the electrolyte and morphology control of the composite electrodes are needed in the future. 7.5

Other electrode materials

As discussed in Sections 5 and 6, PBAs are metal–organic framework materials with an open and tunable crystal structure. The most common PBA in electrochemical studies is hydrated copper hexacyanoferrate (K0.1Cu[Fe(CN)6]0.73.6H2O).225 The solid solution process of Mg2+ intercalation/deintercalation was evidenced through ex situ X-ray diffraction while the 57 Fe Mossbauer spectroscopy and X-ray absorption near edge structure reveal redox of both Cu and Fe. PBA based Mg-ion batteries in aqueous electrolytes usually show long cycle lives and high rate performances.226 Delithiated V2(PO4)3,227 MgFePO4F228 and MgVPO4F229 have also been reported as positive electrodes for Mg-ion batteries. These materials exhibit a discharge capacity near to 200 mA h g1 and a high average working voltage (3.0 V vs. Mg2+/Mg).227 The exhibited voltage by far surpasses other reported electrode materials for Mg-ion batteries. However, it was tested in a three-electrode cell using the Ag/AgCl electrode as a reference electrode. To develop high performance prototype two-electrode Mg batteries, a stable electrolyte possessing a wide voltage window of 0–4 V (vs. Mg2+/Mg) should be developed. Besides inorganic materials, conductive polymers, like poly-2,2 0 -dithiodianiline (PDTDA)230 and polypyrrole (PPy),231 have also been tested as the positive electrode for Mg-ion batteries. However, their electrochemical performances are not good and the present results are only the first step in the exploration of conductive polymers. Further work is needed to increase the specific capacity and cyclability through further modifications.

8. Nanostructured positive electrode materials for Al-ion batteries Aluminum is one of the most widespread chemical elements in the outer 16 km crust of the earth (about 8% by weight), with only oxygen and silicon being more abundant.231 Aluminum has a low atomic weight of 26.98 along with its trivalence, which gives a gram-equivalent weight of 8.99 and a corresponding electrochemical equivalent of 2980 mA h g1, compared with 3860 mA h g1 for lithium, 2200 mA h g1 for magnesium and 820 mA h g1 for zinc.232 Since the 1850s, Hulot described a cell with zinc (mercury) as the positive electrode, aluminum as the negative electrode and aqueous H2SO4 as the electrolyte.233 In 1948, the use of

Energy Environ. Sci., 2016, 9, 3570--3611 | 3595

View Article Online

Published on 26 September 2016. Downloaded by SLUB DRESDEN on 29/11/2016 20:29:33.

Energy & Environmental Science

amalgamated aluminum as a negative electrode in heavy duty chlorine-depolarized batteries was reported, in which the open circuit voltage was as high as 2.45 V.234 The early efforts on Al-based batteries mainly focused on primary batteries, like Al–H2O2 batteries,235a Al–S batteries,235b Al–KMnO4 batteries,235c and Al–NiOOH235d and Al–air batteries.236 Some secondary battery systems (such as Al–FeCl3237a and Al–Cl2237b) were also constructed and tested. Early studies on Al-based rechargeable batteries paid little attention to nanostructured materials. Until 2011, the first Al-ion rechargeable battery was studied preliminarily.238 The battery used V2O5 nanowires as the positive electrode with an aluminium metal as the negative electrode in an ionic liquid-based electrolyte. The open circuit voltage of the Al-ion battery is found to be 1.8 V (vs. Al3+/Al). It demonstrates a well-defined and very stable Al3+ intercalation plateau at 0.55 V (vs. Al3+/Al). In the first cycle, the battery exhibits a capacity of 305 mA h g1 against 273 mA h g1 at the end of 20 cycles. This work initiates a global search for new materials as the positive electrode for Al-ion batteries. Besides V2O5, the Al-ion battery based on VO2 nanorods as the positive electrode was also assembled.239 First-principles calculations were employed to theoretically investigate the crystal structure change and the intercalation–deintercalation mechanism of the Al ions. In the initial cycle, the as-prepared cell delivers a discharge capacity of 165 mA h g1. Obvious charge and discharge voltage plateaus could be seen around 0.7 V and 0.5 V (vs. Al3+/Al). There is still a great space for the improvement of capacity because only a little amount of Al3+ ions participates in the actual redox reaction. Another Al-ion battery comprising fluorinated natural graphite as the positive electrode was obtained successfully.240 A graphite fluoride material is a superior electronic conductor than graphite itself. The charge capacity of the fluorinated graphite nanosheet electrode is approximately 300 mA h g1, but the columbic efficiency of the cell is 75%. Very recently, an ultrafast Al ion battery was presented that uses a 3D graphiticfoam as the positive electrode.241 The cell exhibits well-defined discharge voltage plateaus near 2 V (vs. Al3+/Al) (Fig. 10a) and a stable cycle life up to 7500 charge/discharge cycles without decay at ultrahigh current densities. Another significance of this work is that it unraveled the intercalation/deintercalation of chloroaluminate anions in the graphite. Besides, this Al ion pouch cell is mechanically bendable and foldable. Although the electrolyte (chloroaluminate ionic liquids) of the above Al-ion batteries is nonflammable, the cost of the ionic liquid is high. In this regard, the aqueous solution of inorganic Al salts is used as the electrolyte. A PBA, copper hexacyanoferrate, is a potential positive electrode material for Al-ion intercalation in aqueous electrolytes.242,243 Different from the behavior of monovalent ions (Na+ and K+) in PBAs, it is a twostep reaction for the intercalation/deintercalation process of the Al3+ ion in copper hexacyanoferrate.243 The general formula of PBAs can be rewritten as AxPR(CN)6. P and R ions are separated and bonded by CN ligands to form a face-centred cubic structure with an open framework, which contains large interstitial A sites for introducing guest ions or molecules. Al3+

3596 | Energy Environ. Sci., 2016, 9, 3570--3611

Review

Fig. 10 (a) The charge/discharge profiles of Al//graphite batteries in the Al salt ionic liquid electrolyte and the inset is its voltage verse time curves (modified from ref. 241, copyright permission from Nature Publishing Group). (b) SEM micrograph of the graphite nanosheet and (c) the charge/discharge profile Zn//graphite nanosheet battery in the Al salt aqueous electrolyte. (d) High-magnification TEM micrograph of the original graphite nanosheet electrode, (e) and (f) graphite nanosheet electrode after discharging 0.4 V vs. SCE. (d1)–(f1) are their corresponding SAED images (modified from ref. 244, copyright permission from the American Chemical Society).

ions have a smaller Shannon radius (39 pm) compared to Na+ ions (102 pm), K+ ions (151 pm) and Mg2+ ions (72 pm). However, a hydrated Al3+ ion is larger with a radius of 480 pm. Thus there is a dehydration reaction process during trivalent Al3+ ion insertion. In the discharge process, Al3+ ions are inserted into the CuHCF structure and partly occupy the tetrahedrally coordinated A sites, in which the carboncoordinated Fe3+ ions are reduced to Fe2+, whereas nitrogencoordinated Cu2+ ions remain electrochemically inactive and rather contribute only to maintain the stable framework. A specific capacity of 62 mA h g1 is obtained at 0.05 A g1, which is close to that in the electrolyte containing Na+, K+ and NH4+. So its specific capacity is highly dependent on the Fe3+/Fe2+ redox couple in the framework of CuHCF rather than the guest species. To improve the specific capacity of graphite, ultrathin graphite nanosheets were prepared through a simple electrochemically expanded method.244 Then an aqueous rechargeable aluminum battery was fabricated using these graphite nanosheets as the positive electrode and zinc as the negative electrode in an aqueous Al2(SO4)3/Zn(CHCOO)2 electrolyte. The ultrathin nanosheet morphology (Fig. 10b) facilitates rapid Al ion diffusion and a fast reaction between Al ions and graphite. Thus, this battery could be rapidly charged and discharged at a high current density while maintaining a high capacity. As shown in Fig. 10c, the average charge and discharge voltages are 1.35 and 1.0 V, respectively, which exceed those of the recently reported Al//V2O5238 and Al//VO2239 in ionic liquid electrolytes. The graphite nanosheets show a discharge capacity of 60 mA h g1 even at 2 A g1. The charge time needed is only 110 seconds (corresponding to 33C). The graphite nanosheet

This journal is © The Royal Society of Chemistry 2016

View Article Online

Published on 26 September 2016. Downloaded by SLUB DRESDEN on 29/11/2016 20:29:33.

Review

electrode before and after cycling was tested using highmagnification TEM. The pristine graphite nanosheets have well-ordered nanosheets and a clear lattice fringe of E0.34 nm (Fig. 10d). The selected area electron diffraction (SAED) patterns in Fig. 10d1 clearly shows the (002) and (100) faces of graphite. After discharging to 0.4 V (vs. SCE), the graphene layers were wrinkled and some disordered area was found on the edge of the graphite nanosheets (Fig. 10e). Some new phases emerge in the related SAED patterns (Fig. 10e1). However, in the inner area of the graphite nanosheet after discharging to 0.4 V (vs. SCE), the adjacent graphitic layers are clear (Fig. 10f) and the corresponding SAED image demonstrates the characteristic of crystalline graphite (Fig. 10f1). So we think that the reaction between Al3+ (or [Al(H2O)6]3+) ions and graphite is non-uniform. Or Al3+ (or [Al(H2O)6]3+) ion intercalation into graphite occurs more evidently at the surface of the graphite nanosheet while Al3+ (or [Al(H2O)6]3+) ions are hard to intercalate into the graphite bulk (or interior). Of course, further direct evidence about the reaction mechanism based on some in situ characterization techniques is needed in the future. It is evident that the cost of this aqueous rechargeable aluminum battery is low, which shows potential applications in stationary energy storage systems. Since the storage mechanism is based on the simultaneous utilization of two different ions, a difference in the salt concentration occurs, which may influence the physical properties of the electrolyte (such as conductivity and viscosity). Therefore, this kind of aqueous rechargeable aluminum battery needs a highly concentrated Al salt electrolyte to obtain high energy density. This characteristic is very similar to dualgraphite batteries.245 Dual-graphite batteries use graphite as intercalation hosts for both electrodes, where the energy storage mechanism is based on the intercalation of lithium ions into the negative and the intercalation of anions into the positive electrode. In this set-up, the electrolyte does not only act as a charge carrier but also as a source for intercalation guests. However, for this kind of aqueous rechargeable aluminum battery, the highly concentrated Al salt electrolyte would lead to high acidity of the electrolyte as a result of the hydrolysis of Al ions, which may also be corrosive to the negative electrode (Zn).

9. Nanostructured positive electrode materials for other rechargeable batteries 9.1

Zn-ion batteries

A Zn-ion battery (ZIB) with Zn ions as shuttles has been demonstrated with Zn as the negative electrode and an a-MnO2246a or zinc hexacyanoferrate246b as the positive electrode material. In such a battery, the chemistry is based on the migration of Zn2+ ions between the positive electrode and the Zn negative electrode within a mild aqueous ZnSO4 electrolyte. The open circuit voltage (OCV) of this ZIB is approximately 1.5 V.246a At 126C within 27 s, the ZIB based on a-MnO2 delivers a relatively high capacity of 68 mA h g1. The other ZIB using zinc

This journal is © The Royal Society of Chemistry 2016

Energy & Environmental Science

Fig. 11 (a) The illustration of the Zn metal–Zn0.25V2O5 battery on discharge in the aqueous 1 M ZnSO4 electrolyte, (b) galvanostatic discharge profiles at different C rates (1C = 300 mA g1), and (c) scheme illustration of reversible H2O intercalation into Zn0.25V2O5nH2O immersed in the electrolyte, and the water deintercalation accompanying Zn2+ intercalation upon electrochemical discharge. The red and blue spheres represent O and H, respectively; the H2O molecules interact with the oxygen layers through hydrogen bonding (modified from ref. 246c, copyright permission from Nature Publishing Group).

hexacyanoferrates exhibits an average operation voltage of ca. 1.7 V and a specific energy density of 100 W h kg1 based on the total mass of the active electrode materials.246b Single-crystal Zn0.25V2O5nH2O nanobelts have also been reported as the positive electrode for Zn ion batteries in aqueous electrolytes (Fig. 11a). They display a capacity of 220 mA h g1 at an average voltage of 0.81 V (vs. Zn2+/Zn) at a 15C rate based on the mass of the positive electrode (Fig. 11b). The estimated volumetric energy density is 450 W h l1, with costs below US$65 KW1 h1.246c When cycling in the range of 0.5–1.4 V (vs. Zn2+/Zn) was carried out, it has a capacity retention of more than 80% after 1000 cycles. If cycling was carried out in a wider window (41.4 V vs. Zn2+/Zn), it would lead to higher specific capacities, but the structural stress generated from Zn2+ ion insertion may cause larger capacity fading and V dissolution. Based on operando XRD investigations, it was found that Zn0.25V2O5nH2O permits the insertion of H2O molecules from the electrolyte into the interlayer space, further expanding the galleries for facile Zn2+ intercalation (Fig. 11c). The water molecules play an important role in reversibly expanding and contracting the layered galleries of Zn0.25V2O5 to allow Zn2+ ingress/egress.246c In addition, the compact and flexible electrode architecture favors the release of strain generated during electrochemical Zn2+ ion (de)intercalation. This kind of battery has great potential for the applications where both high power and energy densities are needed. Moreover, zinc is a kind of low cost and non-toxic material with a large amount of production per year. A rechargeable nonaqueous Zn ion battery is also developed applying Zn(TFSI)2 solution in acetonitrile (AN) as the electrolyte.246d The AN–Zn(TFSI)2 electrolyte presents a highly

Energy Environ. Sci., 2016, 9, 3570--3611 | 3597

View Article Online

Published on 26 September 2016. Downloaded by SLUB DRESDEN on 29/11/2016 20:29:33.

Energy & Environmental Science

Review

reversible Zn deposition behaviour on a Zn metal negative electrode (Z99% of Coulombic efficiency), relatively low reversible deposition overpotential (E0.1 V), and high anodic stability (up to 3.6 V vs. Zn2+/Zn). The used positive electrode in the above nonaqueous Zn ion battery is bilayered hydrated V2O5 directly grown on a carbon foam substrate. Different from the above aqueous Zn ion battery, no solvent cointercalation occurs because no evidence of the characteristic –CRN band stretch from the acetonitrile solvent molecules was seen in Raman spectroscopy measurements. 9.2

F- and Cl-anion batteries

Besides the rechargeable batteries based on cation shuttles, a novel type of secondary battery with the anion (fluoride and chloride) as the charge transfer ion was developed and named the F- and Cl-ion batteries. The F-ion battery is operated at 150 1C by using a solid electrolyte of a LaF3/BaF2 composite.247 CuF2, BiF3 (both composite and solid solution), SnF2 and KBiF4 against Ce metal were also investigated. Both of them showed an OCV of above 2.0 V. However, capacity fading is a serious and well-known phenomenon for materials going through conversion reactions. The first concept of Cl-ion batteries was demonstrated by using an ionic liquid electrolyte (IL), a lithium foil as the negative electrode, and CoCl2, VCl3, or BiCl3 as the positive electrode.248 These positive electrodes show initial reversible discharge capacities of 105, 111 and 176 mA h g1, respectively. The positive and negative electrode electrochemical reactions are as follows: Positive electrode: MClx + xe 2 M + Cl

(14)

Negative electrode: M 0 + xCl 2 M 0 Clx + xe

(15)

Bismuth oxychloride (BiOCl) and iron oxychloride (FeOCl) were also investigated as positive electrode materials for Cl-ion batteries.249,250 In the discharge stage, the BiOCl (or FeOCl) loses the Cl ion and transforms to Bi (or Fe) metal and/or amorphous Bi2O3 (or FeO). During the charge process, the return of Cl ions results in the recovery of the BiOCl phase at the positive electrode side. The Li–FeOCl battery possesses a discharge capacity of 158 mA h g1 in the first cycle and 60 mA h g1 after 30 cycles. A large volume contraction (58.6%) or expansion (141.7%) could occur during the phase transformation between FeOCl and FeO, which are much larger than those of the BiOCl electrode during the Cl ion transfer.249 The feasibility of using Mg as a negative electrode for Cl-ion batteries was reported by investigating the electrochemical performance of FeOCl/Mg and BiOCl/Mg electrochemical couples.250 However, a large capacity decay was also observed during cycling due to the large volume change. 9.3

Na, K, Mg and Al–S batteries

Conventional Na–S batteries are based on Al2O3 as a solid ionic conductor or electrolyte of Na+ ions and separators, which work at temperatures of about 300–380 1C with a thickness of 41 mm, and have been used to support stationary energy

3598 | Energy Environ. Sci., 2016, 9, 3570--3611

storage systems for several decades.251 During the discharge process, Na+ ions migrate from the molten Na to the molten S through the solid electrolyte generating a cell voltage of about 2 V, following the cell reaction (16): xS + 2Na 2 Na2Sx

(16)

Na–S chemistry has a high theoretical energy density (760 W h kg1), a high energy efficiency and an acceptable cycle life. Despite their successful applications due to the low cost of raw materials, the high temperature easily causes a series of safety issues. A decrease in operation temperature is of great importance for the safety of the Na–S battery. Low- or room-temperature Na–S batteries using polymers and organic solvents as electrolytes have recently been reported.252,253 For example, Na–S batteries with NaClO4 in EC/DEC as the electrolyte show a good cycling performance when S-containing 1D carbon composites generated from a simple thermal reaction are used as the positive electrodes.253b A novel Na–S/NiCl2 hybrid battery was demonstrated to combine the electrochemical reactions for Na–S and Na–NiCl2 redox couples.254 The positive electrode contains a mixture of Ni, NaCl and Na2S as the active materials and NaAlCl4 as the catholyte. The addition of the NaAlCl4 catholyte allows for lower operating temperatures compared to traditional Na–S batteries. Moreover, the mixed chemistry exhibits a higher theoretical energy density than traditional Na–S chemistry. The feasibility and performance of room-temperature rechargeable Mg–S batteries,255 K–S batteries256 and Al–S batteries257a were also demonstrated very recently. Attempts have been made to convert Al–S primary batteries257b in aqueous electrolytes to Al–S secondary batteries in ionic liquid electrolytes,257a but the discharge capacities rapidly fade. As for Li–S batteries, as discussed in Section 2, these batteries usually face a few common troubles, such as the dissolution of polysulfides, and the volume expansion from sulfur to the discharge product of metal sulfides. Some main strategies have been applied to solve these problems. One is to encapsulate sulfur with various carbon materials, and the other is to wrap sulfur with conductive polymers. These strategies have been discussed in Section 2. 9.4

Na and K–O2 batteries

The first reported Na–O2 battery utilized a polymer electrolyte and a molten Na negative electrode operating at 100 1C.258 Both the peroxide259–262 and superoxide263–265 have been reported as the discharge products of Na–O2 batteries. The electrochemical behavior of the Na–O2 battery at room temperature has been reported based on diamond-like carbon,259 graphene,260 CNTs261 and carbon black.262 Combining with the ex situ TEM, SAED and FTIR results, it was found that Na2O2, Na2CO3 and NaOCO–R were formed as the products of the discharge reaction with the use of carbonate. At the same time, sodium superoxide (NaO2) is also the discharge product of the Na–O2 battery systems in an ether based electrolyte.263,264 Compared to Li2O2, the higher stability of NaO2 results in a lower amount of decomposition products and hence a lower charging overpotential.265 Recently, different charge/discharge mechanisms with various discharge

This journal is © The Royal Society of Chemistry 2016

View Article Online

Published on 26 September 2016. Downloaded by SLUB DRESDEN on 29/11/2016 20:29:33.

Review

products were proposed using carbonate and ether based electrolytes.266 It was found that sodium carbonate and hydrated sodium peroxide were produced as discharge products using carbonate and ether based electrolytes, respectively. Furthermore, theoretical calculations show that Na2O2 is the stable compound in the bulk form and NaO2 is only more stable at the nanoscale below 10 nm in particle size.267 A K–O2 battery was reported in which K+ ions can capture  O2 to form the thermodynamically stable KO2 product.268 In contrast to LiO2 and NaO2, KO2 is thermodynamically stable and commercially available. It shows a low discharge/charge potential gap of less than 50 mV at a modest current density without the use of catalysts, which is the lowest one that has ever been reported in metal–oxygen batteries. The formation and removal of KO2 during cycling were characterized using X-ray diffraction and Raman spectroscopy. Although several fundamental and engineering challenges still remain, the emergence of rechargeable Na and K–O2 batteries creates exciting new opportunities for cost-effective, high energy density and energy efficient electrochemical storage devices.269 9.5

Li–CO2 batteries

To achieve the practical application of Li–O2 batteries, they must be operated in an ambient air environment consisting of O2, CO2, and N2, etc. Thus, many research groups investigated the influence of moistured CO2 on the performance of Li–O2 batteries.270–272 The CO2 contamination was found to increase the discharge capacity of the cell and Li2CO3 is the main discharge product in such a battery. Moreover, it was found that Li2CO3 could be decomposed after mixing with NiO as a catalyst.273 Accordingly, it is plausible that a rechargeable Li–CO2 battery could be developed. A Li–CO2 battery with a discharging specific capacity of 1032 mA h g1 was demonstrated at room temperature when lithium triflate (LiCF3SO3)–TEGDME was used as the electrolyte.274 The polarization of the Li–CO2 battery is a little larger than that of the Li–CO2–O2 (2 : 1) battery and the round trip efficiency is a little low (66.3%). Recently, graphene was introduced into Li–CO2 batteries to yield systems with improved performances.275 The Li–CO2 batteries with a graphene positive electrode deliver a high discharge capacity of up to 14 774 mA h g1 and a stable cyclability over 20 cycles at 0.05 A g1. Although the reaction mechanisms (thermodynamic and kinetic properties) are still not very clear, this finding shows that greenhouse gas CO2 could be captured and utilized as a valuable energy storage medium. 9.6

Zn–air batteries

Aqueous zinc–air is a relatively mature technology and holds the greatest promise for future energy applications. Its primary batteries have been known to the scientific community for over a century.276 Zinc–air secondary batteries mainly consist of a zinc electrode, an electrolyte (usually aqueous alkaline solution) and an air electrode, which can be sub-classified into electrically rechargeable zinc–air batteries and mechanically rechargeable zinc–air batteries.277 Mechanically rechargeable zinc–air batteries are recharged by replacing the zinc electrode and/or the electrolyte. The electrically rechargeable zinc–air

This journal is © The Royal Society of Chemistry 2016

Energy & Environmental Science

batteries are recharged inside the cell by the bifunctional air electrode. The electrochemical reactions that occur in zinc–air secondary batteries with an alkaline electrolyte are as follows: Negative electrode: Zn + 4OH 2 Zn(OH)42 + 2e Zn(OH)42 2 ZnO + H2O + 2OH Positive electrode: 1/2O2 + H2O + 2e 2 2OH Overall reaction: 2Zn + O2 2 2ZnO

(17) (18) (19) (20)

A variety of non-noble catalysts can be used in Zn–air batteries with an alkaline catholyte, including carbon materials, transitionmetal oxides and perovskite oxides.276,277 Some previous reviews have provided some excellent summaries on the conventional stack-type zinc//air batteries from different perspectives.276–279 Even after decades of extensive research, the development of Zn–air is still hindered by several important problems. First, the electrolyte is very easy to leak. Second, the solution is evaporated to dry the electrolyte. Third, CO2 from air tends to react with the alkaline catholyte touching the catalyst layer, producing insoluble Li2CO3 on the catalyst layer. In this section, we present the latest two exciting advances: the cable-type flexible Zn–air battery280 and flowing Zn–air battery,281 which may help to address these two problems. The polymer layer helps not only to block CO2 in the air from penetrating into the positive electrolyte but also to relieve the leakage of the electrolytes. The fabrication process of the cable-type flexible Zn–air battery is a little complicated. The zinc metal plate is coiled on the surrounding stainless steel rod, followed by removing the rod. Then, the spiral zinc metal is located at the center of a cellophane template. After that, the spiral zinc is coated with the KOH-based gel polymer electrolyte (GPE). Next, the electrocatalyst loaded air electrode (FeC/Fe/ carbon) is wound on the surrounding GPE. Such cable-type flexible Zn–air demonstrates successful operation with external strain. There are no differences in the discharge voltage profiles between the bending and non-bending conditions.280 A single flow alkaline battery, a Zn/K2[Zn(OH)]4/O2 battery, was developed in which electrodeposited zinc was utilized as the negative electrode, oxygen from the atmosphere was used as the active material of a high capacity positive electrode, and a flowing KOH–K2[Zn(OH)4] solution stored in a tank and circulated using a pump was used as the electrolyte. The battery was designed to integrate the advantages of Zn//NiOOH single flow batteries with those of traditional Zn–air batteries. The battery displays a discharge voltage of 1.32 V with an average coulombic efficiency of 97.4% and an energy efficiency of 72.2% after 150 cycles.281 The redox flow battery is regarded as one of the most efficient energy storage forms appropriate for large-scale development in the future and will been discussed in the next section. 9.7

Hybrid redox flow batteries

Unlike other rechargeable batteries that store electrical energy within solid electrode materials, redox flow batteries (RFBs) normally employ two soluble redox couples as electroactive

Energy Environ. Sci., 2016, 9, 3570--3611 | 3599

View Article Online

Published on 26 September 2016. Downloaded by SLUB DRESDEN on 29/11/2016 20:29:33.

Energy & Environmental Science

species to realize the reversible conversion between electrical energy and chemical energy.90 During operation, redox-active ions undergo oxidation or reduction reactions when they are in contact or close to the current collector; the membrane allows the transport of non-reaction ions (such as H+ and Na+) to maintain electroneutrality and electrolyte balance.6 The energy densities of the RFBs are dependent on the volume and the concentration of the electrolytes, and the power densities can be made flexible by altering the size of the stack (the electrode size and the number of single cells). Since the 1970s, numerous types of redox flow battery systems have been investigated, such as all-vanadium RFBs282 and Zn//Br2 RFBs283 and Fe//Cr RFBs.284 The development of Fe–Cr RFBs was encountered with several snags and has never come to the stage of commercialization due to the use of chromium, which not only presents low electrochemical dynamics but also brings in environmental issues. The Zn–Br2 RFB presents a high energy density and a high output voltage. However, it shows an unfavorable power density resulting from the sluggish redox process of the Br2/Br couple on the positive electrode. Another big concern for most RFBs is the crosscontamination caused by the diffusion of ions from the positive electrolyte into the negative one through the membrane. It not only leads to an efficiency loss during cycling, but also causes capacity fading. However, few cross-contaminations occur in all-vanadium redox batteries (VFBs) since both electrolytes use vanadiumbased compounds. As shown in Fig. 12a, the typical structure of a VRB has two chambers, a positive chamber and a negative one, separated by an ion-exchange membrane. It employs the V5+/V4+ couple as a positive active electrolyte and the V2+/V3+ couple as a negative active electrolyte. In the case of VRBs, carbon felt (CF) is commonly used for current collectors of the positive electrodes because of its low-cost, good stability, high electronic conductivity and corrosion resistance. However, its serious disadvantages lie in the poor kinetic reversibility and low electrochemical activity. Depositing nanometer metals (such as Pt, Au, Ir, Pd, and Ru) onto the carbon electrode

Fig. 12

Review

surface is an effective way to facilitate the electrochemical kinetics of the vanadium redox reactions.285,286 Besides, the surface functional treatment method, including acid treatment and electro-oxidation, could introduce active functional groups into the carbon materials to enhance the hydrophilicity of the electrode and assist faster reaction of VO2+/VO+.90 There are also many limitations for VRBs, such as low energy density and efficiency, highly corrosive electrolytes, and low operating voltage caused by water electrolysis. A new trend in VRBs is the hybrid redox flow battery (or called a Li-redox flow battery).287 In this battery chemistry, the negative electrode is Li-metal in an organic electrolyte. Its positive electrode is a liquid phase redox reaction from the redox molecule with a relatively high redox potential. Its charge carrier is the Li+ ion and the Li-ion conducting membrane (such as LISICON) is adopted to prevent the cross-over of redox couples in either side of the electrodes (Fig. 12b). One advantage of the Li-redox flow battery is its modular design that provides flexible operation, transportability, and moderate manufacturing cost. Moreover, such design provides a higher cell voltage and energy density as compared with the common redox flow batteries. For example, an aqueous positive electrolyte containing a Fe3+/Fe2+ (3.77 V vs. Li+/Li) or Fe(CN)63/Fe(CN)64 (3.4 V vs. Li+/Li) redox couple is separated from metallic Li by a solid–electrolyte (LISICON plate).288 Actually, this kind of device is a semi-VRB. If the negative electrode uses redox-active liquid or slurry (not Li metal), a full Li-redox flow battery can be achieved. Interestingly, similar to the Zn–K2[Zn(OH)]4–O2 battery discussed in the above section, oxygen gas (O2) could also be employed to this hybrid flow battery and a new concept (Li redox flow air battery) was proposed.287a Besides the O2/OH, several reversible redox couples (like S42/S2, I3/I, Br2/Br, and Sn4+/Sn2+) in the aqueous phase are also better choices for such a cell system in consideration of a lower cost. Moreover, Na, Mg, and Al may be some possible alternatives for Li metal. Very recently, a Na-redox flow battery has been designed using a molten Na–Cs alloy as the negative electrode, flowing aqueous V4+/V5+ as the positive electrolyte, and a Na–b-Al2O3 solid

The schematic diagram of (a) redox flow battery (modified from ref. 6, copyright permission from AAAS) and (b) Li-redox flow battery system.

3600 | Energy Environ. Sci., 2016, 9, 3570--3611

This journal is © The Royal Society of Chemistry 2016

View Article Online

Published on 26 September 2016. Downloaded by SLUB DRESDEN on 29/11/2016 20:29:33.

Review

electrolyte as the separator.289a In spite of the high impedance of the state-of-the-art Na–b-Al2O3 membrane, the Na-redox flow battery has been shown to be capable of good capacity retention for up to 30 cycles with high coulombic efficiency (B90%) and at room temperature or 50 1C. This is the first time a waterresistant b00 -Al2O3 disc has been successfully used at and near room temperature. However, as of yet there have been no particular investigations in Mg and Al-redox flow batteries because no suitable solid state electrolyte possesses the exclusive selectivity for Mg2+ and Al3+ ions between the aqueous electrolyte and the organic electrolyte. Different from such aqueous/organic electrolyte systems, a non-aqueous Mg–Br2 battery was reported as a proof-of-concept demonstration.289b The negative electrolytes consist of Mg(TFSI)2 dissolved in a monoglyme and diglyme mixture while the positive electrolytes are composed of Mg(TFSI)2 in PYR14TFSI ionic liquid mixed with active bromine species. The two electrolytes are separated by a fine porous glass frit (4 mm nominal pore size). The authors claim that porous glass frit limits the diffusion of Br2 to slow down the cross-over but allows for exchange of Mg2+ for small polarization. Polymeric Mg2+ ion conductors such as PVDFbased ones may help further reduce the diffusion of Br2 species for future studies.289b

10. Outlooks In summary, a series of positive electrode materials with various nanostructures have been discussed for different postlithium ion batteries. The fast-developing nanoscience and nanotechnology have already established that the electrochemical properties of nanomaterials are closely linked with their nano-structures. The electrochemical properties of the positive electrode materials for these various rechargeable batteries can be improved through careful design of the electrode architecture. Generally, nanostructured materials help to improve the thermodynamic and kinetic properties of electrochemical reactions for achieving high energy and power densities. 0D nanosized electrode materials are advantageous in terms of kinetics and capacity, but their practical application suffers from low thermodynamic stability and some undesirable side reactions because of their confined size and high surface energy, especially when exposed to the ambient atmosphere. 1D electrode materials usually show good mechanical properties and efficient electron diffusion pathways. For example, careful engineering of the electrode materials and highly conductive CNTs into 1D hybrid nanostructures leads to enhanced ion storage properties. However, besides many similarities to 0D electrode materials with high surface energy architectures, 1D electrode materials are also accompanied by the irreversible capacity loss associated with electrolyte degradation occurring on those high energy sites. The nanosheet geometry (2D) may be the best option to take advantage of both the high capacity and rates offered by these nanoarchitectures while avoiding significant irreversible capacity losses as the surface strain effect would be lessened for the flexible sheets. However, its

This journal is © The Royal Society of Chemistry 2016

Energy & Environmental Science

critical issue is the restacking of sheets during electrode preparation. Electrode materials with 3D nanostructures not only create hierarchical porous channels, but also maintain better structural mechanical stability. Specifically, the catalyst with a 3D ordered pore system for metal–O2 batteries has rapid gas and ion transference compared to the 2D catalysts. From a practical point of view, most 3D electrode materials have two serious disadvantages of (i) high production cost and (ii) poor scale-up ability. A lot of researchers are working on developing low-cost processes for the production of high-performance 3D composites. Hollow nanostructured electrodes also demonstrate enhanced rate capability and cycling stability, but suffer from reduced volumetric energy density because of the presence of the inner cavity. Most nanoarchitectures used in a battery present lower volumetric power or energy density, and a hierarchical assembly of nanomaterials into larger micrometersized structures is required to increase tap and volumetric energy densities. Although the specific capacities of the nanostructured positive electrode materials for post-lithium ion batteries are typically reported in the gravimetric form (mA h g1), the volumetric energy density or capacity (W h l1 or mA h cm3) of the composite materials could be more important when considering commercial applications.6b Progress over the last several years in the development of sulfur positive electrodes is mainly focused on confining the migration of polysulfides, relieving the volume expansion, improving poor ionic and electronic transport. These developments represent promising steps towards a viable Li–S battery. However, the volumetric energy density is low since a more conductive carbon or polymer is needed for the operation of the sulfur positive electrode. The typical sulfur content is 30–80 wt% in the most academic reports. Thus, improving the tap density of the whole electrode for Li–S batteries is very important in the future. Elemental Se, an analog of S, is proposed as a new attractive candidate with high volumetric energy density and good electronic conductivity comparable with S. Admittedly, Se is less abundant in the crust than S. The undesirable high cost is a major drawback of Se-based electrode materials. Regarding cost, aqueous batteries such as ARLBs and RFBs have a great advantage due to the low cost of the aqueous electrolyte and the simple assembling process. Moreover the ionic conductivity of aqueous electrolytes is high which ensures high rate capability and thus high specific power. Therefore, the aqueous rechargeable batteries show potential applications in stationary energy storage systems. However, the energy density is not high because of the low electrochemical window of the aqueous electrolyte. Li–O2 batteries (ultimately Li–air batteries) appear to offer the greatest hope for the future: the specific energies are 10 times greater than those of lithium-ion systems. However, the sluggish kinetics of oxygen reduction/ evolution hinders their application. Highly efficient bifunctional nanostructured catalysts are helpful to improve reaction kinetics, but the safety and the stability of the electrolyte are difficult to guarantee. The use of a room-temperature sodium ion battery is being proposed by some academic community and start-up companies.

Energy Environ. Sci., 2016, 9, 3570--3611 | 3601

View Article Online

Published on 26 September 2016. Downloaded by SLUB DRESDEN on 29/11/2016 20:29:33.

Energy & Environmental Science

Although there are many obstacles to overcome before its commercialization, recent research discovers that some of the positive electrode materials for the Na-ion battery have indeed got indisputable advantages over their Li-ion counterparts. In addition, the road to post-Li ion batteries is also being paved with the development of a variety of multivalent cation systems such as Mg, Al, or Zn ion batteries. Currently, despite the significant advances already achieved, the deployment of post-lithium ion batteries is still at its early stage. A solid foundation for future technical milestones is being established. Great opportunities and huge challenges coexist in this field. From a material perspective, the key relies on the continued efforts to develop advanced designs and simple fabrication methods of nanostructured positive electrodes. The promising nanotechnology can undoubtedly help researchers to fabricate various positive electrodes with welldesigned nanocharacteristics, which promote the development of better batteries. Again, which kind of battery would really fulfill future applications such as electric vehicles? There is no clear answer now and further investigation is needed. It should be noted that energy diversification is also necessary for global sustainability because no battery can do everything and various batteries have to find their appropriate uses. Throughout this review, we hope to generate more interest in them and boost extensive investigation in the related areas. Thus, listed below are the new vital interests that are expected. First, the development of new battery materials should rely on an in-depth fundamental understanding of materials properties at the atomic level. Powerful quantum computational techniques rather than purely Edisonian methods can be utilized to predict many material properties. As an example, the physics of room temperature Mg intercalation in V2O5 was explored in detail using first-principles calculations recently.290 It was found that the d-V2O5 polymorph displayed vastly superior Mg mobility compared to the a-phase, which suggested that better performance could be achieved by cycling Mg in the d-phase. Second, the integration of multiple tools based on in situ and ex situ characterization techniques is necessary. A single tool sometimes gives misleading results. Future demands for electricity storage will require significant research progress in both nanomaterial synthesis and in situ monitoring. In situ characterization techniques, such as transmission electron microscopy (TEM),291 nuclear magnetic resonance (NMR),292 electrochemical X-ray absorption spectroscopy,293a X-ray diffraction (XRD),293b and energy-dispersive X-ray diffraction,294 have improved the understanding of the reaction mechanisms and will thus lead to an optimized design for the electrode. Third, some new nanotechnology (such as a 3D printing technique) may offer great promise in developing novel positive electrode materials for post-Li ion batteries in the future.295–298 For example, 3D microbatteries are printed on a sub-millimeter scale, which exhibit the highest areal energy and power densities to date.295 Miniaturized batteries are a very attractive direction since small-scale energy storage devices can be integrated with microelectronic devices to work as stand-alone power sources or

3602 | Energy Environ. Sci., 2016, 9, 3570--3611

Review

efficient energy storage units complementing energy harvesters, leading to a wider use of these devices in many industries.299,300 Besides those new nanotechnologies, some possible hybrid approaches between post-Li ion batteries and other devices may allow for a reinforcing combination of the properties of various components in a synergic way. As a good example, a ‘‘lithium–air capacitor–battery’’ with an additional capacitor positive electrode successfully presented capacitor characteristics of lithium–air batteries in one device.301 In this hybrid device, the capacitor electrode would play the main role in peak power output when high power is needed; while when high energy is demanded, the air catalytic positive electrode would display its high energy characteristics. Very recently, an aqueous lithium–iodine (Li–I) solar flow battery was also demonstrated by the incorporation of a built-in dye-sensitized TiO2 photoelectrode into a Li–I redox flow battery.302 Compared to conventional Li–I batteries, the hybrid device could achieve energy savings up to 20%. They also proved with an example of the sodium–iodine (Na–I) solar flow battery. Interestingly, Wu’s group reported the Na-ion capacitor303 and the Al-ion capacitor304a assembled by capacitor-type electrodes and active ion (Na ions and Al ions) intercalation electrodes. This kind of device combines energy storage mechanisms of supercapacitors and post-Li ion batteries with anions adsorbing/desorbing onto/from one electrode surface and Na+ (or Al3+) ions intercalating/de-intercalating into/from the bulk of the other electrode simultaneously. The main purpose of this hybridization is to bridge the gap existing between the electrochemical double layer capacitors (EDLCs) and the post-Li ion batteries.304b Additionally, a battery with multi-functions is another that is attracting research driven by the need for modern electronic systems such as wearable devices. The use of current collectors with soft, bendable and elastic properties is the prerequisite for making flexible power devices.305 Currently, plastic (e.g. polyethylene terephthalate (PET) and polydimethylsiloxane (PDMS)) polymers and carbon cloth are the predominant scaffold materials due to their low cost, good processability and flexibility. Papers are also proven to be potential candidates for making attractive bendable powering devices.306 Another interesting work is the fabrication of transparent batteries.307,308 A novel grid-structured electrode design was utilized to make a ‘‘transparent’’ electrode and the principle idea is that the feature dimension in such an electrode is less than the resolution of human eyes.307 A future photovoltaic–electrochemical battery in a transparent smart window may serve to store electricity during the day and to illuminate rooms in commercial buildings at night. Besides, some mechanical-to-electrochemical selfcharging batteries have been fabricated based on a new fundamental mechanism (piezo-electrochemical process).309–311 The piezoelectric potential usually from the PVDF film created via mechanical straining can act as a charge pump to drive active ions to migrate from the positive electrode to the negative one accompanying charging reactions at the electrodes. Energy storage devices with one or several of the following characteristics (smart, stretchable, self-heating, reversible thermoresponsive and self-healing)312,313 may also meet some requirements

This journal is © The Royal Society of Chemistry 2016

View Article Online

Published on 26 September 2016. Downloaded by SLUB DRESDEN on 29/11/2016 20:29:33.

Review

of growing development of portable electronic devices in the future. For example, a battery structure with self-heating at low temperatures creates an ‘all-climate battery’ cell even below 30 1C.312b Finally, the design of post-Li ion batteries will not only rely on positive electrode materials, but also on their integration with other elements such as tailored electrolytes, membrane separators, current collectors, negative electrodes, as well as other practical issues of cell design. For example, a solid or gel electrolyte314,315 will become one of the major research focuses in the area of post Li-ion batteries due to its advantages of safe operation/transportation, high packaging efficiency and long device lifetime. Many metals (i.e. Li, Na, and Zn) have shown great promise as negative materials for high-energy storage systems (i.e. metal–sulfur batteries and metal–air batteries) owing to their high theoretical specific capacity. Unfortunately, uncontrolled dendritic and mossy metal growth during repeated charging/discharging cycles has impeded their use in practical applications.316 In this regard, Wu’s group extensively studied the formation of Zn dendrites during cycling and demonstrated that the growth of Zn dendrites can be suppressed by electrodepositing zinc on carbon fibers (CFs) with a zinc@CF core–shell structure.316e In addition, it should be mentioned that the active electrode materials are less than 50% of the weight of the current batteries due to all of the packaging. Currently, there is very limited research on how the electrode material should be incorporated into a real, stable device. When and by how much will post-Li ion batteries based on nanomaterials start to make a profit? Maintaining close collaboration with downstream application companies is necessary for their development. There is still great room for improvement in the cooperation between chemistry (electrochemistry, inorganic chemistry, materials chemistry, theoretical chemistry, polymer chemistry, and the like) and engineering. When they are fully collaborated, practical applications will be soon due. Currently, the batch consistency of the commercial rechargeable batteries has long been reviled. In order to realize the automatic even intelligent production line of commercial batteries, battery enterprises should work together on equipment and materials manufactures for joint development and production. Moreover, rechargeable batteries after thousands of cycles can be recycled but it is a high-cost process and needs a professional workforce. Every year there will now be over 200 million mobile phones eliminated, which means that each year over 200 million pieces of batteries are abandoned. So the industrialization of post-Li ion batteries also needs more attention to the recovery by minimizing harms to the environment. Additionally, there are still many technical setbacks to conquer in the battery management systems.

Acknowledgements Financial support from the Distinguished Young Scientists Program of the National Natural Science Foundation of China (NSFC 51425301), and NSFC (51502137) and MOST (2016YFB0700600) is gratefully appreciated.

This journal is © The Royal Society of Chemistry 2016

Energy & Environmental Science

Notes and references 1 M. M. Thackeray, C. Wolverton and E. D. Isaacs, Energy Environ. Sci., 2012, 5, 7854–7863. 2 (a) R. M. Dell and D. A. J. Rand, Chapter 1 Energy Storage in Batteries in Understanding Batteries, 2001, pp. 1–8; (b) M. S. Whittingham, Science, 1976, 192, 1126–1127; (c) R. M. Dell and D. A. J. Rand, Chapter 10 Lithium Batteries in Understanding Batteries, 2001, pp. 143–162; (d) K. Mizushima, P. C. Jones, P. J. Wiseman and J. B. Goodenough, Mater. Res. Bull., 1980, 15, 783–789; (e) D. Larcher and J. M. Tarascon, Nat. Chem., 2015, 7, 19–29; ( f ) http://libattery.ofweek. com/ 2013-11/ART-36001-8400-28742584.html. 3 (a) P. G. Bruce, S. A. Freunberger, L. J. Hardwick and J. M. Tarascon, Nat. Mater., 2012, 11, 19–29; (b) X. Gao, Y. Chen, L. Johnson and P. G. Bruce, Nat. Mater., 2016, 15, 882–888. 4 G. Wang, L. Fu, N. Zhao, L. Yang, Y. Wu and H. Wu, Angew. Chem., Int. Ed., 2007, 46, 295–297. 5 J. Muldoon, C. B. Bucur and T. Gregory, Chem. Rev., 2014, 114, 11683–11720. 6 (a) B. Dunn, H. Kamath and J. M. Tarascon, Science, 2011, 334, 928–935; (b) J. W. Choi and D. Aurbach, Nat. Rev. Mater., 2016, DOI: 10.1038/natrevmats.2016.13. 7 (a) K. Zhang, X. Han, Z. Hu, X. Zhang, Z. Tao and J. Chen, Chem. Soc. Rev., 2015, 44, 699–728; (b) Q. Zhang, E. Uchaker, S. L. Candelaria and G. Cao, Chem. Soc. Rev., 2013, 42, 3127–3171; (c) S. Yang, R. E. Bachman, X. Feng and K. Mullen, Acc. Chem. Res., 2013, 46, 116–128; (d) M. V. Reddy, G. V. S. Rao and B. V. R. Chowdari, Chem. Rev., 2013, 113, 5364–5457; (e) Z. Yang, J. Ren, Z. Zhang, X. Chen, G. Guan, L. Qiu, Y. Zhang and H. Peng, Chem. Rev., 2015, 115, 5159–5223. 8 B. Kang and G. Ceder, Nature, 2009, 458, 190–193. 9 (a) A. S. Arico, P. Bruce, B. Scrosati, J. M. Tarascon and W. V. Schalkwijk, Nat. Mater., 2005, 4, 366–377; (b) C. Liu, E. I. Gillette, X. Chen, A. J. Pearse, A. C. Kozen, M. A. Schroeder, K. E. Gregorczyk, S. B. Lee and G. W. Rubloff, Nat. Nanotechnol., 2014, 9, 1031–1039; (c) Y. Wang, H. Li, P. He, E. Hosono and H. Zhou, Nanoscale, 2010, 2, 1294–1305. 10 P. G. Bruce, B. Scrosati and J. M. Tarascon, Angew. Chem., Int. Ed., 2008, 47, 2930–2946. 11 Y. Yang, G. Zheng and Y. Cui, Chem. Soc. Rev., 2013, 42, 3018–3032. 12 Y. X. Yin, S. Xin, Y. G. Guo and L. J. Wan, Angew. Chem., Int. Ed., 2013, 52, 13186–13200. 13 A. Manthiram, Y. Fu, S. H. Chung, C. Zu and Y. S. Su, Chem. Rev., 2014, 114, 11751–11787. 14 M. L. Rao, US Pat., 3413154, 1968. 15 D. Nole, and V. Moss, US Pat., 3532543, 1970. 16 A. Manthiram, Y. Fu and Y. S. Su, Acc. Chem. Res., 2013, 46, 1125–1134. 17 (a) J. Guo, Y. Xu and C. Wang, Nano Lett., 2011, 11, 4288–4294; (b) S. Xin, L. Gu, N. H. Zhao, Y. X. Yin, L. J. Zhou, Y. G. Guo and L. J. Wan, J. Am. Chem. Soc., 2012, 134, 18510–18513. 18 (a) G. Zhou, D. W. Wang, F. Li, P. X. Hou, L. Yin, C. Liu, G. Q. Lu, I. R. Gentle and H. M. Cheng, Energy Environ. Sci.,

Energy Environ. Sci., 2016, 9, 3570--3611 | 3603

View Article Online

Energy & Environmental Science

19

Published on 26 September 2016. Downloaded by SLUB DRESDEN on 29/11/2016 20:29:33.

20 21

22

23 24 25

26

27 28

29

30 31 32 33

34 35

2012, 5, 8901–8906; (b) S. Dorfler, M. Hagen, H. Althues, J. Tubke, S. Kaskel and M. J. Hoffmann, Chem. Commun., 2012, 48, 4097–4099. (a) K. Fu, Y. Li, M. Dirican, C. Chen, Y. Lu, J. Zhu, Y. Li, L. Cao, P. D. Bradford and X. Zhang, Chem. Commun., 2014, 50, 10277–10280; (b) W. Ahn, K. B. Kim, K. N. Jung, K. H. Shin and C. S. Jin, J. Power Sources, 2012, 202, 394–399. L. Ji, M. Rao, S. Aloni, L. Wang, E. J. Cairns and Y. Zhang, Energy Environ. Sci., 2011, 4, 5053–5059. Y. Chen, X. Li, K. S. Park, J. Hong, J. Song, L. Zhou, Y. W. Mai, H. Huang and J. B. Goodenough, J. Mater. Chem. A, 2014, 2, 10126–10130. J. J. Chen, Q. Zhang, Y. Shi, L. L. Qin, Y. Cao, M. S. Zheng and Q. F. Dong, Phys. Chem. Chem. Phys., 2012, 14, 5376–5382. Z. Zhang, H. K. Jing, S. Liu, G. R. Li and X. P. Gao, J. Mater. Chem. A, 2015, 3, 6827–6834. Y. S. Su and A. Manthiram, Chem. Commun., 2012, 48, 8817–8819. K. S. Novoselov, A. K. Geim, S. V. Morozov, D. Jiang, Y. Zhang, S. V. Dubonos, I. V. Grigorieva and A. A. Firsov, Science, 2004, 306, 666–669. (a) D. R. Dreyer, S. Park, C. W. Bielawski and R. S. Ruoff, Chem. Soc. Rev., 2010, 39, 228–240; (b) A. Narita, X. Y. Wang, X. Feng and K. Mullen, Chem. Soc. Rev., 2015, 44, 6616–6643; ¨llen, J. Am. (c) W. Wei, G. Wang, S. Yang, X. Feng and K. Mu Chem. Soc., 2015, 137, 5576–5581; (d) X. Zhuang, D. Gehrig, N. Forler, H. Liang, M. Wagner, M. R. Hansen, F. Laquai, F. Zhang and X. Feng, Adv. Mater., 2015, 27, 3789–3796; (e) X. Zhuang, Y. Mai, D. Wu, F. Zhang and X. Feng, Adv. Mater., 2015, 27, 403–427. J. Z. Wang, L. Lu, M. Choucair, J. A. Stride, X. Xu and H. K. Liu, J. Power Sources, 2011, 196, 7030–7034. (a) J. Q. Huang, X. F. Liu, Q. Zhang, C. M. Chen, M. Q. Zhao, S. M. Zhang, W. Zhu, W. Z. Qian and F. Wei, Nano Energy, 2013, 2, 314–321; (b) Y. Cao, X. Li, I. A. Aksay, J. Lemmon, Z. Nie, Z. Yang and J. Liu, Phys. Chem. Chem. Phys., 2011, 13, 7660–7665. Y. X. Wang, L. Huang, L. C. Sun, S. Y. Xie, G. L. Xu, S. R. Chen, Y. F. Xu, J. T. Li, S. L. Chou, S. X. Dou and S. G. Sun, J. Mater. Chem., 2012, 22, 4744–4750. L. Yin, J. Wang, X. Yu, C. W. Monroe, Y. NuLi and J. Yang, Chem. Commun., 2012, 48, 7868–7870. T. Lin, Y. Tang, Y. Wang, H. Bi, Z. Liu, F. Huang, X. Xie and M. Jiang, Energy Environ. Sci., 2013, 6, 1283–1290. B. Li, S. Li, J. Liu, B. Wang and S. Yang, Nano Lett., 2015, 15, 3073–3079. (a) R. D. Cakan, M. Morcrette, F. Nouar, C. Davoisne, T. Devic, D. Gonbeau, R. Dominko, C. Serre, G. Ferey and J. M. Tarascon, J. Am. Chem. Soc., 2011, 133, 16154–16160; (b) Z. Zhao, S. Wang, R. Liang, Z. Li, Z. Shi and G. Chen, J. Mater. Chem. A, 2014, 2, 13509–13512. H. Wang, Y. Yang, Y. Liang, J. T. Robinson, Y. Li, A. Jackson, Y. Cui and H. Dai, Nano Lett., 2011, 11, 2644–2647. Y. Zhu, S. Murali, M. D. Stoller, K. J. Ganesh, W. Cai, P. J. Ferreira, A. Pirkle, R. M. Wallace, K. A. Cychosz,

3604 | Energy Environ. Sci., 2016, 9, 3570--3611

Review

36 37

38 39 40

41 42

43 44 45 46 47 48 49

50 51 52

53

54 55

56 57 58

M. Thommes, D. Su, E. A. Stach and R. S. Ruoff, Science, 2011, 332, 1537–1541. Y. You, W. Zeng, Y. X. Yin, J. Zhang, C. P. Yang, Y. Zhu and Y. G. Guo, J. Mater. Chem. A, 2015, 3, 4799–4802. L. Ji, M. Rao, H. Zheng, L. Zhang, Y. Li, W. Duan, J. Guo, E. J. Cairns and Y. Zhang, J. Am. Chem. Soc., 2011, 133, 18522–18525. Y. Cao, X. Li, I. A. Aksay, J. Lemmon, Z. Nie, Z. Yang and J. Liu, Phys. Chem. Chem. Phys., 2011, 13, 7660–7665. J. Q. Huang, Q. Zhang, H. J. Peng, X. Y. Liu, W. Z. Qian and F. Wei, Energy Environ. Sci., 2014, 7, 347–353. (a) X. Ji, K. T. Lee and L. F. Nazar, Nat. Mater., 2009, 8, 500–506; (b) N. Jayaprakash, J. Shen, S. S. Moganty, A. Corona and L. A. Archer, Angew. Chem., Int. Ed., 2011, 50, 5904–5908. B. Zhang, X. Qin, G. R. Li and X. P. Gao, Energy Environ. Sci., 2010, 3, 1531–1537. S. R. Chen, Y. P. Zhai, G. L. Xu, Y. X. Jiang, D. Y. Zhao, J. T. Li, L. Huang and S. G. Sun, Electrochim. Acta, 2011, 56, 9549–9555. C. Liang, N. J. Dudney and J. Y. Howe, Chem. Mater., 2009, 21, 4724–4730. J. Schuster, G. He, B. Mandlmeier, T. Yim, K. T. Lee, T. Bein and L. F. Nazar, Angew. Chem., Int. Ed., 2012, 51, 3591–3595. G. He, X. Ji and L. F. Nazar, Energy Environ. Sci., 2011, 4, 2878–2883. H. B. Wu, S. Wei, L. Zhang, R. Xu, H. H. Hng and X. W. Lou, Chem. – Eur. J., 2013, 19, 10804–10808. G. Xu, B. Ding, L. Shen, P. Nie, J. Han and X. Zhang, J. Mater. Chem. A, 2013, 1, 4490–4496. K. Xi, S. Cao, X. Peng, C. Ducati, R. V. Kumar and A. K. Cheetham, Chem. Commun., 2013, 49, 2192–2194. M. Hu, J. Reboul, S. Furukawa, N. L. Torad, Q. Ji, P. Srinivasu, K. Ariga, S. Kitagawa and Y. Yamauchi, J. Am. Chem. Soc., 2012, 134, 2864–2867. W. Hu, H. Zhang, Y. Zhang, M. Wang, C. Qu and J. Yi, Chem. Commun., 2015, 51, 1085–1088. S. Evers and L. F. Nazar, Acc. Chem. Res., 2013, 46, 1135–1143. X. Li, Y. Cao, W. Qi, L. V. Saraf, J. Xiao, Z. Nie, J. Mietek, J. G. Zhang, B. Schwenzer and J. Liu, J. Mater. Chem., 2011, 21, 16603–16610. L. Xiao, Y. Cao, J. Xiao, B. Schwenzer, M. H. Engelhard, L. V. Saraf, Z. Nie, G. J. Exarhos and J. Liu, Adv. Mater., 2012, 24, 1176–1181. ˜a, W. Zhou, Y. Yu, H. Chen, F. J. Disalvo and H. D. Abrun J. Am. Chem. Soc., 2013, 135, 16736–16743. M. Wang, W. Wang, A. Wang, K. Yuan, L. Miao, X. Zhang, Y. Huang, Z. Yu and J. Qiu, Chem. Commun., 2013, 49, 10263–10265. F. Wu, J. Chen, R. Chen, S. Wu, L. Li, S. Chen and T. Zhao, J. Phys. Chem. C, 2011, 115, 6057–6063. Z. Dong, J. Zhang, X. Zhao, J. Tu, Q. Su and G. Du, RSC Adv., 2013, 3, 24914–24917. L. Huang, J. Cheng, X. Li, D. Yuan, W. Ni, G. Qu, Q. Guan, Y. Zhang and B. Wang, J. Mater. Chem. A, 2015, 3, 4049–4057.

This journal is © The Royal Society of Chemistry 2016

View Article Online

Published on 26 September 2016. Downloaded by SLUB DRESDEN on 29/11/2016 20:29:33.

Review

59 W. Li, G. Zheng, Y. Yang, Z. W. Seh, N. Liu and Y. Cui, Proc. Natl. Acad. Sci. U. S. A., 2013, 110, 7148–7153. 60 W. Li, Q. Zhang, G. Zheng, Z. W. Seh, H. Yao and Y. Cui, Nano Lett., 2013, 13, 5534–5540. 61 (a) G. Zheng, Y. Yang, J. J. Cha, S. S. Hong and Y. Cui, Nano Lett., 2011, 11, 4462–4467; (b) E. S. Shin, K. Kim, S. H. Oh and W. I. Cho, Chem. Commun., 2013, 49, 2004–2006; (c) S. Jiang, Z. Zhang, X. Wang, Y. Qu, Y. Lai and J. Li, RSC Adv., 2013, 3, 16318–16321. 62 X. Li, A. Lushington, J. Liu, R. Li and X. Sun, Chem. Commun., 2014, 50, 9757–9760. 63 (a) T. Takeuchi, H. Sakaebe, H. Kageyama, H. Senoh, T. Sakai and K. Tatsumi, J. Power Sources, 2010, 195, 2928–2934; (b) A. Hayashi, R. Ohtsubo, T. Ohtomo, F. Mizuno and M. Tatsumisago, J. Power Sources, 2008, 183, 422–426; (c) Y. Yang, G. Y. Zheng, S. Misra, J. Nelson, M. F. Toney and Y. Cui, J. Am. Chem. Soc., 2012, 134, 15387–15394; (d) Y. Yang, M. T. McDowell, A. Jackson, J. J. Cha, S. S. Hong and Y. Cui, Nano Lett., 2010, 10, 1486–1491; (e) M. Nagao, A. Hayashi and M. Tatsumisago, J. Mater. Chem., 2012, 22, 10015–10020. 64 (a) L. Chen, Y. Liu, N. D. Rago and L. L. Shaw, Nanoscale, 2015, 7, 18071–18080; (b) Y. Fu, C. Zu and A. Manthiram, J. Am. Chem. Soc., 2013, 135, 18044–18047; (c) S. Liang, C. Liang, Y. Xia, H. Xu, H. Huang, X. Tao, Y. Gan and W. Zhang, J. Power Sources, 2016, 306, 200–207; (d) L. Chen, Y. Liu, M. Ashuri, C. Liu and L. L. Shaw, J. Mater. Chem. A, 2014, 2, 18026–18032; (e) F. Wu, J. T. Lee, F. Fan, N. Nitta, H. Kim, T. Zhu and G. Yushin, Adv. Mater., 2015, 27, 5579–5586; ( f ) L. Chen, Y. Liu, F. Zhang, C. Liu and L. L. Shaw, ACS Appl. Mater. Interfaces, 2015, 7, 25748–25756; (g) L. Suo, Y. Zhu, F. Han, T. Gao, C. Luo, X. Fan, Y. S. Hu and C. Wang, Nano Energy, 2015, 13, 467–473. 65 (a) F. Ye, Y. Hou, M. Liu, W. Li, X. Yang, Y. Qiu, L. Zhou, H. Li, Y. Xu and Y. Zhang, Nanoscale, 2015, 7, 9472–9476; (b) H. Noh, J. Song, J. K. Park and H. T. Kim, J. Power Sources, 2015, 293, 329–335; (c) M. Wu, Y. Cui and Y. Fu, ACS Appl. Mater. Interfaces, 2015, 7, 21479–21486; (d) F. Wu, A. Magasinski and G. Yushin, J. Mater. Chem. A, 2014, 2, 6064–6070. 66 (a) K. Zhang, L. Wang, Z. Hu, F. Cheng and J. Chen, Sci. Rep., 2014, 4, 6467–6473; (b) Z. Li, S. Zhang, C. Zhang, K. Ueno, T. Yasuda, R. Tatara, K. Dokko and M. Watanabe, Nanoscale, 2015, 7, 14385–14392; (c) C. Wang, X. Wang, Y. Yang, A. Kushima, J. Chen, Y. Huang and J. Li, Nano Lett., 2015, 15, 1796–1802; (d) Y. Qiu, G. Rong, J. Yang, G. Li, S. Ma, X. Wang, Z. Pan, Y. Hou, M. Liu, F. Ye, W. Li, Z. W. Seh, X. Tao, H. Yao, N. Liu, R. Zhang, G. Zhou, J. Wang, S. Fan, Y. Cui and Y. Zhang, Adv. Energy Mater., 2015, 5, 1501369–1501377; (e) G. Zhou, E. Paek, G. S. Hwang and A. Manthiram, Adv. Energy Mater., 2016, 6, 1501355–1501363; ( f ) F. Wu, J. T. Lee, E. Zhao, B. Zhang and G. Yushin, ACS Nano, 2016, 10, 1333–1340. 67 (a) Z. W. Seh, H. Wang, P. C. Hsu, Q. Zhang, W. Li, G. Zheng, H. Yao and Y. Cui, Energy Environ. Sci., 2014, 7, 672–676; (b) J. Liu, H. Nara, T. Yokoshima, T. Momma

This journal is © The Royal Society of Chemistry 2016

Energy & Environmental Science

68

69 70 71

72 73 74 75 76 77 78 79 80 81 82 83 84 85

86

and T. Osaka, J. Power Sources, 2015, 273, 1136–1141; (c) S. Zhang, M. Liu, F. Ma, F. Ye, H. Li, X. Zhang, Y. Hou, Y. Qiu, W. Li, J. Wang, J. Wang and Y. Zhang, J. Mater. Chem. A, 2015, 3, 18913–18919; (d) L. Wang, Y. Wang and Y. Xia, Energy Environ. Sci., 2015, 8, 1551–1558. A. Abouimrane, D. Dambournet, K. W. Chapman, P. J. Chupas, W. Weng and K. Amine, J. Am. Chem. Soc., 2012, 134, 4505–4508. C. P. Yang, Y. X. Yin and Y. G. Guo, J. Phys. Chem. Lett., 2015, 6, 256–266. Z. Zhang, X. Yang, X. Wang, Q. Li and Z. Zhang, Solid State Ionics, 2014, 260, 101–106. Y. Cui, A. Abouimrane, J. Lu, T. Bolin, Y. Ren, W. Weng, C. Sun, V. A. Maroni, S. M. Heald and K. Amine, J. Am. Chem. Soc., 2013, 135, 8047–8056. Y. Cui, A. Abouimrane, C. J. Sun, Y. Ren and K. Amine, Chem. Commun., 2014, 50, 5576–5579. H. Wang, S. Li, Z. Chen, H. K. Liu and Z. Guo, RSC Adv., 2014, 4, 61673–61678. L. Zeng, W. Zeng, Y. Jiang, X. Wei, W. Li, C. Yang, Y. Zhu and Y. Yu, Adv. Energy Mater., 2015, 5, 1401377–1401386. L. Zeng, X. Wei, J. Wang, Y. Jiang, W. Li and Y. Yu, J. Power Sources, 2015, 281, 461–469. X. Wang, Z. Zhang, Y. Qu, G. Wang, Y. Lai and J. Li, J. Power Sources, 2015, 287, 247–252. Z. Zhang, X. Yang, Z. Guo, Y. Qu, J. Li and Y. Lai, J. Power Sources, 2015, 279, 88–93. J. Zhang, Y. Xu, L. Fan, Y. Zhu, J. Liang and Y. Qian, Nano Energy, 2015, 13, 592–600. H. Ye, Y. X. Yin, S. F. Zhang and Y. G. Guo, J. Mater. Chem. A, 2014, 2, 13293–13298. Y. Lai, F. Yang, Z. Zhang, S. Jiang and J. Li, RSC Adv., 2014, 4, 39312–39315. L. Liu, Y. Wei, C. Zhang, C. Zhang, X. Li, J. Wang, L. Ling, W. Qiao and D. Long, Electrochim. Acta, 2015, 153, 140–148. Z. Li, L. Yuan, Z. Yi, Y. Liu and Y. Huang, Nano Energy, 2014, 9, 229–236. L. Liu, Y. Hou, X. Wu, S. Xiao, Z. Chang, Y. Yang and Y. Wu, Chem. Commun., 2013, 49, 11515–11517. W. Li, J. R. Dahn and D. S. Wainwright, Science, 1994, 264, 1115–1118. (a) F. Wang, S. Xiao, Z. Chang, Y. Yang and Y. Wu, Chem. Commun., 2013, 49(9), 209–9211; (b) X. Wang, M. Li, Y. Wang, B. Chen, Y. Zhu and Y. Wu, J. Mater. Chem. A, 2015, 3, 8280–8283; (c) F. X. Wang, S. Y. Xiao, Y. S. Zhu, Z. Chang, C. L. Hu, Y. P. Wu and R. Holze, J. Power Sources, 2014, 246, 19–23. (a) H. Kim, J. Hong, K. Y. Park, H. Kim, S. W. Kim and K. Kang, Chem. Rev., 2014, 114, 11788–11827; (b) F. X. Wang, S. Y. Xiao, X. W. Gao, Y. S. Zhu, H. P. Zhang, Y. P. Wu and R. Holze, J. Power Sources, 2013, 242, 560–565; (c) X. Wang, M. Li, Z. Chang, Y. Yang, Y. Wu and X. Liu, ACS Appl. Mater. Interfaces, 2015, 7, 2280–2285; (d) F. X. Wang, S. Y. Xiao, Y. Shi, L. L. Liu, Y. S. Zhu, Y. P. Wu, J. Z. Wang and R. Holze, Electrochim. Acta, 2013, 93, 301–306.

Energy Environ. Sci., 2016, 9, 3570--3611 | 3605

View Article Online

Published on 26 September 2016. Downloaded by SLUB DRESDEN on 29/11/2016 20:29:33.

Energy & Environmental Science

87 (a) J. Zheng, Y. Hou, Y. Duan, X. Song, Y. Wei, T. Liu, J. Hu, H. Guo, Z. Zhuo, L. Liu, Z. Chang, X. Wang, D. Zherebetskyy, Y. Fang, Y. Lin, K. Xu, L. W. Wang, Y. Wu and F. Pan, Nano Lett., 2015, 15, 6102–6109; (b) F. Wang, S. Xiao, Y. Hou, C. Hu, L. Liu and Y. Wu, RSC Adv., 2013, 3, 13059–13084. 88 W. Tang, Y. Zhu, Y. Hou, L. Liu, Y. Wu, K. P. Loh, H. Zhang and K. Zhu, Energy Environ. Sci., 2013, 6, 2093–2104. 89 Z. Chang, Y. Yang, M. Li, X. Wang and Y. Wu, J. Mater. Chem. A, 2014, 2, 10739–10755. 90 J. Liu, J. Hu, Q. Deng, J. Mo, H. Xie, Z. Liu, Y. Xiong, X. Wu and Y. Wu, Isr. J. Chem., 2015, 55, 1–16. 91 W. Tang, L. L. Liu, S. Tian, L. Li, Y. B. Yue, Y. P. Wu, S. Y. Guan and K. Zhu, Electrochem. Commun., 2010, 12, 1524–1526. 92 Q. Qu, L. Fu, X. Zhan, D. Samuelis, J. Maier, L. Li, S. Tian, Z. Li and Y. Wu, Energy Environ. Sci., 2011, 4, 3985–3990. 93 (a) W. Tang, X. Gao, Y. Zhu, Y. Yue, Y. Shi, Y. Wu and K. Zhu, J. Mater. Chem., 2012, 22, 20143–20145; (b) W. Tang, L. L. Liu, S. Tian, L. Li, L. L. Li, Y. B. Yue, Y. Bai, Y. P. Wu, K. Zhu and R. Holze, Electrochem. Commun., 2011, 13, 1159–1162. 94 (a) W. Tang, L. Liu, Y. Zhu, H. Sun, Y. Wu and K. Zhu, Energy Environ. Sci., 2012, 5, 6909–6913; (b) W. Tang, Y. Hou, F. Wang, L. Liu, Y. Wu and K. Zhu, Nano Lett., 2013, 13, 2036–2040. 95 (a) F. Wang, Y. Liu, X. Wang, Z. Chang, Y. Wu and R. Holze, ChemElectroChem, 2015, 2, 1024–1030; (b) Y. Liu, Z. Wen, X. Wu, X. Wang, Y. Wu and R. Holze, Chem. Commun., 2014, 50, 13714–13717. 96 X. Wang, Y. Hou, Y. Zhu, Y. Wu and R. Holze, Sci. Rep., 2013, 3, 1401–1405. 97 X. Wang, Q. Qu, Y. Hou, F. Wang and Y. Wu, Chem. Commun., 2013, 49, 6179–6181. 98 Y. Hou, X. Wang, Y. Zhu, C. Hu, Z. Chang, Y. Wu and R. Holze, J. Mater. Chem. A, 2013, 1, 14713–14718. 99 Z. Chang, Y. Yang, X. Wang, M. Li, Z. Fu, Y. Wu and R. Holze, Sci. Rep., 2015, 5, 11931–11938. 100 (a) H. Li, Y. Wang, H. Na, H. Liu and H. Zhou, J. Am. Chem. Soc., 2009, 131, 15098–15099; (b) Y. Zhao, Y. Ding, J. Song, L. Peng, J. B. Goodenough and G. Yu, Energy Environ. Sci., 2014, 7, 1990–1995; (c) Z. Chang, X. Wang, Y. Yang, J. Gao, M. Li, L. Liu and Y. Wu, J. Mater. Chem. A, 2014, 2, 19444–19450; (d) Y. Zhao, L. Wang and H. R. Byon, Nat. Commun., 2013, 4, 1896–1902. 101 K. M. Abraham and Z. Jiang, J. Electrochem. Soc., 1996, 143, 1–5. 102 T. Ogasawara, A. Debart, M. Holzapfel, P. Novak and P. G. Bruce, J. Am. Chem. Soc., 2006, 128, 1390–1393. 103 F. Mizuno, S. Nakanishi, Y. Kotani, S. Yokoishi and H. Iba, Electrochemistry, 2010, 78, 403–405. 104 V. S. Bryantsev and M. Blanco, J. Phys. Chem. Lett., 2011, 2, 379–383. 105 S. A. Freunberger, Y. Chen, Z. Peng, J. M. Griffin, L. J. Hardwick, F. Barde, P. Novak and P. G. Bruce, J. Am. Chem. Soc., 2011, 133, 8040–8047. 106 J. Xiao, J. Hu, D. Wang, D. Hu, W. Xu, G. L. Graff, Z. Nie, J. Liu and J. G. Zhang, J. Power Sources, 2011, 196, 5674–5678.

3606 | Energy Environ. Sci., 2016, 9, 3570--3611

Review

107 F. Li, T. Zhang and H. Zhou, Energy Environ. Sci., 2013, 6, 1125–1141. 108 (a) Y. C. Lu, B. M. Gallant, D. G. Kwabi, J. R. Harding, R. R. Mitchell, M. S. Whittingham and S. H. Yang, Energy Environ. Sci., 2013, 6, 750–768; (b) Y. Shao, F. Ding, J. Xiao, J. Zhang, W. Xu, S. Park, J. G. Zhang, Y. Wang and J. Liu, Adv. Funct. Mater., 2013, 23, 987–1004. 109 Y. C. Lu, Z. Xu, H. A. Gasteiger, S. Chen, K. H. Schifferli and S. H. Yang, J. Am. Chem. Soc., 2010, 132, 12170–12171. 110 C. C. Li, W. Zhang, H. Ang, H. Yu, B. Y. Xia, X. Wang, Y. H. Yang, Y. Zhao, H. H. Hng and Q. Yan, J. Mater. Chem. A, 2014, 2, 10676–10681. 111 Y. Liu, L. J. Cao, C. W. Cao, M. Wang, K. L. Leung, S. S. Zeng, T. F. Hung, C. Y. Chung and Z. G. Lu, Chem. Commun., 2014, 50, 14635–14638. 112 A. Debart, J. Bao, G. Armstrong and P. G. Bruce, J. Power Sources, 2007, 174, 1177–1182. 113 A. Debart, A. J. Paterson, J. Bao and P. G. Bruce, Angew. Chem., Int. Ed., 2008, 47, 4521–4524. 114 K. Gong, F. Du, Z. Xia, M. Durstock and L. Dai, Science, 2009, 323, 760–764. 115 Y. Tan, C. Xu, G. Chen, X. Fang, N. Zheng and Q. Xie, Adv. Funct. Mater., 2012, 22, 4584–4591. 116 F. Li, R. Ohnishi, Y. Yamada, J. Kubota, K. Domen, A. Yamada and H. Zhou, Chem. Commun., 2013, 49, 1175–1177. 117 Y. Li, Z. Huang, K. Huang, D. Carnahan and Y. Xing, Energy Environ. Sci., 2013, 6, 3339–3345. 118 H. D. Lim, H. Song, H. Gwon, K. Y. Park, J. Kim, Y. Bae, H. Kim, S. K. Jung, T. Kim, Y. H. Kim, X. Lepro, R. O. Robles, R. H. Baughman and K. Kang, Energy Environ. Sci., 2013, 6, 3570–3575. 119 H. D. Lim, H. Song, J. Kim, H. Gwon, Y. Bae, K. Y. Park, J. Hong, H. Kim, T. Kim, Y. H. Kim, X. Lepr, R. O. Robles, R. H. Baughman and K. Kang, Angew. Chem., Int. Ed., 2014, 53, 3926–3931. 120 F. Li, Y. Chen, D. M. Tang, Z. Jian, C. Liu, D. Golberg, A. Yamada and H. Zhou, Energy Environ. Sci., 2014, 7, 1648–1652. 121 J. Wu, H. W. Park, A. Yu, D. Higgins and Z. Chen, J. Phys. Chem. C, 2012, 116, 9427–9432. 122 Z. Jian, P. Liu, F. Li, P. He, X. Guo, M. Chen and H. Zhou, Angew. Chem., Int. Ed., 2014, 53, 442–446. 123 Y. Qin, J. Lu, P. Du, Z. Chen, Y. Ren, T. Wu, J. T. Miller, J. Wen, D. J. Miller, Z. Zhang and K. Amine, Energy Environ. Sci., 2013, 6, 519–531. 124 L. Trahey, N. K. Karan, M. K. Y. Chan, J. Lu, Y. Ren, J. Greeley, M. Balasubramanian, A. K. Burrell, L. A. Curtiss and M. M. Thackeray, Adv. Energy Mater., 2013, 3, 75–84. 125 (a) F. Wang, Z. Chang, X. Wang, Y. Wang, B. Chen, Y. Zhu and Y. Wu, J. Mater. Chem. A, 2015, 3, 4840–4845; (b) F. Wang, S. Xiao, M. Li, X. Wang, Y. Zhu, Y. Wu, A. Shirakawa and J. Peng, J. Power Sources, 2015, 287, 416–421; (c) F. X. Wang, S. Y. Xiao, Z. Chang, M. X. Li, Y. P. Wu and R. Holze, Int. J. Electrochem. Sci., 2014, 9, 6182–6190. 126 J. K. Ngala, S. Alia, A. Dobley, V. M. B. Crisostomo and S. L. Suib, Chem. Mater., 2007, 19, 229–234.

This journal is © The Royal Society of Chemistry 2016

View Article Online

Published on 26 September 2016. Downloaded by SLUB DRESDEN on 29/11/2016 20:29:33.

Review

127 L. Zhang, X. Zhang, Z. Wang, J. Xu, D. Xu and L. Wan, Chem. Commun., 2012, 48, 7598–7600. 128 J. J. Xu, D. Xu, Z. L. Wang, H. G. Wang, L. L. Zhang and X. B. Zhang, Angew. Chem., Int. Ed., 2013, 52, 3887–3890. 129 Y. Cui, Z. Wen, X. Liang, Y. Lu, J. Jin, M. Wu and X. Wu, Energy Environ. Sci., 2012, 5, 7893–7897. 130 B. Sun, B. Wang, D. Su, L. Xiao, H. Ahn and G. Wang, Carbon, 2012, 50, 727–733. 131 Y. Cao, Z. Wei, J. He, J. Zang, Q. Zhang, M. Zheng and Q. Dong, Energy Environ. Sci., 2012, 5, 9765–9768. 132 X. Zeng, C. You, L. Leng, D. Dang, X. Qiao, X. Li, Y. Li, S. Liao and R. Adzic, J. Mater. Chem. A, 2015, 3, 11224–11231. 133 W. Zhang, J. Zhu, H. Ang, Y. Zeng, N. Xiao, Y. Gao, W. Liu, H. H. Hng and Q. Yan, Nanoscale, 2013, 5, 9651–9658. 134 (a) D. Xu, Z. L. Wang, J. J. Xu, L. L. Zhang and X. B. Zhang, Chem. Commun., 2012, 48, 6948–6950; (b) Z. Zhang, L. Su, M. Yang, M. Hu, J. Bao, J. Wei and Z. Zhou, Chem. Commun., 2014, 50, 776–778; (c) P. Zhang, X. Lu, Y. Huang, J. Deng, L. Zhang, F. Ding, Z. Su, G. Wei and O. G. Schmidt, J. Mater. Chem. A, 2015, 3, 14562–14566. 135 L. Zhang, S. Zhang, K. Zhang, G. Xu, X. He, S. Dong, Z. Liu, C. Huang, L. Gu and G. Cui, Chem. Commun., 2013, 49, 3540–3542. 136 B. Sun, X. Huang, S. Chen, Y. Zhao, J. Zhang, P. Munroe and G. Wang, J. Mater. Chem. A, 2014, 2, 12053–12059. 137 T. F. Hung, S. G. Mohamed, C. C. Shen, Y. Q. Tsai, W. S. Chang and R. S. Liu, Nanoscale, 2013, 5, 12115–12119. 138 S. Han, D. Wu, S. Li, F. Zhang and X. Feng, Adv. Mater., 2014, 26, 849–864. 139 (a) J. Zhang, P. Li, Z. Wang, J. Qiao, D. Rooney, W. Sun and K. Sun, J. Mater. Chem. A, 2015, 3, 1504–1510; (b) J. Xiao, D. Mei, X. Li, W. Xu, D. Wang, G. L. Graff, W. D. Bennett, Z. Nie, L. V. Saraf, I. A. Aksay, J. Liu and J. G. Zhang, Nano Lett., 2011, 11, 5071–5078. 140 W. B. Luo, S. L. Chou, J. Z. Wang, Y. C. Zhai and H. K. Liu, Small, 2015, 11, 2817–2824. 141 Z. Guo, D. Zhou, X. Dong, Z. Qiu, Y. Wang and Y. Xia, Adv. Mater., 2013, 25, 5668–5672. 142 L. Jin, L. Xu, C. Morein, C. Chen, M. Lai, S. Dharmarathna, A. Dobley and S. L. Suib, Adv. Funct. Mater., 2010, 22, 3373–3382. 143 X. Hu, X. Han, Y. Hu, F. Cheng and J. Chen, Nanoscale, 2014, 6, 3522–3525. 144 S. Chen, G. Liu, H. Yadegari, H. Wang and S. Z. Qiao, J. Mater. Chem. A, 2015, 3, 2559–2563. 145 Y. Hu, T. Zhang, F. Cheng, Q. Zhao, X. Han and J. Chen, Angew. Chem., Int. Ed., 2015, 54, 4338–4343. 146 L. Liu, J. Wang, Y. Hou, J. Chen, H. K. Liu, J. Wang and Y. Wu, Small, 2016, 12, 602–611. 147 J. J. Xu, Z. L. Wang, D. Xu, F. Z. Meng and X. B. Zhang, Energy Environ. Sci., 2014, 7, 2213–2219. 148 K. M. Abraham, Solid State Ionics, 1982, 7, 199–212. 149 S. Miyazaki, S. Kikkawa and M. Koizumi, Synth. Met., 1983, 6, 211–271. 150 N. Yabuuchi, M. Kajiyama, J. Iwatate, H. Nishikawa, S. Hitomi, R. Okuyama, R. Usui, Y. Yamada and S. Komaba, Nat. Mater., 2012, 11, 512–517.

This journal is © The Royal Society of Chemistry 2016

Energy & Environmental Science

151 M. Guignard, C. Didier, J. Darriet, P. Bordet, E. Elkaı¨m and C. Delmas, Nat. Mater., 2013, 12, 74–80. 152 A. Hayashi, K. Noi, A. Sakuda and M. Tatsumisago, Nat. Commun., 2012, 3, 856–860. 153 J. Kim, D. H. Seo, H. Kim, I. Park, J. K. Yoo, S. K. Jung, Y. U. Park, W. A. Goddard and K. Kang, Energy Environ. Sci., 2015, 8, 540–545. 154 N. Wongittharom, T. C. Lee, C. H. Wang, Y. C. Wang and J. K. Chang, J. Mater. Chem. A, 2014, 2, 5655–5661. 155 J. Xu, S. L. Chou, J. L. Wang, H. K. Liu and S. X. Dou, ChemElectroChem, 2014, 1, 371–374. 156 M. Nose, S. Shiotani, H. Nakayama, K. Nobuhara, S. Nakanishi and H. Iba, Electrochem. Commun., 2013, 34, 266–269. 157 P. Serras, V. Palomares, P. Kubiak, L. Lezama and T. Rojo, Electrochem. Commun., 2013, 34, 344–347. 158 M. Moradi, Z. Li, J. Qi, W. Xing, K. Xiang, Y. M. Chiang and A. M. Belcher, Nano Lett., 2015, 15, 2917–2921. 159 F. Sauvage, E. Quarez, J. Tarascon and E. Baudrin, Solid State Sci., 2006, 8, 1215–1221. 160 C. Li, C. Yin, L. Gu, R. E. Dinnebier, X. Mu, P. A. Aken and J. Maier, J. Am. Chem. Soc., 2013, 135, 11425–11428. 161 H. He, G. Jin, H. Wang, X. Huang, Z. Chen, D. Sun and Y. Tang, J. Mater. Chem. A, 2014, 2, 3563–3570. 162 H. Kim, R. H. Kim, S. S. Lee, Y. Kim, D. Y. Kim and K. Park, ACS Appl. Mater. Interfaces, 2014, 6, 11692–11697. 163 H. Kim, D. Y. Kim, Y. Kim, S. S. Lee and K. Park, ACS Appl. Mater. Interfaces, 2015, 7, 1477–1485. 164 D. Su and G. Wang, ACS Nano, 2013, 7, 11218–11226. 165 (a) E. Hosono, T. Saito, J. Hoshino, M. Okubo, Y. Saito, D. N. Hamane, T. Kudo and H. Zhou, J. Power Sources, 2012, 217, 43–46; (b) Y. Cao, L. Xiao, W. Wang, D. Choi, Z. Nie, J. Yu, L. V. Saraf, Z. Yang and J. Liu, Adv. Mater., 2011, 23, 3155–3160; (c) H. Kim, D. J. Kim, D. H. Seo, M. S. Yeom, K. Kang, D. K. Kim and Y. Jung, Chem. Mater., 2012, 24, 1205–1211. 166 Y. H. Jung, S. T. Hong and D. K. Kim, J. Electrochem. Soc., 2013, 160, A897–A900. 167 (a) X. Jiang, S. Liu, H. Xu, L. Chen, J. Yang and Y. Qian, Chem. Commun., 2015, 51, 8480–8483; (b) F. Wang, X. Wang, Z. Chang, Y. Zhu, L. Fu, X. Liu and Y. Wu, Nanoscale Horiz., 2016, 1, 272–289. 168 D. Su, C. Wang, H. Ahn and G. Wang, Chem. – Eur. J., 2013, 19, 10884–10889. 169 H. Kang, Y. Liu, M. Shang, T. Lu, Y. Wang and L. Jiao, Nanoscale, 2015, 7, 9261–9267. 170 W. Zhang, Y. Liu, C. Chen, Z. Li, Y. Huang and X. Hu, Small, 2015, 11, 3822–3829. 171 Y. Shen, X. Wang, H. Hu, M. Jiang, Y. Bai, X. Yang and H. Shu, RSC Adv., 2015, 5, 38277–38282. 172 Y. Zhang, H. Yu and H. Zhou, J. Mater. Chem. A, 2014, 2, 11574–11577. 173 (a) W. Shen, C. Wang, H. Liu and W. Yang, Chem. – Eur. J., 2013, 19, 14712–14718; (b) C. Wang, F. Wang, Y. Zhao, Y. Li, Q. Yue, Y. Liu, Y. Liu, A. A. Elzatahry, A. A. Enizi, Y. Wu, Y. Deng and D. Zhao, Nano Res., 2016, 9, 165–173.

Energy Environ. Sci., 2016, 9, 3570--3611 | 3607

View Article Online

Published on 26 September 2016. Downloaded by SLUB DRESDEN on 29/11/2016 20:29:33.

Energy & Environmental Science

174 J. Mao, C. Luo, T. Gao, X. Fan and C. Wang, J. Mater. Chem. A, 2015, 3, 10378–10385. 175 Y. Jiang, Z. Yang, W. Li, L. Zeng, F. Pan, M. Wang, X. Wei, G. Hu, L. Gu and Y. Yu, Adv. Energy Mater., 2015, 5, 1402104–1402111. 176 W. Huang, J. Zhou, B. Li, L. An, P. Cui, W. Xia, L. Song, D. Xia, W. Chu and Z. Wu, Small, 2015, 11, 2170–2176. 177 W. Huang, B. Li, M. F. Saleem, X. Wu, J. Li, J. Lin, D. Xia, W. Chu and Z. Wu, Chem. – Eur. J., 2015, 21, 851–860. 178 (a) H. Fei, Y. Lin and M. Wei, J. Colloid Interface Sci., 2014, 425, 1–4; (b) W. Murphy, P. A. Christian, F. J. DiSalvo and J. N. Carides, J. Electrochem. Soc., 1979, 126, 497–499; (c) N. A. Chernova, M. Roppolo, A. C. Dillon and M. S. Whittingham, J. Mater. Chem., 2009, 19, 2526–2552. 179 (a) Y. Lu, L. Wang, J. Cheng and J. B. Goodenough, Chem. Commun., 2012, 48, 6544–6546; (b) H. Lee, Y. I. Kim, J. K. Park and J. W. Choi, Chem. Commun., 2012, 48, 8416–8418; (c) T. Matsuda, M. Takachi and Y. Moritomo, Chem. Commun., 2013, 49, 2750–2752; (d) L. Wang, Y. Lu, J. Liu, M. Xu, J. Cheng, D. Zhan and J. B. Goodenough, Angew. Chem., Int. Ed., 2013, 52, 1964–1967; (e) Y. You, X. L. Wu, Y. X. Yin and Y. G. Guo, Energy Environ. Sci., 2014, 7, 1643–1647. 180 Z. Li, D. Young, K. Xiang, W. C. Carter and Y. M. Chiang, Adv. Energy Mater., 2013, 3, 290–294. 181 B. H. Zhang, Y. Liu, X. W. Wu, Y. Q. Yang, Z. Chang, Z. B. Wen and Y. P. Wu, Chem. Commun., 2014, 50, 1209–1211. 182 J. F. Whitacre, T. Wiley, S. Shanbhag, Y. Wenzhuo, A. Mohamed, S. E. Chun, E. Weber, D. Blackwood, E. L. Bell, J. Gulakowski, C. Smith and D. Humphreys, J. Power Sources, 2012, 213, 255–264. 183 (a) M. Minakshi, Mater. Sci. Eng., B, 2012, 177, 1788–1792; (b) M. Minakshi and D. Meyrick, Electrochim. Acta, 2013, 101, 66–70. 184 J. Shao, X. Li, Q. Qu and Y. Wu, J. Power Sources, 2013, 223, 56–61. 185 (a) W. Song, X. Ji, Y. Zhu, H. Zhu, F. Li, J. Chen, F. Lu, Y. Yao and C. E. Banks, ChemElectroChem, 2014, 1, 871–876. 186 Z. Jian, L. Zhao, H. Pan, Y. S. Hu, H. Li, W. Chen and L. Chen, Electrochem. Commun., 2012, 14, 86–89. 187 (a) C. D. Wessells, S. V. Peddada, R. A. Huggins and Y. Cui, Nano Lett., 2011, 11, 5421–5425; (b) C. D. Wessells, S. V. Peddada, M. T. McDowell, R. A. Huggins and Y. Cui, J. Electrochem. Soc., 2011, 159, A98–A103. 188 X. Wu, Y. Cao, X. Ai, J. Qian and H. Yang, Electrochem. Commun., 2013, 31, 145–148. 189 F. Sauvage, E. Baudrin and J. M. Tarascon, Sens. Actuators, B, 2007, 120, 638–644. 190 (a) A. Eftekhari, J. Power Sources, 2004, 126, 221–228; (b) N. Imanishi, T. Morikawa, J. Kondo, Y. Takeda, O. Yamamoto, N. Kinugasa and T. Yamagishi, J. Power Sources, 1999, 79, 215–219. 191 T. D. Gregory, R. J. Hoffman and R. C. Winterton, J. Electrochem. Soc., 1990, 137, 775–780. 192 H. D. Yoo, I. Shterenberg, Y. Gofer, G. Gershinsky, N. Pour and D. Aurbach, Energy Environ. Sci., 2013, 6, 2265–2279.

3608 | Energy Environ. Sci., 2016, 9, 3570--3611

Review

193 P. Saha, M. K. Datta, O. I. Velikokhatnyi, A. Manivannan, D. Alman and P. N. Kumta, Prog. Mater. Sci., 2014, 66, 1–86. 194 M. M. Huie, D. C. Bock, E. S. Takeuchi, A. C. Marschilok and K. J. Takeuchi, Coord. Chem. Rev., 2015, 287, 15–27. 195 D. Aurbach, Z. Lu, A. Schechter, Y. Gofer, H. Gizbar, R. Turgeman, Y. Cohen, M. Moshkovich and E. Levi, Nature, 2000, 407, 724–727. 196 (a) E. Lancry, E. Levi, Y. Gofer, M. D. Levi and D. Aurbach, J. Solid State Electrochem., 2005, 9, 259–266; (b) P. Saha, P. H. Jampani, M. K. Datta, C. U. Okoli, A. Manivannan and P. N. Kumta, J. Electrochem. Soc., 2014, 161, A593–A598. 197 A. Mitelman, E. Levi, E. Lancry and D. Aurbach, ECS Trans., 2007, 3, 109–115. 198 (a) D. Aurbach, G. S. Suresh, E. Levi, A. Mitelman, O. Mizrahi, O. Chusid and M. Brunelli, Adv. Mater., 2007, 19, 4260–4267; (b) G. S. Suresh, M. D. Levi and D. Aurbach, Electrochim. Acta, 2008, 53, 3889–3896. 199 G. G. Amatucci, F. Badway, A. Singhal, B. Beaudoin, G. Skandan, T. Bowmer, I. Plitz, N. Pereira, T. Chapman and R. Jaworski, J. Electrochem. Soc., 2001, 148, A940–A950. 200 L. Jiao, H. Yuan, Y. Wang, J. Cao and Y. Wang, Electrochem. Commun., 2005, 7, 431–436. 201 D. Imamura, M. Miyayama, M. Hibino and T. Kudo, J. Electrochem. Soc., 2003, 150, A753–A758. 202 D. B. Le, S. Passerini, F. Coustier, J. Guo, T. Soderstrom, B. B. Owens and W. H. Smyrl, Chem. Mater., 1998, 10, 682–684. 203 R. Zhang, X. Yu, K. W. Nam, C. Ling, T. S. Arthur, W. Song, A. M. Knapp, S. N. Ehrlich, X. Q. Yang and M. Matsui, Electrochem. Commun., 2012, 23, 110–113. 204 S. Rasul, S. Suzuki, S. Yamaguchi and M. Miyayama, Electrochim. Acta, 2012, 82, 243–249. 205 H. Kurihara, T. Yajima and S. Suzuki, Chem. Lett., 2008, 37, 376–377. 206 C. Yuan, Y. Zhang, Y. Pan, X. Liu, G. Wang and D. Cao, Electrochim. Acta, 2014, 116, 404–412. 207 M. E. Spahr, P. Novka, O. Haas and R. Nesper, J. Power Sources, 1995, 54, 346–351. 208 G. Gershinsky, H. D. Yoo, Y. Gofer and D. Aurbach, Langmuir, 2013, 29, 10964–10972. 209 S. Su, Z. Huang, Y. NuLi, F. Tuerxun, J. Yang and J. Wang, Chem. Commun., 2015, 51, 2641–2644. 210 X. L. Li and Y. D. Li, J. Phys. Chem. B, 2004, 108, 13893–13900. 211 Y. Liang, R. Feng, S. Yang, H. Ma, J. Liang and J. Chen, Adv. Mater., 2011, 23, 640–643. 212 Y. Liu, L. Jiao, Q. Wu, Y. Zhao, K. Cao, H. Liu, Y. Wang and H. Yuan, Nanoscale, 2013, 5, 9562–9567. 213 Y. Liu, L. Jiao, Q. Wu, J. Du, Y. Zhao, Y. Si, Y. Wang and H. Yuan, J. Mater. Chem. A, 2013, 1, 5822–5826. 214 S. Yang, D. Li, T. Zhang, Z. Tao and J. Chen, J. Phys. Chem. C, 2012, 116, 1307–1312. 215 Z. L. Tao, L. N. Xu, X. L. Gou, J. Chen and H. T. Yuan, Chem. Commun., 2004, 2080–2081. 216 N. Amir, Y. Vestfrid, O. Chusid, Y. Gofer and D. Aurbach, J. Power Sources, 2007, 174, 1234–1240. 217 W. Yuan and J. R. Guenter, Solid State Ionics, 1995, 76, 253–258.

This journal is © The Royal Society of Chemistry 2016

View Article Online

Published on 26 September 2016. Downloaded by SLUB DRESDEN on 29/11/2016 20:29:33.

Review

218 Z. Feng, J. Yang, Y. NuLi and J. Wang, J. Power Sources, 2008, 184, 604–609. 219 Z. Feng, J. Yang, Y. NuLi, J. Wang, X. Wang and Z. Wang, Electrochem. Commun., 2008, 10, 1291–1294. 220 Y. NuLi, Y. Zheng, F. Wang, J. Yang, A. I. Minett, J. Wang and J. Chen, Electrochem. Commun., 2011, 13, 1143–1146. 221 Y. NuLi, J. Yang, Y. Li and J. Wang, Chem. Commun., 2010, 46, 3794–3796. 222 Y. NuLi, Y. Zheng, Y. Wang, J. Yang and J. Wang, J. Mater. Chem., 2011, 21, 12437–12443. 223 Y. Li, Y. N. NuLi, J. Yang, T. Yilinuer and J. L. Wang, Chin. Sci. Bull., 2011, 56, 386–390. 224 Y. Orikasa, T. Masese, T. Mori, M. Hattori, K. Yamamoto, T. Okado, Z. D. Huang, Y. Uchimoto, Y. Koyama, T. Minato, C. Tassel, J. Kim, Y. Kobayashi, T. Abe and H. Kageyama, Sci. Rep., 2014, 4, 5622–5627. 225 Y. Mizuno, M. Okubo, E. Hosono, T. Kudo, K. Ohishi, A. Okazawa, N. Kojima, R. Kurono, S. Nishimura and A. Yamada, J. Mater. Chem. A, 2013, 1, 13055–13059. 226 R. Y. Wang, C. D. Wessells, R. A. Huggins and Y. Cui, Nano Lett., 2013, 13, 5748–5752. 227 Z. D. Huang, T. Masese, Y. Orikasa, T. Mori and K. Yamamoto, RSC Adv., 2015, 5, 8598–8603. 228 Z. D. Huang, T. Masese, Y. Orikasa, T. Mori, T. Minato, C. Tassel, Y. Kobayashi, H. Kageyam and Y. Uchimoto, J. Mater. Chem. A, 2014, 2, 11578–11582. 229 J. Wu, G. Gao, G. Wu, B. Liu, H. Yang, X. Zhou and J. Wang, RSC Adv., 2014, 4, 15014–15017. 230 Y. NuLi, Z. Guo, H. Liu and J. Yang, Electrochem. Commun., 2007, 9, 1913–1917. ´n, R. J. K. Wood, R. L. Jones, 231 D. R. Egan, C. P. Leo K. R. Stokes and F. C. Walsh, J. Power Sources, 2013, 236, 293–310. 232 Q. Li and N. J. Bjerrum, J. Power Sources, 2002, 110, 1–10. 233 M. Hulot, Compt. Rend., 1855, 40, 148–152. 234 G. W. Heise, E. A. Schumacher and N. C. Cahoon, J. Electrochem. Soc., 1948, 94, 99–105. 235 (a) C. Marsh and S. Licht, J. Electrochem. Soc., 1994, 141, L61–L63; (b) S. Licht and D. Peramunage, J. Electrochem. Soc., 1993, 140, L4–L6; (c) S. Licht, Electrochem. Commun., 1999, 1, 33–36; (d) S. Licht and N. Myung, J. Electrochem. Soc., 1995, 142, L179–L182. 236 (a) S. Zaromb, J. Electrochem. Soc., 1962, 109, 1125–1130; (b) C. Li, W. Ji, J. Chen and Z. Tao, Chem. Mater., 2007, 19, 5812–5814; (c) R. Mori, RSC Adv., 2013, 3, 11547–11551; (d) Z. Zhang, C. Zuo, Z. Liu, Y. Yu, Y. Zuo and Y. Song, J. Power Sources, 2014, 251, 470–475. 237 (a) F. M. Donahue, S. E. Mancini and L. Simonsen, J. Appl. Electrochem., 1992, 22, 230–234; (b) P. R. Gifford and J. B. Palmisano, J. Electrochem. Soc., 1988, 135, 650–654. 238 N. Jayaprakash, S. K. Das and L. A. Archer, Chem. Commun., 2011, 47, 12610–12612. 239 W. Wang, B. Jiang, W. Xiong, H. Sun, Z. Lin, L. Hu, J. Tu, J. Hou, H. Zhu and S. Jiao, Sci. Rep., 2013, 3, 3383–3388. 240 J. V. Rani, V. Kanakaiah, T. Dadmal, M. S. Rao and S. Bhavanarushi, J. Electrochem. Soc., 2013, 160, A1781–A1784.

This journal is © The Royal Society of Chemistry 2016

Energy & Environmental Science

241 M. C. Lin, M. Gong, B. Lu, Y. Wu, D. Y. Wang, M. Guan, M. Angell, C. Chen, J. Yang, B. J. Hwang and H. Dai, Nature, 2015, 520, 324–328. 242 Z. Li, K. Xiang, W. Xing, W. C. Carter and Y. M. Chiang, Adv. Energy Mater., 2015, 5, 1401410–1401415. 243 S. Liu, G. L. Pan, G. R. Li and X. P. Gao, J. Mater. Chem. A, 2015, 3, 959–962. 244 F. Wang, X. Wang, Z. Chang, Y. Zhu and Y. Wu, ACS Appl. Mater. Interfaces, 2016, 8, 9022–9029. 245 (a) J. A. Read, A. V. Cresce, M. H. Ervin and K. Xu, Energy Environ. Sci., 2014, 7, 617–620; (b) S. Rothermel, P. Meister, G. Schmuelling, O. Fromm, H. W. Meyer, S. Nowak, M. Winter and T. Placke, Energy Environ. Sci., 2014, 7, ¨bke, J. Ufheil, 3412–3423; (c) F. Bordet, K. Ahlbrecht, J. Tu T. Hoes, M. Oetken and M. Holzapfel, Electrochim. Acta, 2015, 174, 1317–1323; (d) H. Fan, J. Gao, L. Qi and H. Wang, Electrochim. Acta, 2016, 189, 9–15. 246 (a) C. Xu, B. Li, H. Du and F. Kang, Angew. Chem., Int. Ed., 2012, 51, 933–935; (b) L. Zhang, L. Chen, X. Zhou and Z. Liu, Adv. Energy Mater., 2014, 5, 1400930–1400934; (c) D. Kundu, B. D. Adams, V. Duffort, S. H. Vajargah and L. F. Nazar, Nat. Energy, 2016, 1, 16119–16126; (d) P. Senguttuvan, S. D. Han, S. Kim, A. L. Lipson, S. Tepavcevic, T. T. Fister, I. D. Bloom, A. K. Burrell and C. S. Johnson, Adv. Energy Mater., 2016, DOI: 10.1002/aenm.201600826. 247 M. Anji Reddy and M. Fichtner, J. Mater. Chem., 2011, 21, 17059–17062. 248 X. Zhao, S. Ren, M. Bruns and M. Fichtner, J. Power Sources, 2014, 245, 706–711. 249 X. Zhao, Z. Z. Karger, D. Wang and M. Fichtner, Angew. Chem., Int. Ed., 2013, 52, 13621–13624. 250 X. Zhao, Q. Li, Z. Z. Karger, P. Gao, K. Fink, X. Shen and M. Fichtner, ACS Appl. Mater. Interfaces, 2014, 6, 10997–11000. 251 J. L. Sudworth and A. R. Tilley, The Sodium Sulphur Battery, Chapman & Hall, London, 1985. 252 X. Lu, B. W. Kirby, W. Xu, G. Li, J. Y. Kim, J. P. Lemmon, V. L. Sprenkle and Z. Yang, Energy Environ. Sci., 2013, 6, 299–306. 253 (a) J. Wang, J. Yang, Y. Nuli and R. Holze, Electrochem. Commun., 2007, 9, 31–34; (b) T. H. Hwang, D. S. Jung, J. S. Kim, B. G. Kim and J. W. Choi, Nano Lett., 2013, 13, 4532–4538. 254 X. Lu, J. P. Lemmon, J. Y. Kim, V. L. Sprenkle and Z. Yang, J. Power Sources, 2013, 224, 312–316. 255 H. S. Kim, T. S. Arthur, G. D. Allred, J. Zajicek, J. G. Newman, A. E. Rodnyansky, A. G. Oliver, W. C. Boggess and J. Muldoon, Nat. Commun., 2011, 2, 427–432. 256 Q. Zhao, Y. Hu, K. Zhang and J. Chen, Inorg. Chem., 2014, 53, 9000–9005. 257 (a) G. Cohn, L. Ma and L. A. Archer, J. Power Sources, 2015, 283, 416–422; (b) S. Licht and D. Peramunage, J. Electrochem. Soc., 1993, 140, L4–L6. 258 E. Peled, D. Golodnitsky, H. Mazor, M. Goor and S. Avshalomov, J. Power Sources, 2011, 196, 6835–6840. 259 Q. Sun, Y. Yang and Z. W. Fu, Electrochem. Commun., 2012, 16, 22–25.

Energy Environ. Sci., 2016, 9, 3570--3611 | 3609

View Article Online

Published on 26 September 2016. Downloaded by SLUB DRESDEN on 29/11/2016 20:29:33.

Energy & Environmental Science

260 (a) W. Liu, Q. Sun, Y. Yang, J. Y. Xie and Z. W. Fu, Chem. Commun., 2013, 49, 1951–1953; (b) Y. Li, H. Yadegari, X. Li, M. N. Banis, R. Li and X. Sun, Chem. Commun., 2013, 49, 11731–11733. 261 Z. Jian, Y. Chen, F. Li, T. Zhang, C. Liu and H. Zhou, J. Power Sources, 2014, 251, 466–469. 262 H. Yadegari, Y. Li, M. N. Banis, X. Li, B. Wang, Q. Sun, R. Li, T. K. Sham, X. Cui and X. Sun, Energy Environ. Sci., 2014, 7, 3747–3757. ¨rr, 263 P. Hartmann, C. L. Bender, M. Vracar, A. K. Du A. Garsuch, J. Janek and P. Adelhelm, Nat. Mater., 2013, 12, 228–232. 264 P. Hartmann, C. L. Bender, J. Sann, A. K. Durr, M. Jansen, J. Janek and P. Adelhelm, Phys. Chem. Chem. Phys., 2013, 15, 11661–11672. 265 B. D. McCloskey, J. M. Garcia and A. C. Luntz, J. Phys. Chem. Lett., 2014, 5, 1230–1235. 266 J. Kim, H. D. Lim, H. Gwon and K. Kang, Phys. Chem. Chem. Phys., 2013, 15, 3623–3629. 267 S. Kang, Y. Mo, S. P. Ong and G. Ceder, Nano Lett., 2014, 14, 1016–1020. 268 X. Ren and Y. Wu, J. Am. Chem. Soc., 2013, 135, 2923–2926. 269 S. K. Das, S. Lau and L. A. Archer, J. Mater. Chem. A, 2014, 2, 12623–12629. 270 T. Zhang and H. Zhou, Nat. Commun., 2013, 4, 1817–1823. 271 H. K. Lim, H. D. Lim, K. Y. Park, D. H. Seo, H. Gwon, J. Hong, W. A. Goddard, I. H. Kim and K. Kang, J. Am. Chem. Soc., 2013, 135, 9733–9742. 272 K. Takechi, T. Shiga and T. Asaoka, Chem. Commun., 2011, 47, 3463–3465. 273 R. Wang, X. Yu, J. Bai, H. Li, X. Huang, L. Chen and X. Yang, J. Power Sources, 2012, 218, 113–118. 274 Y. Liu, R. Wang, Y. Lyu, H. Li and L. Chen, Energy Environ. Sci., 2014, 7, 677–681. 275 Z. Zhang, Q. Zhang, Y. Chen, J. Bao, X. Zhou, Z. Xie, J. Wei and Z. Zhou, Angew. Chem., 2015, 127, 6650–6653. 276 Y. Li and H. Dai, Chem. Soc. Rev., 2014, 43, 5257–5275. 277 P. Pei, K. Wang and Z. Ma, Appl. Energy, 2014, 128, 315–324. 278 M. Xu, D. G. Ivey, Z. Xie and W. Qu, J. Power Sources, 2015, 283, 358–371. 279 J. S. Lee, S. T. Kim, R. Cao, N. S. Choi, M. Liu, K. T. Lee and J. Cho, Adv. Energy Mater., 2011, 1, 34–50. 280 J. Park, M. Park, G. Nam, J. Lee and J. Cho, Adv. Mater., 2015, 27, 1396–1401. 281 J. Pan, L. Ji, Y. Sun, P. Wan, J. Cheng, Y. Yang and M. Fan, Electrochem. Commun., 2010, 11, 2191–2194. 282 X. Wu, J. Hu, J. Liu, Q. Zhou, W. Zhou, H. Li and Y. Wu, Pure Appl. Chem., 2014, 86, 633–649. 283 (a) L. Zhang, H. Zhang, Q. Lai, X. Li and Y. Cheng, J. Power Sources, 2013, 227, 41–47; (b) Q. Lai, H. Zhang, X. Li, L. Zhang and Y. Cheng, J. Power Sources, 2013, 235, 1–4. 284 H. Vafiadis and M. S. Kazacos, J. Membr. Sci., 2006, 279, 394–402. 285 (a) W. H. Wang and X. D. Wang, Electrochim. Acta, 2007, 52, 6755–6762; (b) A. Parasuraman, T. M. Lim, C. Menictas and M. S. Kazacos, Electrochim. Acta, 2013, 101, 27–40.

3610 | Energy Environ. Sci., 2016, 9, 3570--3611

Review

286 G. L. Soloveichik, Chem. Rev., 2015, 115, 11533–11558. 287 (a) Y. Wang, P. He and H. Zhou, Adv. Energy Mater., 2012, 2, 770–779; (b) Y. Zhao, Y. Ding, Y. Li, L. Peng, H. R. Byon, J. B. Goodenough and G. Yu, Chem. Soc. Rev., 2015, 44, 7968–7996. 288 (a) Y. H. Lu, J. B. Goodenough and Y. Kim, J. Am. Chem. Soc., 2011, 133, 5756–5759; (b) Y. H. Lu and J. B. Goodenough, J. Mater. Chem., 2011, 21, 10113–10117. 289 (a) C. Liu, J. S. Shamie, L. L. Shaw and V. L. Sprenkle, ACS Appl. Mater. Interfaces, 2016, 8, 1545–1552; (b) X. Yao, J. Luo, Q. Dong and D. Wang, Nano Energy, 2016, 28, 440–446. 290 G. S. Gautam, P. Canepa, A. Abdellahi, A. Urban, R. Malik and G. Ceder, Chem. Mater., 2015, 27, 3733–3742. 291 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. Fan, L. Qi, A. Kushima and J. Li, Science, 2010, 330, 1515–1520. 292 W. Tang, Y. Liu, C. Peng, M. Y. Hu, X. Deng, M. Lin, J. Z. Hu and K. P. Loh, J. Am. Chem. Soc., 2015, 137, 2600–2607. 293 (a) E. M. Erickson, M. S. Thorum, R. Vasic´, N. S. Marinkovic, A. I. Frnkel, A. A. Gewirth and R. G. Nuzzo, J. Am. Chem. Soc., 2012, 134, 197–200; (b) S. Misra, N. Liu, J. Nelson, S. S. Hong, Y. Cui and M. F. Toney, ACS Nano, 2012, 6, 5465–5473. 294 K. Kirshenbaum, D. C. Bock, C. Y. Lee, Z. Zhong, K. J. Takeuchi, A. C. Marschilok and E. S. Takeuchi, Science, 2015, 347, 149–154. 295 K. Sun, T. S. Wei, B. Y. Ahn, J. Y. Seo, S. J. Dillon and J. A. Lewis, Adv. Mater., 2013, 25, 4539–4543. 296 J. Hu, Y. Jiang, S. Cui, Y. Duan, T. Liu, H. Guo, L. Lin, Y. Lin, J. Zheng, K. Amine and F. Pan, Adv. Energy Mater., 2016, DOI: 10.1002/aenm.201600856. 297 L. Zhang, J. Deng, L. Liu, W. Si, S. Oswald, L. Xi, M. Kundu, G. Ma, T. Gemming, S. Baunack, F. Ding, C. Yan and O. G. Schmidt, Adv. Mater., 2014, 26, 4527–4532. 298 K. Fu, Y. Wang, C. Yan, Y. Yao, Y. Chen, J. Dai, S. Lacey, Y. Wang, J. Wan, T. Li, Z. Wang, Y. Xu and L. Hu, Adv. Mater., 2016, 28, 2587–2594. ¨llen, Nat. Commun., 299 Z. S. Wu, K. Parvez, X. Feng and K. Mu 2013, 4, 2487–2494. 300 Z. S. Wu, K. Parvez, S. Li, S. Yang, Z. Liu, S. Liu, X. Feng and ¨llen, Adv. Mater., 2015, 27, 4054–4061. K. Mu 301 Y. Wang, P. He and H. Zhou, Energy Environ. Sci., 2011, 4, 4994–4999. 302 M. Yu, W. D. McCulloch, D. R. Beauchamp, Z. Huang, X. Ren and Y. Wu, J. Am. Chem. Soc., 2015, 137, 8332–8335. 303 F. Wang, X. Wang, Z. Chang, X. Wu, X. Liu, L. Fu, Y. Zhu, Y. Wu and W. Huang, Adv. Mater., 2015, 27, 6962–6968. 304 (a) F. Wang, Z. Liu, X. Wang, X. Yuan, X. Wu, L. Fu, Y. Zhu and Y. Wu, J. Mater. Chem. A, 2016, 4, 5115–5123; (b) F. Wang, C. Wang, Y. Zhao, Z. Liu, Z. Chang, L. Fu, Y. Zhu, Y. Wu and D. Zhao, Small, 2016, DOI: 10.1002/ smll.201602331. 305 J. Jiang, Y. Li, J. Liu, X. Huang, C. Yuan and X. W. Lou, Adv. Mater., 2012, 24, 5166–5180.

This journal is © The Royal Society of Chemistry 2016

View Article Online

Published on 26 September 2016. Downloaded by SLUB DRESDEN on 29/11/2016 20:29:33.

Review

306 B. Scrosati, Nat. Nanotechnol., 2007, 2, 598–599. 307 Y. Yang, S. Jeong, L. B. Hu, H. Wu, S. W. Lee and Y. Cui, Proc. Natl. Acad. Sci. U. S. A., 2011, 108, 13013–13018. 308 J. Wang, L. Zhang, L. Yu, Z. Jiao, H. Xie, X. W. Lou and X. W. Sun, Nat. Commun., 2014, 5, 4921–4927. 309 X. Xue, S. Wang, W. Guo, Y. Zhang and Z. L. Wang, Nano Lett., 2012, 12, 5048–5054. 310 J. Luo, F. R. Fan, T. Jiang, Z. Wang, W. Tang, C. Zhang, M. Liu, G. Cao and Z. L. Wang, Nano Res., 2015, 8, 3934–3943. 311 X. Pu, L. Li, H. Song, C. Du, Z. Zhao, C. Jiang, G. Cao, W. Hu and Z. L. Wang, Adv. Mater., 2015, 27, 2472–2478. 312 (a) X. Chen, H. Lin, P. Chen, G. Guan, J. Deng and H. Peng, Adv. Mater., 2014, 26, 4444–4449; (b) C. Y. Wang, G. Zhang, S. Ge, T. Xu, Y. Ji, X. G. Yang and Y. Leng, Nature, 2016, 529, 515–518; (c) Z. Chen, P. C. Hsu, J. Lopez, Y. Li, J. W. F. To, N. Liu, C. Wang, S. C. Andrews, J. Liu, Y. Cui and Z. Bao, Nat. Energy, 2016, 1, 15009–15016. 313 (a) C. Wang, H. Wu, Z. Chen, M. T. McDowell, Y. Cui and Z. Bao, Nat. Chem., 2013, 5, 1042–1048; (b) H. Wang, B. Zhu, W. Jiang, Y. Yang, W. R. Leow, H. Wang and X. Chen, Adv. Mater., 2014, 26, 3638–3643. 314 (a) Y. S. Zhu, F. X. Wang, L. L. Liu, S. Y. Xiao, Z. Chang and Y. P. Wu, Energy Environ. Sci., 2013, 6, 618–624; (b) Y. S. Zhu, F. X. Wang, L. L. Liu, S. Y. Xiao, Y. Q. Yang and Y. P. Wu, Sci. Rep., 2013, 3, 3187–3192; (c) S. Xiao, F. Wang, Y. Yang, Z. Chang and Y. Wu, RSC Adv., 2014, 4, 76–81; (d) S. Y. Xiao, Y. Q. Yang, M. X. Li, F. X. Wang,

This journal is © The Royal Society of Chemistry 2016

Energy & Environmental Science

Z. Chang, Y. P. Wu and X. Liu, J. Power Sources, 2014, 270, 53–58; (e) Y. S. Zhu, S. Y. Xiao, M. X. Li, Z. Chang, F. X. Wang, J. Gao and Y. P. Wu, J. Power Sources, 2015, 288, 368–375; ( f ) Y. S. Zhu, S. Y. Xiao, Y. Shi, Y. Q. Yang, Y. Y. Hou and Y. P. Wu, Adv. Energy Mater., 2014, 4, 1300647–1300656. 315 (a) C. Zhang, S. Gamble, D. Ainsworth, A. M. Z. Slawin, Y. G. Andreev and P. G. Bruce, Nat. Mater., 2009, 8, 580–584; (b) J. Li, J. K. Park, R. B. Moore and L. A. Madsen, Nat. Mater., 2011, 10, 507–511; (c) R. Bouchet, S. Maria, R. Meziane, A. Aboulaich, L. Lienafa, J. P. Bonnet, T. N. T. Phan, D. Bertin, D. Gigmes, D. Devaux, R. Denoyel and M. Armand, Nat. Mater., 2013, 12, 452–457; (d) A. Hayashi, K. Noi, A. Sakuda and M. Tatsumisago, Nat. Commun., 2012, 3, 856–860. 316 (a) F. Ding, W. Xu, G. L. Graff, J. Zhang, M. L. Sushko, X. Chen, Y. Shao, M. H. Engelhard, Z. Nie, J. Xiao, X. Liu, P. V. Sushko, J. Liu and J. G. Zhang, J. Am. Chem. Soc., 2013, 135, 4450–4456; (b) W. Luo, L. Zhou, K. Fu, Z. Yang, J. Wan, M. Manno, Y. Yao, H. Zhu, B. Yang and L. Hu, Nano Lett., 2015, 15, 6149–6154; (c) W. Li, H. Yao, K. Yan, G. Zheng, Z. Liang, Y. M. Chiang and Y. Cui, Nat. Commun., 2015, 6, 7436–7443; (d) C. P. Yang, Y. X. Yin, S. F. Zhang, N. W. Li and Y. G. Guo, Nat. Commun., 2015, 6, 8058–8066; (e) X. Wang, F. Wang, L. Wang, M. Li, Y. Wang, B. Chen, Y. Zhu, L. Fu, L. Zha, L. Zhang, Y. Wu and W. Huang, Adv. Mater., 2016, 28, 4904–4911.

Energy Environ. Sci., 2016, 9, 3570--3611 | 3611