Journal of Materials Science: Materials in Electronics (2018) 29:14353–14370 https://doi.org/10.1007/s10854-018-9478-1

REVIEW

Design and modification of cathode materials for high energy density aluminum-ion batteries: a review Changgang Li1 · Xudong Zhang1 · Wen He1 Received: 21 March 2018 / Accepted: 11 June 2018 / Published online: 27 June 2018 © Springer Science+Business Media, LLC, part of Springer Nature 2018

Abstract Aluminum-ion batteries (AIBs) have the advantages of high specific volumetric capacity (8046 mAh cm−3), high safety and low cost. However, extended application of AIBs requires the development of innovative electrode materials with high energy density, which mainly depends on the cathode materials. In this review, the recent efforts to improve the electrochemical performances of AIBs were introduced. In particular, we provided a critical overview of the effective modification methods related to the cathode materials of AIBs, including carbonaceous materials (graphitic carbons, amorphous carbons, porous carbons), metal oxides ­(V2O5, ­VO2, ­TiO2, ­MoO2), metal sulphide (­ FeS2, ­Mo6S8, ­SnS2, NiS, CuS) and so on. The synthesis, structure, morphologies and electrochemical performances of these promising cathode materials have been discussed in detail. We anticipate that this review will shed light on the sustainable development of high-performance and low cost AIBs.

1 Introduction Nowadays, we can say that people deal with electricity every moment, small to the mobile phone battery and large to the electric car. Electricity is a necessary part of our daily life [1–4]. However, the reserves of traditional energy such as oil, coal are dwindling and the renewable energy generating electricity, such as wind, water, solar power, are not enough. Therefore, developing new types of energy storage with high energy density and long lifetime has become an important research topic. Electrochemical batteries can efficiently store and release electricity through chemical reactions, that is why lithium-ion battery (LIB) gets an unprecedented development and wide application in the past decades [5–15]. However, there are some crucial disadvantages limiting the application of LIB, such as stability and safety problems. But more seriously, the sharply increasing demand for LIB faces the challenge of shortages in lithium resources [16–18]. In order to slave those problems, people began to research the new batteries to replace LIB, so the new aluminum-ion * Xudong Zhang [email protected] Wen He [email protected] 1



Institute of Materials Science and Engineering, Qilu University of Technology (Shandong Academy of Sciences), Jinan 250353, China

battery (AIB), magnesium-ion battery, zinc-ion battery and calcium-ion battery appear in people’s view [19–27]. As we know, aluminum is the most abundant metal element in the earth’s crust (Table 1). And an aluminum-based redox coupe can transfer three electrons during the electrochemical charge and discharge reactions, which provides a chance to get higher storage capacity battery. Furthermore, the low reactivity of aluminum makes it much safer and easier to exploit than other metals. As a result, the AIBs have received wide attention [28, 29]. People have been studied AIBs since a long time ago. Hulot [30] was the first person who considered aluminum as a battery electrode in 1850. He assembled a cell with zinc (mercury) as anode, aluminum as cathode and dilute ­H2SO4 as electrolyte. In 1857, aluminum as anode was first used in the Buff cell [31]. An amalgamated aluminum–zinc alloy was proposed for use as anode in a cell with carbon cathode in 1893 [32]. The use of aluminum or amalgamated aluminum as anode in heavy-duty chlorine-depolarized batteries was reported in 1948 [33]. With amalgamated aluminum, the open circuit voltage reached up to 2.45 V. When it reaches to 1950s, aluminum was first used as anode in Leclanche-type dry cells in a system of aluminum/aqueous NaOH + ZnO/porous membrane/MnO2(C) [34]. Zaromb and Trevethan proved aluminum/oxygen system for the first time in 1960s [35, 36]. After the 1970s, the researchers focused on the nonaqueous electrolyte battery system, especially on the molten salt electrolyte system. At present, the researchers

13

Vol.:(0123456789)

14354 Table 1  Comparison of relative atomic mass, cost, volumetric capacity, gravimetric capacity, density at 25 °C and electrode potential vs SHE of Al, Li, Na, Mg, K, Zn and Ca

Journal of Materials Science: Materials in Electronics (2018) 29:14353–14370 Characteristics

Al

Li

Na

Mg

K

Zn

Ca

Relative atomic mass Cost (USD) ­kg−1 Volumetric capacity (mAh m−3) Gravimetric capacity (mAh g−1) Density (g cm−3) Electrode potential (v) versus SHE

26.98 1.9 8056 2981 2.7 − 1.76

6.94 19.2 2042 3861 0.53 − 3.04

22.99 3.1 1050 1166 0.9 − 2.71

24.31 2.2 3868 2205 1.74 − 2.37

39.10 13.1 609 685 0.89 − 2.93

65.39 2.2 5857 820 7.13 − 0.76

40.08 2.4 2061 1340 1.54 − 2.86

concentrate in searching ionic liquid electrolyte and suitable cathode materials for AIBs [37, 38]. Despite these advantageous characters, the development of AIBs technology has not kept pace with that of LIB technology. It is mainly due to the following two factors. Firstly, the easy formation of passive oxide film and intrinsic hydrogenation over the aluminum metal lead to serious reduction of the cell voltage and efficiency in AIBs [39]. Secondly, electrochemical Al intercalation into a host crystal structure can be very difficult due to the strong Coulombic effect and high charge density induced by the three positive charges [40]. The drawbacks of previous rechargeable AIBs are the low-specific capacity (< 70 mAh g−1) of cathodes and its low charging rate, which need to match the high capacity of the Al anode to increase the AIBs’ energy density [41]. So it can be very hard to find a satisfactory cathode material to meet the intercalation reaction which have high reversible capacity and adequate operating voltage under appropriate power output conditions [42]. For the first challenge, a new type of rechargeable AIB based on graphitic foam cathode materials has effectively avoided the hydrogenation and formation of passive oxide layer by using ionic liquid electrolyte, which leads to excellent cycling performance at a very high charging rate [43]. Facts have proved that this kind of ionic liquid electrolyte can achieve reasonable electrochemical performance in AIBs system. Now, using ionic liquid electrolyte is a major trend in the research of AIBs. As for the second challenge, a lot of materials have been used as cathodes in AIBs, which include ­VO2 nanorod, ­V2O5 nanowire, graphite [44, 45], ­Mn2O4 [46] and so on. But their electrochemical performances can’t reach the height of LIB. The top priority is to explore suitable cathode materials if AIBs want to replace LIB battery in the field of large-scale grid and portable devices [47–50]. Exploring suitable electrode materials with desirable electrochemical properties remains a primary challenge for rechargeable AIBs. In this review, we mainly intend to introduce some cathode materials that are reported by previous literature in this field. By summarizing the electrochemical performance in aluminum based systems in light of the known structures which function most effectively as AIBs cathodes, we hope to reveal particular strategies which may facilitate future developments of AIBs.

13

2 Cathode materials 2.1 Carbonaceous materials Due to its good electrochemical performances and high conductivity, carbonaceous materials (graphitic carbons, amorphous carbons, porous carbons) are widely applied in the field of lithium-ion batteries and sodium-ion batteries [51, 52]. Gifford and Palmisano first reported the Al/ graphite batteries using the carbon based cathode in 1988 [53]. AIB using the carbonaceous materials as cathodes exhibit high voltage and long cycle life. Rani et al. [54] first utilized fluorinated natural graphite as cathode material of AIB. Fluorinated natural graphite with non-covalent C–F bonds was prepared by electrochemical method. A mixture of 0.5 M ­A lCl 3 and 1,3-di-n-butylimidazolium bromide[bim][Br] was used as electrolyte. In discharge process of this kind of AIB, ­[AlCl3Br]− anions react with the Al anode to produce [­ Al2Cl6Br]− complex compounds, and the fluorinated natural graphite cathode reacts with ­[Al2Cl6Br]− to form aluminum intercalated C ­ xF discharge product. The CV curves of this battery in 0.5:1 molar ratio ­AlCl3 to [bim] [Br] at room temperature are shown in Fig. 1a. There are three cathodic peaks at 1, 2 and 3A, and three corresponding anodic peaks can be observed at 1, 2 and 3C. After 20 cycles, the peak positions and peak current values are almost identical, which indicate the electrochemical performance of this battery is stable. Furthermore, the capacity of the battery is 225 mAh g−1 and Columbic efficiency is 75% after 40 cycles. Haobo Sun group [55] assembled a new type of AIB using carbon paper as cathode, high-purity Al foil as anode and ionic liquid (the combination of ­AlCl3 and 1-ethyl3-methylimidazolium chloride) as electrolyte. They proposed a new theory of “multi-coordination ion/single ion intercalation and de-intercalation” and proved it with the morphology characterization and electrochemical testing results. Specifically, in the discharge process, the metal aluminum is oxidized to form ­Al3+, ­Al3+ and aluminum chloride coordination anion (­ [Al aCl b] −) insert into the graphite layers to form A ­ l xCly. Then, the intercalated ­AlxCly reacts with adjacent graphite layers due to Van der

Journal of Materials Science: Materials in Electronics (2018) 29:14353–14370

14355

Fig. 1  a Cyclic voltammetry of Al-ion cell using graphite fluoride cathode and aluminum anode in ionic liquid at sweep rate of 10 mV s−1. Reproduced with permission from Ref. [54]. Copyright 2013 The Electrochemical Society. b Cyclic voltammetry curves of an Alion full battery and that of Mo. c Charge and discharge curves under different current densities. b, c Reproduced with permission from Ref. [55]. Copyright 2015 The Royal Society of Chemistry. d A scanning electron microscopy (SEM) image showing a graphitic foam with an open frame structure; scale bar, 300 µm. Inset, photograph of graphitic foam; scale bar, 1 cm Reproduced with permission from Ref. [56]. Copyright 2015 Macmillan Publishers Limited

Waals’ forces. The electrode reactions of this AIB should be written as (discharge process):

Al − 3e− → Al3+ (anode)

(1)

Al3+ + [Ala Clb ]− + e− → Alx Cly (cathode)

(2)

The CV curves of this AIB for the first three cycles are shown in Fig. 1b. There are three reduction peaks at the positions of a (1.65 V), b (1.77 V), c (2 V), and three oxidation peaks can be observed at a′ (2.10 V), b′ (2.24 V), c′ (2.34 V). The second and third curves are almost identical, which indicate that the reversibility of the battery is excellent. The CV results of the current collector (Mo) show that the influence of Mo can be ignored in this system. In Fig. 1c the charge and discharge curves with the different current densities from 50 to 150 mA g−1 in the voltage range from 0.4 to 2.35 V all show the clear voltage plateaus of charge and discharge. Especially the discharge curve at the current density of 50 mA g−1 has two voltage plateaus and the second voltage plateau reach up to 1.8 V. As an aluminum-ion full battery, such high voltage plateau is reported for the first time. This kind of battery have excellent capacity performance, the discharge capacity is 69.92 mAh g−1 at current density of 100 mA g−1 even after 100 cycles. The neighboring graphite intercalating space of the discharged carbon paper is greatly enlarged after 100 cycles (0.455 vs. 0.33 nm) but the

morphology of the surface still remains unchanged after 100 cycles. When the current density reaches to 150 mA g−1, the discharge capacity remains 62.71 mAh g−1 after 50 cycles. The Lin et al. [56] do a heart-stirring work in the AIBs field. They published an article called “An ultrafast rechargeable aluminum-ion battery” on Nature in 2015. This new AIB system is based on aluminum foil anode, three-dimensional graphitic-foam cathode and ionic liquid electrolyte which includes anhydrous aluminum chloride (­ AlCl3) and 1-ethyl-3-methylimidazolium chloride ([EMIm]Cl). This AIB has high cycling stabilities and unprecedented discharge profiles at room temperature. The three-dimensional graphitic-form cathode obtained by chemical vapor deposition method has many layered structure and large inner surface areas which ensure electrode and electrolyte contact more fully to speed up the rate of ion exchange. The SEM images in Fig. 1d show the morphology and porous structure of graphitic foam cathode. The electrochemical performances of this new type of AIB are very encouraging. Under the current density of 5000 mA g−1, the battery can be fully charged in less than 1 min and discharge process is slow with a high capacity about 60 mAh g−1. The AIB after 7500 cycles shows the capacity retention of 100% and the Coulombic efficiency of 97 ± 2.3% (Fig. 2b). This new AIB with ultrafast charge/discharge ability over thousands of cycles is an important breakthrough. Compared with LIB, ­Li+ is the carrier of the charge in the LIB, and the carrier

13

14356

Journal of Materials Science: Materials in Electronics (2018) 29:14353–14370

Fig. 2  a Ex situ X-ray diffraction patterns of PG in various charging and discharging states through the second cycle. b Galvanostatic charge and discharge curves of an Al/ graphitic-foam pouch cell at a current density of 4000 mA g−1. a, b Reproduced with permission from Ref. [56]. Copyright 2015 Macmillan Publishers Limited. c Galvanostatic charge and discharge curves of an Al/ NG cell at a current density of 66 mA g−1. d Ex situ C 1 s XPS spectra of the graphite cathode in Al/NG cell in various charging and discharging states (denoted C and D, respectively) through the second cycle. c, d Reproduced with permission from Ref. [57]. Copyright 2017 Nature Publishing Group

of the charge in this new AIB system is A ­ lCl4−. But, just like carbon containing L ­ i+ in LIB, three-dimensional graphitic-form cathode can also accommodating ­AlCl4−. The ­AlCl4− reacts with metal aluminum to generate ­Al2Cl7− and electrons in the anode during the discharge process. And in the cathode, A ­ lCl4− is de-intercalated between graphite layers. In the charging process, it is just the opposite. These two reactions are highly reversible, so they get a complete recyclable secondary battery system. The cell reactions are described as follows:

4Al2 Cl−7 + 3e− ↔ Al + 7AlCl−4

(3)

Cn + AlCl−4 ↔ Cn AlCl4 + e−

(4)

[

]

the n represents the molar ratio of carbon atoms to intercalated anions in the graphite. The ex situ XRD measurements of graphite show the sharp peak (002) of pristine graphite foil disappear at 2θ = 26.55° on charging while two new peaks appear at 28.25° and 23.5° (Fig. 2a). Meanwhile, analysis of the peak separation indicated spacing between adjacent graphitic host layers is 5.7 Å, and the size of ­AlCl4− anions is 5.28 Å, which confirmed graphite intercalation/de-intercalation by chloroaluminate anions during charging/discharging. Wang et al. [57] explored a rechargeable AIB using a film of SP-1 natural graphite flakes (NG) with a polyvinylidene fluoride (PVDF) binder as the cathode in ionic liquid electrolyte. The liquid electrolyte was prepared by mixing anhydrous ­AlCl3 and [EMIm]Cl with a mole ratio of 1.3:1. The studies using various graphite materials as cathodes of AIBs found that natural graphite

13

is better than synthetic graphite in terms of capacities and voltage plateaus. The results indicated that the graphite materials with high crystallinity and low defect density have higher electrochemical performances for AIB. The charge and discharge curves (Fig. 2c) showed two clear discharge voltage plateaus in the ranges 2.25–2.0 and 1.9–1.5 V and the specific discharge capacity reached 110 mAh g−1 based on the mass of graphite (4 mg cm−2) at a current density of 66 mA g−1. At a high rate of 6 C, the Al/NG battery could still provide a capacity of 60 mAh g−1 with a Coulombic efficiency of 99.5%. And this battery delivered a high stability of without specific capacity decay and a Coulombic efficiency of 99.5% under charge–discharge cycling over 6000 cycles at 6 C. They conducted ex situ XPS measurements to probe the structure change of the intercalated compound in NG cathodes in charging and discharging process. Figure 2d shows the ex situ C 1s XPS spectra of graphite cathode in Al/NG cell in various charging and discharging states (denoted C and D, respectively) through the second cycle. In charging process, when the voltage was increased from C 2.0 to C 2.5 V, the C 1s peak at 284.5 eV shifted to a higher binding energy of 285.1 eV. In discharging process, when the voltage was increased from D 0.5 to D 2.4 V, the C 1s peak fully recovered to the initial binding energy. This shows the highly reversible oxidation/reduction of carbon in NG cathode during the chloroaluminate anion intercalation/ de-intercalation. Huang et al. [58] developed a novel coin-cell type AIB. As shown in Fig. 3a, this AIB applied a poly conductive polymer (3,4-ethylenedioxythiophene, PEDOT) to modify

Journal of Materials Science: Materials in Electronics (2018) 29:14353–14370

14357 Table 2  Average interlayer distance (Å), voltage (V), and binding energy (eV) per ­AlCl4 for all stages with different stoichiometries Stages

No. of ­AlCl4

Interlayer distance

Voltage

Binding energy

Stage-1

4 8 12 16 2 4 6 8 2 4 6 8 1 2 3 4

8.36 8.52 8.69 8.76 6.07 6.09 6.12 6.11 5.11 5.16 5.20 5.22 4.72 4.74 4.76 4.78

3.47 2.30 2.00 1.87 3.64 2.04 1.75 1.66 3.91 2.13 1.74 1.65 3.66 2.01 1.65 1.58

+ 0.39 + 0.77 + 1.08 + 1.20 + 0.57 −1.04 −1.31 −1.41 + 0.24 −0.94 −1.33 −1.43 + 0.59 −1.07 −1.42 −1.50

Stage-2

Stage-3

Stage-4

The average interlayer distance term is used to differentiate the total height of the four stages considered upon intercalation. Reproduced with permission from Ref. [59]. Copyright 2017 the Owner Societies Fig. 3  a The correlation between theoretical weight density and pore diameter of highly porous graphene foam at a fixed surface area of 762 m2 g−1. Inset is the illustration of the application of freestanding monolithic NGF cathode in a coin-cell type AIB. Reproduced with permission from Ref. [58]. Copyright 2017 The Royal Society of Chemistry. ­AlCl4 intercalated geometries: b tetrahedral and c planar. b, c Reproduced with permission from Ref. [59]. Copyright 2017 the Owner Societies

and stabilize the stainless steel cover of coin-cells. The freestanding, monolithic nanoporous graphene foam (NGF) cathode synthesized by hydrothermal hard-templating method has a high pore volume of 2.45 cm3 g−1, a large surface area of 762  m2  g−1 and high weight density of 81.0 mg cm−3 (Fig. 3a). This NGF cathode could achieve a volumetric capacity of 12.2 mAh cm−3, a high gravimetric capacity of 151 mAh g−1 and good low temperature performance, which present new opportunities for the advancement of AIBs. Preeti Bhauriyal et al. [59] simulated four different stages (stage-1 to stage-4) to investigate the staging mechanism of ­AlCl4 intercalation in a graphite electrode for AIB. They first exam the most stable geometry of ­AlCl4 in graphite and distinguish that the tetrahedral A ­ lCl4 structure (Fig. 3b) is more stable than the square-planar A ­ lCl4 structure (Fig. 3c). Then they carefully investigated all possible intercalation sites for ­AlCl4 in graphite. The results show the bridging position between two non-bonded carbon atoms is the most stable site for ­AlCl4 intercalation. In other words, the Al atom takes a bridge position between two non-bonded C-atoms, and four

Cl-atoms nearly occupy the center of the hexagon. Binding energy of four stages (Table 2) indicating that the intercalation into graphite is not smooth at low A ­ lCl4 concentrations, but it becomes easier with an increase in the concentration of ­AlCl4. Moreover, the average voltages (2.01–2.3 V) and the gravimetric capacities of 25.94 mAh g−1 (stage-4) and 69.62 mAh g−1 (stage-1) are in agreement with the value reported by group of Lin [56]. In summary, carbonaceous materials are electrochemically active in rechargeable AIBs by using ­AlCl3/[BMIM]Cl electrolyte. And carbonaceous materials can intercalate/deintercalate chloroaluminate ­AlCl4− anions during charge and discharge process. This phenomenon has been proved by the research results of X-ray photoelectron spectroscopy (XPS), X-ray diffraction (XRD) and in-situ Raman spectroscopy. However, graphite cannot intercalate ­Al3+, and result in that the capacity is limited to about 60 mAh g−1. The monovalent intercalation mechanism impedes further improvement of capacity for carbon-based materials.

2.2 Metal oxides 2.2.1 V2O5 Smyrl et al. [60]. investigated ­Al3+ intercalation into ­V2O5 aerogels in 1998, they found that A ­ l3+ insertion is facile in the aerogel and that up to 3.33 equivalents of A ­ l3+ may be inserted per mole of ­V2O5 (ARG). For 3.33 equivalents

13

14358

Journal of Materials Science: Materials in Electronics (2018) 29:14353–14370

of ­Al3+ intercalation, the specific energy is estimated to be 750 Wh kg−1 at the equilibrium potential of 1.3 V. Archer et al. [61] reported a novel AIB system in 2011. The system consists of ­V2O5 nanowires cathode prepared by a hydrothermal method, acidic ionic liquid electrolyte ­(AlCl3 dissolved in 1-ethyl-3-methylimidazolium chloride and the molar ratio is 1.1:1) and an aluminum metal anode. The ­V2O5 cathode slurry was produced by mixing 85% of the synthesized ­V2O5 nano-wires, 7.5% super-p carbon and 7.5% of PVDF binder in NMP dispersant. Stainless steel (301 type stainless steel, McMaster Carr, USA) was used as the current collector. The discharge capacity of this battery for the first cycle and 20th cycles is 305 and 273 mAh g−1 at a current drain of 125 mA g−1, respectively (Fig. 4b). The CV curves in Fig. 4a shown a cathodic peak at 0.45 V and a corresponding anodic peak at 0.95 V, which may result from the insertion/de-insertion of ­Al3+ ions into and from the orthorhombic crystal lattice structure of ­V2O5 nanowires. But Reed et al. [62] have a different test conclusion about the role of V ­ 2O5 in AIB system. Their battery is composed of ­V2O5 composite cathode, aluminum anode, and

1-ethyl-3-methylimidazolium chloride + AlCl3 ionic liquid electrolyte. They found that ­V2O5 was electrochemically inactive in a potential window of 5 mV to 1.5 V versus Al/ Al3+. The electrochemical behavior of the cells was proved to be independent of the V ­ 2O5 and depend entirely on the stainless steel (a current collector). The CV and chronopotentiograms test of 0 wt% ­V2O5 were similar as 90 wt% ­V2O5, which suggested the V ­ 2O5 composite cathode shows no electrochemical activity toward aluminum. And they believed that the battery-like performance can be attributed to reactions with the iron and chromium in the stainless steel current collector. The ­V2O5 cathode materials mentioned above are mixed with conductive additive and polymer binder to form pasted electrodes on current collectors. As a result, the practical capacity of the electrode is lowered and the electrolyte accessibility to the active material is affected because of the presence of the inactive components, which further influence the electrochemical performance [63]. Researchers have found that electrodes synthesized by directly growing particles on conductive substrates can achieve good

Fig. 4  a Typical CV curves of Al-ion battery using ­ V2O5 nanowire cathode and aluminium anode in 1.1:1 molar ratio of A ­ lCl3 in ([EMIm]Cl) at a sweep rate of 0.2  mV  s−1. b Voltage versus sp. capacity of Al-ion battery containing aluminium anode, V ­ 2O5 nanowire cathode in ionic liquid under the potential window 2.5–0.02  V and at a constant current drain of 125  mA  g−1. a, b Reproduced with permission from Ref. [61]. Copyright 2011 The Royal Society of Chemistry. c GITT at the first cycle; the symbols indicate OCVs; inset: the overpotential as a function of x in A ­ lxV2O5. Reproduced

with permission from Ref. [67]. Copyright 2017 Published by Elsevier B.V. d CV curves of the ­V2O5/C and KB electrodes in a mixed electrolyte solution at a sweep rate of 2 mV s−1. e Cycle performance of the rechargeable Al cell assembled with the V ­ 2O5/C positive electrode at C/40, C/20, and C/10 discharge rates. Charge rate is the same as discharging rate. d, e Reproduced with permission from Ref. [66]. Copyright 2015 American Chemical Society. f GITT at the fourth cycle; inset: same that in c. Reproduced with permission from Ref. [67]. Copyright 2017 Published by Elsevier B.V.

13

Journal of Materials Science: Materials in Electronics (2018) 29:14353–14370

electrical contact and enhanced ion migration kinetics, especially for conductive substrates with three-dimensional networks [64, 65]. Such an electrode helps to migrate and diffuse the electrolyte in the electrode, and may reduce the polarization and enhance the battery voltage. Wang et al. [63] reported a binder-free Ni–V2O5 cathode material, which was synthesized by directly depositing ­V2O5 on a Ni foam current collector. This binder free cathode for rechargeable AIB delivered an initial discharge capacity of 239 mAh g−1, which is much higher than that of the cathode composed of ­V2O5 nanowires and binder. Furthermore, a relatively long discharge voltage plateau appeared at 0.6 V in the discharge curves of the binder free cathode. The study results show that the binders (PVDF and PTFE) have the impact on cell performance. PVDF binder was incompatible with acidic chloroaluminate ionic liquids, whereas the PTFE binder makes for enhancing cell performances. Chiku et al. [66] first used amorphous vanadium oxide/carbon (ketjen black, KB) composite ­(V2O5/C) as cathode in AIB system. In order to eliminate the influence of stainless steel current collector, they replaced stainless steel with a Mo current collector. And the ionic liquid with a mixture of ­AlCl3, dipropylsulfone and toluene (1:10:5 in mole ratio) was used as electrolyte. Figure 4d shows the CV curves of amorphous V ­ 2O5/C and KB electrodes. For KB electrode, the anodic peak was not observed and the cathodic peak was found around 0.5 V versus Al/Al3+. The double layer capacitance of KB was quite small and it will not influence the charge/discharge performance of the ­V2O5/C electrode. For ­V2O5/C electrode, cathodic and anodic peaks were observed at 0.8 and 1.6 V vs Al/Al3+, respectively, which agree with the work of Smyrl. And the rechargeable AIB exhibits the maximum discharge capacity over 200 mAh g−1 in the first discharging at a discharge rate of 1/40 C (Fig. 4e). In this cathode material, the average valence of V changed between 4.38 and 4.91 during discharging and charging. Later, Wu et al. [67] investigated the mechanism of ­Al3+ stored in the lattice of the host electrode material and first demonstrated that ­Al3+ can be inserted into the metal oxide and stored reversibly through intercalation and a phase-transition reaction. The ­V2O5 nanowires were synthesized by a hydrothermal method and its diameter range is 80–120 nm. They used nickel foam as the current collector and gave agalvanostatic cycling and CV tests to ensure that nickel foam current collector does not affect the electrochemical performance. To study the electrochemical reaction mechanisms of the V ­ 2O5–Al cells, galvanostatic intermittent titration technique (GITT) measurements were used for the batteries at first discharge (Fig. 4c) and the fourth cycle (Fig. 4f). The thermodynamic mechanism and equilibrium voltage (Ueq) can be indicated by open-circuit voltages (OCVs). As shown in Fig.  4c, OCVs declined gradually from 1.6 to 1.25 V upon the first GITT discharge. Then, it showed a discharge capacity of

14359

55 mAh g−1 and all the subsequent OCVs basically remained unchanged at 1 V (Fig. 4c). For the reversible cycles, all the OCVs returned to 1.05 V after 24 h relaxation (Fig. 4f). Thus, the inference is that the reaction process of the first discharge is intercalation followed by a phase transition. In the first cycle, the cathode material changed from ­V2O5 to ­AlxV2O5 (x ≤ 0.0792) (Fig. 4c inset) and the discharge overpotential increased from 0.501 V to 0.617 V as the capacity increased. Then, the overpotential quickly declined to 0.461 V and increased to 0.866 V again until the end of discharge (0.0792 ≤ x ≤ 0.181) (Fig. 4c inset). However, in the next reversible charge and discharge process, the overpotentials display a tendency to increase, which implies only a phase-transition process, as shown in the inset of Fig. 4f. The values coming from the ex-situ XPS data of the V ­ 2O5 in various discharge/charge stages and after reversible cycling are shown in Table 2. The changes of valence states of V, O and Al demonstrate the electrochemical interaction between ­Al3+ and ­V2O5. Comparing the original HRTEM images of V ­ 2O 5 nanowire with the fully discharged one (Fig. 5c), there is an amorphous layer of 10 nm width appeared at the edges of the ­Al3+-enriched nanowire after discharge. This amorphous layer can reasonably be attributed to the storage of ­Al3+ into ­V2O5. 2.2.2 VO2 nanorods Wang et al. [68] reported a battery which composed of V ­ O2 nanorods cathode, Al foil anode and ionic liquid electrolyte ­(AlCl3 and [BMIm]Cl in a 1:1 molar ratio with 0.5 wt% ­C14H14OS). This battery presented the performances of long cycle life, low cost and high capacity. The tunnel-structured ­VO2 nanorod was synthesized by a low temperature hydrothermal method. Due to special tunnel structure and large surface area of as-prepared material, it can shorten ion diffusion length effectively. The electrochemical reactions of the battery are shown below. In the discharge process,

xAl − 3xe− → xAl3+ (anode)

(5)

VO2 + xAl3+ + 3xe− → Alx VO2 (cathode)

(6)

In the charge process,

xAl3+ + 3xe− → xAl (anode)

(7)

(8) According to the CV curves of this battery at 0.01–0.9 V for the initial three cycles (Fig. 5a), a pair of redox peaks can be observed at 0.76 and 0.47 V. The CV curves of the first three cycles are almost changeless, which indicate that the AIB is very stable and have a good reversibility. The

Alx VO2 − xAl3+ − 3xe− → VO2 (cathode)

13

14360

Journal of Materials Science: Materials in Electronics (2018) 29:14353–14370

Fig. 5  a Cyclic voltammetry curves of V ­ O2 at the scan rate of 0.1  mV  s−1 in the voltage range of 0.01–0.9  V. b Cycling performances under different current densities over 100 cycles. Reproduced with permission from Ref. [68]. Copyright 2013 Nature Publishing Group. c SEM image of the as-prepared T ­ iO2 nanotube arrays. d Typical CV curves of the as-prepared anatase T ­ iO2 nanotube arrays in 1  M A ­ lCl3, ­MgCl2 and LiCl aqueous solutions at 20  mV  s−1.

Reproduced with permission from Ref. [71]. Copyright 2012 The Royal Society of Chemistry. e Typical CV curve of the as-prepared anatase ­TiO2 nanotube arrays in the electrolyte solutions of 1  M ­AlCl3 at a scan rate of 10 mV s−1. f Typical charge–discharge curve of ­TiO2-NTA at 4 mA cm−2. Reproduced with permission from Ref. [72]. Copy right 2014 Elsevier Ltd.

initial three charge/discharge curves at the current density of 50 mA g−1 show that the first discharge capacity is 165 mAh g−1. In the whole process, clear charge and discharge voltage plateaus can be seen at 0.7 and 0.5 V, which are consistent with the results of CV test. Figure 5b shows the cycling performances under different current densities. The discharge capacity is 116 mAh g−1 after 100 cycles at the current density of 50 mA g−1. When current densities reach to 100 and 200 mA g−1, the discharge capacity is 106 and 70 mAh g−1 respectively. It is worth noting that theoretical capacity can reach to 485 mAh g−1 if two Al ions insert into the cathode material. It indicates that the capacity of this battery can be improved greatly.

for electrolyte species due to the high surface area and unique nano-sized geometry, which is a promising cathode material for AIBs. Liu et al. [71] investigated the electrochemical aluminum storage of ­TiO2-NTA in ­AlCl3 aqueous solution. They firstly proved that ­A l 3+ can be reversibly inserted and extracted inside the ­TiO 2-NTA in ­AlCl3 aqueous solution due to the small radius steric effect of ­Al3+. Tetragonal anatase ­TiO2 was prepared by anodizing the metallic titanium foil. The SEM image in Fig. 5c shown that the as-prepared T ­ iO2 film is composed of nanotube arrays with the inner diameter ranging from dozens to one hundred nanometers and the thickness is about 14 µm. Figure 5d is the CV test of T ­ iO 2-NTA in aqueous solutions. There is a pair of large reversible anodic peaks can be observed at − 1.26 and − 0.84 V (vs. SCE), and cathodic peak at − 1.43 V (vs. SCE). XPS test indicated that ­A l 3+ insert into active material and ­TiO 2 reduce from ­Ti4+ to ­Ti3+/Ti2+ for maintaining the charge balance. Liu et al. [72] investigated the electrochemical behavior of ­C l − assisted ­A l 3+ insertion into ­TiO 2-NTA in aqueous solution for AIB. Their results show that the insertion and extraction of ­A l 3+ into ­TiO 2 is reversible

2.2.3 TiO2 The titanium dioxide nanotube array (­ TiO2-NTA) is often used as active materials for dye-sensitized solar cells, photocatalysis, photoelectrochemistry or electrochromic devices due to short solid state diffusion paths [69, 70]. In addition, ­TiO 2-NTA can ensure good electrode/ electrolyte contact and serve as a fast diffusion pathway

13

Journal of Materials Science: Materials in Electronics (2018) 29:14353–14370

and stable with unchanged morphology and crystal structure of T ­ iO2-NTA. A pair of large reversible redox peaks in Fig. 5e can be observed at near − 0.98 and − 1.00 V (vs. Ag/AgCl), corresponding to the reversible aluminum ion insertion/extraction reactions in anatase ­TiO 2. And Fig. 5f shows that the maximum discharge capacity is about 74.8 mAh g −1 at current density of 4 mA cm −2. These electrochemical features attribute to the assistant effect of ­C l − to ­A l 3+ insertion. Furthermore, Zhong’s work [73] showed that ­Al3+ ions can dope into ­TiO2 lattice to lead a high conductivity of supercapacitor. Koketsu et al. [74] realized the reversible insertion of multivalent ions ­( Mg 2+ and A ­ l 3+) in anatase T ­ iO 2 by defect engineering. They introduced plenty of charge-compensating titanium vacancies (22%) through aliovalent doping. The titanium vacancies can serve as microstructural voids to accommodate ­M g 2+ and ­A l 3+ conveniently, which greatly enhance the reversible capacities of material. The formula of compound with 22% titanium vacancies correspond to T ­ i0.78□0.22O1.12F0.40(OH)0.48, where □ represents a cationic vacancy. And the first discharge capacity of T ­ i 0.78□ 0.22O 1.12F 0.40(OH) 0.48 vs A ­ l 3+ ions is about −1 120 mAh g , while the reversible capacity of stoichiometric ­TiO2 is only 30 mAh g−1. Although the reversible capacity is too low for practical applications, it provides a new strategy to exploit new electrode materials for AIBs.

Fig. 6  a Charge/discharge curves of the ­MoO2@Ni electrode. Reproduced with permission from Ref. [75]. Copyright 2017 Published by ECS. b Dependence of the initial discharge capacity on crystallite size. Galvanostatic measurements were performed at 55  °C and

14361

2.2.4 MoO2 Wei et al. [75] fabricated a dense molybdenum oxide layer on nickel foam (­ MoO2@Ni) as cathode material by simple magnetron sputtering and heat-treatment method. The charge/discharge curves in Fig.  6a show two discharge potential plateaus at about 1.95 and 1.0 V. Meanwhile, the first charge/discharge capacities of the M ­ oO2@Ni is about −1 253 and 90 mAh g , which mean a large irreversible capacity loss occurred in the first cycle. The CV test results are consistent with that of charge/discharge. As to the first ten CV cycles, a pair of redox peaks at around 2.15 and 1.91 V can be found obviously. After long cycling, M ­ oO2 is dissolved and transferred to the separator, resulting in rapid capacity decay. Some previous studies also obtained a similar dissolution phenomenon. Suto et al. [76] constructed a battery with V ­ Cl3 and A ­ lCl3/EMImCl and found that V ­ Cl3 could be dissolved into the electrolyte by X-ray absorption near-edge structure (XANES) examination. But M ­ oO2@Ni cathode obtains a high discharge potential plateau of 1.9 V, which is an encouraging outcome in terms of these type batteries. As for the mechanisms of metal oxides materials, Wu et al. [67] believe that ­Al3+ storage in V ­ 2O5 is not a simple electrochemical intercalation/extraction process but rather a combination of intercalation and phase-transition reactions.

8.94  mA  g−1. c Charge–discharge reaction model of ­FeS2 with aluminum ions in an A ­ lCl3-EMIC ionic liquid at 55 °C. b, c Reproduced with permission from Ref. [80]. Copyright 2016 Elsevier B.V.

13

14362

And Liu et al. [72] consider that ­Cl− assist ­Al3+ insertion into ­TiO2-NTA in ­AlCl3 aqueous solution. We think the electrochemical reactions in metal oxides can be described asxAl + M(metal oxides) → xAlM. However, the storage mechanism is not as simple as described, more work need to be done to further analyze the storage mechanism. Above all, metal oxides materials exhibit batter specific discharge capacity but worse voltage and cycle life performance than carbonaceous materials. Theoretically, transition metal oxides may not be the ideal host of Al because they have a strong electrostatic adsorption effect with Al cation. This factor could prevent the redistribution of the charge of the Al cation in the crystal. But the sulfur element has a lower electronegativity than oxygen element and polarized easily due to its larger atomic radius. Therefore, the charge redistribution in the sulfide anion skeleton should be better than that of the oxide. Metal sulfides are expected to be excellent candidate cathode material for the conversion reaction in AIBs. Next, we will give you a detail introduction.

2.3 Metal sulfide 2.3.1 FeS2 Koura et al. [77–79]. first used ­FeS2 as a cathode material for secondary aluminum batteries at different temperatures, but the charge–discharge reaction mechanism was still not well understood. Mori et al. [80] used ­FeS2 as rechargeable AIB cathodes at 55 °C and investigated the reaction mechanisms of ­FeS2. They established a battery system with ­FeS2 cathode, aluminum wire anode and ionic liquid electrolyte which include ­AlCl3 and 1-ethyl-3-methyl imidazolium chloride (2:1 molar ratio). The working electrode is consist of a composite mixture of the active material, vapor-grown carbon fiber and PTFE binder (70:25:5 weight ratio) pressed on a Mo foil. To find the best milling time, the ­FeS2 nanoparticles are prepared by high-energy ball milling at different times. The results indicate that 24 h milling time correspond to minimum crystallite size, in addition, the crystallite size of ­FeS2 exists a close relationship with discharge capacity (Fig. 6b). The proposed discharge reaction is as follow: (9) A reaction model in Fig. 6c is proposed to explain the discharge/charge reaction mechanism of this battery. The ­FeS2 pyrite structure includes dumbbell-shaped disulfide ions, those ions can react with Al-ions. The reaction S2− ↔ 2S2− 2 mainly occurs in closed surface and is reversible. when sulfur is reduced, the original ­FeS2 phase turns into FeS, and amorphous ­Al2S3 phase is formed at the same time. In contrary, FeS and A ­ l2S3 phases transform into F ­ eS2 after charging. But this kind of Al/FeS2 battery still has the problems of low cell voltage and poor cycle stability owing to

FeS2 + 2∕3Al3+ + 2e− → FeS + 1∕3Al2 S3

13

Journal of Materials Science: Materials in Electronics (2018) 29:14353–14370

the dissolution of sulfides. Its theoretical energy density is 462.7 Wh kg−1, which is far less than 1337.0 Wh kg−1 of Li/FeS2 battery. So Al/FeS2 battery has a large space to improve. 2.3.2 Mo6S8 Aurbach et al. [81] first applied ­Mo6S8 as a cathode material for rechargeable magnesium-ion batteries. Guo et al. [82] developed a new AIB with ­Mo6S8 cathode, Al foil anode and an ionic liquid electrolyte composed of A ­ lCl3 and [BMIm]Cl with molar ratio of 1.5:1. The ­Mo6S8 particles were synthesized through a precipitation method and the reversible electrochemical Al intercalation in ­Mo6S8 was demonstrated for the first time. Figure 7a is SEM images of the synthesized ­Mo6S8 particles, and the size is within the range of 1–2 µm (inset in Fig. 7a). As shown in Fig. 7b, the crystal structure of M ­ o6S8 is stacked by M ­ o6S8 blocks which composed of an octahedral cluster of Mo atoms inside a sulfur anion cubic cell. ­Al1 and ­Al2 are two different sites for Al atoms intercalation. In their works, all the electrochemical studies were performed at 50 °C because the elevated temperature improved the electrochemical reaction kinetics. Figure 7c is the CV curves of ­Mo6S8 at a scan rate of 0.1 mV s−1 and the different cycles. The CV curve at the 5th cycle has two cathodic peaks at 0.50 and 0.36 V and two anodic peaks at 0.40 and 0.75 V. Figure 7d is galvanostatic charge–discharge curves of Al/Mo6S8 battery with a current density of 12 mA g−1 at the different cycles. The first discharge curve has two clear plateaus at 0.55 and 0.37 V, which are correspond to the two cathodic peaks in the CV curve. The Al intercalation capacity based on the chemical formula weight of ­Mo6S8 is 148 mAh g−1 in the first discharge, but its first charging capacity is only 85 mAh g−1. And in the first discharge curve the voltage slope is from 0.75 to 0.55 V. Electrochemical performances depend on Al atoms intercalating into crystal lattice of ­Mo6S8. They concluded that the theoretical formula of fully Al intercalated ­Mo6S8 is ­Al2Mo6S8 and the discharge reaction of this battery is

8[Al2 Cl7 ]− + 6e− + Mo6 S8 ↔ Al2 Mo6 S8 + 14[AlCl4 ]− 2.3.3 SnS2

(10)

For the practical application of AIBs, the charge–discharge reaction should be observed at relatively low temperatures. Wang et al. [83] reported the design of an innovative freeze-dried reduced graphene oxide (RGO) supported ­SnS2 (G-SnS2) as a new type of cathode materials for rechargeable AIB at 25 °C. They compared the electrochemical performances of G-SnS2 and ­SnS2 to emphasize the advantages of G-SnS2. Figure 8a shows the charge–discharge curves of two samples, all exhibiting the charging plateau and discharging

Journal of Materials Science: Materials in Electronics (2018) 29:14353–14370

14363

Fig. 7  a SEM images of the synthesized ­Mo6S8. b Crystal structure sketch map of Al intercalated ­Mo6S8. c The first, second, and fifth CV curves of ­Mo6S8 at 50 °C. d First, second, and 20th GCD curves at 50 °C. a–d Reproduced with permission from Ref. [81]. Copyright 2015 American Chemical Society

Fig. 8  a First/fifth charge–discharge curves of AIB assembled with the G-SnS2 and ­SnS2 electrodes at a specific current of 100 mA g−1. b CV curves of the G-SnS2 and ­SnS2 electrodes at a scan rate of 0.5 mV s−1. c Comparison of the specific discharge capacity of the electrodes at different specific currents. d Ex situ XRD patterns of ­SnS2 electrodes at differernt electrochemical status. a–d Reproduced with permission from Ref. [83]. Copyright 2017 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

plateau at around 1.60 and 0.68 V (vs. Al), respectively. In CV curves (Fig. 8b), there is a main cathodic peak at about 1.65 V and a broad anodic peak around 0.65 V, which agree

with the plateaus from the charge–discharge curves. The first discharge capacity of the ­SnS2 electrode is 246 mAh g−1, but the initial discharge capacity of the G-SnS2 electrode attains

13

14364

392 mAh g−1 (equivalent to energy density of 241 Wh kg−1), which is the highest capacities/energy density among the reported for ionic liquid rechargeable AIBs. Furthermore, at a high specifc current of 1000 mA g−1 (Fig. 8c), the sample G-SnS2 exhibited much better capacity performance (112 mAh g−1) than S ­ nS2 (40 mAh g−1). Thermal gravimetric analysis reveals that the weight ratio of RGO in their hybrid is quite small, ≈ 5.2 wt% (­ SnS2 ≈ 94.8 wt%). The results show that the uniform distribution of S ­ nS2 on RGO framework, large surface area and enhanced electronic property are the important reasons for improving electrochemical performances of the G-SnS2 composite electrode. They also demonstrated that ­AlCl4− could intercalate and de-intercalate into the cathode material during the charge/discharge process. In the ex situ XRD patterns of ­SnS2 electrodes at different electrochemical status (Fig. 8d), the typical (001) peak of the S ­ nS2 at 2θ = 15.03° changed while other peaks remained constant during the electrochemical process. A new peak appears at ≈ 12.81° and gradually increases with deeper discharge stages. On the contrary, the peak assigned to ­SnS2 (2θ = 15.03°) gradually decreased, indicating that a larger interlayer space may be formed probably due to the insertion of large chloroaluminate anions. The TEM Fig. 9  a TEM image of NiS nanobelts. HRTEM image and the corresponding SAED pattern in the inset at the lower right corner. Reproduced with permission from Ref. [89]. Copyright 2013 The Royal Society of Chemistry. b Crystal structures of ­Ni3S2. A is viewed along the X axis and B is viewed along the Z axis. Reproduced with permission from Ref. [90]. Copyright 2016 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

13

Journal of Materials Science: Materials in Electronics (2018) 29:14353–14370

test results show that interlayer space increased from 5.9 to 7.0 Å after charging and reduced to 5.9 Å after discharging processes, further supporting the reversible intercalation and de-intercalation of the large anions. XPS test inferred that S is the main element with varied valence to balance the charge variation. Its discharge reaction is

SnS2 + nAlCl−4 ↔ ne− + SnS2 [AlCl4 ]n

(11)

2.3.4 NiS NiS has large theoretical capacity (590 mAh g−1) and a good conductivity of ~ 2 × 104 S cm−1, which stimulates great research interest as electrode materials in lithium batteries [84–88]. Recently Jiao et al. [89] investigate an AIB with the hexagonal NiS nanobelts as cathode and the high purity Al foil as anode in A ­ lCl3/[EMIm]Cl (molar ratio 1.3:1) ionic liquid electrolyte system. The NiS nanobelt powder was synthesized by temperature hydrothermal method. TEM image in Fig. 9a shws that the synthesized NiS sample has a belt-like morphology with a size of 50–100 nm in width and around 10–25 µm in length. The Al/NiS battery keeps the discharge capacity of 104.7 mAh g−1 in the first cycle at the current density of 200 mA g−1, the corresponding

Journal of Materials Science: Materials in Electronics (2018) 29:14353–14370

Coulombic efficiency is 98.1%. And the discharge capacity at the 100th cycle is 104.4 mAh g−1, with a Coulombic efficiency of 97.66%. Even at a high current density of 300 mA g−1, the charge and discharge capacities could remain at about 90.2 and 92.1 mAh g−1, respectively. In order to get better electrochemical performance, Jiao et al. [90] constructed another novel AIB, using aluminum foil anode, the as-prepared N ­ i3S2/graphene cathode and the ­AlCl3/[EMIm]Cl (molar ratio 1.3:1) ionic liquid electrolyte. The special feature of ­Ni3S2/graphene micro flakes is the low charge-transfer impedance, which represents a high rate of intercalation and de-intercalation of ions. The crystal structure of hazelwoodite ­Ni3S2 (Fig. 9b) has a fairly open framework with large interstitial sites and will allow to accommodate compensating counter ions easily during redox reactions. From the work of Parise [91], hazelwoodite ­Ni3S2 has a monocrystal structure (R32 space group), and unit cell parameter a is 4.07 Å. It is found that the initial cycle of this battery has a discharge voltage plateau (≈ 1.0 V vs. Al/AlCl4−) with the capacity of about 350 mAh ­g−1 at a current density of 100 mA g−1. Besides, the electrode exhibits a reversible capacity of over 60 mAh g−1 after 100 cycles, which corresponds to a Coulombic efficiency of about 99%. Even at a high current density of 200 mA g−1, the charge and discharge capacities can keep as high as about 300 and 235 mAh g−1, respectively. It is worth to mention that there is a transformation from monocrystal to polycrystal of the active material. 2.3.5 CuS Because ­Ni3S2 microflakes cathode exhibited a relative low capacity, Jiao’s group constructed a new 3D hierarchical copper sulfide (CuS)@carbon microsphere cathode material (CuS@C) for AIB. In this composite structure, acetylene black nanoparticles were dispersed uniformly on the surface of CuS@C microsphere [92]. The mass ratio of C in this composite can reach about 30%, which enhance the conductivity of the electrode. Figure 10a shows the crystal structure of hexagonal CuS (space group: P63/mmc), two-thirds of the copper atoms are at the center of a triangle of sulfur atoms, and the rest one-third of copper atoms are at the center of the layer. The SEM images (Fig. 10b) indicate that the asprepared CuS is composed of well-dispersed superstructures which build from nanoplates with an edge length of about 1 µm and an average thickness of about 13 nm. The initial cycle charge and discharge curve has two discharge voltage plateaus (~ 1.0 and 0.4 V vs. Al/AlCl4−) with the capacity of about 240 mAh g−1 at a current of 20 mA g−1. The practical capacity is less than the theoretical capacity (280.3 mAh g−1) due to the incomplete reactions and the impact of internal resistance. Figure 10c shows the cycling performance and corresponding Coulombic efficiency when

14365

the current is 20 mA g−1. With the increase of cycle number, the capacity gradually decreases. The specific capacity is about 90 mAh g−1 at the 100th cycle, corresponding to nearly 100% of the Coulombic efficiency. The schematic illustration of the battery during the discharging process is shown in Fig. 10d. On the cathode, CuS transforms into the low-valence C ­ u2S along with the migration of ­Al3+ in the electrolyte during discharging. At the same time, on the anode side, the dissolution of Al occurred on the surface of the Al foil electrode. At present, the reaction mechanisms of metal sulfide cathode materials are still not well understood, people assume that some materials (­FeS2, ­Mo6S8, G-SnS2, NiS, N ­ i 3S 2/ graphene, CuS@C) allow Al ions insertion and extraction reversibly, and others (G-SnS2) can accommodate chloroaluminate anion ­(AlCl4−). ­FeS2 and NiS react with ­Al3+ cations forming amorphous A ­ l2S3 phase and the charge compensation during discharge is dominant at sulfur atoms [78, 87]. ­Ni3S2/graphene exists a transformation from monocrystal to polycrystal due to the intercalation and extraction of ­Al3+ cations [88]. More works need to do to thoroughly understand the reactions occurring in metal sulfide cathode in the future. The cathode materials of metal sulfide diaplay great electrochemical performances, but this battery was suffers from the low cell voltage and the poor cycle stability due to the dissolution of sulfides. Further investigation to understand the Al trapping mechanism and the complex with carbon material are effective methods to solve this problem.

2.4 Other cathode materials Prussian blue analogues (PBAs) might be a promising material for Al ion insertion because of its open structure. Researchers found that the copper hexacyanoferrate (CuHCF) with Prussian blue structure are able to electrochemically insert ions in aqueous solutions [93–99]. Reed et al. [100] investigated CuHCF as a cathode material for AIB in an electrolyte of aluminum trifluoromethanesulfonate (aluminum triflate) dissolved in diethylene glycol dimethyl. This system displays initial discharge capacities as high as 60 mAh g−1 and reversible capacities between 5 and 14 mAh g−1, but the capacity usually fading after 10–15 cycles due to some structure destruction in the composite electrode. They assume that the CuHCF nanoparticles could be written as KCu[Fe(CN)6]⋅xH2O, where the x represents the amount of zeolitic water residing in the CuHCF cavity. They considered that the zeolitic water in CuHCF helps to shield the charge of ­Al3+ ions and reduce its electrostatic forces from the host structure, which is expected to enhance the diffusion kinetics. Other promising cathode material is conducting polymer. Polymer has been extensively researched as electrode

13

14366

Journal of Materials Science: Materials in Electronics (2018) 29:14353–14370

Fig. 10  a Schematic representation of CuS crystal structures, shown perpendicular to different axes. Brownish-red and yellow spheres represent copper and sulfur atoms, respectively. b FESEM images at different magnification. c Cycling performance with Coulombic efficiency at a current density of 20 mA g−1. d Schematic illustration of an aluminum-ion battery during the discharging process. (−) Al foil/ ionic liquid electrolyte ­([EMIm] AlxCly)/GF/D sepatator/ionic liquid electrolyte (­ [EMIm] AlxCly)/CuS@C composites/Ta foil (+) from bottom to top. a–d Reproduced with permission from Ref. [92]. Copyright 2016 American Chemical Society. (Color figure online)

materials for LIB because of the electrochemically oxidization and reduction in electrolyte solutions [101–106]. Hudak [107] used chloroaluminate-doped conducting polymers (polypyrrole and polythiophene) as active materials in the positive electrodes of rechargeable AIBs at room temperature. The discharge capacity of this kind of cathode material varies from 30 to100 mAh g−1, and the Coulombic efficiencies are both 100% in chloroaluminate ionic liquids ­(AlCl3:[EMIm]Cl molar ratio 1.5:1). But the average low discharge voltages limit the application of polymer cathode material. Jiang et al. [108] explored a ­Li3VO4@C microsphere composite cathode material for AIB in ionic liquid electrolyte, which has been widely researched in LIB filed [109–113]. As shown in Fig. 11a, ­Li3VO4 has a hollow lantern-like 3D structure with ­VO 4 and ­L iO 4

13

tetrahedra shared by ordered corners, and there are many vacancies in the framework to accommodate ions inserting reversibly [114]. The as-prepared composite cathode material belongs to orthorhombic ­Li3VO4 (space group: Pnm21) with lattice parameters a = 5.45 Å, b = 6.33 Å, c = 4.95 Å and V = 170.54 Å3. And the carbon layer coated on the ­Li3VO4 particles is amorphous and porous with the thickness of 5 nm. They give a galvanostatic discharge/ charge measurements in a voltage range of 0.05–0.95 V and at a current density of 20 mA g− 1 to investigate the electrochemical performance. The initial discharge/charge capacity (Fig. 11b) is 137 and 85 mAh g−1, respectively, and the Coulombic efficiency is 62%, similar to that of LIB (64.5%) [115]. The discharge/charge capacity of the second cycle is 98 and 74 mAh g −1 with a Coulombic

Journal of Materials Science: Materials in Electronics (2018) 29:14353–14370

14367

Fig. 11  a Crystal structure of ­Li3VO4 projected along a-axis. Color code: Li green; V dark blue; O red. b Initial discharge/ charge curve of Al/Li3VO4@C batteries at a current density of 20 mA g−1. c V 2p XPS spectra of ­Li3VO4@C initial charged to 0.95 V. a–c Reproduced with permission from Ref. [108]. Copyright 2017 American Chemical Society. (Color figure online)

efficiency of 76%. The discharge capacity is more than 48 mAh g−1 after 100 cycles with Coulombic efficiency close to 100%. The above results may be attributed to the structural advantages of L ­ i3VO4@C composite, the carbon layer covering the sphere not only promotes the electrical conductivity of the material, but also regulates the size of the spherical particles. The XPS measurement demonstrates the truth that Al insert into the orthorhombic ­Li3VO4. As shown in Fig. 11c, all the fitting peaks of the fully discharged electrodes show a significant shift towards the low binding energy regions. At the same time, the peak area of ­V5+ (516.7 eV) decreases and the peak area of V ­ 4+ 5+ (517.3 eV) increases, indicating that ­V is reduced to ­V4+ during the discharge process. After the cathode is initially charged to 0.95 V, all of these fitted peaks move back to their original peak positions, and the peak area ratio of V ­ 5+ 4+ to ­V is close to the original electrode, indicating that V ­ 4+ is oxidized to V ­ 5+ in charge process. On the basis of the above results, it can be concluded that ­Li3VO4@C can well maintain the structural stability during the process of aluminum ion intercalation/de-intercalation to the framework. CuHCF with Prussian blue open structure, polymer and ­Li3VO4@C are new exploring in cathode material field of AIBs, which achieve high cycle life and excellent Coulombic efficiency but suffer from low capacities and low average cell voltages. But increased understanding of the multivalent ion insertion/extraction reactions may provide more informed for design of active electrode materials with increased capacity or more desirable voltage profiles. Table 3 is a summary of the mentioned cathode materials.

3 Conclusions AIBs have received wide attention due to its low cost, high security, excellent reversibility and so on, which is seen as a substitute for LIB. In this paper, we mainly summarize the cathode materials for AIBs, which include graphite, vanadium oxide, titanium oxide, molybdenum oxide, metal sulfide, polymers and Prussian blue. The carboneaous materials exhibit a high operating voltage and long cycle life, and its open structure with high mechanical strength enhanced ion transfer properties. Vanadium oxide can provide a relatively high discharge capacity due to the intercalation of ­Al3+. Transition metal sulfides display a weak electrostatic adsorption effect with Al cations, which can reinforce the redistribution of the charge of the Al cation in the crystal. The composite cathode materials, such as ­V2O5/C, CuS/C, are a new trend for the development of AIBs. But those cathode materials still suffer from plenty of problems, such as low discharge voltage and battery capacity (about a half of LIB), structural decomposition and volume expansion due to the embedding of large size intercalation ions. The development of the AIBs are still in primary stage, but it has pointed clear directions for the future research. The future work may focus on exploring new cathode materials and cheaper electrolyte to improve working voltage and energy density [116]. At present, the AIBs seem unlikely to be used in the field of smart phones and laptops that require high energy density. However, AIBs are very suitable for the electricity grid which needs fast charge transfer and high power to balance the electricity supply. In the long run, if the electrochemical

13

14368

Journal of Materials Science: Materials in Electronics (2018) 29:14353–14370

Table 3  Electrochemical performances of AIBs based on various cathodes Cathode material

Electrolyte (molar ratio)

Current collector

Initial capacity (mAh g−1)

Last capacity (mAh g−1)

Current density (mA g−1)

Discharge Voltage (V)

Cycle no.

References

CFx

0.5:1 ­AlCl3:[BIm]Br

225

225

60

0.65

40

[54]

Carbon paper

1.3:1 ­AlCl3:[EMIm]Cl

50

62

100

1.8

50

[55]

3D-graphitic foam

1.3:1 ­AlCl3:[EMIm]Cl

60

60

4000

2

7500

[56]

Pyrolytic graphite

1.3:1 ­AlCl3:[EMIm]Cl

60

60

66

2

200

[56]

Natural graphite flakes

1.3:1 ­AlCl3:[EMIm]Cl

60

60

660

2.25–2.0 1.9–1.5

6000

[57]

Free-standing graphene foam

1.3:1 ­AlCl3:[EMIm]Cl

154

151

500

~ 1.9–1.5

100

[58]

Ni bar

V2O5 nano-wires

1.1:1 ­AlCl3:[EMIm]Cl

Stainless steel

305

273

125

0.5

20

[61]

Binder-Free ­V2O5

1.1:1 ­AlCl3:[EMIm]Cl

Ni foam

239

180

44.2

0.6

5

[63]

V2O5 (PVDF Binder)

1.1:1 ­AlCl3:[EMIm]Cl

Ni foam

46

30

44.2

0.6

5

[63]

V2O5 (PTFE Binder)

1.1:1 ­AlCl3:[EMIm]Cl

Ni foam

86.5

68

44.2

0.6

5

[63]

V2O5/C

1:10:5 ­AlCl3 dipropylsulfone:toluene

Mo

150

~ 75

0.05C

< 1.0

30

[66]

VO2

1:1 ­AlCl3:[EMIm] Cl with 0.5 wt% ­C14H14OS

Stainless steel

165

116

50

0.5

100

[68]

Mo rods

100

[75]

13

[71] [82]

MoO2

AlCl3 / [EMIm]Cl

TiO2-NTAs

1M ­AlCl3 aqueous solution

Mo6S8

1.5:1 ­AlCl3:[BMIm]Cl

G-SnS2

1.3:1 ­AlCl3:[EMIm]Cl

NiS nanobelts

90

25

100

1.95\1.0

50

75

4 mA cm−2

0.98

Carbon paper

148

70

12

0.55

50

Mo

275

70

200

~ 0.6

100

[83]

1.3:1 AlCl3:[EMIm]Cl

Ta

104.7

104.4

200

1.15

100

[89]

Ni3S2@graphene

1.3:1 ­AlCl3:[EMIm]Cl

Ta

235

50

200

~ 1.0

300

[90]

CuS@C

1.3:1 ­AlCl3:[EMIm]Cl

Ta

240

90

20

~ 1.0

100

[92]

CuHCF

50:1 Al(OTF)3:diglyme

Carbon cloth

60

10

0.1C

~ 0.2–0.5

10

[99]

CuHCF

0.5 M aqueous Al2(SO4)3

Ti mesh

41

23

400

~ 0.55

1000

[100]

Polypyrrole

1.5:1 ­AlCl3:EMIC

Glassy carbon

71

49

20

~ 0.8-2

100

[107]

Polythiophene

1.5:1 ­AlCl3:EMIC

Glassy carbon

89

71.5

16

~ 0.8-2

100

[107]

Li3VO4@C

1.3:1 ­AlCl3:[EMIm]Cl

Stainless steel

137

48

20

~ 0.5

100

[108]

performances of the AIBs can be greatly improved, it will play a significant role in the future production and life. Acknowledgements  Thanks the National Natural Science Foundation of China (Grant Nos. 51672139, 51472127 and 51272144) for financial support.

References 1. M.V. Rocco, A.D. Lucchio, E. Colombo, Appl. Energy 194, 832–844 (2017) 2. Z. Pan, Q. Guo, H. Sun, Appl. Energy 192, 395–407 (2017) 3. C.E. Kinsella, S.M. O’Shaughnessy, M.J. Deasy, M. Duffy, A.J. Robinson, Appl. Energy 114, 80–90 (2014) 4. B. Ji, F. Zhang, X. Song, Y. Tang, Adv. Mater. 29, (2017). https​ ://doi.org/10.1002/adma.20170​0519 5. Z. Yue, T. Gupta, F. Wang, C. Li, R. Kumar, Z. Yang, N. Koratkar, Carbon 127, 588–595 (2018) 6. Y. Tian, D. Li, J. Tian, B. Xia, Electrochim. Acta 225, 225–234 (2017)

13

7. Y. Zhou, P. Luckow, S.J. Smith, L. Clarke, Environ. Sci. Technol. 46, 7857–7864 (2012) 8. J. Hu, J. Zhang, H. Li, Y. Chen, C. Wang, J. Power Sources 351, 192–199 (2017) 9. L. Jiao, Z. Liu, Z. Sun, T. Wu, Y. Gao, H. Li, F. Li, L. Niu, Electrochim. Acta 259, 48–55 (2018) 10. L. Luo, H. Yang, P. Yan, J.J. Travis, Y. Lee, N. Liu, D.M. Piper, S.H. Lee, P. Zhao, S.M. George, J.G. Zhang, Y. Cui, S. Zhang, C. Ban, C.M. Wang, ACS Nano 9, 5559–5566 (2015) 11. Y. Zheng, M. Ouyang, X. Han, L. Lu, J. Li, J. Power Sources 377, 161–188 (2018) 12. M.A. Xavier, M.S. Trimboli, J. Power Sources 285, 374–384 (2015) 13. A.M. Bizeray, S. Zhao, S.R. Duncan, D.A. Howey, J. Power Sources 296, 400–412 (2015) 14. L.F. Cui, Y. Yang, C.M. Hsu, Y. Cui, Nano Lett. 9, 3370–3374 (2009) 15. D.J. Noelle, M. Wang, A.V. Le, Y. Shi, Y. Qiao, Appl. Energy 212, 796–808 (2018) 16. Z. Chen, Y. Ren, A.N. Jasen, C. Lin, W. Weng, K. Amine, Nat. Commun. 4, 1513 (2013) 17. F. Cheng, J. Liang, Z. Tao, J. Chen, Adv. Mater. 23, 1695–1715 (2011)

Journal of Materials Science: Materials in Electronics (2018) 29:14353–14370 18. S.W. Baek, I. Honma, J. Kim, D. Rangappa, J. Power Sources 343, 22–29 (2017) 19. F.M. Donahue, S.E. Mancini, L. Simonsen, J. Appl. Electrochem. 22, 230–234 (1992) 20. J. Song, M. Noked, E. Gillette, J. Duay, G. Rubloff, S.B. Lee, Phys ChemChem Phys. 7, 5256–5264 (2015) 21. C. Drosos, C. Jia, S. Mathew, R.G. Palgrave, B. Moss, A. Kafizas, D. Vernardou, J. Power Sources 384, 355–359 (2018) 22. Y.H. Tan, W.T. Yao, T. Zhang, T. Ma, L.L. Lu, F. Zhou, H.B. Yao, S.H. Yu, ACS Nano. (2018). https​://doi.org/10.1021/acsna​ no.8b018​47 23. I. Hasa, R. Verrelli, J. Hassoun, Electrochim. Acta 173, 613–618 (2015) 24. T. Zhou, W.K. Pang, C. Zhang, J. Yang, Z. Chen, H.K. Liu, Z. Guo, ACS Nano 8, 8323–8333 (2014) 25. P. Lu, Y. Sun, H. Xiang, X. Liang, Y. Yu, Adv. Energy Mater. (2017). https​://doi.org/10.1002/aenm.20170​2434 26. P. Padigi, G. Goncher, D. Evans, R. Solanki, J. Power Sources 273, 460–464 (2015) 27. M. Cabello, F. Nacimiento, J.R. González, G. Ortiz, R. Alcántara, P. Lavela, C.P. Vicente, J.L. Tirado, J. Elecom. 67, 59–64 (2016) 28. X.G. Sun, Z. Bi, H. Liu, Y. Fang, C.A. Bridges, M.P. Paranthaman, S. Daiab, Chem. Commun. 52, 1713–1716 (2016) 29. Y. Hu, D. Ye, B. Luo, H. Hu, X. Zhu, S. Wang, L. Li, S. Peng, L. Wang, Adv. Mater. (2017). https​://doi.org/10.1002/adma.20170​ 3824 30. M. Hulot, Compt. Rend. 1855, 40, 148 31. D. Tommasi, Traite des Piles Electriques (Georges Carre, Paries, 1889), p. 131 32. C.H. Brown, US Patent 503, 1893, 567 33. G.W. Heise, E.A. Schumacher, N.C. Cahoon, J. Electrochem. Soc. 94, 99–105 (1948) 34. D.E. Sargent, US Patent 2, 1951, 447 35. S. Zaromb, J. Electrochem. Soc. 109, 1125–1130 (1962) 36. L. Trevethan, D. Bockstie, S. Zaromb, J. Electrochem. Soc. 110, 267–271 (1963) 37. Z.A. Zafar, S. Imtiaz, R. Razaq, S. Ji, T. Huang, Z. Zhang, Y. Huang, J.A. Anderson, J. Mater. Chem. A. 5, 5646–5660 (2017) 38. J. Li, J. Tu, H. Jiao, C. Wang, S. Jiao, J. Electrochem. Soc. 164, 3093–3100 (2017) 39. S.K. Das, S. Mahapatra, H. Lahan, J. Mater. Chem. A. 5, 6347– 6367 (2017) 40. Y. Gao, C. Zhu, Z.Z. Chen, G. Lu, J. Phys. Chem. C 121, 7131– 7138 (2017) 41. M.M. Huie, D.C. Bock, E.S. Takeuchi, A.C. Marschilok, K.J. Takeuchi, Coord. Chem. Rev. 287, 15–27 (2015) 42. A.S. Childres, P. Parajuli, J. Zhu, R. Podila, A.M. Rao, Nano Energy 39, 69–76 (2017) 43. M. Kazazi, P. Abdollahi, M. Mirzaei-Moghadam, Solid State Ionics 300, 32–37 (2017) 44. S.C. Jung, Y.J. Kang, D.J. Yoo, J.W. Choi, Y.K. Han, J. Phys. Chem. C 120, 13384–13389 (2016) 45. H. Chen, F. Guo, Y. Liu, T. Huang, B. Zheng, N. Ananth, Z. Xu, W. Gao, C. Gao, Adv. Mater. 29, 1605958 (2017) 46. M.P. Paranthaman, G.M. Brown, X. Sun, J. Nanda, A. Manthiram, A. Manivannan, 218th ECS Meeting. 2010, Abstract number 314 47. M.L. Agiorgousis, Y.Y. Sun, S.B. Zhang, ACS Energy Lett. 2, 689–693 (2017) 48. N.S. Pearre, L.G. Swan, Appl. Energy 137, 501–510 (2015) 49. U.K. Debnath, I. Ahmad, D. Habibi, A.Y. Saber, Int. J. Electr. Power Energy Syst. 64, 1017–1024 (2015) 50. D. Larcher, J. Tarascon, Nat Chem. 7, 19–29 (2014) 51. H. Zheng, R. Yang, G. Liu, X. Song, V.S. Battaglia, J. Phys. Chem. 116, 4875–4882 (2017)

14369 52. H. Yu, X. Dong, Y. Pang, Y. Wang, Y. Xia, Electrochim. Acta 228, 251–258 (2017) 53. P.R. Gifford, J.B. Palmisano, J. Electrochem. Soc. 135, 650–654 (1988) 54. J.V. Rani, V. Kanakaiah, T. Dadmal, M.S. Rao, S. Bhavanarushi, J. Electrochem. Soc. 160, A1781–A1784 (2013) 55. H. Sun, W. Wang, Z. Yu, Y. Yuan, S. Wang, S. Jiao, Chem. Commun. 51, 11892–11895 (2015) 56. M.C. Lin, M. Gong, B. Lu, Y. Wu, D.Y. Wang, M. Guan, M. Angell, C. Chen, J. Yang, B.J. Hwang, H. Dai, Nature 520, 324– 328 (2015) 57. D.Y. Wang, C.Y. Wei, M.C. Lin, C.J. Pan, H.L. Chou, H.A. Chen, M. Gong, Y. Wu, C. Yuan, M. Angell, Y.J. Hsieh, Y.H. Chen, C.Y. Wen, C.W. Chen, B.J. Hwang, C.C. Chen, H. Dai, Nat. Commun. 8, 14283 (2017) 58. X. Huang, Y. Liu, H. Zhang, J. Zhang, O. Noonan, C. Yu, J. Mater. Chem. A 5, 19416–19421 (2017) 59. P. Bhauriyal, A. Mahata, B. Pathak, Phys ChemChem Phys. 19, 7980–7989 (2017) 60. D.B. Le, S. Passerini, F. Coustier, J. Guo, T. Soderstrom, B.B. Owens, W. H. Smyrl, Chem. Mater. 10, 682–684 (1998) 61. N. Jayaprakash, S.K. Das, L.A. Archer, Chem. Commun. 47, 12610–12612 (2011) 62. L.D. Reed, E. Menke, J. Electrochem. Soc. 160, A915–A917 (2013) 63. H. Wang, Y. Bai, S. Chen, X. Luo, C. Wu, F.Wu, J. Lu, K. Amine, ACS Appl. Mater. Interfaces 7, 80–84 (2015) 64. T. Katsunori, Y. Wang, G. Cao, J. Phys. Chem. B 109, 48–51 (2005) 65. R.R. Salunkhe, B.P. Bastakoti, C.T. Hsu, N. Suzuki, J.H. Kim, S.X. Dou, C.C. Hu, Y. Yamauchi, Chemistry 20, 3084–3088 (2014) 66. M. Chiku, H. Takeda, S. Matsumura, E. Higuchi, H. Inoue, ACS Appl. Mater. Interfaces 7, 24385–24389 (2015) 67. S. Gu, H. Wang, C. Wu, Y. Bai, H. Li, F. Wu, Energy Storage Mater. 6, 9–17 (2017) 68. W. Wang, B. Jiang, W. Xiong, H. Sun, Z. Lin, L. Hu, J. Tu, J. Hou, H. Zhu, S. Jiao, Sci. Rep. 3, 3383 (2013) 69. H. Liu, G. Xu, J. Wang, J. Lv, Z. Zheng, W. Wu, Electrochim. Acta 130, 213–221 (2014) 70. P.C. Lansåker, J. Backholm, G.A. Niklasson, C.G. Granqvist, Thin Solid Films 518, 1225–1229 (2009) 71. S. Liu, J.J. Hu, N.F. Yan, G.L. Pan, G.R. Li, X.P. Gao, Energy Environ. Sci. 5, 9743–9746 (2012) 72. Y. Liu, S. Sang, Q. Wu, Z. Lu, K. Liu, H. Liu, Electrochim. Acta 143, 340–346 (2014) 73. W. Zhong, S. Sang, Y. Liu, Q. Wu, K. Liu, H. Liu, J. Power Sources 294, 216–222 (2015) 74. T. Koketsu, J. Ma, B.J. Morgan, M. Body, C. Legein, W. Dachraoui, M. Giannini, A. Demortière, M. Salanne, F. Dardoize, H. Groult, O.J. Borkiewicz, K.W. Chapman, P. Strasser, D. Dambournet, Nat Mater. 16, 1142–1448 (2017) 75. J. Wei, W. Chen, D. Chen, K. Yang, J. Electrochem. Soc. 164, A2304–A2309 (2017) 76. K. Suto, A. Nakata, H. Murayama, T. Hirai, J.I. Yamaki, Z. Ogumi, J. Electrochem. Soc. 163, A742–A747 (2016) 77. N. Koura, J. Electrochem. Soc. 127, 1529–1531 (1980) 78. N. Takami, N. Koura, Electrochim. Acta 33, 69–74 (1988) 79. N. Koura, T. Yui, S. Takahashi, J. Jpn. Inst. Light Met. 35, 203– 208 (1985) 80. T. Mori, Y. Orikasa, K. Nakanishi, K. Chen, M. Hattori, T. Ohta, Y. Uchimoto, J. Power Sources 313, 9–14 (2016) 81. D. Aurbach, Z. Lu, A. Schechter, Y. Gofer, H. Gizbar, R. Turgeman, Y. Cohen, M. Moshkovich, E. Levi, Nature 407, 724–727 (2000)

13

14370 82. L. Geng, G. Lv, X. Xing, J. Guo, Chem. Mater. 27, 4926–4929 (2015) 83. Y. Hu, B. Luo, D. Ye, X. Zhu, M. Lyu, L. Wang, Adv. Mater. (2017). https​://doi.org/10.1002/adma.20160​6132 84. K. Liang, K. Marcus, S. Zhang, L. Zhou, Y. Li, S.T.D. Oliveira, N. Orlovskaya, Y.H. Sohn, Y. Yang, Adv. Energy Mater. (2017). https​://doi.org/10.1002/aenm.20170​1309 85. S.C. Han, K.W. Kim, H.J. Ahn, J.H. Ahn, J.Y. Lee, J. Alloys Compd. 361, 247–251 (2003) 86. T.S. Sonia, P. Anjali, S. Roshny, V. Lakshmi, R. Ranjusha, K.R.V. Subramanian, S.V. Nair, A. Balakrishnan, Ceram. Int. 40, 8351–8356 (2014) 87. L. Mi, Y. Chen, W. Wei, W. Chen, H. Hou, Z. Zheng, RSC Adv. 3, 17431–17439 (2013) 88. D. Han, N. Xiao, B. Liu, G. Song, J. Ding, Mater. Lett 196, 119–122 (2017) 89. Z. Yu, Z. Kang, Z. Hu, J. Lu, Z. Zhou, S. Jiao, Chem. Commun. 52, 10427–10430 (2016) 90. S. Wang, Z. Yu, J. Tu, J. Wang, D. Tian, W. Liu, S. Jiao, Adv. Energy Mater. (2016). https​://doi.org/10.1002/aenm.20160​0137 91. J.B. Parise, Acta Cryst. 36, 1179 – 1180 (1980) 92. S. Wang, S. Jiao, J. Wang, H.S. Chen, D. Tian, H. Lei, D.N. Fang, ACS Nano 11, 469–477 (2017) 93. Z. Chang, Y.Q. Yang, M.X. Li, X.W. Wang, Y.P. Wu, J. Mater. Chem. A 2, 10739–10755 (2014) 94. C.D. Wessells, R.A. Huggins, Y. Cui, Nat. Commun. 2, 550 (2011) 95. C.D. Wessells, S.V. Peddada, R.A. Huggins, Y. Cui, Nano Lett. 11, 5421–5425 (2011) 96. M. Pasta, C.D. Wessells, R.A. Huggins, Y. Cui, Nat. Commun. 3, 1149 (2012) 97. C.D. Wessells, S.V. Peddada, M.T. McDowell, R.A. Huggins, Y. Cui, J. Electrochem. Soc. 159, A98–A103 (2012) 98. R.Y. Wang, C.D. Wessells, R.A. Huggins, Y. Cui, Nano Lett. 13, 5748–5752 (2013) 99. S. Liu, G.L. Pan, G.R. Li, X.P. Gao, J. Mater. Chem. A. 3, 959– 962 (2015)

13

Journal of Materials Science: Materials in Electronics (2018) 29:14353–14370 100. L.D. Reed, S.N. Ortiz, M. Xiong, E.J. Menke, Chem. Commun. 51, 14397–14400 (2015) 101. P. Novák, K. Müller, K.S.V. Santhanam, O. Haas, Chem. Rev. 97, 207 – 281 (1997) 102. J. Hassoun, S. Panero, P. Reale, B. Scrosati, Adv. Mater. 21, 4807–4810 (2009) 103. F. Croce, M.L. Focarete, J. Hassoun, I. Meschini, B. Scrosati, Energy Environ. Sci. 4, 921–927 (2011) 104. T. Li, C. Li, X. Hu, X. Lou, H. Hu, L. Pan, Q. Chen, M. Shen, B. Hu, Rsc Adv. 6, 61319–61324 (2016) 105. Z. Zhong, Q. Cao, B. Jing, X. Wang, X. Li, H. Deng, Mater. Sci. Eng. B 177, 86–91 (2012) 106. K. Deng, S. Wang, S. Ren, D. Han, M. Xiao, Y. Meng, J. Power Sources 360, 98–105 (2017) 107. N.S. Hudak, J. Phys. Chem. C 118, 5203 – 5215 (2014) 108. J. Jiang, H. Li, J. Huang, K. Li, J. Zeng, Y. Yang, J. Li, Y. Wang, J. Wang, J. Zhao, ACS Appl. Mater. Interfaces 9, 28486–28494 (2017) 109. H. Gu, Y. Zhang, M. Huang, F. Chen, Z. Yang, X. Fang, S. Li, W. Wang, S. Yang, M. Li, Inorg. Chem. 56, 7657–7667 (2017) 110. C. Zhang, C. Liu, X. Nan, H. Song, Y. Liu, C. Zhang, G. Cao, ACS Appl. Mat. Interfaces 8, 680–688 (2016) 111. L. Shen, H. Lv, S. Chen, P. Kopold, P.A.V. Aken, X. Wu, J. Maier, Y. Yu, Adv. Mater. (2017). https​://doi.org/10.1002/ adma.20170​0142 112. Y. Yang, J. Li, J. Huang, J. Zeng, J. Zhao, Electrochim. Acta 247, 771–778 (2017) 113. C. Zhang, H. Song, C. Liu, Y. Liu, C. Zhang, X. Nan, G. Cao, Adv. Funct. Mater. 25, 3497–3504 (2015) 114. N.P. Stadie, S. Wang, K.V. Kravchyk, M.V. Kovalenko, ACS Nano 11, 1911 – 1919 (2017) 115. Y. Yang, J. Li, X. He, J. Wang, D. Sun, J. Zhao, J. Mater. Chem. A 4, 7165–7168 (2016) 116. X. Zhang, S. Wang, J. Tu, G. Zhang, S. Li, D. Tian, S. Jiao, ChemSusChem 11, 709–715 (2018)