Three dimensional metal oxides–graphene composites and their applications in lithium ion batteries

Abstract

Because of the huge energy and environmental problems caused by the use of fossil fuels, R&D on innovative energy storage systems has never become so important as today. Although there are still some safety, energy, power and cost issues because of which electric vehicles are becoming more and more common in our daily life, such as E-bikes or E-cars. This paper provides an overview of the recent progress on three dimensional (3D) metal oxides–graphene (MOs–G) composites as advanced electrode materials in lithium ion batteries (LIBs). Beginning with a brief description of the importance and preparation methods of 3D MOs–G composites, the effects of the morphology and size of metal oxides (MOs) or graphene on composites for LIBs are then systematically reviewed and discussed. Additionally, important effects of composition and interactions between metal oxides and graphene are also pointed out. Finally, the future challenges of MOs–G composites for lithium ion batteries are discussed.

---

Jiantao Zai

Dr Jiantao Zai received his Bachelor (2007) and Ph.D. (2012) degrees in Applied Chemistry from Shanghai Jiao Tong University, and then worked with Prof. Donghai Wang in Penn State University. He joined SJTU as a Lecturer in 2014. His research interests include the synthetic chemistry of Group IV inorganic materials and their potential applications in energy storage, photovoltaic and photocatalytic areas. He is honored to receive the 2014 Shanghai Chenguang project, 2012 Shanghai outstanding graduates and 2012 Scholarship of Sinopoly Battery Ltd.

Xuefeng Qian

Prof. Xuefeng Qian worked in the School of Chemistry and Chemical Engineering, Shanghai Jiao Tong University. He received his Bachelor degree from HeFei University of Technology (1990), Master (1995) and Ph.D (1998) degrees from USTC. He is interested in the design, synthesis, and applications of materials related with solar utilizations and energy storage. He was awarded the 2008 Outstanding University Young Teachers in Shanghai, 2006 New Century Excellent Talents of Education Ministry of China.

---

1 Introduction

Reports from the Chinese Academy of Social Sciences indicate that there are more than 130 million vehicles in China in 2013, and the number is expected to be 400 million (one car per family) in the future.¹ These vehicles have lead to massive traffic jams and are considered as one of the main reasons of the poor air quality in China. It is hard to image our future life if all of the increasing 270 million vehicles are based on fossil fuels. These problems are not only in China but also throughout the world. It is of no doubt that the development of electric vehicles (EVs) is one of the good choices. EVs have become more and more common in our daily life, for instance, E-bikes are widely used in both the city and countryside in China. However, the popularization of EVs is always limited by the safety, cost, cruising ability and power. To overcome these issues, people have started to expect better electrochemical performances from lithium ion batteries (LIBs).^2–8 However, the development of better safety, higher energies and power densities in LIBs is meeting some technical bottlenecks at the present stage.

Graphene is a two-dimensional (2D) sheet of sp² bonded carbon atoms in a hexagonal honeycomb lattice, which can be viewed as an extra-large polycyclic aromatic molecule. Graphene nanosheets (GNSs) have been known to be composed of traditional carbon materials (e.g. graphite) or components of a “new” class of carbon materials (such as carbon nanotubes). The discovery of monolayer graphene can be tracked to the 1960s and 1970s.⁹ Then, graphene with the size of only tens of nm on an appropriate substrate (crystal surfaces of transition metals and metal carbides) was successfully fabricated by Oshima and Nagashima in 1997 through decomposing hydrocarbon gases at high temperature.¹⁰ Ruoff, R. S. also did pioneer works on tailoring graphite with the goal of achieving single sheets.¹¹ After being first transferred to a SiO₂ substrate by Geim and Novoselov in 2004, the field effect of graphene was demonstrated.^10–12 The ideal GNSs achieved by a mechanical exfoliation technique have proven to be highly ordered, and have outstanding surface areas (2630 m² g⁻¹), high Young's modulus (1 TPa), high thermal conductivity (5000 W mK⁻¹), strong chemical durability and high electron mobility (2.5 × 10⁵ cm² V⁻¹ s⁻¹).^13–15

Since the chemical exfoliation method was developed to produce graphene at a low cost and in large quantities,¹⁶ graphene has been widely applied in polymer composites,¹⁷ transistors,^18,19 optoelectronics,^20,21 memory devices,^22,23 sensors (gas-,^24,25 bio-,²⁶ electrochemical-²⁷ and chemical-²⁸), solar cells,^20,29 field emission devices,^30,31 catalysts,³¹ photocatalysts,^14,32 nanogenerators,³³ hydrogen storage³⁴ and CO₂ capture.³⁵ Especially graphene-based electrochemical storage devices (e.g. high-performance LIBs) have attracted considerable attention in fundamental studies and practical applications with greatly improved electrochemical performances due to its unique 2D structure and excellent physiochemical properties.^3,4,36–44 Furthermore, graphene-based electrochemical storage energy devices do not need high quality graphene without any defects, such as electronics, and graphene produced by the chemical exfoliation method, such as reduced graphite oxide (RGO) with many defects and multiple layers, can also meet the demands of high-performance LIBs.

Detailed descriptions of the properties, synthesis, functionalization and applications in energy storage areas of graphene and its composites can be found in recent papers.^{3–5,17,29,36,38–52} For example, Wu et al. systematically reviewed the pros and cons of graphene and metal oxides, and focused on the synergistic effects of metal oxides-graphene composites on improving the electrochemical properties of LIBs and electrochemical capacitors.⁴⁹ Here, we provide an overview of the recent progress in three-dimensional (3D) metal oxides–graphene (MOs–G) composites as advanced lithium storage materials for high-performance LIBs since 2012, especially focusing on the important effects of morphologies, composition, interactions between metal oxides (MOs) and graphene on the improvement of their electrochemical properties, including capacity, rate capability and cyclic stability.

2 3D metal oxides–graphene composites

2.1 Why use 3D MOs–G composites

Due to its special 2D structure, graphene has a high Li⁺ ion storage capacity, 1116 mA h g⁻¹ based on the LiC₂ model⁵³ or 744 mA h g⁻¹ based on the Li₂C₆ model.⁵⁴ However, graphene based electrodes usually show poor stability for the restacking of graphenes.³⁷ Moreover, MOs with high theoretical reversible capacities (such as 717 mA h g⁻¹ of NiO, 1007 mA h g⁻¹ of Fe₂O₃, 755 mA h g⁻¹ of MnO, 890 mA h g⁻¹ of Co₃O₄) will suffer from huge volume changes during the continuous charge/discharge processes, which would lead to the rapid disintegration of anodes and capacity fading upon cycling. Furthermore, the lower electronic conductivity of MOs is another disadvantage. These problems can be greatly overcome by combining MOs with graphene to form composites with multiple synergistic effects. Wu et al.⁴⁹ summarized these synergistic effects as following: (i) graphene is a novel 2D support for the uniform nucleating, growth or assembling of MOs with a well-defined size, shape and crystallinity; (ii) MOs between the layers of graphene can efficiently suppress the re-stacking of graphene; (iii) graphene can act as a 2D conductive template for building a 3D interconnected conductive porous network to improve the electrical conductivity and charge transport of pure oxides; (iv) graphene can suppress the volume change and particle agglomeration of MOs during the charge–discharge process; (v) oxygen-containing functional groups on graphene ensure good interfacial interactions and electrical contact between graphene and MOs.

3D hierarchical structures in micro or sub-micro sizes, assembled by simple low dimensional nano-sized building blocks, can avoid the aggregation of anode materials and are beneficial for the electrode fabrication process.^55–58 3D hierarchical structures can also provide more sites to connect with a conductive matrix (such as graphene, conductors or current collectors) and maintain the activity of Li storage materials during cycling. In addition, 3D hierarchical structures have additional benefits to greatly improve the electrochemical performance of electrode materials in LIBs, and they also show other special effects:^56–58 the opening porous structures in 3D-hierarchical structures are readily accessible for an electrolyte, facilitating the transportation of Li⁺ ions from a liquid to the active surface of active materials. Second, nanosized building blocks of 3D hierarchical structures can significantly shorten the diffusion distance of Li⁺ ions, and therefore significantly enhance the lithium insertion–extraction kinetics. Third, 3D hierarchical structures with plenty of pores can accelerate phase transitions and restrain the crumbling and cracking of electrodes, leading to a superior cycling performance. In addition, well-connected 3D hierarchical structures with a large surface area can reduce the concentration polarization and facilitate electron transportation, which accounts for the high rate performance.

To date, several structural models of MOs–G composites have been developed, such as anchored, wrapped, encapsulated, sandwich-like, layered and mixed models (Fig. 1).⁴⁹ Because of the large surface area and restacking nature of graphene, most graphene composites are in 3D hierarchical structures. These 3D MOs–G composites can combine the synergistic effects of graphene composites and benefits derived from 3D hierarchical structures together. Thus, they would possess better long-term stability and rate capability than pure 3D hierarchical MOs and graphene composites with simple structure, such as anchored model or sandwich-like model.

Fig. 1 Schematic of structural models of MOs–G composites: (a) anchored model: nanosized MOs particles are anchored on the surface of graphene. (b) Wrapped model: MOs particles are wrapped by graphene. (c) Encapsulated model: MOs particles are encapsulated by graphene. (d) Sandwich-like model: graphene serves as a template for the creation of a MOs/graphene/MOs sandwich-like structure. (e) Layered model: a structure composed of alternating layers of MOs nanoparticles and graphene. (f) Mixed model: graphene and MOs particles are mechanically mixed and graphene forms a conductive network among MOs particles. Red: MOs particles; blue: graphene sheets.⁴⁹ Reproduced from ref. 49, Copyright (2012) with permission from Elsevier.

2.2 How to get 3D MOs–G composites

The fabrication of graphene or graphene oxide is the first stage to prepare 3D MOs–G composites. To date, graphene can be fabricated by various methods, such as chemical vapor deposition (CVD), plasma-enhanced CVD, epitaxial growth on electrically insulating surfaces, electric arc discharge and solution-based chemical oxidation–reduction process.^3,59–61 First, for the aim of industrialization and preparing graphene composites, the fabrication method should be easily scalable for the large-scale production of graphene, and then the obtained graphene or its precursors must be easily processable with other active materials. Thus, the most promising method is the chemical oxidation of graphite, conversion of the resulting graphite oxide to graphene oxide (GO), and the subsequent reduction of GO.^16,62 The method can produce GO and RGO on a large scale and realize the industrialization process. Furthermore, abundant oxygen-containing functional groups, such as carboxyl, hydroxyl, epoxy and keto groups, make GO dispersible in water and organic solvents, interactive with metal ions and various compounds by electrostatic interactions and/or chemical bonds.^{4,5,14,46,49,51,63,64} Furthermore, the structure, electrical conductivity and electrochemical performance of RGO are mainly determined by the reduction methods of GO, including chemical, thermal, electrochemical and photo-irradiation techniques, which have been reviewed well by Kuila et al.⁴⁰

Once the large-scale production of graphene by chemical oxidation of graphite is realized, the physically mixing method is considered to be one of the simplest and most convenient methods to fabricate 3D MOs–G composites. In this method, graphene, usually in RGO, is prepared according to abovementioned methods and dispersed into water or other solvents to form a suspension, and then mixed with the pre-fabricated MOs by ultrasonication or stirring process, and MOs–G composites are obtained followed by flocculation, filtration, centrifugation or freeze–drying process. In the pioneer work of Paek et al.⁶⁵ (Fig. 2), RGO was prepared via the chemical reduction of exfoliated graphite oxide, and SnO₂ nanoparticles were obtained by the controlled hydrolysis of SnCl₄ with NaOH, and then SnO₂/GNS nanocomposites were obtained by reassembling RGO in the presence of SnO₂ nanoparticles.⁶⁵ Since the great developments of MOs nanomaterials with rich compositions and morphologies (0D, 1D, 2D and 3D nanostructures) in the past decades, this method has been widely used to fabricate 3D MOs–G composites. For instance, 3D-hierarchical NiO–GNSs composites were prepared by simply mixing 3D-hierarchical NiO carnations with RGO under ultrasonication.⁶⁶ Recently, graphene nanocomposites based on bi-metal oxides (e.g. MnFe₂O₄ (ref. 67), NiFe₂O₄ (ref. 68), CoFe₂O₄ (ref. 69)) were also obtained by similar methods. On the other hand, the milling method is also a facile industrialized physical mixing method to prepare MOs–G composites.^70,71

Fig. 2 (A) Schematic illustration for the synthesis and the structure of SnO₂/GNS; (B) TEM image of SnO₂/GNS, the white arrows denote the GNSs; (C) charge/discharge profile for SnO₂/GNS; (D) cyclic performances for (a) bare SnO₂ nanoparticle, (b) graphite, (c) GNS, and (d) SnO₂/GNS.⁶⁵ Reproduced with permission from ref. 65, Copyright (2009) American Chemical Society.

Due to the abundant oxygen-containing functional groups, GO and RGO are usually negatively charged, and they can easily form composites with positively charged MOs by a co-assembly process via electrostatic interactions. However, the pre-fabricated MOs are always negatively charged or neutral (e.g. Co₃O₄ and Fe₃O₄ nanoparticles are always terminated by OH functional groups^74,75), and thus MOs are usually modified by grafting method. For example, Feng et al. grafted aminopropyltrimethoxysilane (APS) to OH terminated Co₃O₄ or Fe₃O₄ nanoparticles to render positively charged oxides in an acidic solution.⁷⁴ In addition to APS, various positively charged compounds, such as poly dimethyl diallyl ammonium chloride (PDDA), (3-aminopropyl) trimethoxysilane (APTMS) and poly(allylamine hydrochloride) (PAH), were also used.^76–79 However, the method is not suitable to metal hydroxides due to their dissolution in acid solution. Zhang et al. described a general strategy to fabricate graphene coated large-area Co(OH)₂ heterostructures by assembling the positively charged hydroxide nanosheets and negatively charged functionalized graphene in a nearly neutral solution (Fig. 3A and B).⁷² Then, they developed a modified method to prepare binder free and mechanically robust CoO/graphene electrodes (Fig. 3C and D). The negatively charged RGO or GO can also be positively charged by surface grafting, such as amine-functionalized graphene.⁸⁰ 3D MOs–G composites fabricated via electrostatic interactions are always in encapsulated and layered models. The large surface area of graphene and strong intermolecular forces between MOs and graphene can make MOs nanoparticles disperse well on graphene and prevent their aggregation, which are also beneficial for the stability and high electrical conductivity of composites. Furthermore, graphene is usually in the microscale, which can ensure that the self-assembled MOs nanostructures are effectively covered or supported by graphene, and the porous nature, generated by the co-assembly process, can facilitate ion diffusion and accommodate the volume change during the cycle processes.

Fig. 3 (A) Schematic diagram of the fabrication of sandwich structured GC–Co(OH)₂ heterostructures driven by the mutual electrostatic interactions between the two species; (B) cyclic performances for GC–Co(OH)₂ and pure Co(OH)₂ (ref. 72). Reproduced from ref. 72 with permission from The Royal Society of Chemistry. (C) Schematic diagram of the fabrication of CoO/graphene hybrid on Cu foil; (D) reversible Li extraction capacity of CoO/graphene hybrid electrode at 1 A g⁻¹ for 5000 cycles.⁷³ Reproduced with permission from ref. 73, Copyright (2013) WILEY-VCH.

Metal ions from inorganic and/or organic metal salts can also be absorbed on the surface of negatively charged RGO or GO via electrostatic interactions or coordination bonds in RGO or GO suspensions. Starting from the metal ion–GO/RGO dispersions, wet-chemistry strategies, such as in situ chemical deposition, sol–gel processes and hydro-/solvothermal synthesis, are widely used to fabricate a broad range of MOs–G composites. In these strategies, suspended RGO/GO acts as a 2D precursor to form an integrated support network for discrete metal ions, and then composites are formed by hydrolysis or in situ redox reactions to anchor MOs on the surface of RGO, and further followed by various reducing and annealing processes. Special emphasis is that graphene can suppress the agglomeration of MOs nanoparticles during the preparation process.⁴⁹ Researches indicated that Co₃O₄ nanoparticles with a size of 5–20 nm were homogeneously anchored on the surface of graphene, while sub-microparticles were obtained without graphene (Fig. 4).⁸¹ Furthermore, the presence of graphene can also affect the morphology of MOs. Our research indicated 1D Co₃O₄ nanorods were generated instead of aggregated nanoparticles when GO was added to an alcohol–water mixed solvent.⁸² At present, wet chemistry strategies are widely used to fabricate 3D MOs–G composites, which provide simple and practical methods to obtain a uniform distribution of MOs nanostructures anchored on graphene with controlled size, morphology and crystallinity.⁴⁹

Fig. 4 (A) Schematic representation of the fabrication process of a Co₃O₄/graphene composite. (B and C) TEM and HRTEM images of Co₃O₄/graphene composite; the inset in (B) is the SAED pattern of Co₃O₄ NPs with [110] plane in the Co₃O₄/graphene composite, indicative of the well-textured and single-crystalline nature of Co₃O₄ NPs. (D) SEM image of the as-prepared Co₃O₄, which shows that only micro-sized Co₃O₄ particles can be formed without the presence of graphene sheets. (E) Cycling performance for graphene, Co₃O₄, and the Co₃O₄/graphene composite.⁸¹ Reproduced with permission from ref. 81, Copyright (2010) American Chemical Society.

Surfactants are commonly used in the fabrication of stable graphene dispersions and synthesizing MOs nanostructures.^62,83–85 Anionic sulphate could help the thermal reduced graphene dispersion in aqueous solution and facilitate the self-assembly of the in situ grown nanocrystalline TiO₂.⁸⁶ The method is further developed to fabricate ordered mesoporous MOs–G nanocomposites, such as SnO₂, NiO or MnO₂–graphene composites.⁸⁷ In addition to the stabilization of graphene in aqueous/solvent solutions, surfactants can also control the size and morphology of MOs in the wet chemistry strategy. Co₃O₄–graphene nanocomposites with high loading and highly dispersed Co₃O₄ nanoparticles were fabricated by the co-assembly of polyvinylpyrrolidone (PVP) protected precursors and GO, while Co₃O₄ nanoparticles would agglomerate without a surfactant.⁸⁸ With the help of the in situ formed dehydroascorbic acid, oxidized product of L-ascorbic acid, GNSs decorated with ultra-small Fe₃O₄ nanoparticles were synthesized from an Fe³⁺–GO suspension and L-ascorbic acid via the hydrothermal method.⁸⁹ SnO₂ nanorods/graphene nanocomposites have been synthesized through a simple ultrasonic combined hydrothermal process with the assistance of mercaptoacetic acid.⁹⁰ Moreover, the approach can be extended to produce other MOs nanostructures on the surface of graphene, such as mesoporous spheres,^91–94 nanospindles,^95,96 nanorods,^97–99 nanowires,^100,101 nanotubes,¹⁰² nanoplates,¹⁰³ nanobelts,^104,105 nanosheets^106–108 and 3D hierarchical structures.^109,110

In wet chemistry strategies, the reduction of GO to RGO or graphene can be carried by many methods, such as the metal ion (Fe²⁺, Sn²⁺) in situ reduction process,^111–115 chemical reduction process,^89,116–120 hydro-/solvothermal reduction,^{90,101,116,121–126} and thermal reduction.^127–131 Post thermal reduction under an inert atmosphere is always introduced to increase the electronic conductivity of composites and further enhance the rate capability of LIBs. Furthermore, proper heat treatment can cause graphene crosslinking and wrapping, resulting in a more stable composite structure and leading to good cyclic stability. Based on the wet chemistry strategy, several novel and effective processes have been developed. Considering metal ions–GO/RGO suspension as a sol, MnO–¹³² and Fe₃O₄ (ref. 133)–graphene composites have been prepared by the modified sol–gel method via solvent evaporation and thermal reduction in sequence. Electrostatic spray deposition,¹³⁴ electrostatic induced spread⁷³ and electrospun technology^135,136 were also introduced to synthesize MOs–graphene composites. Binder free electrodes synthesized by these methods always showed good cyclic stability and rate capability because of the well dispersed/wrapped MOs nanomaterials and interconnected high electronic conductive graphene 3D network (Fig. 5).

Fig. 5 Illustration of the synthetic process (A), TEM (B), cyclic (C) and rate (D) performances of SnO₂@G NFs.¹³⁵ Reproduced from ref. 135, Copyright (2014) with permission from Elsevier.

The microwave heating technique was successfully applied to synthesize nanomaterials in past years. Magnetite/graphene¹³⁷ or Co₃O₄–graphene sheet-on-sheet nanocomposites¹³⁸ have also been prepared by this method. However, the obtained composites still needed a post thermal reduction process. Pinna et al. invented a one-pot non-aqueous synthesis method of crystalline SnO₂- and Fe₃O₄-based graphene heterostructures in 5–10 minutes by combining microwave heating and the ‘benzyl alcohol route’ together, which allowed the selective growth of MOs nanoparticles on the surface of GO. To date, microwave based approaches have been utilized to synthesize Fe₂O₃–RGO composites,^139,140 sheet-like and/or fusiform CuO nanostructures grown on graphene, SnO₂–RGO composites,^141,142 Mn₃O₄–G composites,¹⁴³ Cu₂O@Cu–graphene composites,¹⁴⁴ MoO₃/graphene film¹²³ and Zn₂GeO₄/N-doped graphene nanocomposites.⁹⁷

Other special technologies have also been developed to prepare MOs–G composites, such as photocatalytic synthesis,¹⁴⁶ coelectrodeposition,¹⁴⁷ atomic layer deposition,¹⁴⁸ supercritical alcohols/CO₂ (ref. 149–151) and template methods.¹⁴⁵ Especially, the template method is an effective process to generate 3D MOs on graphene (Fig. 6). No matter what is the method, the key points in the synthesis of 3D MOs–G composites are (1) generating well dispersed and tightly fixed MOs on graphene, (2) enhancing the electron conductivity of graphene, (3) producing a porous structure to make electrolyte accessible and facilitate the Li⁺ diffusion, (4) low cost and environmental friendly.

Fig. 6 Schematic illustration of sequential steps for the synthesis of graphene/CuO. GO/CaCO₃ prepared by applying CO₂ gas to Ca²⁺ and GO suspension. Graphene/Cu₂Cl(OH)₃ was formed by the chemical reduction of GO to graphene and transformation of CaCO₃ into Cu₂Cl(OH)₃. Graphene/CuO was synthesized from graphene/Cu₂Cl(OH)₃ by anionic exchange and a hydrothermal process.¹⁴⁵ Reproduced from ref. 145 with permission from The Royal Society of Chemistry.

3 3D MOs–G composites in LIBs

MOs in MOs–G composites were mainly in the 0D morphology in the early studies, such as nanoparticles^65,86 and nanocubes.²³⁴ These nanoparticles usually suffer from aggregation problems even in graphene composites and could be easily peeled off from the surface of graphene due to the weak contact between nanoparticles and graphene during the cycling process, resulting in the fast fading of the electrode capacity. Composites in wrapped, encapsulated, sandwich-like and layered models can show higher cyclic capability than that of other models. However, the rate capability of these composites is usually limited by the diffusion of lithium ions because Li⁺ ions cannot pass through the microsized planar GNSs. If the size of the nanoparticles is not large enough, the pores between GNSs will be too small to access the electrolyte and is harmful for lithium diffusion. It is important to investigate the size effects of 0D nanoparticles and their dispersion on GNSs to improve electrochemical performances. These problems also can be solved well by the graphene composites consisting of 1D, 2D or 3D MOs materials, which have at least one dimension in the micro or sub-micro size. In this situation, MOs with 1D, 2D and 3D structures can afford more sites to connect with graphene and create many meso- or macropores to make it electrolyte accessible, which can further enhance the long term stability and rate capability of LIBs. Furthermore, holey GNSs and stable 3D graphene structures can facilitate the diffusion of lithium ions and increase the rate capability. Thus, the morphology and size of MOs or graphene in composites play important roles in the electrochemical performances of LIBs.

3.1 Morphologies of MOs in composites

The assembling type, pore size/structure, surface area and overall dimensions of 3D MOs–G composites are greatly affected by the morphology and size of MOs. Moreover, the properties of 3D MOs–G composites will affect the stability of the hierarchical structure, the access of electrolytes, the diffusion of lithium ions, and further affect the electrochemical performances of LIBs.

3.1.1 Graphene composites with 0D MOs. 0D nanostructures, including quantum dots, nanoparticles, nanospheres or nanopolyhedra, are point structures with all dimensions in the nanoscale, which are also the simplest and easiest obtained nanostructures. In the early stage of studying 3D MOs–G composites of LIBs, SnO₂ nanoparticles,⁶⁵ TiO₂ nanoparticles⁸⁶ and Cu₂O nanocubes²³⁴ were widely studied. Though many other types of MOs have been developed for LIBs, most of the MOs used in 3D MOs–G composites were still in the 0D nanostructure up to 2013 (Table 1). It is well accepted that the decrease of particle size of MOs can ensure the reversibility of lithium intercalation, increase the rate of lithium insertion/removal, enhance electron transport within particles, and modify the chemical potential for lithium ions and electrons. Additionally, the graphene in the composites can shorten the migration distance of electrons from MOs to the conductive substrate. Thus, the above synergy effects can lead to the high reversible capacity of LIBs. For example, the SnO₂/graphene composite in mixed model showed a reversible capacity of 810 mA h g⁻¹, even higher than the theoretical capacity of SnO₂ (782 mA h g⁻¹).⁶⁵ However, only 570 mA h g⁻¹ of charge capacity of the composite was maintained after 30 cycles. In this case, MOs nanoparticles in the graphene composites in anchored, sandwich-like or mixed models usually suffer from aggregation and can be easily peeled off from the surface of graphene due to the weak contact between the nanoparticles and graphene during the cycling process.

Table 1 Summary of structures and electrochemical properties of MOs–G composites reported in literatures

MOs–G^a Composites	MOs Morphology	Synthesis method	MO content (wt%)	ICE^b	Cyclic performance				Rate performance		Ref.
					Current density (mA g^−1)	Initial capacity (mA h g^−1)	Cycles	Remain capacity (mA h g^−1)	Current density (mA g^−1)	Capacity (mA h g^−1)
a G: graphene.b ICE: initial columbic efficiency.c C: carbon.
Graphene composites with 0D MOs
Co3O4/CoO–G	Nanoparticles	Auto-combustion synthesis	90	76	21	890.4	30	801.3	2100	284	152
Co3O4–G	Nanoparticles	PVP assistant reflux	71	—	40	1300	40	860	1000	400	88
Co3O4–G	Nanocrystals	Hydrothermal method	76	70.1	400	∼900	50	775.2	2000	460	153
Co3O4–G	Polyhedral particles	Hydrothermal method	—	∼65	50	800	80	885	700	395	122
CoFe2O4–G	Nanoparticles	Coprecipitation & thermal reduction	82.3	∼71	100	982	50	985	1600	509	154
CoFe2O4–G	Nanoparticles	Hydrothermal method & thermal reduction	77.5	69	100	899	70	921.8	1600	446.3	130
CoO–G	Octahedral nanocrystals	Thermal decomposition	40.7	60.30	100	1184	60	1401	8000	∼500	155
Cr2O3–C^c–G	Nanoparticles	Hydrothermal method & thermal reduction	41.9	72.50	106	894.5	100	630	1000	315	131
SnO2–C–G	Nanoparticles	Evaporation & thermal reduction	—	65.80	200	656.9	100	633.2	1600	379.5	156
SnO2–C–G	Nanoparticles	Hydrothermal method	50.3	∼51	50	756	150	470			124
CuO–G	Nanoparticles	Spex-milling	90	47.20	0.1 mA cm^−2	785.2	45	496.5	6.4 mA cm^−2	201	70
Fe2Mo3O8–G	Nanoparticles	Hydrolysis & thermal reduction	91.7	72.40	200	923.5	40	835	3000	574.8	157
Fe2O3/SnO2–G	Nanoparticles	Hydrothermal method-in situ reduction	82	63.30	400	746	100	700	8410	139	158
Fe2O3–C–G	Nanoparticles	Evaporation & thermal reduction	—	—	500	∼500	100	504	2000	288.6	156
Fe2O3–C–G	Sub-microparticles	Hydrothermal method & glucose impregnation-pyrolysis process	85	71	200	1097	50	1005			159
Fe2O3–G	Nanoparticles	Hydrothermal method	32	∼60	100	1000	50	810	2500	280	160
Fe2O3–G	Nanoparticles	Spray drying	86.7	72	100	756	25	870.5	1600	660	161
Fe2O3–G	Nanoparticles	Hydrothermal method	68	56.30	200	800	200	852	4000	425	162
Fe2O3–G	Microsphere	Hydrothermal method	70	77.20	50	899	50	1206	1000	534	163
Fe2O3–G	Nanoparticles	Hydrothermal method & freeze–drying process	78	69	100	1045	50	995	2000	624	164
Fe2O3–G	Nanoparticles	Hydrothermal method	73	∼69	100	1095	70	∼950	800	∼700	165
Fe2O3–G	Nanoparticles	Thermal decomposition	68.1	53	100	750	50	900	5000	500	166
Fe2O3–G	Nanoparticles	PVP hydrolysis	80	77	200	1107	50	1052	2000	690	167
Fe3O4–CNF–G	Nanoparticles	Wet immersion method	41.6	∼60	1000	1400	100	1427.5	5000	592	168
Fe3O4–G	Nanoparticles	Solution mixing method	78	63.40	500	902	100	892	2000	672	169
Fe3O4–G	Nanoparticles	Supercritical CO2–ethanol	75	73.50	1000	941	100	838	5000	460	151
Fe3O4–G	Nanoparticles	Solution mixing & thermal reduction	—	66.90	100	908.6	50	1082	6000	145	170
Fe3O4–G	Nanoparticles	Thermal evaporation & thermal reduction	78	67.30	200	1443	100	868	1000	539	133
Fe3O4–G	Nanoparticles	Microwave assisted “benzyl-alcohol route”	58.9	55.70	100	1050			1600	∼500	140
Fe3O4–G	Nanoparticles	Electrostatic self-assembly	55.5	59.90	200	674	100	540	2000	384	76
Fe3O4–G	Nanoparticles	Hydrolysis	—	61	92.8	814			4860	282	171
Fe3O4–G	Nanoparticles	Hydrothermal method	30	59.20	100	1037	200	1130	1600	648	172
Fe3O4–G	Nanoparticles	Hydrothermal method	82.2	65	900	960	133	833	1800	437	89
Fe3O4–G–G	Nanospheres	Electrostatic interactions and hydrothermal method	93.7	∼67.3	93	920.3	150	1059	4800	363	75
FeOOH–G	Nanoparticles	Infrared irradiation	69	67.10	1000	∼800	600	767	2000	608	173
G–Co3O4–G	Nanoparticles	Hydrothermal method & mixing method	46.1	∼69.5	89	820	50	715.3	892	310	174
GeOx–G	Nanoparticle	Thermal Ge/Sn co-evaporation	70	51.10	325	1047	100	1008	5850	723	175
G–NiO–G	Nanoparticles	Hydrothermal method and mixing method	44.1	54.90	72	639.4	50	617.6			174
La2O3–NiO–G	Nanoparticles	Physical mixing	50	53.00	50	458.6	100	418.2	—	—	176
Mn3O4–G	Nanoparticles	Hydrothermal method	64	62.70	200	900	40	800	2000	382	121
MnFe2O4–G	Nanoparticles	Ultrasonic process	90	60	100	949	50	1017	12 000	315	67
MnO2–G	Nanoparticles	Hydrothermal method	—	64	100	781.5	50	750	1000	465.2	177
MoO2–G	Nanoparticles	In situ reduction process	93.5	75.40	0.2 C	1067.7	100	950	10 C	411.7	178
MoO3–G	Nanoparticles	Solution mixing method	73.96	—	800	961.5	50	711			179
NiFe2O4–G	Nanoparticles	Coprecipitation & thermal reduction	87.9	∼72	100	1225	50	1005	1600	758	154
NiO–G	Nanoparticles	Supercritical CO2–ethanol	68.5	65.10	500	629	100	741	2000	350	150
NiO–G	Nanoparticles	Electrochemical	87.3	59.30	359	754	50	586	1440	∼500	180
SnO2–C–G	Nanoparticles	Hydrothermal method & thermal reduction	51.4	54	100	1115	100	∼1000	1000	499	128
SnO2–C–G	Nanoparticle	Hydrothermal method & carbonization process	74.5	60	100	1058	80	703	1000	443	181
SnO2–C–G	Nanoparticles	Evaporation & thermal reduction	—	65.80	200	656.9	100	633.2	1600	379.5	156
SnO2–C–G	Nanoparticles	Hydrothermal method	50.3	∼51	50	756	150	470			124
SnO2–G	Nanoparticles	Refluxed & cross-linking reaction	81.8	—	100	1282	50	521	2000	334	111
SnO2–G	Nanoparticles	Microwave & thermal reduction	85	54.40	100	1329.4	20	618			129
SnO2–G	Nanocrystals	In situ Sn^2+ reduction	72	54	100	1017	50	610	2000	372	182
SnO2–G	Nanoparticles	Polyol reduction	30	73	90	1890	100	1220	1000	602	183
SnO2–G	Quantum dots	Hydrothermal method	50	∼43	200	800	200	720	2000	400	184
SnO2–G	Nanoparticles	Wet chemical method	80	96.4	1000	1923.5	40	1545.7			112
SnO2–G	Nanoparticle	Unzipping CNT and ultrasonication	80	74	100	1129	50	825	2000	580	185
SnO2–G	Nanoparticle	Sn^2+ in situ reduction & self-assembly	—	∼39	100	∼900	60	602	1000	200	186
SnO2–G	Nanoparticles	Template	89	∼51	100	878	40	503			187
SnO2–G	Nanoparticle	Hydrothermal method assembly	53	56	100	1211	70	824	500	621	188
SnO2–G	Nanoparticle	Sn^2+ oxidation–reduction reaction	68.1	45.60	100	1254.6	30	985.5			189
SnO2–G	Nanoparticle	Sn^2+ ultrasonic & oxidation–reduction reaction	38.4	∼62	100	627	50	535			190
SnO2–G	Nanocrystals	Freeze–drying & Vapor reduction process	70	61.30	500	1144	500	1346	20 000	417	191
SnO2–GO–G	Nanoparticles	Electrostatic interactions	72.9	55.90	100	∼1100	200	872	2000	519	80
SnO2–In2O3–G	Nanocrystals	Solvothermal	90.37	57.20	60	907	50	551	600	393	192
SnO2–In2O3–G	Nanocrystals	Hydrolysis-chemical reduction & thermal reduction	50	66.40	75	770	30	570	900	504	193
SnOx–CNF@G–G	Nanoparticles	Electrospinning calcination & mixing method	72	62.3	70	838	180	504	700	300	136
SnWO4–G	Nanoparticle	Hydrothermal method	80	55	50	934	20	500			194
TiO2–G	Nanocrystals	Hydrothermal method	93.7	69.30	200	171	100	137	4000	69	195
TiO2–G	Nanoparticles	In situ hydrothermal method growth	65	58.50	200	237	100	157	2000	122	196
TiO2–G	Nanoparticles	Hydrolysis & hydrothermal method	83		0.2 C	226			20 C	97	197
TiO2–G	Nanoparticle	Gas/liquid interface reaction	—	—	1000	∼150	80	136	5000	109	198
TiO2–G	Nanoparticles	Atomic layer deposition	54.7		2000	100	500	95			199
TiO2–SnO2–G	Nanoparticles	Solvothermal & hydrothermal method	90	49	50	954	50	537	1000	250	200
V2O5–G	Quantum dots	Two-step solution phase synthesis	93.55	97.20	100	280	100	212	1000	118	201
V2O5–G	Nanoparticles	Hydrothermal method	92	—	20	235	100	171			202
Zn2SnO4–G	Nanocrystals	Hydrothermal method	82.6	54	200	911	50	688	1600	439	203
ZnFe2O4–G	Nanoparticles	Hydrothermal method	—	68.70	100	945	50	956	1000	∼600	204
ZnO–G	Quantum dots	Atomic layer deposition	68	∼50	100	700	100	540	1000	400	205
ZnWO4–G	Nanoparticles	Sol–gel method	93	68	50	695	20	585	200	440	206
Graphene composites with 1D MOs
CuGeO3–G	Nanowires	Nanowires	81.2	45.5	100	1157	130	780	2000	550	207
Fe3O4–G	Nanospindles	Hydrothermal method	—	—	100	745	200	558	—	—	95
Fe3O4–G	Hollow nanospindles	Vacuum filtration and thermal reduction	70.3	66.7	200	∼1000	50	940	2000	420	208
Fe3O4–G	Nanorods	In situ growth	75	60.2	928	912	100	867	4640	569	209
GeO2–G	Microtubes	Strain-driven method	88.6	33	110	856	100	919	2200	571	210
MnO2–G	Nanowires	Hydrothermal method route	85	—	60	1150	30	890	12 000	∼600	101
MnO2–G	Nanorods	Solution mixed & thermal annealing method	91	70.8	30.8	170.6	100	158.3	—	—	211
Ni-doped MnO2–G	Nanowires	Electrostatic interactions	70	95	286	∼170	40	122	—	—	212
SnO2@G@G–G	Nanowires	Electrospinning & solution mixed method	72	61.6	80	1050	180	591.9	4000	212.8	135,213
SnO2–C–G	Nanorod	Seed assisted hydrothermal method growth & nanocarbon coating	63	70	100	1285	150	1419	3000	540	214
SnO2–G	Nanorods	Ultrasonic & hydrothermal method	67	∼60	200	1153	100	1107	2000	583.3	90
TiO2–G	Nanotube	Hydrothermal method	—	—	100	334	50	255	8000	80	102
TiO2–G	Nanowires	Hydrothermal method	90.4	—	150	∼200	30	∼160	—	—	215
TiO2–G	Nanotubes	Hydrothermal method	95.91	98.9	150	∼210	30	278	1500	114	216
WO3–G	Nanowires	Hydrothermal method	92.1	67.4	100	622	100	656	290	800	100
Zn2GeO4/G–G	Nanorods	Ion-exchange method	87.9	>60	100	873	90	803	3200	522	217
Zn2GeO4–G	Nanorods	Hydrothermal method & microwave process	72.2	∼60	100	873	100	1044	3200	678	97
Zn2GeO4–G	Hollow nanorods	Hydrothermal method	—	—	100	—	20	900	—	—	218
Graphene composites with 2D MOs
CoO–G (binder free)	Nanosheets	Electrostatic induced spread	87	71.60	500	678	150	640	1000	604	73
CuO–G–G	Nanosheets	Vacuum filtration & hydrothermal method reduction	—	91.60	67	782.3	50	736.8			219
Fe2O3–CNT–g–G	Nanoring	Chemical vapor deposition	—	∼51	74.4	984	100	812	3720	534	220
Fe2O3–G	Hexagonal nanoplatelets	Surfactant assistant limited growth	53		0.2 C	1370	150	1100	5 C	631	221
Fe2O3–G	Nanodisk	Hydrothermal method	75.6	71.60	200	1088	50	931	10 000	337	127
MoO3–G	Nanobelts	Hydrothermal method	—	—	500	289	200	174	1000	238	105
MoO3–G	Nanobelt	Microwave hydrothermal method	70	59	100	291	100	172	2000	151	123
SnO2	Nanosheets	Hydrothermal method	87	71.40	100	975	100	451			222
TiO2/graphene sandwich paper–G	Nanosheets	In situ hydrothermal method	82	—	1680	∼150	100	147	8500	120	223
TiO2/SnO2–G	Hybrid nanosheets	Stepwise growth	95	49	160	841	300	600	4000	300	224
TiO2@Co3O4–G	Coaxial nanobelt arrays	Electrostatic self-assembly	51.1	51	100	364	60	437	800	204	77
TiO2–G	Nanosheets	Hydrothermal method	75.12	—	1 C	232	100	189	10 C	150	107
TiO2–G	Nanoplatelets	Solvothermal	78	92	1 C	404	100	350	10 C	165	108
TiO2–G	Nanobelts	Hydrothermal method	86	82	150	746	100	430	3000	210	104
V2O5–G	Nanosheets	Solvothermal	95	89	600	262	160	102	3000	138	225
VO2–G	Ribbons	Hydrothermal method & chemical reduction process	84	—	190 C	204	1000	∼190			117
VO2–G	Nanosheet	Hydrothermal method	69	—	50	418	50	251	5000	102	226
Zn2SiO4–C–G	Layered structure	Hydrothermal method & vaporization	80	47	50	738	50	778	1000	277	119
Graphene composites with 3D MOs
Co3O4–G	Mesoporous hollow sphere	Solvothermal & immersed methods	76.2	69.30	1000	∼900	200	∼700	5000	259	94
Co3O4–G paper–G	Porous fibers	Electrostatic self-assembly	78.3	71.60	100	1005.7	40	840	100	295	79
Cu2O@Cu–G	Porous nanospheres	Microwave	85.2	—	50	734	50	842	2000	410	144
Fe2O3–G	Porous nanocages	Template method	—	76	200	1239	50	864	5000	587	227
Fe3O4/G–G	Porous nanorods	In situ-growth & hydrogen plasma treatment	79	77.50	500	845	100	890	3000	520	228
Fe3O4/G–G	Interconnected nanoparticles	Template synthesis	74	63.40	500	965	100	1124	10 000	506	229
Fe3O4–C–G	Mesoporous carbon supported	Template synthesis & heterocoagulation method	62.4	63.40	92.4	845	100	660	924	380	230
In2O3/G–G	Mesoporous	Template synthesis & heterocoagulation method	87.4	60.80	58	804.5	100	480	580	290	231
Mn0.5Co0.5Fe2O4–G	Mesoporous nanospheres	Solvothermal method	97.9	59.10	200	846.2	200	886.1	5000	266.3	93
NiO–C–G	Mesoporous carbon supported	Template synthesis & heterocoagulation method	51.8	61.80	71.8	∼820	100	∼570	718	∼310	230
SnO2–G	Mesoporous	Electrostatic interactions	—	—	78	∼780	50	∼460			118
SnO2–G	Mesoporous	Solvothermal method	79.5	69.40	78	1107	50	847.5	780	621.5	91
SnO2–G	Porous nanospheres	Electrostatic interaction	95	43.40	200	750	50	517	1000	423	232
TiO2–G	Mesoporous nanospheres	Hydrothermal method & freeze–drying	89	69.70	0.5 C	219	100	197	20 C	124	92
Zn2SnO4–G	Hollow box	Chemical etching & heat-process	65	69	300	1121.2	60	753	2000	345	233

---

To overcome these problems, surfactants can be induced to prevent the aggregation of MOs in graphene. Cai et al. found the presence of PVP could help the dispersion of Fe₃O₄ nanoparticles on graphene and result a stable capacity of 892 mA h g⁻¹ after 100 cycles, compared with that of 430 mA h g⁻¹ without PVP (Fig. 7A and B).¹⁶⁹ Building strong interactions between MOs and graphene is also a good choice to prevent the aggregation of MO nanomaterials and improve the stability and conductivity of composites, which will be further discussed in the following part. Composites in wrapped, encapsulated or layered models usually show higher cyclic stability than that of other models. For instance, the free-standing flexible nanocomposite films produced from layered graphene with 4 nm nanocrystals show no significant fading over 100 charge/discharge cycles (Fig. 7C and D).⁸⁷ However, the composites in wrapped and encapsulated models suffer from similar lithium ions diffusion problems because lithium cannot effectively pass through the 2D structure of graphene, which further affects the rate capability. Thus a small pore size in composites in the layered model (depending on the size of the nanoparticles) is harmful for electrolyte wetting and lithium diffusion, especially leading to an active process in initial several cycles and limiting rate capability. Hence, these films only retain a capacity of 225 mA h g⁻¹ at 0.02 A g⁻¹, compared to 760 mA h g⁻¹ at 0.008 A g⁻¹.⁸⁷ Thus, the structure model, graphene structure, dispersion and stability of MOs nanomaterials on the surface of graphene should be considered at the same time to fabricate 0D MOs based graphene composites with high electrochemical lithium storage performances.

Fig. 7 Schematic illustration of the synthesis (A) and cyclic performances (B) of Fe₃O₄–G composites prepared with the help of PVP.¹⁶⁹ Reproduced from ref. 169 with permission from The Royal Society of Chemistry. Li-ion battery configuration (C), cross-sectional TEM images (D) and cyclic performances (E) of a freestanding SnO₂–G nanocomposite film as an anode.⁸⁷ Reproduced with permission from ref. 87, Copyright (2010) American Chemical Society.

3.1.2 Graphene composites with 1D MOs. In the last decade, one-dimension (1D) nanostructures, such as nanowires, nanotubes and nanorods, have attracted considerable attention due to their potential magnetic, optical, electrical and catalytic properties.^236,237 Considered as the smallest dimension nanomaterials, 1D MOs nanomaterials are particularly interesting in LIBs, owing to the large surface to volume ratio, their vertical transportation of ions and electrons and apt accommodation of lithiation induced stresses.^235,238 On the other hand, 1D nanostructures have line contact with the substrate, compared with the point contact of 0D materials, thus it is reasonable to believe that graphene composites with 1D MOs nanomaterials would possess high electrochemical performances in LIBs. For example, Qiu et al. successfully synthesized surface-nitridated TiO₂ nanospindle–graphene nanocomposites by an in situ alkyl amines assistant hydrothermal process, and they found that the obtained composites showed enhanced cyclic stability and rate capability due to the ordered dispersion of dense ultrafine TiO₂ nanospindles on GNSs and the thin TiN/TiO_xN_y layer on TiO₂ nanospindles (Fig. 8).²³⁵

Fig. 8 (A) A typical TEM image of a TiO₂ nanospindle/graphene oxide nanocomposite (inset is a close-up view); (B) a TEM image of aTiO₂@TiO_xN_y/TiN–G nanocomposite after annealing in NH₃. (C) cycling performance of the electrode made of SP-20 at a rate of 1 C (inset shows the voltage profiles). (D) cycling specific capacity profiles of SP-20, SP-20@TiO_xN_y/TiN, and TiO₂@TiO_xN_y/TiN–G nanocomposite at different charge/discharge rates.²³⁵ Reproduced with permission from ref. 235, Copyright (2010) American Chemical Society.

The morphologies of MOs also affect the electron conductivity of composites. Graphene composites with Fe₂O₃ nanorice (∼200 nm in length and ∼40 nm in diameters) and cube-like nanoparticles (50–100 nm) were control synthesized by a microwave-assisted hydrothermal process.²³⁹ The electrical conductivities of Fe₂O₃ nanoparticles, Fe₂O₃–GNS particle-on-sheet and Fe₂O₃–GNS rice-on-sheet are 2.8 × 10⁻⁵ S cm⁻¹, 0.038 S cm⁻¹ and 0.049 S cm⁻¹, respectively, indicating that the composite with 1D Fe₂O₃ nanoparticles has the highest electron conductivity.²³⁹ As shown in Fig. 9, Fe₂O₃–GNS rice-on-sheet shows superior electrochemical performances compared with that of the particle-on-sheet at both common and high current rates, because the special 1D Fe₂O₃ nanorice can not only increase electron conductivity, but also facilitate the lithium diffusion rate for their high surface-to-volume ratio. Moreover, the heavy agglomeration of Fe₂O₃ nanoparticles are more inclined to occur compared with that of the Fe₂O₃ nanorice.²³⁹ Then, graphene–Fe₂O₃ nanocomposites with different loadings of Fe₂O₃ nanospindles were prepared by the one-step solvothermal method,⁹⁶ and the obtained composites showed the higher electron conductivity of 0.10–0.13 S cm⁻¹ (depending on the content of graphene), which further led to an improved rate capability compared with the Fe₂O₃–GNS rice-on-sheet used as anodes in LIBs.^96,239

Fig. 9 Schematic illustration of the growth process (A), TEM images (B and C), cyclic (D) and rate (E) performances of Fe₂O₃–GNS particle-on-sheet and rice-on-sheet composites.²³⁹ Reproduced with permission from ref. 239, Copyright (2011) American Chemical Society.

Fig. 10 (A) Schematic illustration of the growth process of NiO–GNS sheet-on-sheet and nanoparticle-on-sheet composites; (B) TEM image of NiO–GNS sheet-on-sheet composite; (C) cycling performances at 0.1 C of various anode materials; (D) TEM image of NiO–GNS sheet-on-sheet anode after 40 cycles, E), cycling performances at stepwise increased current rates.²⁴¹ Reproduced from ref. 241 with permission from The Royal Society of Chemistry.

Co₃O₄ nanorods/GNS nanocomposites were synthesized via a one-pot solvothermal method by simultaneously completing the reduction of GO and the growth of Co₃O₄ nanorods.⁸² The obtained Co₃O₄ nanorods–GNS nanocomposites exhibited approximate 1310 mA h g⁻¹ and 1090 mA h g⁻¹ of capacity at 0.1 A g⁻¹ and 1 A g⁻¹ after 40 cycles, respectively.⁸² The improvement of electrochemical performances of the Co₃O₄ nanorods–GNS nanocomposites can be attributed to the unique structures and properties of GNS and Co₃O₄ nanorods, which can provide an excellent ion diffusion and electronic conduction pathway, and further lead to a superior electrochemical performance. Recently, we further developed a small molecule assistant hydrothermal method to directly grow SnO₂ nanorods on GNSs, which were further assembled to a layered structure.⁹⁰ The as-prepared nanocomposites maintained a reversible capacity of 1107 mA h g⁻¹ within 100 cycles at a current density of 0.2 A g⁻¹, retaining 96.2% of the initial value. The capacity, stability and rate capability of the layered structure are considerably higher than that of single layer monodispersed SnO₂ nanorods growth on GNSs,^98,99 which may be attributed to the additional stability and electron conductivity derived from the interlayered/interconnected graphene in the layered model composite. The concept is further supported by the high electrochemical performances of free-standing layer-by-layer assembled graphene–MnO₂ nanotube thin hybrid films prepared by an ultrafiltration technique.²⁴⁰ In all, the unique physiochemical properties of 1D MOs in tandem with the synergistic effects of graphene composites promise high electrochemical performances in LIBs, which is greatly affected by the size of 1D MOs, graphene quality and the structure models of composites.

3.1.3 Graphene composites with 2D MOs. Unique 2D nanosheets have been extensively studied because they can enhance the intrinsic properties of their bulk counterparts, as well as generate new promising properties. 2D nanostructures can offer a face to face contact model, which can utilize nearly half of the surface area and result in much more strong interactions than that of 0D and 1D nanostructures. Thus, 2D nanostructures would show better mechanical stability and enhanced charge exchange properties between MOs and substrates, and the graphene composites with 2D MOs nanomaterials are expected to have unique electrochemical performances when used as electrode materials in LIBs. The fabricated Co(OH)₂ nanoplates–GNS composites by the hydrothermal process open the door of the application of graphene composites in LIBs, which show a capacity of 910 mA h g⁻¹ after 30 cycles, compared with the initial reversible capacity of 1120 mA h g⁻¹ at 200 mA g⁻¹.²⁴²

In the graphene composites with 2D MOs nanostructures, the lithium diffusion within 2D MOs may decide their lithium storage performances, especially the high rate performances. Wang et al. synthesized Co(OH)₂–graphene sheet-on-sheet composites by a single-mode microwave irradiation method, and layered Co₃O₄–GNS sheet-on-sheet nanocomposites were obtained after being further post thermally treated under N₂ to improve the quality and electron conductivity of graphene.^{68,81,138,204,243} The unique structure shows good stable electrochemical properties after 30 cycles. In composites, plenty of pores in the composite and Co₃O₄ nanosheets can accelerate phase transition, restrain the crumbling and cracking of the electrode, and further lead to superior cycling stability. Furthermore, the porous structure of Co₃O₄ nanosheets is readily accessible for electrolyte, facilitating the transportation of Li⁺ ions from the liquid to the active surface of Co₃O₄ and shorten the transportation length for both lithium ions and electrons.¹³⁸ Thus, the rate performances of composites are significantly improved, and a capacity of 931 mA h g⁻¹ is obtained at a current density of 4450 mA g⁻¹, which is larger than the theoretical capacity of Co₃O₄ (890 mA h g⁻¹). While only half of the capacity for Co₃O₄–graphene nanoparticle-on-sheet is obtained at the same conditions.¹³⁸ Then, they prepared NiO–graphene sheet-on-sheet and nanoparticle-on-sheet nanostructures (Fig. 10), the stable sheet-on-sheet structures showed highly stable reversible capacities (1056–1031 mA h g⁻¹ in 40 cycles at 71.8 mA g⁻¹) and good rate capabilities (492 mA h g⁻¹ at 3590 mA g⁻¹) than those of NiO nanosheets, GNSs, and NiO–graphene nanoparticle-on-sheet.²⁴¹ Both of the above results and other research^106,244,245 indicate that the electrochemical performances of graphene composites with 2D structures are better than the ones of 0D structures, derived from the stable and porous 2D structure.

Fig. 11 Schematic illustration of the formation process (A), TEM images (B and C), cyclic (D) and rate (E) performances of a-Fe₂O₃ hexagonal nanoplatelets sandwiched between graphene sheets (HP–Fe–G).²²¹ Reproduced from ref. 221, Copyright (2013) with permission from Elsevier.

In addition, the porous structure, ultrathin nanosheets can also reduce the diffusion distance of Li ions, thus the graphene composites with ultrathin Fe₂O₃ nanosheets (Fig. 11)²²¹ and VO₂ ribbons¹¹⁷ show remarkable stability and rate capability as porous 2D structures. These graphene composites with special 2D MOs nanostructures show the following inspirations: (1) the face to face contact model can enhance the electron contact, good mechanical and electrochemical stability of composites; (2) the enhanced electron contact with interconnected graphene networks can lead to high electron conductivity; (3) ultrathin and porous 2D nanostructures can provide numerous channels for the access of the electrolyte and facilitate the rapid diffusion of lithium ions.

3.1.4 Graphene composites with 3D MOs. 3D hierarchical materials built by nanosized units have both the advantages of nanomaterials and bulk materials. Furthermore, they can overcome the aggregation and/or separation of nanomaterials. The benefits of 3D hierarchical structures in LIBs have been summarized in the previous Section 2.1. However, the main problem of pure 3D hierarchical MOs is their poor electron conductivity, which can be solved well by compositing with graphene. In 2010, urchin-like CuO/graphene composites synthesized via a wet chemistry strategy showed an improved cyclic stability and rate capability compared with urchin-like CuO particles.¹⁰⁹ The urchin-like SnO₂/graphene nanocomposites prepared via the electrostatic attraction assisted co-assembly process could also significantly increase the capacity and cyclic stability of tin oxides.²⁴⁶ To date, several graphene composites with 3D hierarchical structures have been fabricated by combining the reduction of GO and the formation of hollow/porous MOs through one pot wet-chemistry process, such as porous V₂O₅ spheres,²⁴⁷ hollow Fe₂O₃ spheres,²⁴⁸ hollow porous Fe₃O₄ beads,²⁴⁹ porous TiO₂ nanospheres,²⁵⁰ mesoporous anatase TiO₂ nanospheres,²⁵¹ mesoporous SnO₂,⁹¹ mesoporous Mn_0.5Co_0.5Fe₂O₄ nanospheres⁹³ and Fe₃O₄ nanorods,.²²⁸ These obtained composites usually exhibit improved electrochemical performances in LIBs.

Though the one-pot synthesis method of graphene composites with 3D MOs is facile and easy, it lacks universality. Recently, the co-assembly of graphene and 3D MOs has been widely developed to fabricate composites of graphene and 3D MOs, in which mesoporous MOs were pre-synthesized by template methods^{118,230,231,252,253} or one pot processes.^66,94 Furthermore, the template^145,227,229 and chemical etching²³³ methods were also used directly to prepare graphene nanocomposites with 3D MOs. The electrochemical performances of the obtained graphene and 3D MOs composites are decided by the primary particle size, pore structure and post reduction process. Moreover the lithium storage performances, especially the high rate performance, strongly depend on the size of the building units. As shown in Fig. 12, graphene–mesoporous TiO₂ nanosphere composites with primary particles of 4 nm in size show a capacity of 97 mA h g⁻¹ at 8.4 A g⁻¹, while the nanocomposites with primary particles of 20–30 nm in size maintain a capacity of 97.7 mA h g⁻¹ at 1.68 A g⁻¹.^250,251

Fig. 12 SEM (A and D), TEM (B and E) images and rate capability (C and F) of mesoporous anatase TiO₂ nanospheres–graphene composites (A–C)²⁵¹ and porous TiO₂ nanospheres–graphene composites (D–F).²⁵⁰ A–C, Reproduced with permission from ref. 251, Copyright (2011) WILEY-VCH. D and E, reproduced from ref. 250 with permission from The Royal Society of Chemistry.

3.2 Morphologies of graphene

Due to the 2D nature of graphene, Li⁺ ions are difficult to directly transmit through GNSs. Furthermore, the conductivity between individual graphene nanosheets also affects the rate performances of nanocomposites. Thus, the morphologies, size and structures of graphene in composites are also very important for LIBs. To overcome the above problems, 3D graphene structures (free-standing graphene and graphene oxide sponges) were first synthesized by a special drying process, in which the strong interactions of graphene oxide in water played an important role. After being loaded by MOs nanomaterials and post reduced, the obtained graphene composites showed outstanding electrochemical performances.^92,227,255 For example, magnetite modified GNSs maintained a capacity of 450–500 mA h g⁻¹ at 5 A g⁻¹ during 1000 cycles.²⁵⁶ Co₃O₄ nanoparticles deposited onto porous graphene showed a good cycling performance with 90.6% retention of the original capacity after 50 cycles and enhanced rate capability with 71% capacity retention at 1 A g⁻¹, compared with the capacity at 0.05 A g⁻¹ (Fig. 13).²⁵⁴ Fe₂O₃ coated three-dimensional (3D) graphene composites maintained a capacity of 864 mA h g⁻¹ at 0.2 A g⁻¹ and 587 mA h g⁻¹ at 5 A g⁻¹ after 50 cycles.²²⁷ These 3D frameworks with interconnected porous structures can effectively reduce the diffusion length of electrons and Li⁺ ions, and the cross-linked graphene network can provide multidimensional pathways to facilitate the transport of electrons in the bulk electrode.

Fig. 13 Schematic illustration of the formation process (A), cyclic (B) and rate (C) performances of Co₃O₄/rGO films.²⁵⁴ Reproduced from ref. 254 with permission from The Royal Society of Chemistry.

Recently, amorphous GeO_x coated vertically aligned GNS composites were synthesized by a microwave plasma enhanced chemical vapor deposition system combined with the chemical vapor deposition process,¹⁷⁵ in which GNSs formed a fast electron transport channel due to its superior electron conductivity and the vertically aligned sandwich nanoflakes. The unified orientation of GNSs could effectively reduce the vertical lithium diffusion and offer a short pathway for lithium ions.¹⁷⁵ Additionally, amorphous GeO_x nanoparticles less than 10 nm in size could mitigate the mechanical stress of volume change. Thus, the obtained composites showed a unique cyclic stability (1008 mA h g⁻¹ for 100 cycles with retention of 96%) and rate performance (545 mA h g⁻¹ even at 15 C) (Fig. 14).¹⁷⁵

Fig. 14 Scheme of the lithium ion diffusion mechanism in VAG@GeO_x sandwich nanoflake based electrode (A), SEM (B) and TEM (C) image of the flake edge, clearly showing the sandwich structure of the VAG@GeO_x composites. (D), cycling performance of the electrodes with different loads, 70 wt% and 61 wt%, both were tested at the rate of C/3. Inset in (B) shows the EDS result of post-deposition nanosheets.¹⁷⁵ Reproduced from ref. 175, Copyright (2013) with permission from Elsevier.

Introduction of holes into the planar sheet is also a good choice to improve the electrochemical performance of GNS sheets because the holes can provide a high density of cross plane diffusion channels for Li⁺ ions.^{101,106,138,160,241,244,245} Previous work indicated that holey graphene (HG) prepared by HNO₃ (ref. 257)and KOH etching²⁵⁸ exhibited a significantly improved electrochemical performance as an anode material for LIBs, such as better cycle performance and higher rate capability in comparison with graphene sheets, activated graphene sheets, bare SnO₂ and SnO₂–GNSs composites.¹⁸²

4 Beyond morphologies

4.1 Composition

The composition of MOs also plays a significant role on the capacity, stability and rate performances of MOs–G nanocomposites. Firstly, most metal oxide anode materials are based on conversion reactions and alloy reactions. As a result, their theoretical capacities are tightly related with their compositions. For instance, Zn formed during conversion reactions can form alloys with Li, so ZnFe₂O₄ has a higher theoretical capacity of 1072 mA h g⁻¹ than that of Fe₃O₄ (924 mA h g⁻¹).²⁵⁹ But pure ZnFe₂O₄ has a reversible capacity of 957.7 mA h g⁻¹, while its graphene composites can reach to 1082 mA h g⁻¹.²⁰⁴ The search for other metal compounds with higher capacities is also a hot research topic. Metal carbonates, especially their graphene composites, can show higher capacities than their theoretical capacities.^260–262 For example, CoCO₃–graphene composites exhibited high capacities of over 1000 mA h g⁻¹, because both of the cations (Co²⁺) and anions (CO₃²⁻) are involved in the electron transfer.²⁶³

Recently, our research on MFe₂O₄–GNS (M = Mn,⁶⁷ Co,⁶⁹ Ni⁶⁸) nanocomposites indicated that MFe₂O₄ could transform into the nanosized hybrid Fe₃O₄ (MOs) with the size of about 20 nm after the discharge–charge process (Fig. 15), and the in situ formed hybrid Fe₃O₄ (MOs) combined with GNSs to form a spongy porous structure, which could further accommodate its volume change and result in good stability of the electrode. Additionally, the formed hybrid could also act as a matrix of MOs (Fe₃O₄) to prevent the aggregation and growth of the in situ formed Fe₃O₄ (MOs) nanoparticles, and further lead to good cyclic stability.⁶⁷ Various graphene composites with multiple MOs, including NiFe₂O₄–graphene heteroarchitectures,²⁶⁴ CoFe₂O₄/graphene sandwiched structures,²⁶⁵ ZnSnO₃/graphene,²⁶⁶ hollow Zn₂SnO₄ boxes@graphene and the incorporation of In₂O₃ into SnO₂,^192,193 indicated that multiple MOs would improve the electrochemical activity and reduce the charge transfer resistance of electrodes, leading to an enhanced reversible capacity and rate capability. On the other hand, transferring the surface MOs to the corresponding high electron conductivity compounds, such as TiN,²³⁵ SnS₂ (ref. 56, 57 and 267) or MoC,²⁶⁸ would also improve the rate capability of composites. Furthermore, the doping of N,^97,121,269 F²⁷⁰ and boron²⁷¹ could significantly increase the conductivity of graphene, and further lead to an enhanced reversible capacity and rate capability. For example, N-doped graphene–SnO₂ sandwiched films had a capacity of 504 mA h g⁻¹ at 5 A g⁻¹, while the undoped graphene/SnO₂ films only reached to 526 mA h g⁻¹ at 0.1 A g⁻¹ (Fig. 16).^158,272

Fig. 15 TEM image (A) and cyclic performances (B) of MnFe₂O₄–GNS composite. C–E, TEM image of MnFe₂O₄–GNSs nanocomposite after 70 charge–discharge process.⁶⁷ Reproduced from ref. 67 with permission from The Royal Society of Chemistry.

Fig. 16 Charge/discharge curves of N-doped G–SnO₂ papers (A) and commercial SnO₂ nanoparticles (20–50 nm, (B) at various current densities. Cycling performance of N-doped G–SnO₂ paper (C) and SnO₂ nanoparticles (D) at a current density of 50 mA g⁻¹. Schematic representation (E) showing paths for lithium-ions and electrons in the N-doped G–SnO₂ paper, respectively.²⁷² Reproduced with permission from ref. 272, Copyright (2012) WILEY-VCH.

4.2 Interaction

Based on the above discussions, it can be found that the interactions between MOs nanomaterials and graphene greatly affect the stability and charge transfer resistance of anodes, which are tightly related to the electrochemical lithium storage performances (cyclic stability and rate capability). For example, hollow porous Fe₃O₄ bead–rGO composites showed higher electrochemical performances than that of the mechanically mixed composites due to the in situ formed strong interactions between the Fe₃O₄ beads and rGO.²⁴⁹ Graphene composites fabricated by co-assembly via electrostatic interactions also result in strong interactions between MOs nanomaterials and graphene, which can lead to composites in encapsulated and layered models with high mechanical and electrochemical stability and further reduce the charge transfer resistances.^{72,74–80,245}

Covalent bonds can be generated between MOs nanomaterials and graphene by in situ processes, in which the formation of nanomaterials and reduction of GO occur simultaneously.^{184,191,273,274} Detailed research and density functional theory (DFT) calculations clearly indicate that these bonds, usually with O or N atoms on the surface of graphene act as bridges, such as C–O–Ni,²⁷⁴ C–O–Fe,²⁷³ Sn–N–C and Sn–O–C¹⁹¹ bonds. With the C–O–Ni bond as an example (Fig. 17), composites with this bond showed a lower surface charge-transfer resistance (105.6 Ω) than the composite without bonding (213.3 Ω) or pure NiO nanosheets (279.9 Ω).²⁷⁴

Fig. 17 SEM (A) and TEM (B) images, capacity (C), interactions between graphene with oxygen functional groups and NiO (D), EIS spectra (E) and rate capability (F) of oxygen bridged NiO–graphene composites.²⁷⁴ Reproduced with permission from ref. 274, Copyright (2012) American Chemical Society.

Furthermore, the rich functional groups on graphene can react with a cross-linking agent, such as benzene-1,4-diboronic acid (BDBA), to generate a graphene framework. The graphene framework can increase the overall electron conductivity, confine/fix nanoparticles and avoid their aggregations during the charge–discharge process (Fig. 18).¹¹¹ A polymer coating on the composites can also increase structure stability, electrochemical activity and the interactions between oxides and graphene, and graphene composites with conductive polymers, such as polyaniline^275,276 and PEDOT,²⁷⁷ have been invested and they showed remarkable rate performances. Even the un-conductive polymers, such as polydopamine, can also reduce the charge transfer resistances and further lead to stable high rate performances (Fig. 19).⁶⁸ These polymer coated graphene composites can also be transformed to carbon coated graphene composites by the carbonization process, and the formed carbon shells tackle the deformation of the MOs nanoparticles, and keep the overall electrode highly conductive and active in lithium storage.²⁷⁸

Fig. 18 (A) Schematic diagram depicting the different behaviors between rSG and SGF during Li alloying/dealloying. (B) Ultramicrotomed cross-sectional view of SGF. HRTEM images of rSG (C) and SGF (D) after 50 cycles.¹¹¹ Reproduced from ref. 111, Copyright (2013) with permission from Elsevier.

Fig. 19 Schematic illustration of the possible formation mechanism (A), TEM image (B) and cyclic performances at 1 A g⁻¹ (C), Nyquist plots (D) and TEM images after 50 cycles (E) of GNSs–PDA–NiFe₂O₄ nanocomposites.⁶⁸ Reproduced from ref. 68, Copyright (2014) with permission from Elsevier.

Gel mixtures of GO, metal precursors and carbon precursors have been widely used to prepare graphene composites. After being electrospun and carbonized, the obtained carbon coated graphene composites have unique lithium storage properties.^{135,213,279,280} Moreover, Ren et al. proved that the simultaneous formed carbon coating by the carbonization of glucose on the surface of the intermediate products (G–CrOOH) would provide more protection compared with the carbon layer post-formed by the carbonization of glucose on the surface of graphene–Cr₂O₃ composites. The simultaneously formed carbon layers could prevent the aggregation of Cr₂O₃ nanoparticles and limit their growth, whereas the latter could not effectively prevent the aggregation of nanoparticles because of mere coverage. As a result, the former carbon coated composites showed improved electrochemical properties compared to the latter composites, such as higher reversible capacity, better cycle performance and rate capability.¹³¹

5 Conclusions

We have reviewed the applications of 3D MOs–G composites in LIBs, especially on the effects of morphology and size of MOs or graphene on the properties of graphene composites. The core idea to improve the electrochemical performances of graphene composites in LIBs is how to enhance the electron conductivity and the lithium diffusion. Thus, MOs must have at least one dimension in the nanoscale to reduce the lithium diffusion distance. These nanomaterials also need to have enough contact sites and interactions with graphene, which will decide the charge transfer resistance, mechanical and electrochemical stability. Furthermore, proper pore size distribution is important for electrolyte access. To meet all of the above, one of the MOs or graphene must be in 3D structure. On the other hand, the composition of graphene composites and the interactions between graphene and MOs also affect their electrochemical performances.

Although considerable research and breakthroughs have been achieved, the challenges of using graphene composites for lithium storage still remain. The fundamental question is the cost-effective, environmentally friend and sustainable approach to the large-scale production of high quality graphene or GO. At present, starting from graphite is still the most ideal route due to abundant reserves of graphite. In past years, it is believed that the physical exfoliation method is not suitable for mass production even though it can produce high quality graphene. However, good news came from Europe²⁸¹ and China²⁸² recently. Defect free graphene with few layers can be prepared on a large-scale by the high-shear mixing of graphite in suitable stabilizing liquids or using supercritical CO₂ combined with the ultrasound approach.^281,282 Second, graphene composite anodes always show a low initial efficiency and few of them can reach to 75%,^{163,167,228,273} which causes many issues in the full battery. The low initial efficiency is mainly derived from the formation of an SEI film on the surface of graphene composites with a large surface area and the side reactions of functional groups on graphene, thus high quality graphene is required and the surface area of the composites should be optimized. Third, the capacity of transition MOs–G composites usually increases with increasing cycles, and is always considerably higher than their initial and theoretical capacity, which is considered as the decomposition of the electrolyte catalyzed by MOs.^{68,81,204,243} This is also a very important problem for safety issues. The aim of final industrial implementation, large scale, low cost and simple production of graphene composites with high electrochemical lithium storage performances is one of the most important challenges. The road of realizing 3D MOs–graphene composites in LIBs is tortuous. However, with continued exploitations the future is bright!

Acknowledgements

The work was supported by the National Basic Research Program of China (2014CB239700), Shanghai Nano-Project (12nm0503502), Program of National Natural Science Foundation of China (21371121 and 21331004), Minhang District Developing Project, Science and Technology Commission of Shanghai Municipality (14DZ2250800).

Notes and references

1. L. Zhang, China's urban development report No.7, http://www.china.com.cn/opinion/think/2014-09/17/content_33532822.htm, accessed 17th September, 2014.
2. J. W. Kim, J. D. Ocon and D.-W. Park, et al., ChemSusChem, 2014, 7, 1265.
3. V. Chabot, D. Higgins and A. Yu, et al., Energy Environ. Sci., 2014, 7, 1564.
4. C. Xu, B. Xu and Y. Gu, et al., Energy Environ. Sci., 2013, 6, 1388.
5. D. Wei and J. Kivioja, Nanoscale, 2013, 5, 10108.
6. L. W. Su, Y. Jing and Z. Zhou, Nanoscale, 2011, 3, 3967.
7. B. Li, J. Zai and Y. Xiao, et al., CrystEngComm, 2014, 16, 3318.
8. R. Yi, J. Zai and F. Dai, et al., Electrochem. Commun., 2013, 36, 29.
9. J. C. Shelton, H. R. Patil and J. M. Blakely, Surf. Sci., 1974, 43, 493.
10. C. Oshima and A. Nagashima, J. Phys.: Condens. Matter, 1997, 9, 1.
11. X. K. Lu, M. F. Yu and H. Huang, et al., Nanotechnology, 1999, 10, 269.
12. K. S. Novoselov, A. K. Geim and S. V. Morozov, et al., Science, 2004, 306, 666.
13. J. Zhu, M. Chen and Q. He, et al., RSC Adv., 2013, 3, 22790.
14. Q. Xiang, J. Yu and M. Jaroniec, Chem. Soc. Rev., 2012, 41, 782.
15. R. S. Edwards and K. S. Coleman, Acc. Chem. Res., 2013, 46, 23.
16. D. Chen, H. Feng and J. Li, Chem. Rev., 2012, 112, 6027.
17. Y. Sun and G. Shi, J. Polym. Sci., Part B: Polym. Phys., 2013, 51, 231.
18. A. Alarcon, N. Viet-Hung and S. Berrada, et al., IEEE T Electron Dev, 2013, 60, 985.
19. J. Cayssol, B. Huard and D. Goldhaber-Gordon, Phys. Rev. B: Condens. Matter Mater. Phys., 2009, 79, 075428.
20. J. Liu, M. Durstock and L. Dai, Energy Environ. Sci., 2014, 7, 1297.
21. P. Avouris, Proceedings 2010 10th IEEE International Conference on Nanotechnology and Joint Symposium with Nano Korea 2010 KINTEX (IEEE-NANO 2010), 2010, p. 6.
22. D. Gunlycke, D. A. Areshkin and J. Li, et al., Nano Lett., 2007, 7, 3608.
23. D. Yong-Joo and Y. Gyu-Chul, Nanotechnology, 2010, 21, 105204.
24. R. Arsat, M. Breedon and M. Shafiei, et al., Chem. Phys. Lett., 2009, 467, 344.
25. H. Song, L. Zhang and C. He, et al., J. Mater. Chem., 2011, 21, 5972.
26. T. Kuila, S. Bose and P. Khanra, et al., Biosens. Bioelectron., 2011, 26, 4637.
27. D. Zhao, Z. Li and L. Liu, et al., Acta Chim. Sin., 2014, 72, 185.
28. H. Jiang, Small, 2011, 7, 2413.
29. P. V. Kamat, J. Phys. Chem. Lett., 2011, 2, 242.
30. S.-S. Yu and W.-T. Zheng, Nanoscale, 2010, 2, 1069.
31. C. Huang, C. Li and G. Shi, Energy Environ. Sci., 2012, 5, 8848.
32. L.-L. Tan, S.-P. Chai and A. R. Mohamed, ChemSusChem, 2012, 5, 1868.
33. H. Tian, S. Ma and H.-M. Zhao, et al., Nanoscale, 2013, 5, 8951.
34. G. K. Dimitrakakis, E. Tylianakis and G. E. Froudakis, Nano Lett., 2008, 8, 3166.
35. S. Yang, L. Zhan and X. Xu, et al., Adv. Mater., 2013, 25, 2130.
36. L. Dai, Acc. Chem. Res., 2013, 46, 31.
37. S. Guo and S. Dong, Chem. Soc. Rev., 2011, 40, 2644.
38. H. Gwon, H.-S. Kim and K. U. Lee, et al., Energy Environ. Sci., 2011, 4, 1277.
39. H. Junbo, S. Yuyan and M. W. Ellis, et al., Phys. Chem. Chem. Phys., 2011, 13, 15384.
40. T. Kuila, A. K. Mishra and P. Khanra, et al., Nanoscale, 2013, 5, 52.
41. I. V. Lightcap and P. V. Kamat, Acc. Chem. Res., 2013, 46, 2235.
42. B. Luo, S. Liu and L. Zhi, Small, 2012, 8, 630.
43. M. Pumera, Energy Environ. Sci., 2011, 4, 668.
44. D. Zhou, Y. Cui and B. Han, Chin. Sci. Bull., 2012, 57, 2983.
45. J. Liu, Y. Xue and M. Zhang, et al., Mrs Bull., 2012, 37, 1265.
46. Y. Sun, Q. Wu and G. Shi, Energy Environ. Sci., 2011, 4, 1113.
47. L. Grande, V. T. Chundi and D. Wei, et al., Particuology, 2012, 10, 1.
48. X. Wan, Y. Huang and Y. Chen, Acc. Chem. Res., 2012, 45, 598.
49. Z.-S. Wu, G. Zhou and L.-C. Yin, et al., Nano Energy, 2012, 1, 107.
50. M. Xu, Y. Gao and X. Yang, et al., Chin. Sci. Bull., 2012, 57, 3000.
51. G. Zhao, T. Wen and C. Chen, et al., RSC Adv., 2012, 2, 9286.
52. Q. Tang, Z. Zhou and Z. F. Chen, Nanoscale, 2013, 5, 4541.
53. V. Z. Mordkovich, Synth. Met., 1996, 80, 243.
54. G. Wang, X. Shen and J. Yao, et al., Carbon, 2009, 47, 2049.
55. P. G. Bruce, B. Scrosati and J. M. Tarascon, Angew. Chem., Int. Ed., 2008, 47, 2930.
56. J. Zai, K. Wang and Y. Su, et al., J. Power Sources, 2011, 196, 3650.
57. J. Zai, X. Qian and K. Wang, et al., CrystEngComm, 2012, 14, 1364.
58. Y. Li, Z.-Y. Fu and B.-L. Su, Adv. Funct. Mater., 2012, 22, 4634.
59. C. N. R. Rao, H. S. S. R. Matte and K. S. Subrahmanyam, Acc. Chem. Res., 2013, 46, 149.
60. S. Han, D. Wu and S. Li, et al., Small, 2013, 9, 1173.
61. V. Georgakilas, M. Otyepka and A. B. Bourlinos, et al., Chem. Rev., 2012, 112, 6156.
62. Y. Zhu, D. K. James and J. M. Tour, Adv. Mater., 2012, 24, 4924.
63. W. I. Park, C.-H. Lee and J. M. Lee, et al., Nanoscale, 2011, 3, 3522.
64. X. Huang, X. Qi and F. Boey, et al., Chem. Soc. Rev., 2012, 41, 666.
65. S.-M. Paek, E. Yoo and I. Honma, Nano Lett., 2009, 9, 72.
66. L. Tao, J. Zai and K. Wang, et al., RSC Adv., 2012, 2, 3410.
67. Y. Xiao, J. Zai and L. Tao, et al., Phys. Chem. Chem. Phys., 2013, 15, 3939.
68. Y. Xiao, J. Zai and X. Li, et al., Nano Energy, 2014, 6, 51.
69. Y. Xiao, X. Li and J. Zai, et al., Nano-Micro Lett., 2014.
70. A. K. Rai, L. T. Anh and J. Gim, et al., J. Power Sources, 2013, 244, 435.
71. Z. Chen, M. Zhou and Y. Cao, et al., Adv. Energy Mater., 2012, 2, 95.
72. X.-l. Huang, J. Chai and T. Jiang, et al., J. Mater. Chem., 2012, 22, 3404.
73. H. Xiao-lei, W. Ru-zhi and X. Dan, et al., Adv. Funct. Mater., 2013, 23, 4345.
74. S. Yang, X. Feng and S. Ivanovici, et al., Angew. Chem., Int. Ed., 2010, 49, 8408.
75. W. Wei, S. Yang and H. Zhou, et al., Adv. Mater., 2013, 25, 2909.
76. T. Yoon, J. Kim and J. Kim, et al., Energies, 2013, 6, 4830.
77. Y. Luo, J. Luo and W. Zhou, et al., J. Mater. Chem. A, 2013, 1, 273.
78. R. Yi, J. Zai and F. Dai, et al., Nano Energy, 2014, 6, 211.
79. X. Yang, K. Fan and Y. Zhu, et al., ACS Appl. Mater. Interfaces, 2013, 5, 997.
80. S. J. R. Prabakar, Y.-H. Hwang and E.-G. Bae, et al., Adv. Mater., 2013, 25, 3307.
81. Z.-S. Wu, W. Ren and L. Wen, et al., ACS Nano, 2010, 4, 3187.
82. L. Tao, J. Zai and K. Wang, et al., J. Power Sources, 2012, 202, 230.
83. Y.-J. Hu, J. Jin and H. Zhang, et al., Acta Phys.-Chim. Sin., 2010, 26, 2073.
84. P. Lue, Y. Feng and X. Zhang, et al., Sci. China: Technol. Sci., 2010, 53, 2311.
85. J. N. Coleman, Acc. Chem. Res., 2013, 46, 14.
86. D. Wang, D. Choi and J. Li, et al., ACS Nano, 2009, 3, 907.
87. D. Wang, R. Kou and D. Choi, et al., ACS Nano, 2010, 4, 1587.
88. L. Pan, H. Zhao and W. Shen, et al., J. Mater. Chem. A, 2013, 1, 7159.
89. Y. Chen, B. Song and L. Lu, et al., Nanoscale, 2013, 5, 6797.
90. Q. Han, J. Zai and Y. Xiao, et al., RSC Adv., 2013, 3, 20573.
91. Y. Sheng, Y. Wenbo and Z. Jia, et al., Adv. Funct. Mater., 2013, 23, 91.
92. S. X. Yu, L. W. Yang and Y. Tian, et al., J. Mater. Chem. A, 2013, 1, 12750.
93. Z. Zhang, Y. Wang and D. Li, et al., Ind. Eng. Chem. Res., 2013, 52, 14906.
94. H. Sun, X. Sun and T. Hu, et al., J. Phys. Chem. C, 2014, 118, 2263.
95. J. Zhang, S. Wan and B. Yan, et al., J. Nanosci. Nanotechnol., 2013, 13, 4364.
96. S. Bai, S. Chen and X. Shen, et al., RSC Adv., 2012, 2, 10977.
97. F. Zou, X. Hu and Y. Sun, et al., Chem–Eur. J., 2013, 19, 6027.
98. X. Chaohe, S. Jing and G. Lian, Nanoscale, 2012, 4, 5425.
99. X. Chaohe, S. Jing and G. Lian, J. Mater. Chem., 2012, 22, 975.
100. M. Yu, H. Sun and X. Sun, et al., Mater. Lett., 2013, 108, 29.
101. Y. Zhang, H. Liu and Z. Zhu, et al., Electrochim. Acta, 2013, 108, 465.
102. J. Wang, Y. Zhou and B. Xiong, et al., Electrochim. Acta, 2013, 88, 847.
103. Y. Chen, M. Zhuo and J. Deng, et al., J. Mater. Chem. A, 2014, 2, 4449.
104. H. Huang, J. Fang and Y. Xia, et al., J. Mater. Chem. A, 2013, 1, 2495.
105. Y. Dong, S. Li and H. Xu, et al., Phys. Chem. Chem. Phys., 2013, 15, 17165.
106. S. Yongming, H. Xianluo and L. Wei, et al., Adv. Funct. Mater., 2013, 23, 2436.
107. Z. Zhang, Q. Chu and H. Li, et al., J. Colloid Interface Sci., 2013, 409, 38.
108. S. Yaodong, C. Jun Song and Z. Jixin, et al., J. Nanopart. Res., 2013, 15, 1913.
109. W. Bao, W. Xing-Long and S. Chun-Ying, et al., J. Mater. Chem., 2010, 20, 10661.
110. L. Q. Lu and Y. Wang, Electrochem. Commun., 2012, 14, 82.
111. H. Yun-Hwa, B. Eun Gyoung and S. Kee-Sun, et al., J. Power Sources, 2013, 240, 683.
112. G. Qi, Z. Zhe and G. Hailing, et al., J. Power Sources, 2013, 240, 149.
113. B. Zhao, G. Zhang and J. Song, et al., Electrochim. Acta, 2011, 56, 7340.
114. J. Zhu, T. Zhu and X. Zhou, et al., Nanoscale, 2011, 3, 1084.
115. Z. Ming, Q. Baihua and D. Lei, et al., J. Mater. Chem., 2012, 22, 3868.
116. S.-K. Park, S.-H. Yu and N. Pinna, et al., J. Mater. Chem., 2012, 22, 2520.
117. S. Yang, Y. Gong and Z. Liu, et al., Nano Lett., 2013, 13, 1596.
118. J. Shuhua, Y. Wenbo and G. Ziqi, et al., J. Mater. Sci., 2013, 48, 3870.
119. J. Qu, Y. Yan and Y.-X. Yin, et al., ACS Appl. Mater. Interfaces, 2013, 5, 5777.
120. D. W.-P. Pang, F.-W. Yuan and Y.-C. Chang, et al., Nanoscale, 2012, 4, 4562.
121. S.-K. Park, A. Jin and S.-H. Yu, et al., Electrochim. Acta, 2014, 120, 452.
122. X. Chi, L. Chang and D. Xie, et al., Mater. Lett., 2013, 106, 178.
123. L. Noerochim, J.-Z. Wang and D. Wexler, et al., J. Power Sources, 2013, 228, 198.
124. F. Xiao-Yong, S. Xiao-Yuan and W. Jing, et al., J. Solid State Electrochem., 2013, 17, 201.
125. C. Nethravathi, B. Viswanath and J. Michael, et al., Carbon, 2012, 50, 4839.
126. L. Tian, Q. Zhuang and J. Li, et al., Electrochim. Acta, 2012, 65, 153.
127. J. Qu, Y.-X. Yin and Y.-Q. Wang, et al., ACS Appl. Mater. Interfaces, 2013, 5, 3932.
128. P. Lian, J. Wang and D. Cai, et al., Electrochim. Acta, 2014, 116, 103.
129. Y. Zhu, C. Li and C. Cao, RSC Adv., 2013, 3, 11860.
130. M. Zhang, M. Jia and Y. Jin, et al., J. Alloys Compd., 2013, 566, 131.
131. W. Yue, S. Tao and J. Fu, et al., Carbon, 2013, 65, 97.
132. D. Qiu, L. Ma and M. Zheng, et al., Mater. Lett., 2012, 84, 9.
133. D. Yucheng, M. Ruguang and H. Mingjun, et al., Phys. Chem. Chem. Phys., 2013, 15, 7174.
134. Y. Jiang, T. Yuan and W. Sun, et al., ACS Appl. Mater. Interfaces, 2012, 4, 6216.
135. J. Zhu, G. Zhang and X. Yu, et al., Nano Energy, 2014, 3, 80.
136. Z. Jian, L. Danni and Z. Guanhua, et al., Nanoscale, 2013, 5, 5499.
137. M. Zhang, D. Lei and X. Yin, et al., J. Mater. Chem., 2010, 20, 5538.
138. S. Q. Chen and Y. Wang, J. Mater. Chem., 2010, 20, 9735.
139. X. Zhu, Y. Zhu and S. Murali, et al., ACS Nano, 2011, 5, 3333.
140. S.-H. Yu, D. E. Conte and S. Baek, et al., Adv. Funct. Mater., 2013, 23, 4293.
141. C. Zhong, J. Wang and Z. Chen, et al., J. Phys. Chem. C, 2011, 115, 25115.
142. T. Chen, L. Pan and X. Liu, et al., RSC Adv., 2012, 2, 11719.
143. L. Li, Z. Guo and A. Du, et al., J. Mater. Chem., 2012, 22, 3600.
144. N. Li, Y. Xiao and C. Hu, et al., Chem–Asian J., 2013, 8, 1960.
145. J. W. Ko, S.-W. Kim and J. Hong, et al., Green Chem., 2012, 14, 2391.
146. J. Qiu, P. Zhang and M. Ling, et al., ACS Appl. Mater. Interfaces, 2012, 4, 3636.
147. K. Gil-Pyo, N. Inho and K. Nam Dong, et al., Electrochem. Commun., 2012, 22, 93.
148. X. Meng, D. Geng and J. Liu, et al., Nanotechnology, 2011, 22, 165602.
149. E. B. Nursanto, A. Nugroho and S.-A. Hong, et al., Green Chem., 2011, 13, 2714.
150. L. Zhuo, Y. Wu and W. Zhou, et al., ACS Appl. Mater. Interfaces, 2013, 5, 7065.
151. L. Zhuo, Y. Wu and L. Wang, et al., J. Mater. Chem. A, 2013, 1, 3954.
152. A. K. Rai, J. Gim and A. Ly Tuan, et al., Electrochim. Acta, 2013, 100, 63.
153. D. Li, D. Shi and Z. Chen, et al., RSC Adv., 2013, 3, 5003.
154. J. Xiao, G. Xu and S.-G. Sun, et al., Part. Part. Syst. Charact., 2013, 30, 893.
155. Y. Qi, H. Zhang and N. Du, et al., J. Mater. Chem. A, 2013, 1, 2337.
156. H. Xin, C. Jing and Y. Hong, et al., J. Mater. Chem. A, 2013, 1, 6901.
157. Y. Sun, X. Hu and W. Luo, et al., J. Mater. Chem. A, 2013, 1, 4468.
158. G. Xia, N. Li and D. Li, et al., ACS Appl. Mater. Interfaces, 2013, 5, 8607.
159. G. Wang, H. Wang and S. Cai, et al., J. Power Sources, 2013, 239, 37.
160. L. Li, G. Zhou and Z. Weng, et al., Carbon, 2014, 67, 500.
161. G.-W. Zhou, J. Wang and P. Gao, et al., Ind. Eng. Chem. Res., 2013, 52, 1197.
162. C. Zhao, S.-L. Chou and Y. Wang, et al., RSC Adv., 2013, 3, 16597.
163. W. Xiao, Z. Wang and H. Guo, et al., Appl. Surf. Sci., 2013, 266, 148.
164. L. Xiao, D. Wu and S. Han, et al., ACS Appl. Mater. Interfaces, 2013, 5, 3764.
165. D. Meng, X. Chaohe and S. Jing, et al., J. Mater. Chem. A, 2013, 1, 7154.
166. B. Jang, O. B. Chae and S.-K. Park, et al., J. Mater. Chem. A, 2013, 1, 15442.
167. D. Chen, H. Quan and J. Liang, et al., Nanoscale, 2013, 5, 9684.
168. X. Huang, B. Sun and S. Chen, et al., Chem–Asian J., 2014, 9, 206.
169. M. Cai, H. Qian and Z. Wei, et al., RSC Adv., 2014, 4, 6379.
170. M. Zhang and M. Jia, J. Alloys Compd., 2013, 551, 53.
171. A. Hu, X. Chen and Y. Tang, et al., Mater. Lett., 2013, 91, 315.
172. Y. Chang, J. Li and B. Wang, et al., J. Mater. Chem. A, 2013, 1, 14658.
173. Y. Sun, X. Hu and W. Luo, et al., ACS Appl. Mater. Interfaces, 2013, 5, 10145.
174. W. Yue, S. Jiang and W. Huang, et al., J. Mater. Chem. A, 2013, 1, 6928.
175. S. Jin, N. Li and H. Cui, et al., Nano Energy, 2013, 2, 1128.
176. S. Wu, Y. Zhu and C. Li, et al., J. Alloys Compd., 2014, 582, 289.
177. Z. Xiaoyi, L. Jianjiang and S. Xilin, et al., Adv. Mater. Res., 2013, 709, 157.
178. Y. Chen, X. Di and C. Ma, et al., RSC Adv., 2013, 3, 17659.
179. D. Zhou, H. Wu and Z. Wei, et al., Phys. Chem. Chem. Phys., 2013, 15, 16898.
180. G.-P. Kim, I. Nam and S. Park, et al., Nanotechnology, 2013, 24.
181. C. Jianli, X. Huolin and Z. Haimei, et al., J. Power Sources, 2013, 232, 152.
182. Y. Yang, X. Ji and F. Lu, et al., Phys. Chem. Chem. Phys., 2013, 15, 15098.
183. B. P. Vinayan and S. Ramaprabhu, J. Mater. Chem. A, 2013, 1, 3865.
184. H. Song, N. Li and H. Cui, et al., J. Mater. Chem. A, 2013, 1, 7558.
185. J. Lin, Z. Peng and C. Xiang, et al., ACS Nano, 2013, 7, 6001.
186. J. Liang, Y. Liu and L. Guo, et al., RSC Adv., 2013, 3, 11489.
187. Z. Junsheng, W. Dianlong and W. Lin, Ionics, 2013, 19, 1223.
188. Y. Huang, D. Wu and S. Han, et al., ChemSusChem, 2013, 6, 1510.
189. Q. Guo and X. Qin, ECS Solid State Lett., 2013, 2, M41.
190. C. Gu, H. Zhang and X. Wang, et al., Mater. Res. Bull., 2013, 48, 4112.
191. X. Zhou, L.-J. Wan and Y.-G. Guo, Adv. Mater., 2013, 25, 2152.
192. H. Yang, T. Song and S. Lee, et al., Electrochim. Acta, 2013, 91, 275.
193. W. Xing, X. Wang and L. Song, et al., Mater. Chem. Phys., 2013, 140, 441.
194. R. Huang, H. Ge and X. Lin, et al., RSC Adv., 2013, 3, 1235.
195. M. Zhen, L. Su and Z. Yuan, et al., RSC Adv., 2013, 3, 13696.
196. H. Tao, S. Xiang and S. Hongtao, et al., Carbon, 2013, 51, 322.
197. D. Li, D. Shi and Z. Liu, et al., J. Nanopart. Res., 2013, 15.
198. C. Dandan, L. Dongdong and W. Suqing, et al., J. Alloys Compd., 2013, 561, 54.
199. C. Ban, M. Xie and X. Sun, et al., Nanotechnology, 2013, 24.
200. X. Jiang, X. Yang and Y. Zhu, et al., New J. Chem., 2013, 37, 3671.
201. C. Han, M. Yan and L. Mai, et al., Nano Energy, 2013, 2, 916.
202. H. Zhao, L. Pan and S. Xing, et al., J. Power Sources, 2013, 222, 21.
203. W. Song, J. Xie and W. Hu, et al., J. Power Sources, 2013, 229, 6.
204. H. Xia, Y. Qian and Y. Fu, et al., Solid State Sci., 2013, 17, 67.
205. X. Sun, C. Zhou and M. Xie, et al., J. Mater. Chem. A, 2014, 2, 7319.
206. L. Zhang, Z. Wang and L. Wang, et al., Mater. Lett., 2013, 108, 9.
207. Z. Chen, Y. Yan and S. Xin, et al., J. Mater. Chem. A, 2013, 1, 11404.
208. R. Wang, C. Xu and J. Sun, et al., J. Mater. Chem. A, 2013, 1, 1794.
209. A. Hu, X. Chen and Y. Tang, et al., Electrochem. Commun., 2013, 28, 139.
210. Y. Chen, C. Yan and O. G. Schmidt, Adv. Energy Mater., 2013, 3, 1269.
211. F. Tu, T. Wu and S. Liu, et al., Electrochim. Acta, 2013, 106, 406.
212. J. G. Radich, Y.-S. Chen and P. V. Kamat, ECS J. Solid State Sci. Technol., 2013, 2, M3178.
213. J. Zhu, D. Lei and G. Zhang, et al., Nanoscale, 2013, 5, 5499.
214. D. Wang, J. Yang and X. Li, et al., Energy Environ. Sci., 2013, 6, 2900.
215. H. Juan, W. Rong and Z. Pengjun, et al., Mater. Lett., 2013, 100, 173.
216. H. Juan, W. Rong and Z. Pengjun, et al., Integr. Ferroelectr., 2013, 144, 66.
217. F. Zou, X. Hu and L. Qie, et al., Nanoscale, 2014, 6, 924.
218. Y. Li, Y. Wang and Y. Zhang, Sci. Adv. Mater., 2013, 5, 523.
219. Y. Liu, W. Wang and L. Gu, et al., ACS Appl. Mater. Interfaces, 2013, 5, 9850.
220. S. Chen, P. Bao and G. Wan, Nano Energy, 2013, 2, 425.
221. X. Wang, W. Tian and D. Liu, et al., Nano Energy, 2013, 2, 257.
222. H. Liu, J. Huang and C. Xiang, et al., J. Mater. Sci.: Mater. Electron., 2013, 24, 3640.
223. N. Li, G. Zhou and R. Fang, et al., Nanoscale, 2013, 5, 7780.
224. Y. Tang, D. Wu and S. Chen, et al., Energy Environ. Sci., 2013, 6, 2447.
225. J. Cheng, B. Wang and H. L. Xin, et al., J. Mater. Chem. A, 2013, 1, 10814.
226. C. Nethravathi, C. R. Rajamathi and M. Rajamathi, et al., ACS Appl. Mater. Interfaces, 2013, 5, 2708.
227. X. Cao, B. Zheng and X. Rui, et al., Angew. Chem., Int. Ed., 2014, 53, 1404.
228. Q. Zhou, Z. Zhao and Z. Wang, et al., Nanoscale, 2014, 6, 2286.
229. J. Zhu, D. Yang and X. Rui, et al., Small, 2013, 9, 3390.
230. S. Tao, W. Yue and M. Zhong, et al., ACS Appl. Mater. Interfaces, 2014, 6, 6332.
231. L. Zhao, W. Yue and Y. Ren, Electrochim. Acta, 2014, 116, 31.
232. G. Ji, B. Ding and Z. Sha, et al., Nanoscale, 2013, 5, 5965.
233. Y. Zhao, Y. Huang and W. Zhang, et al., RSC Adv., 2013, 3, 23489.
234. C. Xu, X. Wang and L. Yang, et al., J. Solid State Chem., 2009, 182, 2486.
235. Y. Qiu, K. Yan and S. Yang, et al., ACS Nano, 2010, 4, 6515.
236. Y. Xia, P. Yang and Y. Sun, et al., Adv. Mater., 2003, 15, 353.
237. M. Kwiat, S. Cohen and A. Pevzner, et al., Nano Today, 2013, 8, 677.
238. Z. L. Wang, Adv. Mater., 2000, 12, 1295.
239. Y. Zou, J. Kan and Y. Wang, J. Phys. Chem. C, 2011, 115, 20747.
240. A. Yu, H. W. Park and A. Davies, et al., J. Phys. Chem. Lett., 2011, 2, 1855.
241. Z. Yuqin and W. Yong, Nanoscale, 2011, 3, 2615.
242. Y.-S. He, D.-W. Bai and X. Yang, et al., Electrochem. Commun., 2010, 12, 570.
243. H. Xia, D. Zhu and Y. Fu, et al., Electrochim. Acta, 2012, 83, 166.
244. J. Zai, C. Yu and L. Tao, et al., CrystEngComm, 2013, 15, 6663.
245. H. Yun, H. Xiao-lei and L. Jian-she, et al., J. Mater. Chem., 2012, 22, 2844.
246. K. Haegyeom, K. Sung-Wook and P. Young-Uk, et al., Nano Res., 2010, 3, 813.
247. X. Rui, J. Zhu and D. Sim, et al., Nanoscale, 2011, 3, 4752.
248. Q. Zou, J.-T. Zai and P. Liu, et al., Chem. Res. Chin. Univ., 2011, 32, 630.
249. Y. Chen, B. Song and X. Tang, et al., J. Mater. Chem., 2012, 22, 17656.
250. C. Huagiang, L. Baojun and Z. Jingxian, et al., J. Mater. Chem., 2012, 22, 9759.
251. L. Na, L. Gang and Z. Chao, et al., Adv. Funct. Mater., 2011, 21, 1717.
252. Y. Wenbo, L. Zhenzhen and J. Shuhua, et al., J. Mater. Chem., 2012, 22, 16318.
253. X. Yang, K. Fan and Y. Zhu, et al., J. Mater. Chem., 2012, 22, 17278.
254. B. G. Choi, S.-J. Chang and Y. B. Lee, et al., Nanoscale, 2012, 4, 5924.
255. X. Zhu, G. Ning and X. Ma, et al., J. Mater. Chem. A, 2013, 1, 14023.
256. J. Zai, C. Yu and Q. Zou, et al., RSC Adv., 2012, 2, 4397.
257. X. Zhao, C. M. Hayner and M. C. Kung, et al., ACS Nano, 2011, 5, 8739.
258. J. Zhongging, P. Bo and A. Manthiram, J. Mater. Chem. A, 2013, 1, 7775.
259. Y. Ding, Y. F. Yang and H. X. Shao, Electrochim. Acta, 2011, 56, 9433.
260. S. Mirhashemihaghighi, B. Leon and C. P. Vicente, et al., Inorg. Chem., 2012, 51, 5554.
261. M. J. Aragon, C. Perez-Vicente and J. L. Tirado, Electrochem. Commun., 2007, 9, 1744.
262. M. J. Aragon, B. Leon and C. P. Vicente, et al., J. Power Sources, 2011, 196, 2863.
263. L. W. Su, Z. Zhou and X. Qin, et al., Nano Energy, 2013, 2, 276.
264. Y. Fu, Y. Wan and H. Xia, et al., J. Power Sources, 2012, 213, 338.
265. S. Liu, J. Xie and C. Fang, et al., J. Mater. Chem., 2012, 22, 19738.
266. W. Song, J. Xie and S. Liu, et al., J. Mater. Res., 2012, 27, 3096.
267. J. Zhu, D. Wang and L. Wang, et al., Electrochim. Acta, 2013, 91, 323.
268. H.-J. Zhang, K.-X. Wang and X.-Y. Wu, et al., Adv. Funct. Mater., 2014, 24, 3399.
269. A. L. M. Reddy, A. Srivastava and S. R. Gowda, et al., ACS Nano, 2010, 4, 6337.
270. H. Tachikawa, T. Iyama and H. Kawabata, Jpn. J. Appl. Phys., 2010, 49, 01ah02.
271. X. Wang, Z. Zeng and H. Ahn, et al., Appl. Phys. Lett., 2009, 95, 183103.
272. X. Wang, X. Cao and L. Bourgeois, et al., Adv. Funct. Mater., 2012, 22, 2682.
273. M. Zheng, D. Qiu and B. Zhao, et al., RSC Adv., 2013, 3, 699.
274. G. Zhou, D.-W. Wang and L.-C. Yin, et al., ACS Nano, 2012, 6, 3214.
275. L. Renlong, C. Huaqiang and Q. Dong, et al., J. Mater. Chem., 2011, 21, 17654.
276. F. Zhang, H. Cao and D. Yue, et al., Inorg. Chem., 2012, 51, 9544.
277. W. Wang, W. Lei and T. Yao, et al., Electrochim. Acta, 2013, 108, 118.
278. Y. Su, S. Li and D. Wu, et al., ACS Nano, 2012, 6, 8349.
279. C. Zhang, X. Peng and Z. Guo, et al., Carbon, 2012, 50, 1897.
280. X. Zhang, P. S. Kumar and V. Aravindan, et al., J. Phys. Chem. C, 2012, 116, 14780.
281. K. R. Paton, E. Varrla and C. Backes, et al., Nat. Mater., 2014, 13, 624630.
282. Y. Gao, W. Shi and W. Wang, et al., Ind. Eng. Chem. Res., 2014, 53, 2839.

---