Iron-Oxide-Based Advanced Anode Materials for Lithium-Ion Batteries



Iron oxides, such as Fe2O3 and Fe3O4, have recently received increased attention as very promising anode materials for rechargeable lithium-ion batteries (LIBs) because of their high theoretical capacity, non-toxicity, low cost, and improved safety. Nanostructure engineering has been demonstrated as an effective approach to improve the electrochemical performance of electrode materials. Here, recent research progress in the rational design and synthesis of diverse iron oxide-based nanomaterials and their lithium storage performance for LIBs, including 1D nanowires/rods, 2D nanosheets/flakes, 3D porous/hierarchical architectures, various hollow structures, and hybrid nanostructures of iron oxides and carbon (including amorphous carbon, carbon nanotubes, and graphene). By focusing on synthesis strategies for various iron-oxide-based nanostructures and the impacts of nanostructuring on their electrochemical performance, novel approaches to the construction of iron-oxide-based nanostructures are highlighted and the importance of proper structural and compositional engineering that leads to improved physical/chemical properties of iron oxides for efficient electrochemical energy storage is stressed. Iron-oxide-based nanomaterials stand a good chance as negative electrodes for next generation LIBs.

1 Introduction

With the rapid development of the global economy, the fast depletion of fossil fuels, and the increasing environmental concerns, there is an urgent need for efficient, clean, and sustainable sources of energy in addition to new technologies associated with energy conversion and storage. Rechargeable lithium-ion batteries (LIBs) are one promising technology of choice for upcoming large-scale applications such as stationary energy storage and electric vehicles because of the advantages of high energy density, long lifespan, no memory effect, and environmental benignity.[1-3] Typically, a LIB cell consists of a negative electrode (anode, e.g., graphite) and a positive electrode (cathode, e.g., LiCoO2), separated by an insulating separator. The cell is filled with a liquid/solid electrolyte that facilitates the transport of ionic species of the electrochemical reaction inside the cell and forces the electrons to transport through the external circuit.[4] During the charging process, Li+ ions de-intercalate from the cathode, pass through the electrolyte, and intercalate into the anode. Conversely, Li+ ions migrate from the anode to the cathode through the electrolyte upon discharge. Thus, Li+ ions shuttle between the two electrodes, enabling the reversible conversion between the stored chemical energy and the electrical energy to power the appliance. It would come as no surprise that the development of high-performance batteries is ultimately dependent on the electrode materials. However, the commercially available anode materials are commonly limited to graphite materials with low specific capacity (theoretically 372 mA h g−1) and poor safety, which are not like the multiplicity of cathode materials (such as layered LiCoO2, spinel LiMn2O4, and olivine LiFePO4).

A series of transition metal oxides has received growing interests as potential anode materials for next-generation LIBs.[5-15] Among these candidates, iron oxides, in particular Fe2O3 and Fe3O4, have drawn particular attention because of their high theoretical capacity of ≈1000 mA h g−1, non-toxicity, high abundance, high corrosion resistance, and low processing cost.[16, 17] The improved safety associated with the higher lithium insertion potential and the non-flammable nature is another advantage of iron-oxide-based anodes for large-scale applications. The lithium storage mechanism of iron oxides is based on a redox conversion reaction, where the iron oxides are reduced to metallic nanoclusters dispersed in a Li2O matrix upon lithiation and are then reversibly restored to their initial oxidation states during delithiation. Using Fe2O3 as an example, the reaction mechanism can be described as follows: Fe2O3 + 6Li+ + 6e↔ 3Li2O + 2Fe.[6, 7] The forward displacement reaction of Fe2O3 with Li+ is thermodynamically feasible, and the formation of Fe0 from Fe3+ involves multiple-electron transfer per metal atom, leading to a high theoretical lithium storage capacity. However, the reverse extraction of Li+ from the Li2O matrix seems to be thermodynamically unattainable.[6, 7] Moreover, iron oxides suffer from poor cyclability, which is partly caused by the drastic volume change during the charge/discharge process. The low conductivity of iron oxides also induces additional performance degradation, particularly when charging and discharging at high current densities. Strategies have been proposed to facilitate the reverse extraction of Li+ from the Li2O, mitigate the pulverization, and further enhance the structural stability of electrode materials. One effective method relies on the design and synthesis of nanostructured electrode materials with various morphologies, configurations, porosity, and a combination of micro- and nanostructures, which are all expected to promote the electrochemical processes and maintain the good structural integration. Of particular note, hollow structures, such as hollow spheres and nanotubes, stand out as a remarkable category of micro-/nanostructures with several important merits, as discussed shortly.[18] Another approach is to integrate carbonaceous matrix into the metal oxide active materials to form hybrid nanostructures, which can better buffer the volume strain due to the elastic feature of carbon supports and increase the electrical conductivity. As some excellent reviews have provided comprehensive description and discussion of advanced metal-oxide-based electrode for LIBs,[5-15] this article focuses mainly on the recent development in the structural design, synthetic methodology, formation mechanism, and lithium storage properties of nanostructured iron oxides and iron-oxide-based hybrid materials. Specifically, we will main focus on the proper nanostructure design and enhanced electrochemical performance of iron-oxide-based anodes. Finally, future trends and prospects in the development of advanced iron-oxide-based anodes are highlighted.

2 Nanostructured Iron Oxides

Constructing nanostructured materials has been considered as a promising avenue towards the development of high-performance LIBs with high energy density, high power density, long cycle life, and improved safety.[2-15] In general, nanostructures can provide reduced distance for ion and electron transport, and larger electrode/electrolyte contact area. Thus the electrochemical processes in nanostructured material-based electrodes are significantly boosted. Therefore, faster charge/discharge capability (i.e., higher rate capability/power density) and higher specific capacity (i.e., higher energy density) can be expected for LIBs based on advanced nanostructured electrode materials.[5-13] Additionally, some of the structural features, such as low-dimensional morphologies and hollow spaces, are able to better withstand the huge volume change during the charge/discharge process, thus leading to enhanced cycling performance.[18-20]

To maximize the advantages of nanostructured materials, the morphology, composition, porosity, and surface characteristics need to be optimized. Therefore, a wide variety of nanostructured iron oxides with diverse geometric shapes and morphologies has been extensively explored, such as porous particles,[21, 22] 0D nanoparticles,[23] 1D nanowires and nanotubes,[24, 25] 2D nanosheets and nanoflakes,[26-30] and 3D hollow structures.[31-37] Among them, hollow structures, including tubular structures, hollow spheres/cubes, etc. have been recognized as a unique class of micro-/nanostructured materials with unusual characteristics, which are discussed separately in the following section. In this section, recent advances in fabrication of various nanostructured iron oxides except for hollow structures are surveyed and discussed.

Porosity is one key factor that influences the electrochemical performance of solid-state electrode materials. Recently, we have demonstrated an interesting top-down method to fabricate melon-like α-Fe2O3 microparticles with controllable porosities.[21] Oxalic acid is chosen as the etchant that could dissolve α-Fe2O3 in a very gentle but highly efficient manner while maintaining its overall structural integrity. At the same time, H2PO4 ions are introduced into the system to control the specific etching direction. As shown in Figure 1a–c, the morphology of the microparticles, especially the surface texture, changes dramatically with the etching duration. The pristine microparticles with smooth surfaces become highly porous after etching. The textural characterization confirms that the total pore volume and the average pore size increase significantly with the etching time. The microparticles with contrasting porosities exhibit distinct cycling stability when evaluated as anode materials for LIBs (Figure 1d). Specifically, despite the similar initial capacity, a much higher reversible capacity of 662 mA h g−1 can be retained by the porous microparticles at the end of 100 charge/discharge cycles, whereas the non-etched sample can only deliver a capacity of 341 mA h g−1. The improved cycling performance is ascribed to the high porosity that provides void space to buffer the volume variation during the repeated charge/discharge cycles, thus suggesting the significance of proper porosity in the electrode materials. In related work, Cho and co-workers reported a simple synthesis of mesoporous spindle-like porous α-Fe2O3 by calcination of an iron-based metal organic framework template. The mesoporous α-Fe2O3 material exhibits significantly improved electrochemical performance, which further demonstrates the feasibility of enhancing material properties through porosity design.[22]

Figure 1.

FESEM and TEM (inset) images of the melon-like α-Fe2O3 microparticles obtained with different oxalic acid etching durations: a) 0 h, b) 36 h, and c) 42 h. d) Comparative cycling performance of the α-Fe2O3 microparticles shown in a (I) and b (II). e) FESEM and TEM (inset) images and f) cycling performance of iron oxide nanowires. g) FESEM image and h) cycling performance of α-Fe2O3 flowers constructed by nanosheets. i,j) FESEM images of α-Fe2O3 nanosheets grown directly on Ni foam. k) FESEM image and l) cycling performance of α-Fe2O3 nanosheets grown on Cu foil. Panels (a–d) reproduced with permission.[21] Copyright 2010, American Chemical Society. Panels (e,f) reproduced with permission.[24] Copyright 2009, Elsevier. Panels (g,h) reproduced with permission.[26] Copyright 2011, Elsevier. Panels (i,j) are reproduced with permission.[28] Copyright 2012, Royal Society of Chemistry. Panels (k,l) reproduced with permission.[29] Copyright 2007, Wiley.

It is also suggested that the strain associated with the volume expansion upon lithiation could be well accommodated in low-dimensional nanostructures (e.g., 1D nanowires and 2D nanosheets).[19, 20, 24, 25] These low-dimensional nanostructures might be able to expand and contract in particular directions (e.g., change in diameter of nanowires or thickness of nanosheets) without destroying the structure. On the other hand, lithium-ion transport is also enhanced because of the reduced diffusion length and large electrolyte/electrode contact area.[4-14] Wang and co-workers reported a facile hydrothermal synthesis of iron-containing precursor nanowires in the mixture of distilled water, isopropyl alcohol, and nitrilotriacetic acid.[24] The hydrolysis of Fe3+ ions in the presence of nitrilotriacetic acid produces 1D long-chain polymer precursors, which are converted to α-Fe2O3 nanowires after calcination at 500 °C in air. The field-emission scanning electron microscopy (FESEM) image shows that the nanowires possess a length as long as 100 μm with a diameter of around 200 nm (Figure 1e). The transmission electron microscopy (TEM) image indicates the polycrystalline feature of an individual nanowire, which is constructed by small α-Fe2O3 nanoparticles (inset of Figure 1e). When tested as anode materials at a current density of 0.1 C within the voltage window of 0.01−3 V, these α-Fe2O3 nanowires show high initial discharge capacity of 1303 mA h g−1. After 100 cycles, a discharge capacity of 456 mA h g−1 is retained (Figure 1f). The capacity retention is not particularly remarkable, which is possibly due to the structural disintegration of the porous nanowires upon prolonged cycling.

The 2D nanostructures of iron oxides present a finite lateral size and enhanced open-edges, which facilitate lithium-ion and electron diffusion through active materials and better withstand the large volume change during the charge/discharge process.[19, 20] Han et al. reported the synthesis of microflower-like α-Fe2O3 constructed by sheet-like subunits via simple heat treatment of a iron(III)-oxyhydroxide precursor, which is obtained by the hydrolysis of FeCl3 solution in the presence of NaClO.[26] The flower-like architectures with a size of 7–10 μm are built from tens of self-assembled nanosheets with a width of 2–3 μm and a thickness of about 20 nm (Figure 1g), giving rise to a high Brunauer-Emmett-Teller (BET) surface area of 115 m2 g−1. The flower-like Fe2O3 electrode shows very high initial capacity under a current density of 100 mA g−1, and the value stabilizes at 929 mA h g−1 after 10 cycles as shown in Figure 1h. The better capacity retention of the flower-like α-Fe2O3 compared with commercial nanoparticles would be related to the robust hierarchical architecture and 2D subunits with high surface area, which can accommodate the volume change due to the Li+ insertion/extraction and prevent the peeling off of active materials from the current collector.

Later, a template-free hydrothermal method was developed to synthesize vertically aligned, single crystalline α-Fe2O3 nanosheets grown directly on Ni foam.[28] During the reaction, uniformly and densely distributed α-FeOOH nanosheets are first grown on the Ni foam through the hydrolysis of Fe3+, which are transformed to α-Fe2O3 nanosheets by calcination at 400 °C for 3 h in Ar. The FESEM image shows that the α-Fe2O3 nanosheets have an open-up network structure assembled by interconnected nanosheets (Figure 1i,j). The high-resolution FESEM image also shows that the open space between nanowalls is relatively large and the nanowalls are about 20 nm in thickness with height of 2 μm (Figure 1j). Because of the 2D feature and the intimate contact with the current collector, the α-Fe2O3 nanosheets exhibit stable cycling performance with reversible capacity of 518 mA h g−1 after 50 cycles in the potential range 0.05−2.5 V at 0.1 C.

Furthermore, Chowdari and co-workers reported a surprisingly simple method to synthesize single-crystalline α-Fe2O3 nanosheets on conductive substrates by directly heating Fe-coated substrates in air using a hotplate.[29, 30] The FESEM image of the α-Fe2O3 nanosheets shows that the nanometer-sized flakes are generally pointed perpendicular to the Cu substrate (Figure 1k). The α-Fe2O3 nanoflakes on Cu foil could be directly used as the working electrode without the use of any ancillary components, such as carbon black and polymer binder. As anticipated, such an integrated electrode composed of α-Fe2O3 nanoflakes exhibits stable capacity of ≈700 mA h g−1 with no noticeable capacity fading up to 80 cycles (Figure 1l), when cycled in the potential range of 0.005−3.0 V at 65 mA g−1.

3 Iron Oxide Hollow Structures

Hollow structures refer to a thriving family of micro-/nanostructured materials with well-defined shell and interior void, which have aroused tremendous interest because of their widespread applications in nanoreactors, drug delivery, gas sensors, and energy storage/conversion.[38-43] The improvement of electrochemical performance brought by the hollow structures can be rationalized in the following aspects: a) the relatively high surface area endows the metal oxide particles with more lithium storage sites and large electrode-electrolyte contact area for high Li+ ions flux across the interface, which is beneficial for enhancing specific capacity of the active materials; b) the permeable thin shells provide significantly reduced paths for both Li+ ions and electrons diffusion, leading to a better rate capability; and c) the hollow interior provides extra free space for alleviating the structural strain and accommodating the large volume variation associated with repeated Li+ ions insertion/extraction processes, giving rise to improved cycling stability.[18, 43] Various strategies have been developed to fabricate hollow structures in the past decades. Here we highlight some advanced approaches to synthesize iron-oxide-based hollow structures in an efficient manner.

Compared with typical hard-templating methods, which involve the controlled deposition of the designed materials on various removable templates, one-pot synthesis of hollow particles is more appealing in view of the simplicity.[44-46] We have recently reported the synthesis of hierarchical α-Fe2O3 hollow spheres with sheet-like subunits by a facile quasi-emulsion-templated method.[32] In this system, glycerol is first mixed with water to form oil-in-water quasi-emulsion microdroplets, which serve as soft templates for the subsequent deposition and crystal growth of the shell structure. As shown in Figure 2a, the resultant α-Fe2O3 hollow structures have a uniform diameter of about 1 μm and a large BET surface area of 103 m2 g−1. When the sample is evaluated as the anode material for LIBs (Figure 2b), it demonstrates excellent lithium storage properties with very good cycling stability and a high reversible capacity of 710 mA h g−1 at the end of 100 charge/discharge cycles tested at a constant current density of 200 mA g−1 between 0.05 and 3 V. On the other hand, microparticles of similar size can only deliver a much lower capacity of 340 mA h g−1 at the end of the test.

Figure 2.

a) TEM image and b) cycling performance of the α-Fe2O3 hollow spheres. c) FESEM images and d) cycling performance of Fe3O4 hollow microspheres constructed by nanoplate building blocks. e,f) TEM images of Fe(OH)x nanoboxes (e) and octahedral α-Fe2O3 nanocages (f). g) FESEM and TEM (inset) images, and h) cycling performance of α-Fe2O3 nanotubes. i–k) FESEM and TEM (inset) images and l) comparative cycling performance of Fe2O3 microboxes (i), porous microboxes (j), and hierarchical microboxes (k). Panels (a,b) reproduced with permission.[32] Copyright 2011, American Chemical Society. Panels (c,d) reproduced with permission.[33] Copyright 2013, Wiley. Panels (e,f) reproduced with permission.[34] Copyright 2010, American Chemical Society. Panels (g,h) reproduced with permission.[35] Copyright 2011, Royal Society of Chemistry. Panels (i–l) reproduced with permission.[36] Copyright 2012, American Chemical Society.

Meanwhile, several novel template-free approaches towards hollow structures have been developed on the basis of different principles, including the Kirkendall effect,[47] inside-out Ostwald ripening,[48] self-assembly,[49] and thermal decomposition.[50] Recently, we developed a simple ethylenediamine-mediated solvothermal method together with a post heat treatment to prepare hierarchical hollow Fe3O4 microspheres constructed from nanoplate building blocks (Figure 2c).[33] At the initial stage of solvothermal reaction, small iron-containing nanoparticles initially formed self-assemble into solid spheres comprised of platelike subunits. As the reaction proceeds, the small crystallites in the core region of the spheres are selectively dissolved and re-crystallized according to a well-known inside-out Ostwald ripening process.[48] The hollowing process continues with longer reaction duration until well-defined hollow spheres comprised of nanoplates are formed. These iron-containing hollow microspheres are subsequently converted into magnetic Fe3O4 hollow microspheres without significant structural alteration by thermal decomposition under N2 atmosphere. In view of their unique structural advantages, these as-prepared Fe3O4 hollow microspheres show significantly improved cycling stability with a high reversible capacity of 580 mA h g−1 after 100 cycles at a constant current rate of 200 mA g−1 between 0.05 and 3 V (Figure 2d). The hierarchical hollow structures provide a better system of choice because they can take the advantages of both nanosized building blocks and microsized assemblies. Specifically, while the former provides significantly reduced diffusion distance and excellent lithium storage capability, the latter guarantees good stability and high packing density.

Nevertheless, the above discussed one-pot approaches for hollow structures usually require precise control of the synthesis conditions and encounter some difficulties to manipulate the shape and size. In this regard, the hard-templating method still shows certain advantages.[44] To address the drawbacks of conventional hard-templating methods, such as the tedious procedure and the incompatibility issue between the desired materials and the template surfaces, sacrificial templating (or template-engaged) approaches have been extensively explored in recent years.[51-54] Such an approach is realized based on the different physicochemical or chemical properties between the sacrificial templates and the shell materials and thus is very effective at forming hollow particles with various shapes. For example, the template-engaged redox etching of sacrificial Cu2O crystals templates with non-spherical shapes (i.e., cubic, octahedral) with FeCl3 solution gives rise to a broad family of iron-based hollow structures, such as single-/multi-shelled Fe(OH)x nanoboxes/octahedral cages, and their derived hollow structures of iron oxides (Figure 2e,f).[34] In this reaction, the hydrolysis of Fe3+ ions leads to the formation of iron-based shells, while the hollow cavities are created via simultaneous redox etching of Cu2O templates to form soluble Cu2+ ions.

Importantly, this template-engaged chemical etching approach for hollow structures can be extended to fabricate tubular structures when the proper sacrificial template is used. We further demonstrated the synthesis of polycrystalline α-Fe2O3 nanotubes with a diameter of 50–200 nm and wall thickness of 10–20 nm by combining the controlled hydrolysis of Fe3+ ions in aqueous solution and etching of pre-formed Cu nanowires by Fe3+ and Cl ions.[35] During the synthesis, Cu nanowires are oxidized by Fe3+ into Cu(I)-based soluble complexes and gradually consumed. At the same time, Fe3+ will be reduced to Fe2+ and precipitate as Fe(OH)x on the surface of the dissolving Cu nanowires, giving rise to a Fe(OH)x nanotube structure. After controlled annealing of Fe(OH)x nanotubes in air, polycrystalline α-Fe2O3 nanotubes are obtained without apparent alteration in structure (Figure 2g). The tubular structure could not only well accommodate the volume variation upon lithium insertion/extraction, but also protect the active materials from severe aggregation during repeated discharge/charge cycling. The nanoscale wall thickness and large surface area of the nanotubes also endow the material with high specific capacity and favourable response to high rate cycling due to enhanced reactivity and extremely reduced diffusion path. When evaluated as an anode material for LIBs, the α-Fe2O3 nanotubes manifest stable capacity retention of over 1000 mA h g−1 for 50 cycles (cycled between 0.01–3.0 V at 0.5 C, see Figure 2h) and exceptional high rate capability (500–800 mA h g−1 at 1–2 C). The cycling stability of the α-Fe2O3 nanotubes is much superior to that of α-Fe2O3 particles as reference under identical testing conditions.

Another intriguing approach towards hollow structures is based on the controllable decomposition or transformation of proper solid particles. During the synthesis, the pre-formed solid particles serve as both the template and precursor for the formation of hollow structures of desired materials. The available solid precursors of hollow metal oxides consist of a wide range of metal-containing salts, hydroxides, and even metal-organic frameworks (MOFs).[36, 37, 55-60] This simple strategy has been successfully applied to fabricate iron oxide hollow structures as well.[36, 37] A scalable synthesis of Fe2O3 microboxes with hierarchically structured shells was recently developed based on simple annealing of pre-formed Prussian blue (PB) microcubes.[36] By controlling the simultaneous oxidative decomposition of PB microcubes and crystal growth of iron oxide shells, scalable syntheses of anisotropic Fe2O3 hollow structures with various shell architectures are realized. The continuous decomposition of the PB microcubes is accompanied by outward gas flow, which eventually results in the formation of a relatively dense iron oxide shell and a large interior cavity (Figure 2i). With an increase in the annealing temperature to 550 °C, highly porous Fe2O3 microboxes constructed from enlarged Fe2O3 nanoparticles are obtained as a result of crystal growth (Figure 2j). Interestingly, further increasing the annealing temperature to 650 °C causes the formation of hierarchical shells consisting of Fe2O3 nanoplatelets (Figure 2k). When examined as anode materials for LIBs, all three samples display excellent cycling stability with almost no fading of capacity over the first 30 cycles (Figure 2l). In particular, the hierarchical Fe2O3 microboxes obtained at 650 °C exhibit the highest reversible capacity of 945 mA h g−1 in the 30th cycle at a constant current density of 200 mA g−1 between 0.01 and 3.0 V, while comparable capacities of 802 and 871 mA h g−1 are also obtained for Fe2O3 microboxes (Figure 2i) and porous Fe2O3 microboxes (Figure 2j), respectively. This is another good demonstration of enhancing electrochemical properties by hierarchical nanostructures. Furthermore, the Fe2O3 microboxes with high crystallinity are likely to exhibit good structural stability, which would help to retain the pristine nanostructure upon cycling. Most recently, we employed the same PB precursor to prepare various complex hollow structures, including multishelled and iron-oxide-based composite hierarchical microcubes, via a solution-based synthetic route.[37] The capability to engineering the structure and composition of the materials offers the opportunity to further optimize the electrochemical properties of these iron-oxide-based hollow structures.

4 Hybrid Nanostructures Based on Iron Oxides and Carbon

Carbonaceous materials have been widely used to improve the electrochemical performance of electrode materials for LIBs by forming desirable nanocomposites. Hybrid nanostructures of iron oxides and carbon, in which the nanostructured iron oxides are assembled onto or embedded into the conductive carbon matrix, are under intensive investigations to achieve high capacity and high rate capability. Typically, the carbon components in the hybrid materials are anticipated to serve dual functions: as conducting additives to promote the electron transport in the poorly conductive metal oxides and as elastic buffer layers/supports to enhance the stability of the electrode materials. By combining various iron oxide nanostructures and carbon materials in different ways, a wide variety of iron oxide/carbon nanocomposites has been synthesized and examined as electrode materials for LIBs, such as carbon-coated Fe3O4 nanospindles,[61] nanorods,[62] nanospheres,[63] 1D hybrid iron oxide/carbon nanowires,[64] 2D iron oxide/carbon hybrid nanosheets,[65, 66] 3D iron oxide/mesocellular carbon foam and mesoporous carbon spheres,[67, 68] Fe3O4/single-walled carbon nanotubes,[69] carbon-coated Fe2O3 hollow nanohorns on carbon nanotube backbones,[70] conformal Fe3O4 sheath on aligned carbon nanotubes,[71] graphene-wrapped iron oxides,[72-77] and graphene foams cross-linked with pre-encapsulated Fe3O4 nanospheres.[78]

Carbon nanocoating is one of the most popular and effective surface modification approaches to improve the electrochemical performance of active materials.[7-14] The carbon layers can significantly enhance the electrical conductivity of iron oxide particles, which results in improved rate performance. More importantly, the conformal and elastic carbon coating is anticipated to prevent the collapse of nanostructures and aggregation of nanoparticles during repeated charging/discharging. Carbon nanocoating can be simply achieved by hydrothermal formation of glucose-derived carbon-rich polysaccharide (GCP) layer with subsequent low-temperature carbonization. Wan and co-workers synthesized carbon-coated Fe3O4 nanospindles (Figure 3a) by in situ partial reduction of monodispersed Fe2O3 nanospindles with uniform carbon coating.[61] The TEM image (Figure 3a, inset) shows that the carbon layer is uniform and continuous with a thickness of 2–10 nm. When tested as anode materials for LIBs, the carbon-coated Fe3O4 nanospindles show a high initial reversible capacity of 745 mA h g−1 at C/5 (Figure 3b). The use of a higher rate of C/2 results in slightly reduced capacity, which remains as 530 mA h g−1 after 80 charge/discharge cycles. Obviously, the carbon-coated Fe3O4 spindles manifest much improved cycling stability and rate capability compared with α-Fe2O3 spindles and commercial Fe3O4 particles. This example evidently demonstrates that carbon nanocoating could be an effective way to improve the lithium storage properties, as the uniform and continuous carbon layers help to maintain the integrity of metal oxide particles, protect the inner active material, and increase the electrical conductivity of the composite electrodes.

Figure 3.

a) FESEM and TEM (inset) images, and b) cycling performance of the carbon-coated Fe3O4 nanospindles. c,d) FESEM and TEM (inset) images of carbon-coated Fe3O4 nanorods (c) and nanospheres (d). e) FESEM and TEM (inset) images, and f) comparative cycling performance of carbon-Fe3O4 composite nanofibers. g) FESEM and TEM (inset) images, and h) cycling performance of 2D ferrite/carbon hybrid nanosheets. i) FESEM image and j) cycling performance of hybrid Fe3O4/SWNT composite. k) TEM images and l) cycling performance of the carbon-coated CNT@α-Fe2O3 nanohorns. m) FESEM and TEM (inset) images, and n) cycling performance of GNS/Fe3O4 composite. o) FESEM and TEM (inset) images, and p) cycling performance of 3D graphene foams cross-linked with pre-encapsulated Fe3O4 nanospheres. Panels (a,b) reproduced with permission.[61] Copyright 2008, Wiley. Panel (c) reproduced with permission.[62] Copyright 2011, American Chemical Society. Panel (d) reproduced with permission.[63] Copyright 2011, American Chemical Society. Panels (e,f) reproduced with permission.[64] Copyright 2008, Elsevier. Panels (g,h) reproduced with permission.[65] Copyright 2012, American Chemical Society. Panels (i,j) reproduced with permission.[69] Copyright 2010, Wiley. Panels (k,l) reproduced with permission.[70] Copyright 2012, Royal Society of Chemistry. Panels (m,n) reproduced with permission.[72] Copyright 2010, American Chemical Society. Panels (o,p) reproduced with permission.[78] Copyright 2013, Wiley.

The above-mentioned carbon coating method usually requires an additional step to deposit a layer of carbon precursor on the surface of the pre-synthesized iron-containing materials, which complicates the synthesis procedure. It is thus more desirable to develop a one-pot strategy to prepare iron-oxide-carbon composites. We reported a simple one-step synthesis of GCP-coated ferric oxyhydroxide (FeOOH) nanorods using inexpensive iron chloride hexahydrate and glucose as precursors.[62] The corresponding carbon-coated magnetite (Fe3O4@C) nanocomposites are easily obtained by carbonizing the as-prepared GCP-coated FeOOH nanorods under inert atmosphere (Figure 3c and inset). These Fe3O4@C nanorods exhibit good cycling performance for lithium storage with reversible capacity as high as 808 mA h g−1 after 100 charge/discharge cycles at a current density of 924 mA g−1. Furthermore, a one-pot poly(acrylic acid) (PAA)-mediated solvothermal method was reported to synthesize uniform Fe3O4 nanospheres within carbon matrix support.[63] During the synthesis, the PAA molecules serve as the structure coordinating agent and also the carbon source. After carbonization at 500 °C under inert atmosphere, a thin layer of amorphous carbon with a uniform thickness of ≈6 nm is coated around Fe3O4 nanospheres with uniform diameter of 150–200 nm (Figure 3d). From the electrochemical measurement, the carbon matrix-supported Fe3O4 nanospheres demonstrate much better lithium storage properties than the carbon-free counterpart, with a high reversible capacity of 712 mA h g−1 after 60 charge/discharge cycles.

Embedding the iron oxides into a conductive carbon matrix is also effective to enhance the electrochemical properties. Wang et al. reported the synthesis of carbon-Fe3O4 composite nanofibers by electrospinning of iron-containing polyacrylonitrile (PAN)-based nanofibers and subsequent carbonization.[64] After the electrospinning and pre-oxidation step, a carbonization treatment is conducted at various temperatures. The carbon-Fe3O4 nanofibers exhibit a smooth surface and a diameter of 350 nm (Figure 3e). From the TEM image of composite nanofibers treated at 600 °C, it can be clearly seen that nanosized Fe3O4 particles with an average diameter of 20 nm are dispersed in the matrix of the carbon nanofiber (inset of Figure 3e). The carbon-Fe3O4 nanofibers carbonized at 600 °C show the highest capacity and excellent cycling stability. Specifically, after the 80th cycle, the carbon-Fe3O4 nanofibers can retain a high reversible capacity of about 1000 mA h g−1 at a current rate of 200 mA g−1 (Figure 3f). The enhanced electrochemical performance could be mainly attributed to the following reasons. First, the carbon matrix provides sufficient electrical conductivity and buffers the volume change of the Fe3O4 nanoparticles during charging/discharging. Moreover, the fibrous morphology and the small Fe3O4 nanoparticles enable high electrochemical activity by facilitating the lithium-ion diffusion.

Other low-dimensional hybrid structures of iron oxides and carbon, such as 2D nanosheets, have been also studied, which are anticipated to well alleviate the structural strain and accommodate the large volume variation upon cycling. Hyeon and co-workers reported a direct synthesis of 2D ferrite/carbon hybrid nanosheets composed of uniform-sized ferrite nanocrystals and carbon through an interesting salt-templated process.[65] In this synthesis, the surface of thermally stable salt particles is used as the template for the 2D nanostructure, while some metal-oleate complex is used as the precursor of both ferrite and carbon. In Figure 3g, it is shown that well-ordered iron oxide nanocubes with highly uniform size of 16 nm are embedded in a carbon sheet. The electrochemical tests clearly demonstrate the contribution of 2D nanosheet morphology to capacity retention during cycling. As shown in Figure 3h, stable reversible capacities of around 600 mA h g−1 are obtained for the ferrite/carbon nanosheet electrodes after 50 cycles (cycled between 0.05–3.0 V at a current density of 100 mA g−1). On the contrary, the capacity of the 3D composite electrode (carbon agglomerates embedded with spherical iron oxide nanocrystals with a diameter of ≈40 nm) drops rapidly to 323 mA h g−1. With the assistance of similar salt particle templates, Zhao and co-workers synthesized Fe3O4 nanoparticles homogeneously embedded in 2D porous graphitic carbon nanosheets.[66] The 2D composite of Fe3O4 and graphitic carbon nanosheets exhibits superior high-rate capability and excellent cycling performance. Moreover, embedding the iron oxides into 3D interconnected open pores of carbon matrix (such as mesocellular carbon foam and mesoporous carbon spheres) could also accommodate the volume expansion of iron oxides upon Li+ insertion, thus resulting in excellent electrochemical performance.[67, 68]

Carbon nanotubes (CNTs) with high conductivity and flexibility can serve as an excellent conductive network to facilitate the charge transfer and are therefore widely used to synthesize iron-oxide-based nanocomposites. For example, Dillon and co-workers reported a simple two-step process to embed Fe3O4 nanoparticles evenly in an interconnected single-walled carbon nanotube (SWNT) “net”.[69] FeOOH nanorods are employed as the precursor and assembled with SWNTs via a vacuum-filtration method. After subsequent annealing process, Fe3O4 nanoparticles embedded in an interconnected SWNTs network are obtained (Figure 3i). When tested as a binder-free electrode, the hybrid Fe3O4/SWNTs film shows a high reversible capacity of ≈1000 mA h g−1 at 1 C rate and excellent rate capability (Figure 3j). Even for 5 μm Fe3O4 microparticles incorporated into SWNTs network via the same method, the as-prepared composite electrode still shows a stable capacity of 600 mA h g−1 for 50 cycles, which is much superior to that of the pure Fe3O4 microparticle electrode without SWNTs. This simple yet versatile approach demonstrates that a conductive and flexible network can effectively improve the electrochemical performance of iron oxides.

Our recent study further integrates several design rationales, namely hollow structure, carbon nanocoating and conductive network, to prepare iron-oxide-based hybrid nanostructured materials. As an example, we reported a novel hierarchical nanostructure composed of carbon-coated α-Fe2O3 hollow nanohorns grafted on CNT backbones (carbon-coated CNT@Fe2O3, see Figure 3k) via the bottom-up assembly of β-FeOOH nanospindles on CNTs, followed by subsequent thermal treatment and carbon nanocoating.[70] Such a hierarchical structure possesses several merits for efficient lithium storage. Specifically, the highly conductive and flexible CNT backbones provide a 3D conducting network to facilitate the charge transfer. Moreover, the hollow α-Fe2O3 nanohorns grafted on the CNT backbone exhibit enhanced electrochemical activity and mechanical integrity of the electrode due to their large surface area, nanoscale diffusion length, and sufficient internal void space. Furthermore, the outmost continuous carbon nanocoating around the overall architecture serves as a structural buffering layer to further cushion the internal strain associated with lithium uptake while preventing the Fe2O3 nanostructures from being electrically isolated upon cycling. With these beneficial structural features, carbon-coated CNT@Fe2O3 manifests stable capacity of ≈800 mA h g−1 for 100 cycles at a current density of 500 mA g−1 (Figure 3l). The comparative studies on bare CNT@Fe2O3 without carbon coating and α-Fe2O3 nanoparticles clearly demonstrate the advantageous effects of CNT backbones and carbon nanocoating, which not only enhance the electrochemical activity of iron oxide, but also improve the capacity retention upon prolonged cycling. Recently, Wang and co-workers reported the synthesis of conformal Fe3O4 sheath on aligned carbon nanotube scaffolds by magnetron sputtering methods, which also exhibits remarkable lithium storage properties.[71] These examples evidently demonstrate that hybridization of iron oxides and CNTs could effectively improve the lithium storage properties of iron-oxide-based anode materials.

Graphene has become the most popular carbon material in recent years due to its exceptional physical/chemical properties, such as high electrical and thermal conductivity, large specific surface area, light weight, high mechanical strength, and flexibility. These properties of graphene have driven much effort to design and synthesize hybrid nanostructures based on graphene for various applications.[79-83] When hybridized with iron oxides for reversible lithium storage, graphene facilitates electron transport, and provide large surface area and possibly a porous framework to anchor active materials, thus enhancing the electrochemical activity. Moreover, the “flexible confinement” function could compensate for the volume change and prevent the detachment and agglomeration of pulverized iron oxide, thus effectively prolonging the cycle life of the electrode. For example, Cheng and co-workers reported the synthesis of a well-organized flexible interleaved composite of graphene nanosheets (GNSs) and decorated Fe3O4 particles through in situ reduction of iron hydroxide between GNSs.[72] A typical cross-section FESEM image shows that the GNS/Fe3O4 composite possesses a layer-by-layer assembled structure consisting of GNSs and Fe3O4 particles (Figure 3m). It can also be seen that each Fe3O4 particle is wrapped by GNSs, which helps to prevent Fe3O4 from agglomeration and enables a good dispersion of these oxide particles over the graphene support. As shown in Figure 3m (inset), the short-range disordered structure observed at the interfacial region of GNS and Fe3O4 nanoparticle suggests the possible formation of interfacial bonds for stabilization of the oxide particles on the graphene support. The GNS/Fe3O4 composite shows a reversible specific capacity over 1000 mA h g−1 after 30 cycles at 35 mA g−1 (Figure 3n). The beneficial effect of graphene has also been proved by the composite material of Fe2O3 and graphene, in which enhanced electrochemical performance could be achieved by hybridization of Fe2O3 nanoparticles and graphene.[73]

Very recently, Müllen and co-workers reported a novel strategy to fabricate 3D graphene foams cross-linked with Fe3O4 nanospheres (NSs) encapsulated with graphene.[78] The Fe3O4 NSs are wrapped by graphene sheets (denoted as Fe3O4@GS) and further confined within continuous graphene foam networks (denoted as Fe3O4@GS/GF, see Figure 3o). In such hierarchical Fe3O4/graphene hybrids, the graphene shell suppresses the aggregation of Fe3O4 NSs and buffers the volume expansion, while the interconnected 3D graphene network acts to reinforce the core-shell structure of Fe3O4@GS and further enhances the electrical conductivity of the whole electrode. As a result, the Fe3O4@GS/GF electrode delivers a high reversible capacity of 1059 mA h g−1 over 150 cycles at a current density of 93 mA g−1 (Figure 3p). The flexible, ultrathin, and hierarchical graphene matrix provides double protection against the aggregation and volume change of iron oxides and ensures favorable transport kinetics for both electrons and lithium ions.

5 Conclusions and Outlook

This article highlights the recent advances on the rational design and synthesis of nanostructured iron oxides and hybrid nanomaterials composed of iron oxides and carbon as negative electrodes for lithium-ion batteries. The advantageous effects of proper nanostructure engineering on their electrochemical performance are discussed in detail. Specifically, we have reviewed various nanostructured iron oxides with different shapes, morphologies and textures, such as 1D nanowires and nanotubes, 2D nanosheets and nanoflakes, 3D porous/hollow/hierarchical architectures, and their hybrids with different carbon materials (including amorphous carbon, carbon nanotubes, and graphene). The representative iron-oxide-based anode materials for lithium-ion batteries are summarized in Table 1. Our ultimate aim is to provide some general guidelines for designing advanced electrode materials for high-performance LIBs, although this may be very challenging. In general, these iron oxide nanostructures possess the advantages of enhanced electrochemical activity, reduced ionic/electronic diffusion length, and capability to withstand the internal strain, which mainly originate from the nanometer-sized building blocks, unique geometric shape, and the presence of porosity/void space. Importantly, the lithium ions uptake and extraction in nanostructured iron oxides become much more facile and reversible compared with their bulk counterparts. Therefore, high specific capacity, improved rate performance, and prolonged cycling stability can be achieved in iron oxide nanomaterials. In particular, hollow structured iron oxides with nanoscale building blocks stand out as an appealing family of micro-/nanostructured materials for efficient lithium storage. Further hybridizing the nanostructured iron oxides with various carbonaceous supports provide additional electrical conductivity and structural stability, which are beneficial for high rate charging/discharging and long cycle life.

Table 1. Summary of the representative iron oxide-based anode materials for lithium-ion batteries
StrategiesTypical examplesElectrochemical propertiesRef.
NanostructuringPorous α-Fe2O3 microparticles662 mA h g−1 after 100 cycles at a current density of 200 mA g−1[21]
 α-Fe2O3 nanowires456 mA h g−1 after 100 cycles at 0.1 C[24]
 Microflower-like α-Fe2O3 constructed by sheet-like subunits929 mA h g−1 after 10 cycles at a current density of 100 mA g−1[26]
 Single crystalline α-Fe2O3 nanosheets grown directly on Ni foam518 mA h g−1 after 50 cycles at 0.1 C[28]
 Single-crystalline α-Fe2O3 nanosheets on conductive substrates700 mA h g−1 after 80 cycles at a current density of 65 mA g−1[29]
Hollow structuresHierarchical α-Fe2O3 hollow spheres with sheet-like subunits710 mA h g−1 after 100 cycles at a current density of 200 mA g−1[32]
 Hierarchical hollow Fe3O4 microspheres580 mA h g−1 after 100 cycles at a current density of 200 mA g−1[33]
 Polycrystalline α-Fe2O3 nanotubes1000 mA h g−1 after 50 cycles at 0.5 C, 500–800 mA h g−1 at 1–2 C[35]
 Hierarchical Fe2O3 microboxes945 mA h g−1 after 30 cycles at a current density of 200 mA g−1[36]
Carbon nanocoatingCarbon-coated Fe3O4 nanospindles530 mA h g−1 after 80 cycles at 0.5 C[61]
 Carbon-coated Fe3O4 nanorods808 mA h g−1 after 100 cycles at a current density of 924 mA g−1[62]
 Fe3O4 nanospheres within carbon matrix support712 mA h g−1 after 60 cycles at a current density of 200 mA g−1[63]
Embedding into carbon matrixCarbon-Fe3O4 composite nanofibers1000 mA h g−1 after 80 cycles at a current density of 200 mA g−1[64]
 2D ferrite/carbon hybrid nanosheets600 mA h g−1 after 50 cycles at a current density of 100 mA g−1[65]
Hybridization with carbon nanotubesFe3O4 nanoparticles embedded in an interconnected SWNTs network1000 mA h g−1 after 50 cycles at 0.5 C[69]
 Carbon-coated α-Fe2O3 hollow nanohorns grafted on CNT backbones800 mA h g−1 after 100 cycles at a current density of 500 mA g−1[70]
Hybridization with grapheneInterleaved composite of graphene nanosheets and decorated Fe3O4 particles1000 mA h g−1 after 30 cycles at a current density of 35 mA g−1[72]
 3D graphene foams cross-linked with Fe3O4 nanospheres encapsulated with graphene1059 mA h g−1 after 150 cycles at a current density of 93 mA g−1[78]

Meanwhile, one should be also aware of some drawbacks of nanostructured iron oxides as electrode materials, such as the relatively low thermodynamic stability, possible surface side reactions, high processing cost, and low volumetric energy density. Some of these detrimental effects could be mitigated by using proper hierarchical micro-/nanostructures with micrometer size, reduced external surface, and optimized porosity. For example, multishelled hollow structures allow effective utilization of the interior void space compared with conventional single-shelled hollow particles. Nevertheless, certain compromise would still be necessary since the nanoscale building blocks and high porosity are the main origins of the exceptional electrochemical performance of nanostructured materials. Additionally, the nature of carbonaceous materials and the configuration of hybrid structures should be properly chosen and optimized since they would also affect the electrochemical properties of iron oxide/carbon nanocomposites, such as the conductivity and structural stability. Considering that these carbonaceous additives are mostly inactive during the electrochemical processes, their content should be carefully controlled in order to obtain optimal performance.

To promote the development of high-performance iron-oxide-based electrodes, the elegant combination of structural and compositional engineering will be essential. In addituion to forming hybrid nanostructures with carbonaceous materials, complex metal oxides, either as ternary metal oxides or heterostructures of two simple metal oxides, would be another promising avenue for compositional engineering. Developing ternary ferrites and iron oxide-based heterostructures could possibly tune the inherent physical/chemical properties of iron oxides, some of which could be hardly achieved if merely relying on structural engineering. Furthermore, demonstration of full LIB cells based on iron-oxide-based anodes with truly durable lifespan would be highly anticipated to manifest the feasibility of practical applications. Some obstacles that hinder the realization of full cells should be addressed as well, such as the relatively high voltage plateau and low initial Coulombic efficiency. Proper compositional engineering would probably modify the working potential and reduce the polarization of the iron-oxide-based anodes. Nevertheless, LIB full cells with high energy density would depend highly on the development of suitable high voltage/capacity cathodes. Meanwhile, more efforts need to be devoted to develop facile, feasible, and efficient methods to reduce the initial capacity loss, such as pre-lithiation and surface modification. With the rapid development of high-performance nanomaterial-based electrodes, one can confidently anticipate that these intriguing anode materials will greatly boost the development of advanced rechargeable LIBs with high energy/power density, long cycle life, low cost, and improved safety for the upcoming large-scale applications. Among the many choices, it is believed that iron-oxide-based materials stand a good chance.