Electrospun Nanostructured Iron Oxides for High-Performance Lithium Ion Batteries

Abstract

The versatility and ability of formation of one-dimensional nanostructures make electrospinning one of the promising methods to fabricate the components of energy storage devices. Electrospinning is widely explored in lithium ion batteries (LIBs) for the development of electrodes as well as electrolytes. Electrospun electrodes, both anodes and cathodes, are widely studied because they can deliver excellent electrochemical performance and are stable and freestanding even without using binders. For the best performance of LIBs, different materials were examined as electrodes, especially as the anodes, which serve as the negative electrode in LIBs. Metal oxides and carbon are considered to be the most favored anode material in LIBs. Among the different metal oxides studied, iron oxide-based anodes are emerging materials under serious research due to its superior properties which make them to deliver better properties than other materials commonly used. High specific capacity and large lithium storage through a conversion mechanism make iron oxide a promising anode than that of carbon-based ones. However, poor capacity retention, higher operating potential, and large polarization limit its application in LIBs. Electrospinning can resolve these issues by the formation of nano-sized structures. For achieving the better and best electrochemical properties, different structural and morphological modifications were carried out using electrospinning. Porous/hollow iron oxide-based nanotubes and nanorods are found to be delivering good electrochemical performance than the pristine-based anodes. Further enhancement is also done by the formation of composites with conducting dopants. The significance of iron oxide-based anodes and the methods for the enhancement of its electrochemical properties will be discussed in detail in this chapter.

10.1 Introduction

The rechargeable Li-ion battery (LIB) has attracted intensive research interest due to their large spectrum of applications as energy storage devices for electric, electric/hybrid electric vehicles, and intermittent renewable energy sources [1,2,3,4,5]. The LIB is referred to as a rocking-chair battery, because Li⁺-ions “rock” back and forth between the anode and the cathode during cycling and they possess high energy and power densities, no memory effect and long cycle life. The current generation commercial LIBs are utilizing the electrode materials, which could store Li⁺-ions by insertion between their structural layers during charging and extracted out from the layers during discharging (Fig. 10.1) without any significant structural change leading to excellent cycling performance. In recent years, globally great efforts have been paid by the researchers and battery technologists to develop and design high-performance electrode materials in terms of energy density, cycling stability and rate capability. Among different classes of anode materials such as intercalation, conversion reaction, alloying/de-alloying reaction-type materials, transition metal oxides, Co₃O [6,7,8,9,10], FeOx [11,12,13,14,15,16,17], TiO₂ [18, 19], MnO₂ [20,21,22], and SnO₂ [23,24,25], iron oxide micro-/nanomaterials, such as hematite (α-Fe₂O₃) [26,27,28] and magnetite (Fe₃O₄) [29,30,31], have been extensively studied as potential electrode materials in LIBs. Iron oxide is popular in their higher theoretical capacities (1004 mAh g⁻¹ for α-Fe₂O₃ and 924 mAh g⁻¹ for Fe₃O₄), has low toxicity, and is economically viable.

Fig. 10.1

Schematic representation and operating principle of lithium ion batteries

In spite of their lower cost and better safety, the capacity retention of Fe₂O₃/Fe₃O₄ remains a major drawback, due to the huge volumetric expansion/contraction during the lithiation/delithiation process which ultimately leads to pulverization of the electrode from the current collector, resulting in loss of electrical contact and loss of morphological structure of the active material [32]. The electrochemical performance is highly dependent on their diverse morphologies and micro-/nanostructure of an anode material [33,34,35,36,37] The confining dimension effect and high surface area of the nanostructured materials lead to the short lithium diffusion lengths and increased active sites for Li⁺-ion insertion/extraction reactions [38]. Also, the cyclability of the nanostructured electrodes significantly improved due to the sufficient free spaces to relax the large volume changes during the continuous charge–discharge process [29]. Hence, various types of nanostructures have been employed as anode materials for LIBs. In order to mitigate the large-volume variation problem of transition metal oxides and to increase the electronic conductivity, carbon coatings have been extensively explored [39,40,41,42,43,44,45,46,47,48,49,50]. However, the compact carbon coatings on nanoparticles cannot allow residual buffer space to accommodate the large volume change of Fe₃O₄ nanoparticles during Li⁺ insertion/extraction. Thus, it remains necessary to exploit an approach for the fabrication of suitable carbon matrix to accommodate volume expansion upon Li⁺-ion insertion as well as to increase the electronic conductivity [51,52,53]. One-dimensional (1D) structure can efficiently improve the performance of Fe₂O₃ as anode material in LIBs due to its excellent electron transport along the lengthways direction and large surface–volume ratio [54,55,56,57,58]. Electrospinning is now a convenient, inexpensive, simple, and versatile method to manufacture the 1D structure including polymer, metal oxide, and organic–inorganic composites [59,60,61,62], especially the robust electrode for LIBs. The material obtained by electrospinning can take full advantage of 1D architectures as well as the material can form metal oxide nanoparticles/carbon nanofibers (CNFs) after being calcined under an inert atmosphere. Uniformly dispersing the nano-sized metal oxide into CNFs matrix can significantly enhance the electronic conductivity, buffer the large volume change and pulverization of the electrode, and prevent the agglomerates of nanoparticles [63,64,65,66,67,68,69,70,71]. The electrospinning technique facilitates to develop 1D metal oxide nano-/microstructure with various morphologies including porous nanowires, nanotubes, nanorods, and tube-in-tube by using non-coaxial electrospinning. This chapter is presenting a detailed overview on the facile fabrication and electrochemical performance of hierarchal Fe₂O₃/Fe₃O₄ nanostructured anode for LIBs.

10.2 Principle of Lithium Ion Batteries

The global lithium-ion battery market is forecasted to grow from USD 36.20 billion in 2018 to USD 109.72 billion by 2026, at a compound annual growth rate (CAGR) of 13.4%, during the forecast period. In recent years, lithium-ion batteries are increasingly being used as the power source for hybrid (HEV) and full-battery electric vehicle (BEV). Over the past couple of years, maximum sales of electric vehicles have accounted by China, the USA, and the European region, which are primarily high-end electric vehicles. Roughly, 1.6 million electric cars were on the roads in China in 2018, followed by 810,000 in the USA. By March 2018, BEV production and sales in China reached 27,673 and 24,127 units, rising 88.35 and 69.21% year-on-year; such figures for plug-in hybrid electric vehicle (PHEV) were 11,210 and 11,171 units, rising 291.21 and 201.47% year-on-year [72]. The industry produced about 660 million units of cylindrical lithium-ion cells in 2012; the 18650 size is by far the most popular for cylindrical cells. The Tesla’s Model S electric cars SUVs under 40,000 USD with 85 kWh battery uses 7,104 of lithium ion cells. A 2014 study projected that the Model S alone would use almost 40 percent of estimated global cylindrical battery production during 2014. Production of the cell was gradually shifted to higher-capacity 3000+ mAh cells.

Based on the use type, LIBs are categorized as primary LIBs and secondary or rechargeable LIBs. A primary battery is one-direction galvanic device designed to be used once and discarded when it is fully discharged, and not recharged with electricity and reused like a secondary or rechargeable battery; i.e., the electrochemical reaction occurring in the cell is not reversible, or it has only discharging process. Lithium primary battery has metallic lithium as anode. Hence, these types of batteries are also referred to as lithium-metal batteries. Presently represent the primary EES systems, with a production higher than 100 million cells/month and about 1500 tons/month of electrode materials. Lithium–manganese dioxide, lithium iron disulfide, lithium thionyl chloride, and lithium iodine batteries are the common lithium primary batteries. Among different lithium primary batteries, lithium thionyl chloride battery has the highest energy density of all lithium-type cells and has a service life of 15–20 years, while lithium iodine batteries provide excellent safety and long service life. In batteries, during discharging, reduction happens on the cathode gaining electrons and oxidation happens on the anode, which is losing electrons, as per the electrochemical reaction shown below [73].

$$ \begin{array}{*{20}l} {{\text{Cathode:}}} \hfill & {{\text{MS}}_{2} + {\text{Li}}^{ + } + {\text{e}}^{ - } } \hfill & {\xrightarrow{{{\text{discharge }}}}} \hfill & {{\text{LiMS}}_{2} } \hfill \\ {{\text{Anode:}}} \hfill & {{\text{Li}}} \hfill & {\xrightarrow{{{\text{discharge }}}}} \hfill & {{\text{Li}}^{ + } + {\text{e}}^{ - } {\text{ }}} \hfill \\ {{\text{Full cell:}}} \hfill & {{\text{Li}} + {\text{MS}}_{2} } \hfill & {\xrightarrow{{{\text{discharge }}}}} \hfill & {{\text{LiMS}}_{2} } \hfill \\ {} \hfill & {} \hfill & {} \hfill & {({\text{M}} = {\text{Ti or Mo}})} \hfill \\ \end{array} $$

In contrast to lithium primary batteries, lithium secondary batteries, referred as lithium-ion batteries, are rechargeable batteries in which lithium ions move from the negative electrode to the positive electrode during discharge and opposite action happens during charging. Research on LIBs started in the early 1980s, and the principle of the current LIB was completed in 1985 and then first commercialized in 1991 by Sony. Most of the technological developments to date have been directed toward the needs of portable electronics, but now the focus tends to be on the performance demands of medium- and large-scale applications. As shown in Fig. 10.1, typically, LIB consists of three layers: (i) cathode or positive electrode which commonly consists of LiCoO₂ [74, 75], LiNiO₂ [76], LiMn₂O₄ [77], etc., (ii) anode or negative electrode consists of graphitic carbon [78], TiO₂ [79], Fe₂O₃/Fe₃O₄ [80], etc., and (iii) a separating cum electrolyte called gel polymer electrolytes (GPEs) which is permeable to the ions and the electrolyte (e.g., LiPF₆ in an organic solvent). GPEs are prepared by immobilization of organic liquid electrolytes, e.g., a 1M solution of LiPF₆, LiClO₄, or LiTFSI into polymer structures [81,82,83]. Polymers such as polyethylene oxide (PEO) [84, 85], polyacrylonitrile (PAN) [86, 87], polyvinylidene difluoride (PVdF) [88, 89] and its copolymer polyvinylidene difluoride-co-hexafluoropropylene (PVdF-co-HFP) [90, 91], and polymethyl methacrylate (PMMA) [92, 93] are among the well-studied materials. As the name implies, the working of a lithium-ion battery mainly relies on repeated transfer of lithium ions between the anode and the cathode. The electrochemical properties of the electrodes are strongly influenced by the physical and chemical properties of the electrode active material, such as particle size, homogeneity, morphology, and surface area. Lithium-ion polymer batteries (LiPo batteries) are by far the most common commercialized secondary cell polymer battery, with leading technology among other types of metal-ion polymer batteries. A LiPo battery is a rechargeable battery of lithium-ion technology using a polymer electrolyte instead of a liquid electrolyte. LIBs are able to supply continuous energy due to the spontaneous oxidation–reduction reactions occurring at the electrodes. During the charging process (delithiation), Li⁺ ions are extracted from the cathodic material by supplying energy by an external source. The extracted Li + ions diffuse in the electrolyte and enter the anodic material (according to the reaction C_x + LiMO₂ →Li_(1−y)MO₂ + C_xLi_y, in the case of a traditional LIB [94, 95]), while electrons are simultaneously the electrons transferred to the positive electrode through the external circuit. In the discharge process (lithiation), the opposite process takes place (i.e., Li_(1−y)MO₂ + C_xLi_y →C_x + LiMO₂, in the considered example): i.e., Li⁺-ions, extracted from the anodic material, are re-inserted into the cathodic material, and the cell provides energy. The oxidation and reduction process occurred at two electrodes in the lithium rechargeable batteries as shown below [96, 97].

$$ \begin{array}{*{20}l} {{\text{Cathode:}}\;{\text{LiMn}}_{2} {\text{O}}_{4} \to {\text{Li}}_{1 + x} {\text{Mn}}_{2} {\text{O}}_{4} + x{\text{Li}}^{ + } + x{\text{e}}^{ - } } \hfill \\ {{\text{Anode:}}\;x{\text{Li}}^{ + } + x{\text{e}}^{ - } + {\text{C}}_{6} \to {\text{Li}}_{x} {\text{C}}_{6} } \hfill \\ {{\text{Full Cell:}}\;{\text{LiMn}}_{2} {\text{O}}_{4} + {\text{C}}_{6} {\text{LiC}}_{6} + {\text{LiMn}}_{2} {\text{O}}_{4} } \hfill \\ \end{array} $$

Each combination of the aforementioned materials and compound will slightly influence cost, voltage, cycle durability, and other characteristics of the LiPo batteries. The secondary lithium-ion batteries, in general, operate 3.7 V and demonstrate a capacity of 150 mAh g⁻¹ [98].

10.3 Electrode Materials for Lithium Ion Batteries

LIB primarily contains four essential components, namely the anode, the electrolyte, the separator, and the cathode. Typically, LIBs use an intercalated lithium compound as the positive electrode and graphite as the negative electrode. However in the earliest configuration of LIBs metallic lithium or Li–Al alloys was used as the negative electrode, with a variety of chalcogenides (TiS₂, MoS₂, etc.), [73] as the positive electrode in several prototypes and commercial products. Due to safety concerns, lithium metal as an anode material in rechargeable batteries was ultimately rejected. The unavoidable dendrite growth on the lithium metal surface during the repeated cycling cause lithium plating that leads to internal short circuits.

10.3.1 Positive Electrode (Cathode) for Lithium Ion Batteries

The cathode material requires a stable crystalline structure over wide ranges of composition because during the process of lithiation, the oxidation reaction leads to large compositional changes and therefore to unfavorable phase changes [99]. Also, the cathode performance directly depends not only on the electrode microstructure and morphology, but also on the inherent electrochemical properties of the cathode material due to the fact that Li⁺-ion exchange with the electrolyte only happens at the electrode–electrolyte interface [100, 101]. The development of electrochemically stable LiCoO₂ as a positive electrode leads to the commercialization of the lithium ion battery by Sony, Japan, paved a path to hunt novel electrode materials which provided a step change in the approach to the development of advanced energy storage based on lithium technology [102, 103]. Combining LiCoO₂ with graphitic carbon which provided a host for Li⁺-ions at low potential thus successfully removed metallic lithium from the LIBs. The domination of LIBs as the power source in the portable electronic and automobile market leads to the subsequent improvements in the LIBs, which forced the scientific community to focus on developing new cathode materials; thus, LMO, NMC, LFP, etc., were introduced. But recently, the introduction of nanocomposites comprised of Sn (theoretical capacity 992 mAh g⁻¹) or Si (theoretical capacity 4200 mAh g⁻¹) led to major developments in anode materials, which require higher-capacity cathode materials to provide optimum utilization of the storage properties.

In LIBs, cathode materials can store energy through two different electrochemical reaction mechanisms, (i) intercalation and (ii) conversion reaction [100]. Conversion-type cathodes undergo a solid-state redox reaction during lithiation/delithiation process, in which there is a change in the crystalline structure, accompanied by the breaking and recombining chemical bonds, while the intercalation cathode materials act as a host for Li⁺-ions, so that the ions can insert in or extracted out from the cathode material reversibly. Metal halides such as FeF₂, CoFe, and NiF₂ are examples of conversion-based cathode materials. Due to the high volume expansion, poor electronic conductivity, and hysteresis issues, development of conversion-based cathode materials has faced a lot of challenges [100].

Intercalation-based cathode materials are mainly divided into three categories: chalcogenides, transition metal oxides, and polyanion compounds. Due to the higher operating voltage and higher specific capacity, most of research on intercalation cathode materials is focused on transition metal oxides [100]. LiCoO₂, LiNiO₂ (LNO) [104], LiNi_xMn_yCo_zO₂ (with x + y + z = 1 or NMC) [105, 106] are some of the examples for the transition metal oxide-based cathodes in LIBs. Even though the layered crystal structure of LNO is similar to LCO, LNO delivers 20–30% more reversible capacity than LCO, but due to its inherent electrochemical properties [104].

Over-lithiated oxides (OLOs) are relatively new replacement material for the cathode in high-capacity LIBs due to their very high capacity over 250 mAh g⁻¹ at high-voltage charge over 4.5 V along with many other enhanced properties. OLOs have been developed with a variety of stoichiometric variations of the general composition: Li₂MnO₃-(1−x)LiMO₂ [107]. The major limitation of OLOs is their initial irreversibility caused by Li₂O formation [105]. Another class of important cathode materials is spinel oxides with a general formula of AB₂O₄. The most famous cathode in this group is LiMn₂O₄ (LMO), which is a low-cost, reliable, non‐toxic, and a high electrochemical potential material, and delivers a practical capacity of 100–120 mAh g⁻¹ (theoretical specific capacity is 148 mAh g⁻¹); however, its reversible capacity is less than that of LCO or LNO. The spinel structure of LMO creates a 3D framework which promotes the easy movement of Li⁺-ions [105, 108] and undergoes less damage during continuous charge–discharge cycles in comparison with the anode materials having layered structure, because the continuous two-way transportation of Li⁺-ions in the spinel structure does not make large volume change at room temperature. However, at high-temperature LMO shows poor cycle as well as calendar life [105]. The manganese dissolution, structural fatigue, and microcracks are other issues associated with LMO, which lead to poor cycling stability and capacity fading. Because of its low specific capacity and low practical capacity, LMO is not recommended to be used as a single cathode material. It also suffers severely from the Jahn–Teller effect, which is a geometric distortion of a nonlinear molecular system that reduces its symmetry and energy. Hence, the cathode material in the new generation of commercially produced batteries consists of a complementary blend of spinal LMO and layered NMC [109].

LiFePO₄ (LFP) olivines are polyanionic compounds developed by John B. Goodenough, who received Nobel Prize, in 2019, at the University of Texas in 1996, and that have attracted a lot of attentions due to their thermal stability, environmental friendliness, very flat potential during charge–discharge processes, and high-power capabilities [105]. Again even in harsh environments, the release of oxygen from the active cathode material is inhibited due to the strong P–O bond in phosphate and this structurally stable material guarantees their safety [110, 111].

10.3.2 Anode Materials for Lithium Ion Batteries

Traditionally in commercial LIB anode, the negative electrode from which electrons flow out toward the external part of the circuit is constructed from graphite and other carbon materials coated on a thin copper foil current collector. However, they suffer from serious safety problems, which have hampered their further development [98, 112]. To replace conventional carbon-based anodes, several studies have focused on exploiting novel anode materials. As a result of extensive research, the anode in LIBs can be fabricated from three distinguished groups of materials having very different electrochemical energy storage mechanisms broadly categorized as (i) intercalation-based materials, (ii) conversion-reaction-based materials, and (iii) alloying-reaction-based materials. The properties of most common anode materials used for lithium-ion batteries are summarized in Table 10.1.

Table 10.1 Summarization of most common anode materials used for lithium-ion batteries

10.3.2.1 Intercalation-Based Materials as Anodes in Lithium Ion Batteries

Graphite in its natural or artificial form is the best representative material in the intercalation-based anode materials first introduced by Rachid Yazami, a Moroccan scientist and engineer in 1980. He established the reversible intercalation of lithium into graphite in an electrochemical cell using a polymer electrolyte. Eventually, his discovery led to the lithium–graphite anode now used in commercial lithium-ion batteries, over US$20B value market. In the intercalation-based anode materials, the lithium ions are electrochemically intercalated into the space between the layers of the active materials. Lithium-ion intercalation in graphite can be described by the equation

$$ x{\text{Li}}^{ + } + {\text{C}}_{6} ({\text{in graphite}}) + x{\text{e}}^{ - } \longleftrightarrow {\text{Li}}_{x} {\text{C}}_{6} $$

The reversible lithiation/delithiation (intercalation/deintercalation) reaction proceeds less than 0.25 V versus Li/Li⁺, with a practical reversible capacity greater than 360 mAh g⁻¹ (theoretically at 372 mAh g⁻¹ or 975 mAh cm⁻³) with high Coulombic efficiency approaching 100% [113, 114]. However, one of the drawbacks with graphitic anode is that some irreversible reactions happen during the first charge (lithiation) process causing a cathodic decomposition of some constituents of the electrolyte. Another major downside of graphite anodes is their low specific capacity, which is addressed in great extent by increasing the surface area of the carbonaceous materials; therefore, the active material can provide more space for the intercalation of Li⁺-ions between the graphitic layers leading to the higher specific capacity. Different carbon allotropes such as carbon nanotubes (CNTs), buckminsterfullerene (buck balls), and graphene nanosheets (GNSs) [115] or carbon nanofibers (CNFs) [116] are vastly studied as an alternative to graphite due to their larger surface area as well as higher electronic conductivity which makes them suitable for high rate charging/discharging [117]. Single-wall CNTs are expected to exhibit reversible capacities somewhere around 300–600 mAh g⁻¹ [118], and for graphene the theoretical capacity is about 744 mAh g⁻¹. A reversible specific capacity is as high as 1264 mAh g⁻¹ at a current density of 100 mA g⁻¹, and a capacity retention of 718 mAh g⁻¹ is reported even at a high current density of 500 mA g⁻¹ [115, 116, 119].

10.3.2.2 Conversion Reaction-Based Materials as Anodes in Lithium Ion Batteries

The simple conversion electrochemistry of transition metal oxides (TMOs), sulfides, phosphides, and similar compounds of p-block metalloids shares some interesting and useful electrochemical features with other anode materials [120]. It was already well understood, and different types of anode electrochemistries that qualify the required norms set for the battery application, namely the operating potential of <2 versus Li/Li⁺, are categorized as intercalation, alloying, and conversion types. The conversion reaction-based materials are based on the Faradaic reaction represented as follows.

$$ {\text{M}}_{a} X_{b} + (b \cdot n){\text{Li}}^{ + } + a{\text{e}}^{ - } \longleftrightarrow a{\text{M}} + b{\text{Li}}_{n} X $$

where M is the transition metal such as Ti, V, Cr, Mn, Fe, Co, Ni, Cu, Mo, W, and Ru, X is the anion such as O, N, F, S, and P, and n is the number of negative charges of X [121]. A variety of anode materials are possible in this group by a simple combination of candidates of M and X even without considering multiple oxidation states of M. The theoretical capacity of the conversion-reaction-based anode materials ranges from 350 mAh g⁻¹ for Cu₂S to 1800 mAh g⁻¹ for MnP₄ [121]. The relatively high theoretical capacity of conversion reaction-based compounds as compared to graphite (372 mAh g⁻¹) makes these materials as ideal anode materials. However, compared to graphitic carbon, these materials have some major down steps including lower Coulombic efficiency, electronic conductivity, inferior cycling stability, and rate capability, which must be dealt with before being used as anode material. In addition, the conversion-reaction-based anode materials undergo large volume changes during lithiation and the following delithiation, which could lead to pulverization or electric isolation leads to the fast capacity fade under cycling. This means that the conversion reactions in conversion-reaction-based anodes would have intrinsically limited reversibility [122, 123]. Again the reaction potentials of conversion-reaction-based anode materials (E_conv), at which the potential profiles reach a plateau, are relatively higher than that of graphite. Higher E_conv leads to lower cell potential which results in a lower energy density than expected only from capacities [121]. Although intercalation and alloying materials have received the preferred focus by the battery technologist, conversion materials have thus far been left out for any type of practical industrial applications owing to the aforementioned specific shortcomings [124].

10.3.2.3 Alloying Reaction-Based Materials

The third category of anode materials after intercalation and conversion reaction-based materials is the alloying reaction-based materials. This group consists of metals that can be alloyed with lithium such as silicon (Si), germanium (Ge), tin (Sn), and their alloys [114, 125]. In this category, the Li ions are inserted into the structure of anode material during the charge cycle, making an alloy with the anode. The reversible alloying reaction is shown in Eq. 4, where M is the anode material [125].

$$ {\text{M}} + x{\text{Li}}^{ + } + x{\text{e}}^{ - } \longleftrightarrow {\text{Li}}_{x} {\text{M}} $$

Alloying reaction-based materials are most famous for their high theoretical capacity: 4200 mAh g⁻¹ for Si in Li_4.4Si, 1600 mAh g⁻¹ for Ge in Li_4.4Ge, 993 mAh g⁻¹ for Al in LiAl, 992 mAh g⁻¹ for Sn in Li_4.4Sn, and 660 mAh g⁻¹ for Sb in Li₃Sb [126]. However, the major disadvantage of these materials is their extremely large volume change during charge and discharge [118]. They experience serious pulverization resulting in electrical isolation of the active materials from electric contact with the conducting agent (carbon black) and the current collector. Among the alloying elements which can be used in this group of anode materials, the vast majority of research and development has been focused on silicon because of its highest capacity and its most serious detrimental volumetric change [126,127,128,129].

10.4 Electrospun-Based Iron Oxide Anodes for Lithium Ion Batteries

Iron oxide-based nanocomposites including Fe₃O₄/carbon nanocomposite [43, 49, 130, 131] and Fe₂O₃/carbon nanocomposites [132,133,134] have been fabricated and studied extensively as electrode materials for lithium-ion batteries. Iron oxide is a paramagnetic material which has only two unpaired electrons. Because the lower number of unpaired electrons irons oxide is less magnetic than iron, which has four unpaired electrons. Iron (III) oxide or ferric oxide is the inorganic compound with the formula Fe₂O₃. It is one of the three main oxides of iron, the other two being iron (II) oxide (FeO), which is rare, and iron (II,III) oxide (Fe₃O₄), which also occurs naturally as the mineral magnetite. Fe₂O₃ can be obtained in various polymorphs. In the main ones, α and γ, iron adopts octahedral coordination geometry; i.e., each Fe center is bound to six oxygen ligands. α-Fe₂O₃ has the rhombohedral, corundum (α-Al₂O₃) structure and is the most common form, which occurs naturally as the mineral hematite while γ-Fe₂O₃ has a cubic structure and occurs naturally as the mineral maghemite. γ-Fe₂O₃ is metastable and converted from the alpha phase at high temperatures. γ-Fe₂O₃ is ferromagnetic, and however ultrafine particles smaller than 10 nm are superparamagnetic. α-Fe₂O_3, γ-Fe₂O₃, and Fe₃O₄ are explored as electrode materials in LIBs. The crystal structure of Fe₂O₃ is displayed in Fig. 10.2 [162]. In principle, lithium storage capacity of iron oxides is mainly delivered through the reversible conversion reaction between lithium ions and metal oxide forming metal nanocrystals dispersed in a Li₂O matrix. Even though this material is abundant, inexpensive, and environmentally friendly and has high theoretical capacity (Fe₃O₄: 924 mAh g⁻¹ [135]; Fe₂O₃: 1005 mAh g⁻¹ [136]), they are fraught with several problems such as poor electronic conductivity, higher volume expansion, and poor cycling stability. Among the major disadvantages, the cycle deterioration is the most important one which is caused by several factors such as the decomposition of the electrolyte solution and loss of the conductive path owing to electrode collapse [137]. In addition, large volume change of the anode material during the continuous charge–discharge cycling causes the cracking of the electrode, loose the contact with current collector, and destruction of the solid electrolyte interface (SEI), which result in the breaking of the continuous conductive path. Reducing the extent of volume expansion and contraction would suppress these problems in a great extent. Nanoscale processing of active material particles has been studied for this purpose [138,139,140]. Carbon coating [141], mixing with electronically conducting materials such as conducting carbon [142], carbon nanotubes [143], and graphene [144], is the commonly adopted technique to control the extent of volume change and for improving the electronic conductivity of iron oxides. Although the volume change ratio depends on material-specific quantities, it is possible to control the extent of volume change by controlling the particle size of the active material. Owing to the formation of short Li⁺-ion diffusion length by the size reduction of the active material to nanoscale, the chemical reaction resistance also get reduced. However, there are various processing difficulties associated in the use of nano-sized active material particles and nanoscale processing such as poor dispersibility due to an increase in the van der Waals force, low initial Coulombic efficiency due to high specific surface area, and low safety due to high chemical activity. To solve out these issues, electrodes prepared using composites of nanoscale active materials and carbon materials have been reported extensively [140], owing to the reduction in interfacial area between the electrolyte and active material by the composite effect, which suppresses the SEI growth, thereby improving the Coulombic efficiency.

Fig. 10.2

Adapted and reproduced from Ref. [162]. Copyright 2016 Springer

The surface atomic configurations in a the (001) plane and schematic hematite structure project along {001} and b the (010) plane and schematic hematite structure projected along {010}.

Carbon coating on iron oxide nanoparticles can form 0D nanospheres [145, 146] 1D nanowires [36, 147], 2D nanoflakes [148, 149], and 3D structures of porous carbon foam loading iron oxide [43, 45]. Different methods such as solid-state reaction [150], hydrothermal process [151], solgel method [152], spray-drying technique [153], vacuum sintering [154], and vacuum decomposition are some of the commonly used methods for doing the carbon coating. Electrospinning is one of the unique carbon coating methods and has been used to fabricate 1D hybrid carbon coating iron oxide nanofibrous composites [79, 155,156,157,158], 1D iron oxide nanofibers [12, 159], and 1D carbon nanofibers for Li-ion batteries [160, 161]. The method can embed the iron oxides into a conductive carbon by subsequent heat treatment which enhances the electrochemical properties effectively due to the increased electronic conductivity. The uniformly coated nanometer-thick carbon layer on the active material (iron oxide) acts as the mechanical buffer, which prevents or minimizes the large volume expansion and cracking of electrode during the continuous charge–discharge cycles.

10.4.1 Electrospun Fe₂O₃/Fe₃O₄ Nanostructures

Different coating methods for the preparation of nanostructures such as chemical vapor deposition (CVD) [163], atomic layer deposition (ALD) [164, 165], electrochemical deposition (ECD) [166], and chemical bath deposition (CBD) [167,168,169] or the conventional synthetic strategies such as hydrothermal method [170] and scaffold-assisted synthesis method [170] are reported for the fabrication of nanostructures of Fe₂O₃/Fe₃O₄. Unfortunately, these methods need to involve the multi-step growth of designed shell materials on various removable or sacrificial templates, which suffers from severe drawbacks. These methods are time consuming; tedious, high-temperature, and complicated processes need expensive equipment. Compared to the other synthesis methods, electrospinning is a facile, cost-effective, and flexible platform for one-dimensional (1D) robust Fe₂O₃/Fe₃O₄ nanostructures. This versatile method offers several advantages, such as high yield, mass production, easy control over the morphology, and high degree of reproducibility of the obtained materials [171,172,173,174,175,176,177]. More importantly, the electrospun nanofibers possess high surface-to-volume ratios due to the formation of super long scale in length of thin fibers, the completely interconnected pore structure, the porous substructure formed on the fiber during annealing, nano- to submicron size fiber diameters, and the great control over the morphology by simply changing the electrospinning process, solution, or ambient parameters [172]. When used these electrospun Fe₂O₃/Fe₃O₄ nanostructured materials as electrode in lithium-ion batteries, the large specific surface area and sufficient void spaces not only tolerate the volume change during the Li⁺-ion intercalation, but also endow with more open channels for ions and electrons to migrate rapidly, resulting in the improved electrochemical performances. There are significant number of studies reported on the preparation of F₂O₃-/Fe₃O₄-based nanostructured anodes having different morphologies such as nanotubes [159, 178], nanorods [162, 179,180,181,182], porous structure [158, 183], microbelt [9], or hollow fibers [12, 184].

10.4.1.1 Porous Fe₂O₃/Hollow Fe₃O₄ Nanotube

A nanotube is a nanoscale material that has a seamless tubelike structure. Among the various nanostructures of Fe₂O₃ such as nanoparticles [185], nanorods [26, 186] and nano fibers [187, 188] etc., (1D) Fe₂O₃/Fe₃O₄ nanotubes [159, 170, 189] gained munch attraction due to their advantages properties, including the increase contact surface area between electrolyte and active materials, shorten migration path for Li+ and electron, and accommodate the volume variations via additional void space during cycling.

Sun et al. [170] prepared 1D porous Fe₂O₃ nanotubes with 2-µm length, 220-nm outer diameter, and 65-nm wall thickness via a low-temperature hydrothermal method followed by thermal treatment. The resulted porous Fe₂O₃ nanotubes exhibited enhanced electrochemical properties in terms of lithium storage capacity (1050 mAh g⁻¹ at 100 mA g⁻¹ rate), initial Coulombic efficiency (78.4%), cycle performances (90.6% capacity retention at 50th cycle), and rate capability (613.7 mAh g⁻¹ at 1000 mA g⁻¹ rate). 1D porous Fe2O3 nanostructures have also been synthesized via a SiO2 scaffold method, exhibiting the initial discharge and charge capacities of 1304.3 and 950.9 mAh g⁻¹ at a current density of 100 mA g⁻¹, respectively [170]. Also, the porous Fe₂O₃ nanorods deliver a capacity of 671 and 541 mAh g⁻¹ at current densities of 1000 and 2000 mA g⁻¹, respectively, showing good rate capability. Although the aforementioned porous Fe2O3 nanotubes showed the enhancement of lithium storage capacities, to fabricate porous Fe2O3 nanotubes with satisfied properties via a facile technology is still an appealing challenge. Hence, the robust electrospun Fe₂O₃ nanotubes are prepared and studied its electrochemical properties in LIBs.

Porous Fe₂O₃ [159] and Fe₃O₄/C [178] nanotubes were prepared by electrospinning a solution of iron (III) acetylacetonate and PVP (for porous Fe₂O₃ nanotubes) or iron (III) acetylacetonate and polyacrylonitrile (PAN) (10:8 wt./wt.) along with a 40% mineral oil on weight of iron (III) acetylacetonate (for Fe₃O₄/C nanotubes). First, the Fe₂O₃ or Fe₃O₄/C precursor fiber was prepared and then the fibers are heat-treated to transform to the porous hollow nanotubes. Fe₂O₃ precursor nanofibers (Fig. 10.3a, b) having an average fiber diameter of 520 nm possess continuous fibrous geometry with a relatively smooth surface and without any pores or hierarchical nanostructure were first prepared by electrospinning and then annealed at 500 °C. After annealing, the as-prepared sample inherits the continuous 1D nanostructures from the precursor fibers and shows a large quantity of tubelike structures having average diameter of 400 nm (Fig. 10.3c, d); i.e., the diameter of the precursor fiber gets reduced by the shrinkage at higher temperature. For the production of Fe₃O₄/C nanotubes, the as-spun nanofibers were pre-oxidized in air at 250 °C for 2 h to follow by the carbonization at 600 °C for 2 h under high-purity argon atmosphere. The resultant nanotubes have outer diameter range from 200 to 400 nm and length of several millimeters.

Fig. 10.3

Adapted and reproduced from Ref. [159]. Copyright 2015 Elsevier

a, b Low- and high-resolution images of precursor fibers, c, d FE-SEM images of porous Fe₂O₃ nanotubes.

During the electrospinning process, when the fibers are spun out from the spinning needle, the solvent will evaporate rapidly from the surface of fibers. This leads to the formation of a concentration gradient of solvent along the radial direction of the fibers, as illustrated in Fig. 10.4a; hence, the concentration at the center of the fibers is usually high than that of the surface. Due to the rapid evaporation of the solvent, the PAN will solidify on the surface of fiber. During the solidification process, the AAI and PAN will be extracted and reside at the shell of the fiber due to the rapid phase separation result from their poor solubility in mineral oil. Hence, the rapid evaporation of solvent produces a region near the fiber surface enriched in PAN/AAI, and the mineral oil would diffuse from the surface to the core of the fiber. With the continuing evaporation of the solvent, the concentration of PAN/AAI decreases continuously from the surface to the center of fiber, and the mineral oil tends to congregate at the center of fibers, as displayed in Fig. 10.4a. Further, the mineral oil at the center of fiber might evaporate through the wall of the nanotubes and PAN/AAI nanotubes were collected on the target. During the carbonization at 600 °C in Ar atmosphere, PAN would be decomposed and carbonized completely leading to the formation of Fe₃O₄/C nanotubes obtained. The tubes have a wall thickness of 40 and 150 nm hollow cores as shown in Fig. 10.4b, and it shows a roughness surface that can be distinctly observed from Fig. 10.4c.

Fig. 10.4

Adapted and reproduced from Ref. [178]. Copyright 2014 Elsevier

a Schematic illustration of one step method to fabricate 1D Fe₃O₄/C nanotubes by electrospinning, b SEM image on the surface morphology of the final Fe₃O₄/C nanofibers, and the lower-left corner inset is an enlarged view of a single nanotube, and the scale bar is 200 nm, c TEM images of the Fe₃O₄/C nanotube.

On contrary, during the production of porous Fe₂O_3, the electrospun precursor nanofibers showed a smooth surface without porous or hollow structures. When the precursor fiber is sintering in air atmosphere, the degradation of the side chain of PVP formed through the intermolecular cross-linking reaction and the decomposition of iron acetylacetonate occur simultaneously. Upon increasing the temperature, both PVP and iron acetylacetonate would be oxidized. The oxidation of PVP results in the volatilization and evolution of CO₂, while the oxidation of iron acetylacetonate resulted in the formation of Fe₂O₃ nanoparticles. The outward diffusion of CO₂ generates a pressure to compress the Fe₂O₃ nanoparticles on the surface region of the fiber and forms a porous shell. Then, the porous shell allows CO₂ effusion from the regions below the shell in the fiber and the iron precursor in the inner part of the fibers would move toward the surface which is presumably accelerated by gaseous species that are produced by the oxidation of PVP, and crystallized into Fe₂O₃ nanoparticles. Finally, these Fe₂O₃ nanoparticles get connected together to generate porous Fe₂O₃ nanotubes. However, the formation of porous annotate greatly depends on the concentration iron precursor (iron acetylacetonate) in the electrospunned precursor fiber. To get more insight into the actual evolution process of the Fe₂O₃ nanostructures, we carry out a series of concentration-dependent experiments which samples are prepared by adjusting the concentration of iron acetylacetonate in the electrospun precursor solution. The precursor fiber having iron acetylacetonate higher than about 45% in the precursor electrospun fibers resulted in the formation of Fe₂O₃ nanotubes. If the concentration of iron precursor is less than 45% obviously, it results in the formation of a few Fe₂O₃ nanoparticles, which is difficult to keep the robust frame of the nanotube shell. As a result, the collapse of the tube structure results in the formation of the nanobelts (Fig. 10.5a). When the iron precursor concentration is lower than 37%, only nanobelts are formed (Fig. 10.5b), while the increase in concentration to 47% forms both nanobelts and nanotubes (Fig. 10.5c). When the iron precursor concentration is further extended to about 50%, the increase progressively encourages the formation of a large number of Fe₂O₃ nanoparticles; as a result, the Fe₂O₃ porous nanotubes appear (Fig. 10.5d). Hence, the evolution of Fe₂O₃ nanostructures could be controllably synthesized by adjusting the concentration of the electrospun precursor solution (Fig. 10.5e).

Fig. 10.5

Adapted and reproduced from Ref. [159]. Copyright 2015 Elsevier

SEM images on the surface morphology of Fe₂O₃ nanostructures obtained by adjusting the quantity of iron acetylacetonate (g): a 0.2, b 0.3 c 0.4, and d 0.5, respectively in the electrospun precursor solution. e Possible formation mechanism of Fe₂O₃ nanostructures.

The cyclic voltammetry studies showed three cathodic peaks correspond to the potentials 1.55, 0.89, and 0.55 V, indicating the different lithiation steps [12, 156, 190, 191]. The peaks at 1.55 and 0.89 V correspond to the intercalation of Li + ions into the crystal structure of porous Fe₂O₃ nanotubes and the transformation from hexagonal α-Li_xFe₂O₃ to cubic Li₂Fe₂O₃ without any crystal structure destruction. The high intensity peak observed at 0.55 V corresponds to the crystal structure destruction accompanied by the complete reduction of iron from Fe(III) to Fe(0) and the decomposition of electrolyte. The anodic polarization peaks observed at 1.85 V correspond to the oxidation of Fe(0) to Fe(II) and Fe(III) to re-form Fe₂O₃. Compared to the first cycles, the subsequent cycles are significantly different, which is due to irreversible phase transformation during lithium insertion and extraction in the initial cycle. During the second cathodic process, the peaks at 1.55 and 0.89 V disappear, which indicates lithium insertion and irreversible phase transformation of hexagonal α-Li_xFe₂O₃ to cubic Li₂Fe₂O₃. Also, a decrease in the peak intensity with the number of cycles is observed, which indicates that the capacity is decreased during cycling. However, the CV curves starts to overlap after the third cycle cathodic scan, which indicate the reversibility and capacity stability the continuous charge-discharge processes. The galvanostatic charge–discharge profiles of the porous Fe₂O₃ are consistent with CV curves [159].

The porous Fe₂O₃/Li metal half cells assembled by using Celgard 2400 membrane as the separator and 1M LiPF₆ in ethylene carbonate/dimethyl carbonate (EC/DMC,1:1 v/v) as the electrolyte delivered an initial charge and discharge capacities of around 1045 mAh g⁻¹, at a current density of 100 mA g⁻¹. Particularly, the porous Fe₂O₃ nanotubes still exhibit an excellent cyclic performance at a much higher current density of 200 mA g⁻¹, and the capacity reaches 988 mAh g⁻¹ after 250 discharge and charge cycles. The Coulombic efficiency rises rapidly in the subsequent cycles, reaching up to 95% at the 5th cycle, and remains above 98% after 50 cycles, suggesting an excellent reversible Li⁺-ion intercalation/extraction performance. The specific reversible charge and discharge capacity decrease slightly up to the initial 50 cycles and reach 513 and 524 mAh g⁻¹, respectively, and then increase significantly and reach over 995 and 988 mAh g⁻¹ by 250th cycle. The porous Fe₂O₃ also showed good rate capability [159]. Similarly, the Fe₃O₄/C nanotubes also showed good charge–discharge cycling stability and rate capability. In the half cell studies using M LiPF₆ in ethylene carbonate, diethyl carbonate and ethyl methyl carbonate (EC/DMC/EMC, 1:1:1 vol) as the electrolyte and Celgard 2400 polypropylene as separator showed an initial discharge and charge capacity of 1102 and 727 mAh g⁻¹, with a Coulombic efficiency of only 66% at a current density of 0.15 °C. The significantly higher capacity loss (34% loss) during the first cycle [178] is corresponding to the formation of SEI layer and the incomplete conversion reaction [192, 193]. When the cell is cycled at 1600 mA g⁻¹, the cell delivers a very stable capacity of 350 mAh g⁻¹, and when the current density is switched back to 0.15 °C, a capacity about 600 mAh g⁻¹ is delivered, which is about 100 mAh g⁻¹ less compared to the first cycles at 0.15 °C [178].

In both porous Fe₂O₃ and Fe₃O₄/C electrodes, an initial capacity loss (25–35%) is observed. Similar irreversible capacity loss was noted with other metal oxides or metal oxide combinations reported in previous literature [12, 193, 194].

Compared to Fe₃O₄/C nanofibers prepared by a controlled fabrication process, sample fabricated according to Wang’s reports [157] good electrochemical performance. The charge–discharge studies at a current density of 0.15 °C displayed a quick drop-down in discharge capacity to 300 mAh g⁻¹ after 100 cycles, which is about 50% discharge capacity of the F₃O₄/C. More clearly, F₃O₄/C nanotube shows negligible capacity decreases from the 2nd cycle onward and displayed a discharge capacity as high as 600 mAh g⁻¹ (85% of the second cycle capacity) after 100 cycles at a current density of 0.15 °C [178]. The lackluster cycling performance of Fe₃O₄/C nanofibers is due to the solid structure of the nanofiber; hence, there is no enough space to accommodate the mechanical stress of Fe₃O₄ nanoparticles along the fiber axis of the nanofibers. The substantially subdued capacity fade in Fe₃O₄/C nanotubes is due to the hollow nanotube structure having larger surface-to-volume ratio than that of 1D nanofibers, which could effectively accommodate large volume changes associated with Li⁺-ions insertion/extraction. In addition, the tubular structure can increase the surface area accessible to the electrolyte facilitating the diffusion of Li⁺ ions at the interior and exterior of the nanotube. The charge–discharge capacity fade under continuous cycling is mainly due to the pulverization of original aggregation of Fe₂O₃ particles by the huge volume expansion and contraction during the Li⁺-ion intercalation/extraction process and resulting in the loss of electrical connectivity between the particles and current collector.

A novel process for the preparation of aggregate-free metal oxide nanopowders with spherical (0D) and non-spherical (1D) hollow nanostructures was introduced. Carbon nanofibers embedded with iron selenide (FeSe) nanopowders with various nanostructures are prepared via the selenization of electrospun nanofibers. Ostwald ripening occurs during the selenization process, resulting in the formation of a FeSe-C composite nanofiber exhibiting a hierarchical structure. These nanofibers transform into aggregate-free hollow Fe₂O₃ powders via the complete oxidation of FeSe and combustion of carbon. Indeed, the zero-dimensional (0D) and one-dimensional (1D) FeSe nanocrystals transform into the hollow-structured Fe₂O₃ nanopowders via a nanoscale Kirkendall diffusion process, thus conserving their overall morphology. The discharge capacities for the 1000th cycle of the hollow-structured Fe₂O₃ nanopowders obtained from the FeSe–C composite nanofibers prepared at selenization temperatures of 500, 800, and 1000 °C at a current density of 1 A g⁻¹ are 932, 767, and 544 mAh g⁻¹, respectively; their capacity retentions from the second cycle are 88, 92, and 78%, respectively. The high structural stabilities of these hollow Fe₂O₃ nanopowders during repeated lithium insertion/desertion processes result in superior lithium-ion storage performances.

Figure 10.6 outlines the mechanism of the formation of Fe₂O₃ nanopowders exhibiting hollow nanostructures of different dimensions via the nanoscale Kirkendall diffusion process. Following the selenization processes at different temperatures (i.e., 500, 800, or 1000 °C), the electrospun nanofibers (Fig. 10.6a) were transformed into the hierarchical nanostructures. Selenization of the iron components located close to the nanofiber surface resulted in the formation of FeSe nanocrystals during the early stages of the process. Ostwald ripening then occurred during further selenization to yield the hierarchical FeSe–C composite nanofiber. In this process, the ultrafine FeSe nanocrystals formed inside the carbon nanofiber diffused to the surface to produce FeSe crystals via Ostwald ripening. Complete selenization at 500 °C resulted in the carbon nanofiber being uniformly embedded with ultrafine FeSe nanocrystals (Fig. 10.6b). However, at higher selenization temperatures (Fig. 10.6c), crystal growth occurred via the segregation of nanocrystals and spheroidization due to melting. Finally, the hierarchical FeSe–C nanofibers transformed into hollow aggregate-free Fe₂O₃ nanopowders (Fig. 10.6d, e) via the complete combustion of carbon and oxidation of FeSe. Furthermore, as shown in Fig. 10.7, FeSe nanocrystals with 0D and 1D structures transformed into the hollow-structured Fe₂O₃ nanopowders via a nanoscale Kirkendall diffusion process, thus conserving their overall morphology. For simplicity, the hollow Fe₂O₃ nanopowders obtained from the FeSe-C composite nanofibers prepared at 500, 800, and 1000 °C are referred to as “Sel.500-Oxi.600,” “Sel.800-Oxi.600,” and “Sel.1000-Oxi.600,” respectively [1].

Fig. 10.6

Adapted and reproduced from Ref. [182]. Copyright 2016 Springer

Formation mechanism of the hollow-structured Fe₂O₃ nanopowders with 0D and 1D structure.

Fig. 10.7

Adapted and reproduced from Ref. [182]. Copyright 2016 Springer

Conversion reaction of the FeSe filled structure into Fe₂O₃ hollow structure by nanoscale Kirkendall diffusion effect, a hollow-structured Fe₂O₃ nanopowder with 1D and b hollow-structured Fe₂O₃ nanopowder with 0D.

10.4.1.2 α-Fe₂O₃ Nanorods

Nanorods are one of the nanostructured entities, reported as electrodes in LIBs. Different synthesis techniques such as forced hydrolysis, solgel synthesis, template methods, molten salt process, spray pyrolysis, hot plate method, hydrothermal method, and co-precipitation technique have been adopted to prepare various nanostructures of hematite [34, 134, 191, 195,196,197,198,199,200]. Nanoscale Kirkendall diffusion and Ostwald ripening processes, in which filled structures are transformed into hollow structures during heat treatment, have been applied recently for the preparation of hollow nanospheres (0D) in the absence of templates [182, 201,202,203,204,205] α-Fe₂O₃nanorods [156] and bubble-nanorod-structured Fe₂O₃-carbon nanofibers [179] prepared by electrospinning are reported as high performance anode in LIBs.

The α-Fe₂O₃ nanorods are synthesized by electrospinning of polyvinylpyrrolidone (PVP)/ferric acetylacetonate (Fe(acac) ₃) composite precursors and subsequent annealing at 500 °C for 5 h. The phase separation of Fe(acac)₃ and PVP process involving a polymer/precursor forms “islands” that account for the generation of α-Fe₂O₃ nanorods having average diameter of the nanorods is found to be 150 nm upon heat treatment. The annealed electrospun α-Fe₂O₃ nanorods are composed of agglomerates of nano-sized α-Fe₂O₃ particles. The electrospun α-Fe₂O₃ nanorods exhibit a high reversible capacity of 1095 mAh g⁻¹ at 0.05 °C, are stable up to 50 cycles (with capacity retention of 93% between 2 and 50 cycles), and also show high rate capability, up to 2.5 °C. At a current rate of 2.5 °C, α-Fe₂O₃ nanorods deliver a discharge capacity of 765 mAh g⁻¹ and when the current rate is reduced from 2.5 to 0.1 °C after 70 cycles, still a reversible capacity of 1090 mAh g⁻¹ is obtained showing the good rate capability of the material. The high rate capability and excellent cycling stability can be attributed to the unique morphology of the macroporous nanorods comprised of interconnected nano-sized particles [156].

The structure denoted as “bubble-nanorod composite” is synthesized by introducing the Kirkendall effect into the electrospinning method. Bubble-nanorod-structured Fe₂O₃/C composite nanofibers, which are composed of nano-sized hollow Fe₂O₃ spheres uniformly dispersed in an amorphous carbon matrix, are synthesized as the target material using Fe(acac)₃-PAN composite solution as the precursor. Post-treatment of the electrospun precursor nanofibers at 500 °C under 10% H₂/Ar mixture gas atmosphere produces amorphous FeO_x/carbon composite nanofibers, and the further post-treatment at 300 °C under air atmosphere produces the bubble-nanorod-structured Fe₂O₃/C composite nanofibers. The solid Fe nanocrystals formed by the reduction of FeO_x are converted into hollow Fe₂O₃ nanospheres during the further heating process by the well-known Kirkendall diffusion process [179].

The formation mechanism of bubble-nanorod-structured Fe₂O₃/C composite nanofibers is schematically displayed in Figs. 10.8 and 10.9. During the carbonization at 500 °C, the PAN gets decomposed to form the carbon matrix and the decomposition of iron acetylacetonate produced FeO_x, which is uniformly dispersed in carbon matrix resulting in the carbon composite nanofibers. The significantly large amount of carbon in the fiber disturbed the crystal growth of FeO_x. The subsequent post-treatment of the FeO_x/carbon composite nanofibers at 300 °C under air atmosphere produced the bubble-nanorod-structured Fe₂O₃C composite nanofiber. Reduction of FeO_x crystals surrounded by the carbon matrix into Fe metal occurred during the post-treatment under air atmosphere by the following equation:

Fig. 10.8

Adapted and reproduced from Ref. [179]. Copyright 2015 American Chemical Society

Formation mechanism of bubble-nanorod-structured Fe₂O₃/C composite nanofiber by Kirkendall-type diffusion.

Fig. 10.9

Adapted and reproduced from Ref. [179]. Copyright 2015 American Chemical Society

a Formation mechanism of hollow Fe₂O₃ nanosphere in the bubble-nanorod-structured Fe₂O₃ carbon composite nanofiber by Kirkendall effect, b, c its chemical conversion process in the surface region of a sphere.

$$ {\text{FeO}}_{x} ({\text{s}}) + x{\text{C}}({\text{s}}) \to {\text{Fe}}({\text{s}}) + x{\text{CO}}({\text{g}}) $$

The crystal growth of Fe formed ultrafine Fe nanocrystals uniformly dispersed within the carbon nanofibers during the early stage of post-treatment by the consumption of some amount of carbon. The solid Fe nanocrystals were converted into hollow Fe₂O₃ nanospheres during the subsequent heating process by the well-known Kirkendall effect. The Kirkendall effect, a vacancy flux, and subsequent void formation process resulting from diffusivity differences at inorganic interfaces were first reported by Aldinger [206]. The Kirkendall effect results in the formation of a thin Fe₂O₃ layer on the Fe metal surface (Fig. 10.8c), followed by simultaneous outward diffusion of Fe cations through the oxide layer and inward diffusion of oxygen into the nanospheres, creating an intermediate Fe@Fe₂O₃ core–shell structure (Fig. 10.9b). Fe cations diffused outward more quickly than oxygen diffused inward, which is consistent with the larger ionic radius of oxygen anions (140 pm) than Fe cations (Fe²⁺ is 76 pm, and Fe³⁺ is 65 pm). Accordingly, Kirkendall voids were generated near the Fe/Fe₂O₃ interface during vacancy-assisted exchange of the material via bulk interdiffusion (Fig. 10.9c), which gave rise to coarsening and enhancement of pore growth in the spheres (Fig. 10.8d). Complete conversion of Fe metal into Fe₂O₃ by Kirkendall-type diffusion resulted in the bubble-nanorod-structured Fe₂O₃C composite nanofibers (the highly crystalline structure of the hollow Fe₂O₃ nanofibers). The figure shows the TEM image and elemental mapping of the nanofibers post-treated at 500 °C under H₂/Ar mixed gas atmosphere. The elemental mapping images shown in Fig. 10.10 show the trace amounts of carbon present in the nanofibers.

Fig. 10.10

Adapted and reproduced from Ref. [179], Copyright 2015 American Chemical Society

Morphologies, SAED, and elemental mapping images of the nanofibers post-treated at 500 °C under H₂/Ar gas atmosphere: a, b TEM images, c SAED pattern, and d elemental mapping.

The cell studies displayed discharge capacities of 812 and 285 mAh g⁻¹, respectively, for bubble-nanorod-structured Fe₂O₃/C composite nanofibers and hollow bare Fe₂O₃ nanofibers for the 300th cycles at a current density of 1.0 A g⁻¹, and their capacity retentions measured from the second cycle are 84 and 24%, respectively. The initial (first cycle) Coulombic efficiencies of hollow bare Fe₂O₃ nanofibers and bubble-nanorod-structured Fe₂O₃/C composite nanofibers were 81 and 69%, respectively. The initial irreversible capacity loss of the hollow bare Fe₂O₃ nanofibers is ascribed to the formation of an SEI layer on the surface of the nanofibers and the incomplete restoration of metallic Fe into the original oxide during the initial charging process [156, 207]. The high amount of amorphous carbon, which has a low lithium storage capacity and a large initial irreversible capacity loss, decreased the initial discharge capacity and Coulombic efficiency of the bubble-nanorod-structured Fe₂O₃/C composite fibers [208, 209]. The rate capability studies showed that the stable reversible discharge capacities of the bubble-nanorod-structured Fe₂O₃C nanofibers decreased from 913 to 491 mAh g⁻¹ as the current density increased from 0.5 to 5.0 A g⁻¹. Furthermore, the discharge capacity recovered to 852 mAh g⁻¹ as the current density was restored to 0.5 A g⁻¹. The electrochemical studies showed that the bubble-nanorod-structured Fe₂O₃C nanofibers showed superior electrochemical properties as an anode material for LIBs as compared with the hollow bare Fe₂O₃ nanofibers. The synergetic effect of hollow nanospheres and a carbon matrix resulted in the superior cycling and rate performance of the bubble-nanorod-structured Fe₂O₃/C nanofibers. The hollow nanospheres accommodate the volume change that occurs during the continuous charge–discharge cycling which leads to the long-term cycling stability. The unique structure of the bubble-nanorod-structured Fe₂O₃/C composite nanofibers results in their superior electrochemical properties by improving the structural stability during long-term cycling.

10.4.1.3 Metal/Fe₂O₄ or Fe₃O₄ Composite Nanofibers

In spite of their lower cost and better safety, the capacity retention of transition metal oxides especially Fe₂O₃/Fe₃O₄ and turning of their energy/power density remain as major drawbacks. However, transition metal oxide spinels (AB₂O₄) with two transition elements (both A- and B-sites) provide the feasibility to tune the energy density and working voltages by varying the transition metal content [210, 211]. Owing to it, zinc [212,213,214,215], Co, Cu, etc., have been substituted into iron oxide to construct a ternary metal ferrite (ZnFe₂O₄/CuFe₂O₄/CoFe₂O₄, etc.) and possess a lower working voltage to effectively enhance the total output voltage of the LIBs. CuFe₂O₄ nanoparticles, nanorods, and hollow spheres have been fabricated through wet chemical routes, and their electrochemical properties have been investigated [216,217,218]. Cobalt ferrite (CoFe₂O₄) has been well regarded as a distinguished anode due to its low cost, high chemical stability, and good environmental benignity [219,220,221,222]. Especially, CoFe₂O₄ can deliver a high theoretical capacity of 916 mAh g⁻¹ based on an eight-electron conversion reaction. Similarly, zinc ferrite (ZnFe₂O₄) generates high capacity as lithium ions form an alloy with Zn and dealloy, while Fe and Zn react with Li₂O to absorb/release Li during lithiation/delithiation. Thus, ZnFe₂O₄ implements both conversion [6, 223] and alloy/dealloy [224, 225] reaction, simultaneously. Also, fabrication of nanostructured binary metal oxides could buffer the mechanical strain during the cycling process. In particular, 1D hollow nanostructure could not only provide sufficient void spaces to tolerate the volume change during cycling process, but also allow for efficient electron transport along the longitudinal direction, resulting in the improved electrochemical performance. Hence, CoFe₂O₄ nanotubes [226], CuFe₂O₄ nanotubes and nanorods [227], CoFe₂O₄/C composite fibers [228], interwoven ZnFe₂O₄ nanofibers [229], and N-doped amorphous carbon-coated Fe₃O₄/SnO₂ coaxial nanofibers [230] were prepared by electrospinning and reported as anode in LIBs having enhanced electrochemical properties. When evaluated as anode materials for LIBs, the CoFe₂O₄ nanotubes exhibited good electrochemical performance with high specific capacity of 1228 and 693.9 mAh g⁻¹ at a current density of 50 and 200 mA g⁻¹, respectively, long cycling stability over 160 cycles, and good rate capability (214.7 mAh g⁻¹ at 2 A g⁻¹) [226], while the CuFe₂O₄ nanotubes delivered a high reversible capacity of 1399.4 mAh g⁻¹ at a current density of 200 mA·g⁻¹ and a capacity retention of ∼816 mAh g⁻¹ after 50 cycles with a good rate capability (450 mAh g⁻¹ at 2.5 A g⁻¹) [227] which is much higher compared to CoFe₂O₄ nanotubes (wall thickness of ∼50 nm, presented diameters of ∼150 nm, and lengths up to several millimeters) [226]. The CuFe₂O₄ nanorods showed a discharge capacity of only ∼489 mAh g⁻¹ after 50 cycles for the same current density of 2.5 A g⁻¹ which clearly suggest that the nanotubes have superior electrochemical properties than their counterpart nanotubes [227] which can be attributed to the continuous one-dimensional (1D) hollow nanostructure and their higher surface area. However, compared to CoFe₂O₄ nanotubes [226], CoFe₂O₄/C composite fibers consist of CoFe₂O₄ nanoparticles with a diameter of about 42 nm well dispersed in the carbon matrix as anode material prepared by electrospinning and thermal technique displayed a stable and reversible capacity of over 490 mAh g⁻¹ after 700 cycles at a rate of 2.0 C and good rate capability [228].

The (1D) CuFe₂O₄ nanotubes and nanorods were fabricated by a single spinneret electrospinning method followed by thermal decomposition for removal of polymers from the precursor fibers. It was found that phase separation between the electrospun composite materials occurred during the electrospinning process, while the as-spun precursor nanofibers composed of polyacrylonitrile (PAN), polyvinylpyrrolidone (PVP), and metal salts might possess a core–shell structure (PAN as the core and PVP/metal salt composite as the shell) and then transformed to a hollow structure after calcinations as shown in Fig. 10.11 [227]. Based on the literature and the above analysis, the electrochemical reactions involved in the cycling process are believed to proceed as follows:

Fig. 10.11

Adapted and reproduced from Ref. [227], Copyright 2014 Elsevier

Scheme of the procedure for producing CuFe₂O₄ nanotubes.

$$ \begin{aligned} & {\text{CuFe}}_{2} {\text{O}}_{4} + {\text{nLi}} + + {\text{ne}}^{ - } \to {\text{Li}}_{n} {\text{CuFe}}_{2} {\text{O}}_{4} \\ & {\text{Li}}_{n} {\text{CuFe}}_{2} {\text{O}}_{4} + (8 - n){\text{Li}}^{ + } + (8 - n) \to x{\text{Cu}} + (2 - x){\text{Fe}} \\ & \quad + {\text{Fe}}_{x} {\text{Cu}}(1 - x) + 4{\text{Li}}_{2} {\text{O}}(0 \le 1 \le 4) \\ & {\text{Cu}} + {\text{Li}}_{2} {\text{O}} \to {\text{CuO}} + 2{\text{Li}}^{ + } + 2{\text{e}}^{ - } \\ & {\text{Fe}} + 3{\text{Li}}_{2} {\text{O}} \to {\text{Fe}}_{2} {\text{O}}_{3} + {\text{Li}}^{ + } + {\text{e}}^{ - } \\ \end{aligned} $$

As a heavy metal, Co, is a toxic material, hence nanowebs consisting of interwoven ZnFe₂O₄ nanofibers are synthesized by a simple electrospinning technique, to be employed as an environmentally friendly anode in lithium-ion batteries. The morphological studies showed self-assembly of electrospun ZnFe₂O₄ nanofibers into intertwined porous nanowebs with a continuous framework. Benefitting from the one-dimensional functional nanostructured architecture, the application of electrospun nanowebs with ZnFe₂O₄ nanofiber (ZFO-NF) anodes in LIBs delivers the first charge capacity of 925 mAh g⁻¹, exhibits excellent cyclability, and retains a reversible capacity of 733 mAh g⁻¹ up to 30 cycles at 60 mA g⁻¹ as compared to ZnFe₂O₄ nanorods (ZFO-NR) with a capacity of 200 mAh g⁻¹. The ZnFe₂O₄ nanowebs also displayed a high capacity of 400 mAh g⁻¹ at 800 mA g⁻¹ (1C). The enhanced capacity releases at higher current densities have shown the importance of having a well-connected electronic wiring during lithium insertion/extraction especially in prolonged cycling [229].

N-doped amorphous carbon-coated Fe₃O₄/SnO₂ coaxial nanofibers were prepared via a facile approach and studied as binder-free self-supported anode for lithium-ion batteries and their electrochemical performance in LIBs. In the process of preparation, the core composite nanofibers were first made by electrospinning technique, and then the shells were conformally coated using the chemical bath deposition and subsequent carbonization using polydopamine as a carbon source (Fig. 10.12). The coaxial nanofibers displayed an enhanced electrochemical storage capacity of 1223, 1030, 862, and 640 mAh g⁻¹ at 100, 200, 400, and 800 mAg⁻¹, respectively. Even after the current density is increased to 1600 mA g⁻¹, it still maintains a high charge capacity of 402 mAh g⁻¹. When the current density is restored to the initial setting of 100 mA g⁻¹, the carbonized polydopamine-coated Fe₃O₄/SnO₂ coaxial nanofiber electrode (Fig. 10.13) leads to a reversible capacity of 1070 mAh g⁻¹ displaying not only its superior capacity retention but also excellent capacity recovery performance. Also, it was found that the morphology of the interwoven nanofibers was maintained even after the rate cycle test. The superior electrochemical performance originates from the structural stability of the N-doped amorphous carbon shells formed by carbonizing polydopamine [230]. On the basis of the literature, the electrochemical reactions of the carbonized polydopamine-coated Fe₃O₄/SnO₂ coaxial nanofibers can be described as follows

Fig. 10.12

a, b TEM, c high-resolution TEM images of the prepared single Fe₂O₃/SnO₂ composite nanofiber, d, e TEM, f high-resolution TEM images of the carbonized polydopamine-coated single Fe₃O₄/SnO₂ coaxial nanofiber. Adapted and reproduced from Ref. [230]. Copyright 2014 American Chemical Society

Fig. 10.13

a, b TEM images of the carbonized polydopamine-coated Fe₃O₄/SnO₂ coaxial nanofibers after rate performance test. Adapted and reproduced from Ref. [230]. Copyright 2014 American Chemical Society

$$ \begin{aligned} & {\text{SnO}}_{2} + 4{\text{Li}}^{ + } + 4{\text{e}}^{ - } \to {\text{Sn}} + 2{\text{LiO}} \\ & {\text{Sn}} + x{\text{Li}}^{ + } + x{\text{e}}^{ - } \to {\text{Li}}_{x} {\text{Sn}}(0 \le x \le 4.4) \\ & {\text{Fe}}_{3} {\text{O}}_{4} + 8{\text{Li}}^{ + } + 8{\text{e}}^{ - } \to 3{\text{Fe}} + 4{\text{Li}}_{2} {\text{O}} \\ \end{aligned} $$

10.5 Conclusion

For the achievement of best performing lithium-ion batteries, different materials were greatly explored that can deliver a best performing system. Metal oxides possess great significance in lithium-ion batteries since they are capable of exhibiting better electrochemical properties. Similar to carbon base materials, currently, metal oxides are also widely explored as anodes in LIBs. Iron oxide-based anode materials such as Fe₂O₃ and Fe₃O₄ are promising due to its high capacity and lithium storage properties. Even though, despite of its high capacity, the poor cycling stability and high polarization during lithiation and delithiation process limit its practical use in LIBs. Modifications of structural and surface characteristics are considered to be the best method for the enhancement of electrochemical properties of these anodes. Electrospinning is considered to be the best method for this. Electrospinning is considered to be the most versatile and simple method that provides a flexible platform for the fabrication of nanostructures that can deliver better battery performance. Porous and hollow iron oxide structure results by the electrospinning technique facilitate the fast lithium-ion transfer as well as it accommodates the volume change that results in unique electrochemical properties. Similarly, the hollow nanotubes and nanorods that results by this technique can deliver high performance in battery owing to its high surface area, controlled shape, and a low density. Further performance can be enhanced by the structural modification by incorporating different metals to form the composite structures which can result an enhancement in conductivity which is significant for a better performing electrode material. These potential features make iron oxide-based materials as the best performing anode material in LIBs.