1. Introduction
With the transition from fossil fuel consumption to renewable energy, extensive research is being conducted on the development of energy storage and conversion systems that permit the effective use of intermittent renewable energy sources. Among the several types of en- ergy storage and conversion systems in use, lithium-ion batteries (LIBs), which convert chemical energy to electrical energy and vice versa through electrochemical reactions between active materials, have become the primary option owing to their valuable properties, such as high energy density, high operating voltages, limited self-discharging, and low maintenance requirements[1-3]. In conventional LIBs, the cells are assembled with a carbonaceous anode and lithiated metal oxide cathode (LiCoO2 and LiNiO2) with a porous membrane separator immersed in a lithium salt and mixed with liquid alkyl carbonates[4-6]. Currently, graphite is the most widely used anode material in LIBs because of its abundance, low production cost, and charge-discharge cy- cle stability[7]. However, as the use of LIBs expands from portable electronic devices to large-scale energy storage systems (ESSs) and electric vehicles (EVs), the current commercial graphite anode cannot meet the increasing demand for such applications due to its moderate intrinsic specific capacity (372 mAh g-1), sluggish Li+ diffusion coefficient, and safety issues that arise from the formation of lithium dendrites, especially under the operating conditions of high current rates resulting from its high polarization and low operating voltage of ~0.1 V vs. Li/Li+[8,9]. Consequently, the development of new generation anode materials that possess characteristics of high power and energy density, stable cyclability, and can operate with sufficient safety at high current density is urgently required[10,11].
Among the several candidate groups for use as next-generation anode materials such as lithium metal and Si and graphene-based materials, titanium-based compounds including Li4Ti15O12 and TiO2 are regarded as promising alternative anode materials for high power density LIBs owing to their advantages of high rate capability, safety, environmental friendliness, and low cost[11,12]. Compared with graphite, titanium-based anode electrodes possess a three-dimensional crystal lattice in which Li+ ions are easily (de)intercalated. They also exhibit a high operating potential of ~1.5~1.7 V vs. Li/Li+ within the electrochemical window of typical LIB electrolytes, which ensures the excellent safety of LIBs by suppressing the formation of lithium dendrites even under fast charge rates above 20 C[13,14]. Furthermore, owing to their outstanding structural stability, they exhibit negligible or little volume expansion (0.1% lattice strain) during cycling, significantly less than that exhibited by graphite (~10%), leading to excellent cycling properties. Notably, Li4Ti15O12-based electrodes have previously been used for fast charging applications in Toshiba batteries (SCiB) and Mitsubishi electric vehicles (i-MiEV) [15]. Among the eight known polymorphs of TiO2 (rutile, anatase, brookite, TiO2-B, TiO2-R, TiO2-H, TiO2-II, TiO2-III), rutile, anatase, and TiO2-B have been most widely investigated as anode materials for LIBs because of their thermodynamically (meta)stable nature[ 1,16,17]. The typical lithium intercalation-deintercalation reaction that occurs during the electrochemical process on the TiO2 anode is described by [12,18].
where the value x representing the molar fraction is largely influenced by the crystallinity, particle size, morphology, surface area, and the involved polymorph[19].
There are numerous approaches to the preparation of TiO2 micro and nanostructures as anode materials for LIBs, such as anodization, sol-gel, templating, and hydrothermal methods, microwave irradiation, and alkaline synthesis[20-25]. Among these, electrochemical anodization is the simplest, most straightforward, and cost-effective strategy to obtain diverse TiO2 micro and nanostructures on Ti substrates. This method involves controlling experimental conditions such as applied voltage, temperature, and electrolyte composition to alter the dimensions of the obtained structural morphologies. In this review, we will briefly discuss: the mechanism of formation of TiO2 micro and nanostructures through electrochemical anodization, with a focus on nanotubes and microcones, and the parameters that determine the type or morphology of the oxide structures. An overview of the electrochemical performance of anodic TiO2 micro and nanostructures as anode materials for LIBs is also provided.
2. Anodic TiO2 Nanotubes
2.1. Mechanism for the preparation of TiO2 nanotubes via anodization
Anodization is a conventional technique in use for more than a century to form an oxide layer on the surface of a metal substrate. It is generally performed in a two-electrode system where the valve metal functions as the working electrode (anode) and platinum or carbon is the counter electrode (cathode), by applying a specific voltage in aqueous or organic electrolytes. After early reports on the anodization of pure Ti or Ti-6Al-4V in a chromic acid-based electrolyte by Zwilling et al. in 1999[26] and the synthesis of well-defined nanotube arrays on a Ti substrate in 0.5 to 3.5 wt% aqueous hydrofluoric acid (HF) by Gong et al. in 2001[27], intensive studies have been carried out to optimize the parameters affecting the dimensions of the fabricated oxide structures and to elucidate the formation mechanism[28-32]. During anodization, barrier- or porous-type anodic TiO2 is formed on the surface of the Ti substrate depending on the nature of the electrolyte[33]. Typically, barrier-type oxides are formed when electrolytes containing no fluoride, perchlorate, chloride, or bromide ions are used, and the thickness of the oxide layer is proportional to the applied voltage (2.5 nm V-1) owing to the movement of ions under an applied electric field, known as field-assisted oxidation[33,34]. However, in an electrolyte containing fluoride ions (F-), as shown in Figure 1, field-assisted oxidation [Eq. (2)~(4)] and field-assisted dissolution [Eq. (5)~(7)] play important roles in forming a porous or tubular structured oxide layer.
Oxygen anions (O2-) generated by the deprotonation of H2O by the induced electric field migrate from the electrolyte to the interface between the metal and the oxide layer, where they combine with Ti4+ ions dissolved from the Ti surface resulting in the growth of an additional oxide layer. Furthermore, because F- exists in the electrolyte, the highly water-soluble [TiF6]2- complex is produced through the complexation reaction of F- with Ti4+ as well as the chemical dissolution of TiO2 by F- ions[35,36]. The large number of pits formed by the dissolution of the [TiF6]2- complex work as preferential sites where the electric field is concentrated, further improving the dissolution of the oxide layer to form porous or tubular TiO2 structures[37]. In particular, the formation of nanotubes can be attributed to the generation of a fluoride- rich layer at the oxide-metal interface induced by ion migration, followed by the displacement and dissolution of the layer toward the cell boundaries by a flow mechanism[38,39]. Therefore, 1D highly ordered TiO2 nanotubes are formed, indicating that each nanotube is separated by chemical etching at the top and bottom ends, aligned in a close-packed arrangement. In an organic electrolyte containing a small amount of oxygen/water with a fluorine source such as NH4F at high pH, longer TiO2 nanotubes with a better self-organized arrangement are obtained, whereas TiO2 nanotube arrays with tube lengths below a few micrometers are formed in aqueous HF electrolytes[40,41].
Current-time transient, which is divided into three-stages, provides us important information about the formation of TiO2 nanotube arrays during the anodization of Ti. In stage (I), the current density rapidly decreased to a minimum value as barrier oxide layer acting as high resistance is formed. Then, in stage (II), the current density subsequently rises to a maximum point as resistance of anodic layer decreases by nucleation of the pores. Finally, in stage (III), the current density is maintained or decreased slightly, indicating that a stead state where the formation and dissolution at the oxide/electrolyte interface are equal is reached[33].
2.2. Anode materials based on TiO2 nanotubes
2.2.1. Pristine TiO2 nanotubes
As shown in Equation (1), the lithium intercalation-deintercalation reaction occurring in the potential range of 1.4~1.8 V vs. Li/Li+ results in a maximum theoretical capacity of 335 mAh g-1, where 1 M Li+ is intercalated per 1 mol TiO2, and is comparable to that of graphite at 372 mAh g-1[42]. However, the typical amount of intercalated Li+ for anatase is 0.5 mol per 1 mol TiO2[43]. The electrochemical properties of TiO2 as an anode are significantly influenced by its morphology. In nanotubular structures, the porous structure permits the facile approach of Li+ ions to the electrode by reducing the movement pathway. Therefore, the polarization of the anode electrode decreases, allowing fast charging and discharging at a high current rate[1]. The perpendicular orientation of TiO2 nanotubes grown on a Ti substrate also has a positive effect on the electrochemical performance of LIBs. In 2012, Han et al. systematically explored the effect of the contact resistance between the active material and current collector by fabricating electrodes comprising vertically aligned and randomly oriented TiO2 nanotubes[ 44]. The aligned TiO2 nanotubes show a six-fold increase in capacity at a 10 C rate compared to randomly oriented TiO2 nanotube electrodes with a 10 wt% conducting agent because of the good electrical contact between TiO2 as the active material and Ti foil as the current collector. Additionally, crystallinity is an important factor that affects the electrochemical performance of TiO2 anode electrodes for LIBs. The anatase and TiO2(B) phases offer more facile uptake of Li+ ions than rutile structures[42,43]. The anatase phase transformation develops from a tetragonal to an orthorhombic phase via spontaneous phase separation[45]. Several studies have demonstrated that amorphous TiO2 nanotubes exhibit superior specific capacity and rate capability compared to crystalline TiO2 (anatase) nanotubes with similar dimensions due to the high lithium-ion diffusivity owing to the larger number of disordered defects in the amorphous state[46-48].
2.2.2. TiO2 nanotube composites
Despite the advantages of structural stability, low volume expansion during lithium intercalation-deintercalation, and stable capacity retention of TiO2 nanotube composites which make it suitable for EV and stationary storage applications, the commercialization of LIBs based on TiO2 anode materials is still hindered by several drawbacks, such as poor electric conductivity, low ionic diffusivity, the strong repulsive force of Li+ ions, and low theoretical capacity. Extensive investigations on TiO2 nanotube composites with high electronic conductivity (e.g., carbon-based materials such as graphene) or high capacity (e.g., Si- or transition metal oxide-based materials) have been carried out to overcome the limitations of pristine TiO2 nanotubes as anode electrodes for LIBs. In 2020, Gao et al. synthesized boron-doped graphene/TiO2 nanotubes via anodization following an electron-assisted hot-filament plasma chemical vapor deposition route, by controlling the morphology of graphene and oxygen contamination, which improved the conductivity of TiO2 nanotubes[49]. Additionally, Menéndez et al. reported an electrophoretic approach that allowed the preparation of self-organized TiO2/ graphene heterostructures with a capacity of over 200 mAh g-1 after 100 cycles, higher than that of pristine TiO2 nanotubes[50]. The obtained electrode exhibited a capacity above 50 mAh g-1 at extremely high rates close to 300 C owing to an apparent reduction in charge transfer resistance within the heterostructure.
In 2018, Kim et al. prepared and studied Co3O4/CuO@TiO2 composites by spray coating an anodic TiO2 substrate using CuCl2 and CoCl2 solutions at various molar concentrations[51]. Co3O4 has a high theoretical specific capacity of 890 mAh g-1 with a significant volume expansion rate of ~300% during the charge/discharge process. In contrast, CuO has relatively good conductivity and an adequate theoretical specific capacity of 670 mAh g-1 with a relatively low volume expansion of ~174%. As shown in Figure 2, after five repeated spray and calcination steps, crystalline Co3O4 and CuO were uniformly decorated and anchored onto the TiO2 nanotubes, imparting extraordinary rate capability and volumetric capacity at optimized molar ratios of Co3O4 and CuO by the synergetic effects of enhanced capacity and cycling stability. However, owing to the limitations of the spray coating technique, TiO2 nanotubes shorter than 1 μm were required. To overcome this drawback, in 2021 Heo et al. succeeded in preparing 10-μm-thick TiO2 nanotubes decorated with MoO3 via cyclic voltammetry (CV), as shown in Figure 3(a)~(g)[52]. In this work, an ammonium molybdate tetrahydrate aqueous electrolyte with various concentrations was used for CV at a scan rate of 5 mV s-1. The as-prepared electrode, which was annealed at 450 ℃ for 2 h under an Ar atmosphere, exhibited a high specific capacity of 451 mAh g-1 at a rate of 1 C after 150 cycles and capacity retention of 97% at a rate of 5 C over 500 cycles owing to the superior theoretical capacity of MoO3 (~1117 mAh g-1) and the role of TiO2 nanotubes in buffering the large volume expansion of MoO3 [Figure 3(h)~(i)].
3. Anodic TiO2 Microcones
3.1. Parameters affecting the formation of TiO2 microcones
The physical properties and electrochemical performances of TiO2 structures are significantly influenced by their morphology: 0-dimensional nanoparticles, 1-dimensional nanorods or nanowires, and 3-dimensional porous structures. In anodization, diverse morphological TiO2 structures can be obtained, including mesosponge and fishbone types, by adjusting experimental conditions such as the electrolyte, applied voltage, and temperature[53,54]. In 2016, Park et al. reported a new anodization approach for the preparation of hollow crystalline TiO2 microcones composed of multilayered nanofragments in H2SO4 (or H3PO4) based aqueous electrolytes containing HF[55]. The formation mechanism of this method is slightly different from that of nanotubes because of the presence of H2SO4; although not critically affected by concentration, H2SO4 is essential for the formation of microcone structures. The presence of SO42- ions accelerates the formation of TiO2, which makes the surface inhomogeneous and contributes to the formation of a thick oxide layer[56,57]. Furthermore, tiny cracks in the oxide surface occur because of the competitive dissolution reaction of F- and SO42- ions. Therefore, considerable compressive stress is generated in the oxide layer resulting from volume expansion; this force pushes the oxide layer upwards and downwards for compensation, leading to the formation of TiO2 microcone structures [Figure 4(a)].
The primary parameters that lead to the formation of TiO2 microcones instead of barriers or nanotubes were investigated with respect to the applied voltage, concentration, and composition of the electrolyte. As shown in Figure 4(b)~(f), when the applied voltage increases, the morphology of the TiO2 transforms from the barrier oxide layer to nanotubes and finally to microcones. Interestingly, the threshold voltage at which microcones are obtained is lower in the H2SO4-based electrolyte (30 V) than in the H3PO4-based electrolyte (45 V) because of the lower pH of the H2SO4 electrolyte. The critical factor for the formation of microcones is the concentration of HF. As the F- ions render the oxide layer porous by forming a [TiF62-] complex and dissolving the oxide layer, a specific suitable range of HF concentration is required to obtain microcone structures[58]. Additionally, a minimum concentration of H2SO4 is required to initiate the nucleation of microcones [Figure 5(a)~(c)].
As opposed to the electrolytes based on H2SO4 or H3PO4, an electrolyte based on (COOH)2 results in the formation of TiO2 microcones with a larger mean height and diameter of 15 and 18 μm, respectively. In (COOH)2-based electrolytes the microcone formation is linearly proportional to the applied voltages, whereas they converge to approximately 8 μm using the other electrolytes [Figure 5(d)~(e)][59]. Notably, the ratio of the surface area occupied by well-dispersed and fully grown TiO2 microcones is approximately 20%, regardless of the applied voltage or type of electrolyte. In particular, the electrolyte based on (COOH)2 can be used in a wider range of applied voltages than other electrolytes, resulting in well-dispersed TiO2 microcones, and demon strating that (COOH)2 is more suited to prepare TiO2 microcones than H2SO4 or H3PO4 [Figure 5(f)]. The phase of the obtained anodic TiO2 microcones is another unique physical characteristic. Typically, the phase of TiO2 structures prepared by anodization is strongly dependent on the specific electrochemical conditions during fabrication, including applied voltage, time, and temperature; however, in many cases, they are amorphous[60,61]. Therefore, an additional annealing process is required to transform the phases from amorphous to anatase at approximately 300~400 ℃, and from anatase to rutile at temperatures of 500 ~700 ℃. However, as shown in Figure 5(g)~(i), anodic TiO2 microcones prepared in H2SO4, H3PO4, or (COOH)2-based aqueous electrolytes containing HF are already composed of anatase phases without a subsequent annealing step, which is likely attributable to the higher applied voltages than that required for the formation of nanotubes.
3.2. Anode materials based on TiO2 microcones
The first application of TiO2 microcones as anode electrodes for LIBs was reported by Rhee et al. in 2016[55]. As demonstrated in their pioneering report, anodic hollow crystalline microcone structures were prepared by anodization using a mixture of 1 M H3PO4 and 0.5 wt% HF. The obtained microcones were perpendicularly oriented on Ti foil with an average diameter of 7.2 μm and 8.74 μm [Figure 6(a)]. Notably, the sidewall of the microcones was composed of several multilayered nanofragments from the bottom upwards, leading to a large surface area where Li+ ions could be stored [Figure 6(b)~(c)]. Electrochemical measurements showed that the as-prepared TiO2 microcones exhibited a much higher areal capacity with excellent rate capability up to 50 C and reliable capacity retention compared to anodic TiO2 nanotubes prepared by anodization at 20 V for 4 h in an aqueous solution containing 1 M H3PO4, 1 M NaOH, and 0.1% HF, as well as other types of TiO2 electrodes [Figure 6(d)~(e)]. These results can be ascribed to the large surface area and facile diffusion of Li+ ions through the hollow multilayered nanofragment structures. In addition, in 2020, Kim et al. reported the inverse-direction growth of TIO2 microcones via subsequent anodization in trace concentrations of HClO4 without altering the overall microcone structure and crystallinity[62]. As shown in Figure 6(f), after the second anodization step, the underlying hemispherical morphology of the bottom surface of the TiO2 microcones transformed from a bumpy to a plump surface with a large number of protruding nanoparticles, permitting the storage of large amounts of Li+ ions. They observed that these protruding nanoparticles were fabricated by the reaction of trace concentrations of perchlorate ions with Ti4+ at the interface between the metal and oxide layer. Therefore, owing to the increased active surface area, the HClO4-treated TiO2 electrode shows 1.6 times higher areal capacity than that of pristine TiO2 microcones with excellent cycling stability and capacity retention [Figure 6(g)~(h)].
Other strategies to overcome the limitations of pristine TiO2 as anode materials have also been developed by combining them with foreign materials with high conductivity or specific capacity. In 2018, Park et al. demonstrated the synthesis of rGO-coated TiO2 microcones via electrophoretic deposition followed by CV for use in LIBs[63]. They attempted several approaches to reduce GO to rGO coated on TiO2 microcones by electrophoretic deposition, including hydrothermal, cathodic reduction, and annealing treatment in a H2/Ar atmosphere. These approaches to reduce GO to rGO have several drawbacks, such as the need for hazardous hydrazine hydrate, the aggregation of GO, and detachment of TiO2 microcones from the Ti substrate. Therefore, they applied CV to successfully reduce GO to rGO without damaging the TiO2 microcone aggregation of GO [Figure 7(a)]. In an optimized potential range of -1.0~0.8 V, where the reduced GO was reoxidized to GO during the backward scan, TiO2 microcones evenly coated with rGO were obtained by one CV cycle. As anode materials for LIBs, the obtained rGO-coated TiO2 microcones showed a higher discharge capacity of 235 mAh g-1 compared to that of pristine TiO2 microcones (181 mAh g-1). Additionally, the rGO-coated TiO2 microcone electrode delivered a discharge capacity of 157 mAh g-1 even at an ultrafast rate of 10 C (and of 88 mAh g-1 at 50 C) with a capacity fading rate of only 0.02% per cycle for 1000 cycles owing to the decreased charge transfer resistance from 64.65% to 25.13 Ω [Figure 7(b)~(d)].
In 2019, Yoo et al. succeeded in preparing a SnO2-TiO2 composite by decorating SnO2 on a TiO2 microcone substrate by the potential shock method using a 0.4 Na2SnO3 H2O electrolyte in the potential range of 10~80 V for 10 s[64]. SnO2 is known as a potential anode material for LIBs with a high theoretical capacity of 780 mAh g-1 and a volume expansion ratio of ~150%. It was confirmed that amorphous SnO2 was deposited in the valleys between the crystalline microcones and their hollow cores, demonstrating that SnO2 is encapsulated by the TiO2 shell. Therefore, the large volume expansion of SnO2 during the charge/discharge process is effectively suppressed by counteracting the TiO2 microcone shell, achieving a specific capacity of over 500 mAh g-1 with excellent cycling stability over high current densities.
4. Others
Plasma electrolytic oxidation (PEO) is another type of anodization that involves a complex process of plasma-assisted electrochemical conversion of a metal surface to produce thick, hard, and well-adhered solid products of electrolysis and adsorbed gel layers at the surface of metals at high discharge temperatures and pressures[65,66]. In 2017, Lee et al. reported partially crystalline anodic TiO2 where SiO2 was uniformly composited over the entire oxide through the PEO of Ti foil using an aqueous electrolyte containing H2SO4 and Na2SiO3 with ionic conductivity below ~50 mS cm-1[67]. During the PEO process, during the formation of the microporous TiO2 layer, SiO32- ions migrated and were incorporated into the oxide layer owing to the strong electric field and subsequently oxidized to SiO2 at high temperatures. These microporous structures caused by plasma bubbles during the PEO process facilitated the movement of Li+ and suppressed massive volume expansion during cycling [Figure 8(a)~(c)]. As a promising anode electrode, as shown in Figure 8(d), the binder-free SiO2/TiO2 composite film exhibits twice the capacity with stable cycling stability over more than 250 cycles owing to the high content of SiO2 (≈ 25%) and durable structural properties with a low volume expansion rate of TiO2.
More recently, in 2019, Jie et al. fabricated TiO2/SiO2 composite by the one-step plasma discharge in an alkaline electrolyte for 2 min without post annealing process[68]. The core temperature ranging from 3000 to 5000 K facilitates the formation of composite film composed of crystalline TiO2 with uniformly distributed amorphous SiO2 within a short period of time. Due to the porous characteristic offering sufficient diffusion paths for Li+ ions, intrinsic properties of TiO2 with high stability and SiO2 with high specific capacity, the obtained electrode exhibits an excellent cycling stability and rate capability with capacity above 400 mAh g-1 at the current density of 100 μA cm-2.
5. Conclusions
TiO2 micro and nanostructures prepared by anodization are promising alternatives for carbon-based anode electrodes for developing next-generation LIBs with high power density and safety owing to their intrinsic characteristics of large surface area and short diffusion path of Li+ ions. TiO2 nanotubes grown in an electrolyte containing F- ions are obtained as highly ordered tubular structures via field-assisted oxidation and dissolution along with a plastic flow model. The dimensions of the nanotubes are determined by experimental conditions, including applied potential, temperature, and type of electrolyte. However, for the preparation of TiO2 microcones, the existence of other anions such as (COOH)-, SO42-, or PO43- with higher applied potential is essential to form a thick oxide layer that contributes to the formation of conical structures by pushing the oxide layer up and down to compensate for the high compressive stress generated. Typically, unlike anodic TiO2 nanotubes that exhibit an amorphous phase, anodic TiO2 microcones exhibit an anatase phase owing to their higher applied potentials.
Anodic TiO2 micro and nanostructural electrodes without binders show moderate electrochemical performance in LIBs using a Ti substrate as a current collector. Specifically, their capacity using microcones is three times higher than that using nanotubes because of their large surface area and structural stability. However, to overcome the intrinsic drawbacks of TiO2 including low electrical conductivity and capacity, several intensive studies have been performed to develop TiO2 composites with materials having high conductivity or specific capacity. In this review, we summarize recently reported anode electrodes comprising anodic TiO2 structures in which foreign materials were introduced to improve the specific capacity of TiO2 and suppress the volume expansion of these high-specific-capacity materials.