Skip to content
naulon

No retraction, correction or concern found when last checked on Oct 22, 2026.

Vanadium Oxide: Phase Diagrams, Structures, Synthesis, and Applications

Peng Hu, Ping Hu, Tuan Duc Vu and 6 more

Chemical Reviews · 2023 · Review · CC BY

doi:10.1021/acs.chemrev.2c00546

Abstract

. We begin with a tutorial on the phase diagram of the V-O system. The second part is a detailed review covering the crystal structure, the synthesis protocols, and the applications of each vanadium oxide, especially in batteries, catalysts, smart windows, and supercapacitors. We conclude with a brief perspective on how material and device improvements can address current deficiencies. This comprehensive review could accelerate the development of novel vanadium oxide structures in related applications.

Full text

From PubMed Central, U.S. National Library of Medicine, as current on Oct 22, 2026. Paragraph numbers are ours, so an answer can point at one.

Vanadium

Vanadium was first discovered by Andrés Manuel del Rio in Mexico City from Pb5(VO4)3Cl in 1801. However, it was wrongly identified as a form of chromium by Hippolyte Victor Collet-Descotils in 1805. Until 1831, Swedish chemist Nil Gabriel Self-ström in Stockholm named the element vanadium, which is from the Norse Goddess Vanadis and means beauty and fertility. In the Earth’s crust, vanadium is the 20th most abundant element and the sixth most abundant element among the transition metals.− However, some literature indicates that vanadium is the fourth most abundant transition metal after iron, titanium, and manganese.− High purity vanadium (about 99.7%) was first produced in 1925 by reducing vanadium pentoxide (V2O5) with calcium metal. Pure vanadium exhibits a transition metal feature, which shows a high melting point and good corrosion resistance at low temperatures. Vanadium can be dissolved in nitric and sulfuric acids but is insoluble in hydrochloric acid. In nature, vanadium is difficult to exist in metal form because it easily reacts with oxygen, even nitrogen and carbon at elevated temperatures., Vanadium is an important component of specific steel alloys, which provides additional tensile strength and extra protection against rust and corrosion of these materials.

Vanadium Oxides

Vanadium has the electronic configuration [Ar]4s23d3. Therefore, the oxidation state of vanadium can range from +5 to −3, and the valences of +5, +4, +3, and +2 are most commonly observed., Four vanadium oxides feature single oxidation states (+2 for VO, +3 for V2O3, +4 for VO2, and +5 for V2O5), and others have mixed oxidation states. Different oxidation states exhibit various colors: +5 (orange to yellow), +4 (blue), and +3 (green). The vanadium oxides exhibit crystalline structures with different oxygen coordinations, which result in the formation of octahedral, pentagonal bipyramids, square pyramids, and tetrahedral sharing corners, edges, or faces. The oxidation state of the vanadium cations dramatically affects the physicochemical properties of the vanadium oxides with different phases.

Due to the multioxidation states and various crystalline structures, the vanadium oxides exhibit excellent intercalation properties to host–guest molecules or ions, giving excellent catalytic activities, strong electron–electron correlations, outstanding phase transitions (metal–insulator transition), and high electrical conductivity. Furthermore, the abundant nanostructures of vanadium oxides can be achieved by different preparation methods, which not only shorten the transportation distance of ions or electrons and yield a faster solid-state diffusion in electrochemical energy conversion systems,, but also provide more active positions for the interaction with other molecules or ions and more exposed active crystal facets for catalysis applications., Thus, the vanadium oxides provide promising applications in energy conversion/saving fields,, such as ion batteries,− water splitting, smart windows,, supercapacitors,, sensors, and so on.

A series of vanadium oxides with strong electron–electron correlations exhibit metal–insulator transition (MIT). The V2O3, VO2, and V2O5 with single oxidation undergo MIT at 160 K, 340 K, and 530 K, respectively. These phase transitions are reversible and accompanied by a change of crystallographic, magnetic, optical, and electrical properties. The mixed-valence vanadium oxides belong to either Magnéli series (VnO2n-1) or Wadsley series (VnO2n+1). For the Wadsley series, V3O7 and V6O13 exhibit the phase transition at 5.2 and 155 K, respectively., Except for V7O13 (metallic), all the Magnéli series show a transition from a paramagnetic to an antiferromagnetic state and consequently exhibit an antiferromagnetic ground state at low temperatures, including V3O5 (430 K), V4O7 (250 K), V5O9 (135 K), V6O11 (170 K), and V8O15 (70 K) with different phase transition temperatures, respectively.

Scope of the Review

Vanadium oxides have a long history and rapid development in recent years as they are one of the most promising candidates in versatile applications in batteries, energy-saving smart windows, sensors, catalysts, optoelectronic devices, and so on. Therefore, many research papers and reviews have been published. Pioneering reviews on the chemistry of oxovanadium were published in 1965. The synthesis of vanadium oxides through hydrothermal and gas phase was reviewed by Whittingham, Livage, and Bahlawane. The atomic layer deposition of vanadium oxides was summarized by Papakonstantinou. The synthesis, properties, and applications of vanadium oxide nanotube were described by Kianfar. The catalytic applications of vanadium oxides have been described by Delferro, Hess, Carrero, and Granozzi. The Raman spectroscopy of vanadium oxides was recently reviewed by Shvets. The sensing properties of vanadium oxide nanostructures were described by Sheikhi and Madanagurusamy. The energy-related applications of vanadium oxides were reviewed by Xie,, Jiang, and Streb. A large number of reviews described the progress in metal ions batteries, including those by Chen, O’Dwyer, Mai,,, Lowe, Whittingham, Rashada, Lee, Yang, Zheng, Kim, Liang, and Cao. The vanadium oxides in supercapacitors were reviewed by Chen, Dutta, and Li. Furthermore, our group also reviewed the multistimuli responsive properties and energy applications of some vanadium oxides.,,,

The great potential of vanadium oxides for new applications and accelerated industrialization led to the dramatically increase of the importance of vanadium oxides over the last 5–10 years. Furthermore, due to the complexity of various oxidation states of vanadium, vanadium oxides show a large variety of stable and metastable structures, which pose an inevitable challenge to synthesize vanadium oxides with high purity, well controlled stoichiometry, and meticulously designed nanostructures, a must for high performance devices. Even though lots of reviews have been published, most of them focus on a specific kind or application of vanadium oxides, such as batteries, spectra, supercapacitors, and so on. No comprehensive reviews illustrate the different vanadium oxides with different applications. In this review, we focus on the most recent progress on the structure, synthesis, and applications of five thermodynamically stable vanadium oxides (V2O3, V3O5, VO2, V3O7, V2O5) and some metastable vanadium oxides (V2O2, V6O13, V4O9, etc.), which can provide a better understanding of a specific vanadium oxide phase and their process-structure–property interrelationships. Figure 1 summarizes the main contents of this review. This review begins with the phase diagram of the V–O system to show the different vanadium oxide phases, followed by the vanadium oxides with different stoichiometries. For each vanadium oxide, the structures, synthesis methods, and applications in several fields will be covered. The last section presents the future prospects and a summary of this review.

Phase Diagram of the V–O System

The V–O binary phase diagram was compiled according to previously reported experimental data in 1989. The oxygen-rich phases are well-defined, which reveal more than 20 compounds. However, the vanadium-rich phases exhibit broad homogeneity ranges and high nonstoichiometry.,

Several groups have calculated the V–O binary phase diagram.−Figure 2a shows a calculated phase diagram of the V–O system in the entire composition range at 1 atm. In the V-rich range, four types of solid solutions exist. The α and β solid solutions are formed by a certain amount of oxygen dissolved in the vanadium. The maximum solubilities of oxygen in α-V and β-V phase are up to 17.9 atom % and 27.4 atom %, respectively. The β-phase exhibits a wide range of homogeneities. With the increase of oxygen content, the γ and δ solid solutions phase can be formed. The γ-phase is monoclinic and δ-phase has the stoichiometry of VO with NaCl-type structure.

For the stoichiometric phases, only five phases are thermodynamically stable compositions as pure compounds, which include divanadium trioxide (V2O3, cubic, Ia3), trivanadium pentoxide (V3O5, monoclinic, P2/c), vanadium dioxide (VO2, tetragonal, P42/mnm), trivanadium heptaoxide (V3O7, monoclinic, C2/c), and divanadium pentoxide (V2O5, orthorhombic, Pmnm). Other metastable phases or unstable phases are likely to decompose to stable phases. All the stable and metastable phases are listed in the enlarged phase diagram of the V2O3–V2O5 system, which is shown in Figure 2b.

There are two types of vanadium oxides with a mixed valence of vanadium. One is the Magnéli series, which is defined by the general stoichiometric formula:

This type of homologous series has been reported for molybdenum oxides for the first time by Magnéli. The other homologous series is the Wadsley series, which has the general formula of VnO2n+1 (n ≥ 3). All the Magnéli phases maintain a triclinic symmetry (P1) and are metastable. They are expected to yield VO2 and V3O5 after lowering the system entropy. For the Wadsley series, V3O7 exhibits lower formation entropy compared with other Wadsley phases, which indicates the phase is more stable. Other Wadsley phases, such as V4O9 and V6O13, can be decomposed into a mixture of VO2 and V3O7. V4O9 and V6O13 show multiple metastable crystalline structures due to the close formation energies. Therefore, these compounds are candidates for polymorphism, where three crystalline structures were identified for V6O13 and two for V4O9.

Structures and Synthesis

V2O3 has a typical corundum-type hexagonal structure (space group: R3̅c) with lattice parameters of a = b = 4.9492(2) Å, c = 13.988(1) Å at room temperature., The crystal structures of V2O3 viewed in different directions are illustrated in Figure 3. It is interesting that after being calcined at 600 °C for several hours, vanadium vacancies (red circle in Figure 3a) formed, which is suitable for aqueous zinc metal batteries. From another viewing direction, V2O3 possesses an open tunnel structure consisting of a 3D V–V framework (Figure 3b). Such tunnel structures could efficiently facilitate the insertion of alkali metal ions, which provide potential application in metal ion batteries.

V2O3 with different morphologies can be obtained by several methods, including reduction, oxidation, and hydrothermal approaches. For the reduction pathway, V2O5 was employed as initial materials, and hydrogen or ammonia gas was used as reducing agents, whereas vanadium metal was used as a starting material for the oxidation method. In the hydrothermal route, vanadium alkoxides or sulfides with some small organic molecules (e.g., thiourea, benzyl alcohol, etc.) were added together to the autoclave to grow V2O3.

Li et al. designed a plasma hydrogen reduction system to synthesize V2O3 nanocrystals via a single precursor of V2O5 powders. The coarse-grained V2O5 powders are injected into the hydrogen plasma by a powder feeder, reducing the V2O5 powders into V2O3 by hydrogen. Such single crystalline V2O3 nanocrystals have a spherical shape with average sizes in the range of 35–50 nm (Figure 4a). The morphology of the raw materials determines the V2O3 morphology. Seshadri et al. first synthesized the V2O5 nanorods via a hydrothermal reaction followed by reducing in 5% H2:95% N2 (reduction time = 3 h and reduction temperature = 600 °C) to obtain V2O3 nanorods (Figure 4b). The ammonia gas (NH3) is another agent to reduce V2O5 to V2O3. V2O3 shows a much larger size with several micrometers, and the morphologies of the V2O3 particles were micrometer layered structures that were assembled by nanometer or micrometer sheets. Tao et al. reported that the V2O3 nanoparticles had been synthesized by supercritical ethanol fluid reduction of VOC2O4 with an average size of 50 μm. Madanagurusamy et al. employed the oxidation way to obtain the V2O3 nanosheets. High-density vertically aligned V2O3 nanosheets on glass substrates were obtained via a simple one-step sputtering technique. Vanadium metal was first deposited on the well-cleaned glass substrates followed by oxidation in the argon and oxygen mixture gas with a ratio of 3:1. Well-ordered, ultrathin, vertically aligned V2O3 nanosheets with voids were obtained (Figure 4c).

Hydrothermal is one of the most popular methods for crystal growth, which is conducted under moderate temperature and a high vapor pressure environment in sealed containers with the unique advantages of being easy to handle and environmentally friendly. Compared with the reduction and oxidation methods, the hydrothermal reaction is more convenient for synthesis the V2O3 with various morphologies. Niederberger et al. adopted vanadium alkoxides and benzyl alcohol as precursors to synthesize V2O3 nanocrystals through hydrothermal reaction sizes ranging from 20 to 50 nm with good yields (Figure 4d). Without using any surfactant and template, Su et al. successfully synthesized dandelion-like V2O3 microspheres with core–shell structures. With increasing reaction time, the morphology of V2O3 can be tuned from a solid sphere to dandelion-like. If the reaction time is further increased, some broken V2O3 microspheres with core–shell structures could be observed with an average diameter of the core–shell microspheres of 2 μm.

Batteries

As the low valence of V3+ in vanadium oxide, corundum-type V2O3 with a metallic behavior shows that the electrons in the V-3d orbital travel along the V–V chains. The 3D V–V framework provides an intrinsic tunnel structure, which is suitable for ion transport and intercalation. Mai et al. reported uniform nitrogen-doped carbon-confined V2O3 (V2O3@NC) hollow spheres (Figure 5a). The in situ carbonization hollow structure can provide high ion/electron conductivity, short diffusion distance, and excellent structure adaptability, which is beneficial for lithium-ion storage (Figure 5b). The V2O3@NC delivers an average discharge capacity of 785, 599, 528, and 361 mAh g–1 at a current density of 100, 500, 1000, and 5000 mA g–1, respectively (Figure 5c). Furthermore, 811 mAh g–1 can be maintained after 120 cycles at a current density of 200 mA g–1 (Figure 5d). In addition, V2O3 can also be used as other ions (Na+/K+/Zn2+) storage material. Jiao’s group fabricated a flexible and self-standing electrode of V2O3 nanoparticles embedded in porous N-doped carbon nanofibers (V2O3@PNCNFs) through electrospinning assisted high-temperature sintering method, which can directly be used as a KIB anode material (Figure 5e). V2O3@PNCNFs deliver a capacity of 240 and 134 mAh g–1 at a current density of 50 and 1000 mA g–1, respectively. High-capacity retention of 94.5% can be maintained after 500 cycles (Figure 5f). The density functional theory (DFT) results demonstrate that 1 mol K+ can insert into the V2O3 crystal forming KV2O3, with the K+ occupying the 6e sites of the KV2O3 crystal. The main capacity of V2O3 in KIB is main contribution from capacitance (surface or near-surface redox reactions), which is very different from the conversion reaction of conventional transition metal oxides for ion storage. When K+ inserts into the tunnels of V2O3@PNCNFs, the material exhibits no structural phase transitions, which indicates that V2O3@PNCNFs can exhibit excellent K+ rate/cycling performance. This work suggests that the pseudocapacitive electrode materials could be suitable for a large-scale energy storage system. Generally, the conventional low valent V2O3 cannot effectively accommodate Zn2+ intercalation during the discharging process due to its inherently unsuitable structure and inferior physicochemical properties. Luo et al. attempted to utilize V2O3 in aqueous ZIBs by the in situ anodic oxidation strategy, and the hierarchical microcuboid structure V2O3 can accommodate nearly 2 electrons intercalation. Moreover, they found H2O is a reactant that participates in the first charge oxidation process of V2O3. Furthermore, the porous structure with a higher specific surface area of the V2O3 leads to more reaction sites and a faster phase transition from V2O3 to V2O5–x·nH2O (Figure 5h). Meanwhile, the high surface and the small size of V2O3 nanoparticles benefit the first charge oxidation reaction process. The V2O3 delivers a Zn2+ discharging capacity of 625 mAh g–1 at 0.1 A g–1, corresponding to 1.75-electron intercalation (Figure 5g). Specifically, the capacities can maintain 87% and 78% when the current increases to 10 and 20 A g–1, respectively. The V2O3 can maintain 100% after 10000 cycles at 10 A g–1, which is better than some previously reported ZIB cathode materials.

Catalysts

V2O3 and its composites have been widely used as catalysts for chemical looping dry reforming of methane, ammonium perchlorate decomposition, the hydrogen evolution reaction (HER),, the oxygen evolution reaction (OER), water splitting, etc.

Propane Dehydrogenation and Ammonium Perchlorate Decomposition

Very recently, Zhu et al. examined the catalytic properties of propane dehydrogenation through single transition metal atom doping of a V2O3 (0001) surface by self-consistent DFT calculation. The results indicated that the single atoms act as promoters and active sites, and Mn–V2O3 is a good candidate as a catalyst for propane dehydrogenation. Huang et al. synthesized V2O3 and V2O3/carbon composites by a facile hydrothermal route, which exhibited excellent performance for ammonium perchlorate decomposition. The decomposition temperature decreased by 49 and 73 K for V2O3 and V2O3/carbon composites, respectively.

HER, OER, and Water Splitting

Electrocatalytic water splitting consisted of two half-reactions: HER and OER. Electrocatalysis is used to accelerate the rate of a chemical reaction through lowering the activation energy to reduce the electrochemical overpotentials.

V2O3 composited with other functional materials can exhibit efficient electrocatalysis performance. Zhang et al. reported a self-supported MoSx/V2O3 heterostructure for the HER. Two steps were involved in the synthesis process (Figure 6a). V2O3/carbon cloth (CC) was first obtained by a hydrothermal method, and the dense V2O3 was uniformly distributed on CC fibers. The as-prepared V2O3/CC was immersed in an ammonium thiomolybdate solution and dried under a vacuum. Then, followed by a thermal decomposition process, the final MoSx/V2O3/CC was successfully achieved with the same morphology as V2O3/CC. The XRD pattern (Figure 6b) shows the composite contains the MoSx and hexagonal V2O3. MoSx/V2O3/CC displayed an overpotential of 146 mV to achieve a 10 mA cm–2 HER current density (Figure 6c), which is lower than that of MoSx/CC (221 mV). The Tafel slope of MoSx/V2O3/CC is around 45 mV dec–1, suggesting the HER process obeys the Volmer-Heyrovsky mechanism. Meanwhile, the MoSx/V2O3/CC presents a higher double-layer capacitance (Cdl: 85 mF cm–2) than MoSx/CC (4 mF cm–2), indicating V2O3 can create more active sites for the HER (Figure 6d). Furthermore, the MoSx/V2O3/CC electrocatalyst has great stability in the acid electrolyte for the HER. Two reasons for the MoSx/V2O3/CC electrocatalyst were given to understand the improved HER activity: (1) V2O3 enhanced the electrochemically active surface area with more active sites for the HER; (2) better electron transfer between MoSx and V2O3. Qiu et al. synthesized V2O3 nanosheets anchored with NiFe nanoparticles as a bifunctional electrode for overall water splitting. A self-templated strategy was employed to synthesize the NiFe@V2O3 (Figure 6e). By a calcined reduction process, ZnO@NiFe@V2O3 nanosheets can be obtained. In-situ alkaline media corrosion was performed to dissolve the ZnO NPs and produce the porous NiFe@V2O3 nanosheets, which exhibit a clear V2O3 porous matrix and NiFe nanoparticles (Figure 6f). The porous NiFe@V2O3 exhibits good OER performance in an alkaline medium with an overpotential of 255 mV at 10 mA cm–2 and good stability (Figure 6g,h). Meanwhile, the NiFe@V2O3 also gave good HER performance in the same alkaline medium. It shows an overpotential of 84 mV at 10 mA cm–2 and good stability (Figure 6i,j). Furthermore, the porous NiFe@V2O3 delivers a small Tafel slope of 51 mV dec–1 for OER and a Tafel slope of 85.4 mV dec–1 for the HER, respectively. Considering the excellent HER and OER performance, two identical NiFe@V2O3 electrodes are integrated into a two-electrode cell to investigate the water splitting performance. The catalyst shows a cell potential of 1.56 V to reach 10 mA cm–2 with an ignorable cell voltage increase during 20 h measurements (Figure 6k).

The stability of the catalyst is a key parameter for practical applications, which is crucial to determine whether the catalyst can be commercialized. Several factors limit the catalytic stability, such as poor chemical/electrochemical stability of the catalysts under operation, abandoned gas evolution leading to the physical detachment of catalysts, dissolution of the catalysts in the electrolytes with different pHs, poor mechanical stability, and so on.− For example, Schmidt et al. proved that the metal atoms are thermodynamic unstable during the OER process due to metal oxide released lattice oxygen, which leads to the low stability of the catalysts. From the theoretical perspective, there are three critical criterion of the catalysts that should be considered: excellent water dissociation performance, suitable Gibbs adsorption energy of H* (ΔGH*), and faster H2 desorption. Such properties enable the fast release of the active site without destroying the active site and further improve the stability of the catalysts.

Supercapacitors and Electromagnetic Wave Absorber

Supercapacitors (SCs) have attracted attention in recent years to bridge between a classic electrolytic capacitor and rechargeable battery, which are characterized by high power density and good cyclic stability. SCs are capable of storing electrical energy via two mechanisms: the electrochemical double layer capacitors having a nonfaradaic charge character and the pseudocapacitors based on faradaic electrochemical redox reactions. The theoretical specific capacitance of a pseudocapacitive electrode is proportional to the number of electrons involved in a specific redox reaction, and vanadium oxides possess four readily accessible valence states, making vanadium oxides especially promising for high pseudocapacitance. The conductivity of V2O3 (∼103 Ω–1 cm–1) is higher than monoclinic VO2 (∼4 Ω–1 cm–1) and comparable with Ru (∼104 Ω–1 cm–1). Meanwhile, V2O3 is stable in both acid and base mediums. Thus, V2O3 is a suitable material for energy storage, especially for SCs. However, the reported specific capacitances of V2O3-based materials are not good enough, suggesting that design and synthesis of new structured V2O3-based materials with high performance are required. Cao et al. synthesized V2O3 nanoparticles highly dispersed in amorphous carbon composites through the calcination of the (NH4)2V3O8/C precursor, which was fabricated through the hydrothermal reaction by using commercial NH4VO3 and glucose (Figure 7a). The as-prepared V2O3-based composites exhibit a specific capacitance of the electrode of 458.6 F g–1 at a current density of 0.5 A g–1 (Figure 7b), which is higher than other reported V2O3-based composites. Meanwhile, the asymmetric supercapacitors assembled by the as-prepared V2O3-based composites display good flexibility properties (Figure 7c). The capacitances are almost constant with the bending from 180° to 30°. The V2O3-based composites are one kind of high-efficiency electromagnetic wave absorber. Yan et al. produced the hierarchical Co/C@V2O3 hollow spheres by hydrothermal, aging, and annealing methods (Figure 7d). The VO2 spheres were first fabricated through a hydrothermal method followed by coating with ZIF-67 by precipitation. After that, the ZIF-67@VO2 composite was calcined in H2/Ar to form the hierarchical Co/C@V2O3 hollow spheres. The Co/C@V2O3 hollow spheres exhibited excellent electromagnetic wave absorption performance with a reflection loss of @40.1 dB and a broad bandwidth of 4.64 GHz at a small thickness of only 1.5 mm (Figure 7e). The good performance is mainly due to the impedance matching and low density, which come from the combination of hollow V2O3 spheres and porous Co/C.

Structures and Synthesis

V3O5 crystal structure was first reported in 1954, which possesses a monoclinic symmetry (space group P2/c). The crystal structure is shown in Figure 8. The oxygen atoms are occupied in the octahedral sites, and vanadium atoms are located in the center of octahedron. There are two types of octahedra chains along the c axis: face-shared octahedra via edge-sharing and corner-sharing octahedra. Such chains formed a framework with many large open spaces, which is capable of accommodating lithium ions.

It is difficult to stabilize the V3O5 phase by using solid-state chemistry and to control the stoichiometry between oxygen and vanadium by using a solution method, which makes V3O5 an uncommon phase used for electrochemical and other applications., Therefore, the reported synthetic method for V3O5 is very limited. Reduction of V2O5 by different reductants is commonly used to obtain the V3O5 polycrystalline powders. Reisner et al. selected vanadium metal as a reductant to reduce V2O5, and the V3O5 polycrystalline powders were synthesized according to the following chemical reaction: V + V2O5 → V3O5. Vanadium powder and V2O5 powder were mixed together and pressed into bars, which were sealed in a quartz ampule and heated for 24 h at 870 K and for 100 h at 1220 K. The V3O5 polycrystalline powders exhibit a size around 10 μm. Alternatively, Yu et al. used sulfur powders as a reductant. The mixture of sulfur and V2O5 powders were vigorously grounded and calcined at 1023 K in a tube furnace under a vacuum for 2 h. The V3O5 microcrystals can be obtained by washing the calcined powders with nitrogen tetrasulfide several times. The obtained V3O5 microcrystals range from 1 to 3 μm. The large size single crystal of V3O5 was grown by chemical vapor transport in the 1970s., V3O5 powders or V2O3–VO2 mixed powders were used as starting materials, and TeCl4 was used as a transport agent. Both starting materials and transport agent were sealed in a quartz tube with low pressure (∼1 × 10–5 mbar), and then the tube was put into a two-zone furnace. Generally, the source zone has higher temperature, while the crystallization zone has a lower temperature, and the temperature gradient is around 100 to 150 K. The size of final V3O5 single crystal is as large as 1 cm.

Batteries

V3O5 is relatively less-studied though it exhibits a three-dimensional open-framework structure, which is due to the strict synthesis condition. Yu’s group successfully synthesized V3O5 microcrystals via vacuum calcination and first employed it as an LIB anode material. The oxygen atoms are closely arranged in a hexagonal shape, and the vanadium atoms take up three-fifths of the octahedral interstices (Figure 9a). The 3D open-framework structure of V3O5 is formed with chain connections, which contain distorted, sharing corners, edges, and faces of VO6 octahedral, respectively (Figure 9b). The 3D framework of V3O5 with much large open space endows the capacity of Li+ intercalation/deintercalation. No impurity peaks from the XRD pattern are observed (Figure 9c), which indicates that the V3O5 powder shows a single-phase nature with a monoclinic structure within the P2/c space group (JCPDS card 72-0977). It delivers a high capacity of 628 mAh g–1 at 100 mA g–1, a good rate (125 mAh g–1 at 50 A g–1), and a long stable cycling performance (117 mAh g–1 after 2000 cycles) (Figure 9d,e). There is no obvious crystal structure change of the V3O5 with Li+ intercalation, which is the main reason for the good rate and cycling performance (Figure 9f).

Other Applications

V3O5 thin films exhibit a photoinduced insulator-to-metal phase transition, which results in a strong nonlinear optical response.− Fernández et al. deposited the V3O5 directly on the SiO2 substrates by DC magnetron sputtering to form thin films. The ultrafast nonlinear optical response was probed by using a pump–probe scattering technique. A reduction in the transient relative scattered light signal was observed, which showed an ∼10% decrease within 800 fs. Such a response is due to the changes in the material’s optical constants and very likely related to the photoinduced insulator-to-metal phase transition. The photoinduced screening of electron correlations followed by melting of polaronic Wigner crystal and coalescence of V–O octahedra is the main reason for the order–disorder structural transition.

Structures and Synthesis

Figure 10a shows the crystal structure of V3O7. The unit cell contains 36 vanadium atoms (12 vanadium atoms are inside the octahedra and 24 vanadium atoms are five-coordinated). The V3O7 consists of VO6 octahedra and VO5 polyhedra, which are linked by corners and edges to form a three-dimensional framework. The crystal structure of V3O7·H2O shows a two-dimensional structure compared with V3O7 (Figure 10b). Each V3O8 layer consists of corner- or edge-shared VO6 octahedra and VO5 polyhedra. The water molecules are located at the sides of the V3O8 layer, where the hydrogen atoms are directly bonded to the vanadium atom of a VO5 polyhedron and hydrogen bonds connect two neighboring V3O8 layers.,

Bulk V3O7 single crystal can be grown by a typical chemical vapor transport method using V3O7 polycrystalline powders as starting materials and NH4Cl as a transport agent. The mixed V3O7 polycrystalline powders and NH4Cl were pressed into a pellet and heated at 823 K in an evacuated silica tube for 7 days. The black needle-like single crystals of V3O7 were formed with a length of 2 mm (Figure 11a). Nanostructured V3O7 was generally obtained by a typical hydrothermal method. The precursor solution dramatically affects the morphology of the V3O7 nanostructures. Wen et al. combined a soft chemical topotactic synthesis and hydrothermal process to prepare V3O7 nanobelts. In the beginning, layered structured KV3O8 plate-like particles were first prepared as the precursor by a hydrothermal method of V2O5 and KOH. The H+-form vanadate (HVO) nanobelt colloidal solution was subsequently obtained by reacting KV3O8 plate-like particles with HNO3. Finally, the one-dimensional (1D) pure single crystal V3O7 nanobelts were successfully prepared after hydrothermally treating the colloidal solution at 180 °C for 12 h (Figure 11b). By using NH4VO3 and HCl as starting materials, the nest-like V3O7 self-assembled by porous V3O7 nanowires on Ti foil was also synthesized through a hydrothermal method. In the beginning, a layer of V3O7 nanosheets was deposited on the Ti substrate, which was subsequently placed in the NH4VO3–HCl solution and kept at 160 °C for 10 h for a hydrothermal method. With the increase of hydrothermal time, the nest-like V3O7 is self-assembled by nanowires. The schematic synthesis process and SEM images for nest-like V3O7 are shown in Figure 11c–f. The V3O7 fibers can be obtained by thermal treatment of the electrospun NH4VO3/PVP nanofibers in the presence of reductant. The thermal treatment condition dramatically affects the final products, and V2O5 may be obtained together with V3O7.

The hydrothermal reaction is also a facile way to synthesize V3O7·H2O nanostructures. The V3O7·H2O nanobelts can be achieved by using V2O5, phenolphthalein and distilled water as starting materials through a hydrothermal method for 4 days of reaction at 180 °C. By changing phenolphthalein to ethanol or glucose, V3O7·H2O nanobelts were also successfully synthesized at 180 °C, while the reaction time was shortened to 12 h., In above-mentioned the methods, phenolphthalein, ethanol, or glucose is used as the reductant. Without the reductant, the V3O7·H2O nanobelts or nanowires can be obtained by a hydrothermal reaction of only V2O5 or NH4VO3., The key point is the pH of the precursor solution. After adjusting the pH to 3 by adding concentrated HCl, V3O7·H2O nanobelts and nanowires were obtained by a reaction at 190 °C for 24 h and 160 °C for 3 h, respectively. Furthermore, ultrafine V3O7·H2O nanogrids can be obtained through electrochemical oxidation. In the beginning, the nsutite-type VO2 black powder was synthesized by a hydrothermal method. Then, a three electrode system was employed to electrochemically transform the VO2 precursor to V3O7·H2O. The slurry, consisting of VO2 nanoplates, was coated on the working electrode. With a constant current density of 50 mA cm–2 and a cutoff potential of 1.7 V, V3O7·H2O nanogrids were obtained (Figure 12).

Batteries

V3O7 is a mixed-valence vanadium oxide for metal-ion storage. Yan et al. designed and synthesized a V3O7 nanowire templated graphene scroll (VGS) via an “oriented assembly” and “self-scroll” strategy. They used joint experimental-MD simulation to investigate the construction and formation mechanisms of VGS. The systemic energy, the curvature of nanowires, and the reaction time determined the length and formation process of the semihollow bicontinuous structure. Through this strategy, the VGS with a length up to 30 μm has interior cavities between the nanowire and scroll (Figure 13a). The unique structure of VGS with the nanowire templated graphene scroll offers a continuous Li+/ion transfer channel and free volume expansion space during Li+ de/intercalation (Figure 13b). The VGSs exhibit a high capacity of 321 mAh g–1 and good cycle stability (87.3% after 400 cycles), which is better than the pure V3O7 nanowire and V3O7 nanowire/graphene structure (Figure 13c).

In addition, Nazar et al. synthesized layered V3O7·H2O nanobelts with single crystalline via a microwave solvothermal method and applied it to a ZIB cathode in nonaqueous and aqueous electrolytes. The electrochemical studies and in situ XRD results demonstrate the different electrochemical behaviors of layered V3O7·H2O in nonaqueous and aqueous electrolytes. Combining the DFT calculations, ∼2 mol Zn2+ can insert into V3O7·H2O per the formula in the ZnSO4/H2O aqueous electrolyte (Figure 13d). The V3O7·H2O delivers a capacity of 400 mAh g–1 (>2 mol Zn2+ insertion) with an average voltage of ∼0.65 V at the first discharge process, and maintains 375 mAh g–1 at the subsequent charge process (Figure 13e), while in the Zn(CF3SO3)2/acetonitrile nonaqueous electrolyte, the V3O7·H2O exhibits a poor Zn2+ storage performance (59 and 175 mAh g–1 for the first and 50th cycles at 5 mA g–1, respectively) (Figure 13f).

Magnesium-ion battery (MIB) as another multivalent ion battery has been attracting more attention due to its high abundance in the Earth and low redox potential (−2.37 V vs. SHE). V3O7·H2O with high electronic conductivity (V+4.67) has been widely used as a cathode material in LIB/NIB and hybrid Li+/Mg2+ batteries. Hong et al. synthesized V3O7·H2O nanowires via a one-step hydrothermal method and applied them to a high-energy MIB cathode (Figure 13g). The electrochemical tests and structural characterization results demonstrate that the structured water in V3O7·H2O will remain stable during the cycling. 0.97 mol Mg2+ inserts into V3O7·H2O, accompanying the formation of Mg0.97H2V3O8 at the first discharged state. V3O7·H2O exhibits an initial discharge capacity of 231 mAh g–1 at 10 mA g–1 with an average discharge voltage of ∼1.9 V, and the energy density can reach 440 Wh kg–1 (Figure 13h). Meanwhile, V3O7·H2O delivers a 171 mAh g–1 and maintains 132 mAh g–1 (77%) after 100 cycles at 40 mA g–1 (Figure 13i). The excellent Mg2+ storage performance is attributed to the unique crystal structure with direct bonding. This strategy of applying water-metal bonding and hydrogen bonding provides a new idea to search for new oxide-based MIB materials with stable and high energy density.

Ammonium Perchlorate Decomposition

Ammonium perchlorate, a common oxidizer, plays a key role in the combustion of composite solid propellants. Furthermore, the catalyst greatly affected the performance of composite solid propellants by the thermal decomposition of ammonium perchlorate., Huang et al. found that the thermal decomposition temperatures of ammonium perchlorate in the presence of V3O7·H2O nanobelts and V3O7·H2O@C core–shell structures can be dramatically reduced. Both V3O7·H2O nanobelts and V3O7·H2O@C core–shell structures were synthesized by a hydrothermal method. Especially, the core–shell structures are synthesized by using V3O7·H2O nanobelts as the cores and glucose as the source of carbon. A well-defined nanobelt morphology with a length up to several micrometers can be observed (Figure 14a), which consists of a V3O7·H2O core and carbon shell (Figure 14b). The thermogravimetric analysis (TGA) indicated that the addition of V3O7·H2O or V3O7·H2O@C in ammonium perchlorate exhibited a significant reduction in the decomposition temperature of ammonium perchlorate (Figure 14c). The thermal decomposition temperature was lowered by 70 and 89 °C by adding V3O7·H2O or V3O7·H2O@C, respectively. The V3O7·H2O@C core–shell structures exhibited a higher catalytic activity than V3O7·H2O, with two possible mechanisms proposed. First, the partially filled 3d orbit in the vanadium atom promoted the electrotransfer process by accepting electrons from ammonium perchlorate and further accelerated the thermal decomposition of ammonium perchlorate. Second, the amorphous carbon shell possessed lots of active groups (such as C=C, C=O), which could facilitate the thermal decomposition of ammonium perchlorate.

Supercapacitors

V3O7 and V3O7·H2O are promising supercapacitor materials due to their layered structure and mixed oxidation states of +4 and +5. V3O7 can be converted to V6O13 at the lowest potential of −0.6 V and V2O5 at the highest potential of 0.2 V. Thus, the working potential window is in the range of −0.6 and 0.2 V. Huang et al. fabricated porous V3O7 nanowire self-assembled nest-like V3O7 and investigated the supercapacitor properties. The pure nest-like V3O7 exhibits worse supercapacitor performance compared with N-doped carbon coated nest-like V3O7 composites. The N-doped carbon coated nest-like V3O7 electrode showed a higher specific capacity of 660.63 F g–1 at 0.5 A g–1 compared to V3O7 (362.63 F g–1). Even at a higher current density of 50 A g–1, the N-doped carbon - V3O7 electrode still exhibits a better performance (187.72 F g–1) than V3O7 (33.18 F g–1), as shown in Figure 15a. Furthermore, the N-doped carbon coated nest-like V3O7 electrode also delivers better stability (80.47% capacitance retention after 4000 cycles) compared with pure V3O7 (23.16% capacitance retention after 4000 cycles). The superior performance of N-doped carbon coated nest-like V3O7 is mainly due to the unique three-layer structure: V3O7 core/carbon/nitrogen doped carbon (Figure 15b). Such a unique three-layer structure can not only stabilize V3O7, but also provide high-speed ionic and electronic transmission channels, which is responsible for the good supercapacitor performance of N-doped carbon coated nest-like V3O7. Yu et al. reported the growth of V3O7 nanowires on a carbon fiber cloth through a hydrothermal method. The obtained V3O7/carbon fiber cloth composites show a spider web-like morphology, which exhibits robust adhesion. The composite electrode gives a maximum specific capacitance of 151 F g–1 at a current density of 1 A g–1 with ultrahigh cycling stability of 97% (after 100000 cycles) in a full cell configuration (Figure 15c). Meanwhile, the V3O7/carbon fiber cloth composites reveal maximum power and energy densities of 5.128 kW kg–1 and 24.7 Wh kg–1, respectively by using 1-ethyl-3-methylimidazolium trifluoromethanesulfonate as the electrolyte. Furthermore, coin cell-type configuration with the V3O7- carbon fiber cloth composites electrode was assembled. The symmetric supercapacitors successfully and effectively power light-emitting diodes to produce blue light (Figure 15d). Huang et al. reported that V3O7·H2O nanobelts exhibited a capacitance of 447.6 F g–1. However, the cycling performance is limited by the poor conductivity and high solubility in an aqueous electrolyte. Therefore, composing with another conductive phase could be an alternative way to fabricate high-performance V3O7·H2O based materials. When the V3O7·H2O nanobelt is incorporated with carbon nanotube and reduced graphene, the formed 3D hierarchical porous composites exhibit outstanding electrochemical performance with a high specific capacitance (685 F g–1 at 0.5 A g–1) and excellent cycle stability (99.7% after 10,000 cycles) (Figure 15e,f). Meanwhile, the composites also give relatively high energy densities and power densities of 34.3 Wh kg–1 and 150 W kg–1, respectively. The better electrochemical performance can be attributed not only to the highly conductive carbon materials, but also to the 3D hierarchical porous structure. The carbon materials offer the transport pathway bridges, leading to the rapid transfer of charges. Meanwhile, the 3D porous structure minimizes the diffusion distance and supplies a large surface area with abundant active sites.

In general, vanadium oxides have received massive interest as supercapacitor electrodes that exhibit high theoretical specific capacity than most of the other transition metal oxides due to their variable valence state from V2+ to V5+. In addition, the layered structure of vanadium oxides facilitates the intercalation/deintercalation of electrolyte ions during the charging/discharging process. However, vanadium oxides-based supercapacitor electrode materials still suffer from poor long-term cycling stability, which is usually caused by the collapse of a layered crystal structure, severe agglomeration of particles, and low electrical conductivity. The electrochemical stability of vanadium oxide-based supercapacitors can be improved by material modification, optimization of the structure, or combining with other materials with excellent electrical conductivity. Developing vanadium oxide nanomaterials with suitable micro-/nanostructures is an important factor to improve the cycling stability. 3D vanadium oxides with micro-/nanostructures are often employed, including microspheres and hollow spheres, which can inherit the superior high surface area characteristics of nanobuilding blocks, and simultaneously possess a decent structural stability.− Furthermore, the 3D structure can effectively reduce the agglomeration of particles, which is beneficial for the cycling and rate performance. Integrating vanadium oxides with carbon materials has been demonstrated to be effective to suppress the structural degradation upon cycling. The carbon materials, such as porous carbons and graphene, can serve as elastic buffering layers to release the strain within metal oxides during cycling. To some extent, carbon materials can avoid loose attachment between the electrode material and current collector, which helps improve both the conductivity and the stability of the supercapacitor.

Electrochromism

Nanostructured V3O7 thin films showed electrochromic properties by using lithium perchlorate as the electrolyte, which was prepared by a nebulizer spray pyrolysis technique. The color of the films is changed from yellow to pale blue by applying an external potential of 1.5 V for intercalation of Li+ ions, while the color is reversed by applying an external potential of −1.5 V for deintercalation of Li+ ions. Such results indicate that the nanostructured V3O7 thin films could be effectively used for smart window applications. In general, several models are proposed to explain the electrochromic phenomenon, such as the color center model (Deb model), electrochemical redox model, and so on. The “Deb’s color center model” is in nature related to the defects (such as oxygen vacancies) induced change of the visible light absorption, which is generally independent of the external electric field, while in the “electrochemical redox model” it is believed that the injection and trapping of a large density of electron and hole lead to coloration. The mechanism of electrochromism for V3O7 thin films is the variation of the band structure caused by intercalation/deintercalation of Li+ ions, which belongs to the electrochemical redox model. During the application of negative bias, Li+ ions are absorbed onto the surface of V3O7 and diffuse into the lattice of V3O7. The intercalated Li+ ions react with O2– ions to introduce oxygen vacancies in the lattice and reduce the V5+ in the mixed state V3O7 to V4+. As the result, the optical transparency of the film changes. Moreover, the intercalation of Li+ ions upshifts the Fermi level close to the conduction band, which leads to an increased transmittance of V3O7. The process can be reversed by deintercalating of Li+ ions through applying positive bias. Furthermore, the V3O7·H2O thick films also show good electrochromic properties (good reversibility and good color switching between reduced and oxidized states). By using lithium bis-trifluoromethanesulfonimide (LiTFSI) as the electrolyte, the color of the V3O7·H2O films changes from green to orange by applying a positive potential of 1.9 V, while it changes from green to blue by applying a negative potential of −1.2 V (Figure 16a). Such film presented a maximum optical reflectance modulation of 29% at 590 nm (Figure 16b). The V3O7·H2O films still exhibited color changes by using a Na-based electrolyte (Figure 16c) and a maximum optical reflectance modulation of 10% at 590 nm (Figure 16d). The maximum optical reflectance modulation of Na-based electrolyte is lower than that in the Li-based electrolytes, which is mainly due to a larger faradaic contribution resulting from the larger cation size of Na+ ions.

Structures and Synthesis

Vanadium dioxide (VO2) can exist in various polymorphic phases, including but not limited to VO2 (B), VO2 (A), VO2 (M), VO2 (R), VO2 (D), and VO2 (P). Some of these phases and their corresponding lattice parameters are shown in Figure 17. The discussion will focus more on VO2 (R) and its reversible MIT to VO2 (M) as well as on VO2 (B) due to its application as cathode materials in electrochemical devices.

VO2 was first demonstrated to undergo MIT in 1959 by Morin. At the critical temperature (τC) of about 340 K, VO2 transforms from high-temperature, conducting rutile VO2 (R) to low-temperature, insulating monoclinic VO2 (M). There are two mechanisms that have been used to describe this ultrafast phase transition phenomenon in VO2, namely the Peierls model and Mott-Hubbard model. The Peierls model describes the nature of MIT as the change from a shared d-orbital between all vanadium atoms to localized d-orbital in the V–V dimer, which is a result of the change in the V–V distance from 2.88 Å in VO2 (R) to 2.65 and 3.12 Å in VO2 (M). Therefore, the Peierls model stated that the structural distortion is the cause of MIT. Wentzcovitch et al. applied the local-density approximation to study the electronic and structural change of VO2 during the MIT. From the view of band theory, a monoclinic distorted state is in good agreement with the experiment result. Meanwhile, the structural distortion enhances the bonding between neighboring V atoms, which is expected in the Peierls model. Moreover, Baum et al. utilized four-dimensional (4D) femtosecond electron diffraction to visualize the phase transformation process of VO2. They pointed out that during the MIT, the displacement of atoms happened within picoseconds, and followed by the sound wave shear motion of the crystal in the time scale of nanoseconds. The observation indicates the occurrence of fast structural distortion during the MIT.

On the other hand, the Mott-Hubbard model states that MIT would occur when the electron density (ne) and Bohr radius (aH) satisfy ne1/3aH ≈ 0.2. Compared with the Peierls model, the Mott-Hubbard model has the advantage in explaining the phenomenon such as the anomalously low conductivity in the metallic phase. Whittaker et al. summarized multiple experiment cases and pointed out that the metallic phase of VO2 might be introduced without the structural phase transformation if the excitation of carriers reaches a threshold density. Their observation provides key evidence in revealing the nature of MIT. While there remains a debate on which mechanism best describes the MIT, the usage of both models is strongly encouraged due to the transition kinetics as well as the stimuli involved during the transition. Shao et al. reported recent progress in understanding the mechanism and kinetics of MIT, including the lattice distortion and electron correlations (Peierls phase transition, Mott phase transition) and modulation methods (elemental doping, external electric field, light irradiation, and strain engineering).

Upon application of suitable external stimuli (i.e., photons, heat, electric, magnetic, electrochemical, and stress) to initiate the MIT, physical properties of VO2, such as electrical resistance, optical transmittance, and thermal conductivity, are reversibly and drastically changed. Long et al. have summarized the connections between the stimuli and responses of VO2-based devices in detail. This flexibility in external stimuli and corresponding responses is the main reason why VO2 is a material of interest in multiple novel devices spanning from thermally,, electrically,− to optically, activated devices. The functional performance of these devices is thus highly dependent on, but not limited to, the physical attributes, such as dimension, morphology, doping level, and crystallinity, of the fabricated VO2. Multiple techniques have been used to fabricate functional VO2 devices, each with their own strengths and weaknesses. Some of these fabrication processes are reviewed with polymer-assisted deposition methods, hydrothermal, sol–gel, chemical vapor deposition (CVD), and physical vapor deposition (PVD) methods discussed in detail regarding the controlled synthesis of VO2 for thermochromic application.

Applications

In recent years, multiple omnibus reviews− have been done in attempts to give the most encompassed view of VO2 research progress, often including a combination of the following topics: MIT mechanism, kinetics, fabrication techniques, and applications. While these reviews could offer wide coverage of VO2 research progress in multiple topics, in-depth discussion of each topic, applications in this case, is in much needed demand. In subsequent sections, reported applications of VO2 in functional devices both based and not based on MIT in recent years, especially in the last three years, are classified and discussed in different categories: optical, electrical, and mechanical applications.

Optical Applications

The operations of these VO2-based optical devices are often based on the changes in refractive index, n, and extinction coefficient, k of VO2 upon crossing the MIT. The changes to n and k at different temperatures are shown in Figure 18. These optical functional devices are further divided into two main groups: infrared regulators and optical switches.

Infrared Regulators

Since its first introduction in 1985, the smart window has attracted much attention due to rising energy consumption in the commercial building sector, which contributes up to 40% of total consumption globally and leads to 30% of global greenhouse gas emissions today. VO2 is the prime candidate for smart window materials due to its ability to seamlessly and rapidly regulate the amount of infrared (IR) across MIT with miniscule side effects to the visible transmission. However, intrinsic limitations (high τC ≈ 68 °C, low luminous transmittance (Tlum) < 40%, poor solar modulation (ΔTsol) < 10%, and poor durability) prevent pristine VO2 from meeting the requirements for commercial smart window applications. The most common way to reduce τC is by elemental (i.e., W, Mo, Ti, F, Mg, etc.) doping as summarized by Cui et al. Different approaches exist to enhance Tlum and ΔTsol via advanced device morphological engineering such as multilayered VO2, biomimetic structure,− nanothermochromism, porous, and gridded structure.− Zhou et al. recently reported a new customized VO2 composite structure in which a new factor, the incident angle, was considered in the development of smart window devices. Figure 19a shows a schematic of how different incident angles in the summer and winter can be taken advantage of with the reported customizable composite structure. The VO2 composite structure was fabricated through a 3D printing procedure, enabling the flexibility of device dimensions. Several samples of different thicknesses, widths, and spacing are demonstrated in Figure 19b. The incident-angle dependency properties of this structure are shown in Figure 19c,d in which modulating performance of the 3D printed device improves significantly, and ΔTsol improves from 9.7% to 25.8% with respect to an increase in the incident angle from 0° to 45°.

Aside from NIR transmittance regulation in smart windows, VO2 also has an unusual ability to change its long-wave infrared emissivity (εLWIR) upon crossing the MIT. An ideal smart window should have high transparency in the visible region (380–780 nm), while having a transparent state in the winter and an opaque state in the summer (Figure 20a). Moreover, the ideal smart window should have a high εLWIR at a high temperature to promote radiative cooling (RC) and a low εLWIR at a low temperature to suppress RC. Based on this concept, Long et al. fabricated a VO2-based multilayer structure which was able to regulate NIR transmittance and εLWIR spontaneously (Figure 20b). Through forming a Fabry–Perot resonator, the passive RC regulating thermochromic (RCRT) smart window possessed an εLWIR of 0.21 at 20 °C, while the εLWIR increased to 0.61 above τc. In addition, the RCRT window kept a promising Tlum of 27.8% and a ΔTsol of 9.3% (Figure 20c). With the actual building energy consumption simulation conducted with a 12-story building, the RCRT window yielded a higher energy savings compared with a commercial low-E window across different climate zones (Figure 20d). Meantime, Wu et al. designed a flexible temperature-adaptive radiative coating (TARC) through embedding lithographically patterned W-doped VO2 in a dielectric BaF2 layer on top of a reflective gold layer (Figure 20e). The TARC film had a low εLWIR (∼0.2) in the insulation state and a high εLWIR (∼0.9) at the metallic state (Figure 20f), and the observation agreed with the simulation (Figure 20g). Long et al. further expanded the concept of RC regulation from window to wall by preparing a switchable interwoven structure. As shown in Figure 20h, through pulling the block of interwoven structure, the original exposed block on the top side becomes concealed, and the underneath block becomes exposed. As a result, the structure switches its phase from phase 0 to phase 1. Taking into account different requirements of windows and walls, Long’s group designed on-demand interwoven structures. Figure 20i shows the interwoven structure for window and wall applications. As discussed in Figure 20a, a window requires high visible transmittance and dual-band regulation for NIR and LWIR ranges. An ITO/VO2/PVC combination was employed for windows. In this structure, VO2 was used to regulate NIR, while the εLWIR was regulated by alternatively exposing low-E ITO and high-E PVC. On the other hand, an ideal wall has high solar absorption and low εLWIR in the winter, and low solar absorption and high εLWIR in the summer. An ITO/black paint/PVDF-HFP combination was utilized to cater to this demand. On cold days, visible transparent ITO is exposed, and sunlight will be absorbed by black paint underneath, while on hot days, the highly solar reflective high-E PVDF-HFP is exposed to prompt RC. Hence, compared with a conventional performance index Tlum and ΔTsol, the newly proposed εLWIR needs to be included to gauge the real energy-saving performance. Moreover, VO2 is notoriously known for poor durability, which is the bottleneck for the applications in smart windows. There are two recent reports to embed VO2 in the V2O5 matrix, and such a strategy could increase the lifetime up to 33 years.,

Besides the application in building, the unique property of emissivity switching makes VO2 the material of interest for IR camouflage and passive radiator for military and aerospace applications because VO2-based devices can function entirely on the thermal trigger with no additional sources, electrical or otherwise, required. VO2-based camouflage devices work by reducing the amount of IR emitted into the environment, shrouding the user from being detected with an IR detector such as most night-vision technologies. Examples include VO2/graphene/CNT heterostructure by Xiao et al., VO2/carbon hybrid by Wang et al., and VO2/ZnS core–shell structure (Figure 21a) by Ji et al. As seen in Figure 21b, under a similar IR detector and temperature, the VO2/ZnS core–shell pallets exhibit the ability to control their IR radiation intensity and lower their detected temperature as compared to V2O5 pallets with constant emissivity.

Different from a camouflage device, a VO2-based passive radiator requires modification to the device structure to counter the lower emissivity at the higher temperature. This intrinsic problem could be overcome by depositing VO2 on a highly reflective metal substrate with, or without a spacer layer. Figure 22a is a multilayer Si/VO2/BaF2/Au structure reported by Kim et al., which was designed and tested specifically for simulated space (vacuum) applications. Figure 22b demonstrates the radiated thermal power of the multilayer device. Experimental data were compared with simulated ones for both high and low-temperature operations. The measured radiated power at 300 and 373 K was 72 W/m2 and 552 W/m2, respectively, showing a massive jump in emitted radiation upon crossing the MIT threshold. Due to the typical multilayer design of a VO2-based passive radiator, factors such as functional emissivity difference between low and high temperature as well as the wavelength of emitted light can be further fine-tuned by adjusting the substrate/spacer/film combination. The electrochromic setup has also been shown to also result in IR regulating behavior in VO2-based devices.Figure 22c is the schematic of a three-terminal thin-film-transistor-type electrochromic device by Katase et al. Upon application of external voltage (+12 V according to literature), the VO2 channel undergoes protonation, and the device becomes IR opaque, similar to smart window applications. When a reversed voltage is applied, deprotonation happens, and the device becomes IR transparent once again. Figure 22d shows optical transmittance spectra measured during this transition with +12 V stimulus. The optical transmittance modulation ratio at λ of 3000 nm was 49%. Electrical input into a VO2-based device can also be utilized as a Joule heating source for MIT. To realize this, VO2 can be combined with transparent conductive electrode materials such as ITO, Ag NWs, CNT, etc., to become an electro-optic modulator. An example of such a modulator is a VO2+Au/GaN/Al2O3 device by Fan et al., which has the ability to change the transmission step-by-step according to the applied voltage.

Optical Switches

Based on the sudden change in the optical constants n and k of VO2 across MIT, radio frequency (RF) switches or waveguides can be fabricated to control the flow of electromagnetic waves (i.e., microwave and radiowave). Even though the design of each device is largely dependent on whether it is thermally or electrically activated, the mechanism for turning from the ON to OFF state is still entirely based on the transition from VO2(M) to VO2(R) respectively. Examples of RF-switches are the thermally configurable hybrid Al nanoholes/VO2 photonic switch by Sun et al., metamaterial design by Ding et al. which can act as an absorber from 0.562 to 1.232 THz at room temperature and a high-efficiency halfwave plate at high temperature, and temperature controlled asymmetric optical switch by Liu et al.Figure 23a shows the working principal of this design in which different output of the same polarized electromagnetic wave input can be achieved at low temperature by physically reverting the device by 180° while achieving similar output at high temperature. This asymmetrical mechanism is demonstrated in Figure 23b,c. A large asymmetry exceeding 20% was detected at 23 °C, while it disappeared almost entirely at 87 °C (Figure 23b). With incident x-polarized waves, the device gave y-polarized waves output at 23 °C, and this can be considered the ON state. Upon heating to 87 °C, output waves returned to approximately x-polarized, similar to input waves, turning the switch OFF (Figure 23c). While the thermal- and electrical-activated optical switches mainly depend on the change of optical constant upon MIT, the optical-activated switches of VO2 focus on the speed of the transition as the defining factor. However, the details of the ultrafast induced phase transition of VO2 are not the focus of this review; it can be found in a summary and discussion by Wegkamp et al. VO2 ability to switch between the insulator and metal state within picoseconds is promising for the field of nanophotonics as well as all-optical integrated circuits (i.e., switches, modulators, and data-storage devices). VO2 integrated metamaterials have been reported to exhibit nonlinear transmittance by Liu et al. in the THz range as well as broadband responses spanning from the visible to mid-infrared range by Guo et al. VO2/Au nanoplate memory device was also reported by Lei et al., giving a stepwise tuning ability with the use of successive laser pulses.

Plasmonic Applications

The reversible crystal phase transition makes VO2 very unique among plasmonic materials. It undergoes a crystal phase transition from the monoclinic semiconductor state to rutile metallic state with significantly promoted conductivity and free carrier density, leading to a significant difference in its plasmonic property. Recently, VO2 nanoparticles (NPs) have been reported with thermal-responsive localized surface plasmonic resonance (LSPR) in the NIR region., Based on the colloidal lithography method, Long’s group successfully produced the hexagonally patterned VO2 NPs on quartz with controllable average diameters from ∼70 to ∼280 nm. It was observed that the LSPR position of metallic VO2 shifts to the longer wavelength on the larger NPs (∼1120 to ∼1220 nm) or under the increasing reflective index of the surrounding medium (∼1120 to ∼1360 nm) (Figure 24a). Besides, the NIR LSPR is temperature-responsive that is quenched on a low-temperature semiconductor state and can be gradually switched on from 20 to 100 °C (Figure 24b). They further investigated the LSPR-induced absorbance and scattering effects of VO2 plasmonics through a finite-difference time-domain method. On a single VO2 NP, it is revealed that both the absorbance and scattering are low at the semiconductor state (monoclinic, M), while a strong absorbance emerges in the metallic state (rutile, R) (Figure 24c). This result suggests that the LSPR in metallic VO2 is characterized as a strong absorbance enhancement and a relatively weak scattering effect. Moreover, they reported the dispersity- and strain-induced LSPR response on VO2 NPs in the polydimethylsiloxane (PDMS) elastomer matrix (Figure 24d,e). The dispersity-induced LSPR position can be attributed to the changes in average gaps among VO2 NPs in the matrix, which is consistent with the simulation result (Figure 24d), while the strain-dependent LSPR position change can be explained by the local reflective index change induced by the delamination between the NP and matrix under applied strains as being demonstrated by the finite element method (Figure 24e,f). A more recent report used a similar approach to tailor the VO2 surface plasmon by manipulating its atomic defects and establishing a universal quantitative understanding. Record high tunability is achieved for LSPR energy from 0.66 to 1.16 eV and a transition temperature range from 40 to 100 °C. The Drude model and DFT calculation reveal that the charge of cations plays a dominant role in the numbers of valence electrons to determine the free electron concentration. It is believed the investigation of VO2 LSPR is still in its early stage. The reversible crystal transition makes VO2 an intrinsic active plasmonic material, which is unique among the plasmonic field. It is expected for researchers to further understand the VO2 plasmonic and to explore its potential applications.

Electrical Applications

Not only optical constants, but the electrical conductivity of VO2 is also altered dramatically upon transitioning from insulating VO2 (M) to metallic VO2 (R). This measurable electrical response to various external stimuli makes VO2 the prime candidate for electrical applications such as sensors or transistors. The following section discusses the electrical applications of VO2 and their corresponding devices.

Sensors

A sensor is defined by its ability to measure physical input and translate these measurements into interpretable data. Based on the significant conductivity changing of VO2 across the MIT, it is possible to convert physical environmental input into a readable electrical signal. Some examples of VO2-based sensors include temperature sensors, photodetectors, flexible strain sensors, and gas sensors. Intrinsically, VO2 is not suitable for temperature sensing applications because the change in its electrical conductivity only happens at 68 °C, even though Kim et al. managed to fabricate a programmable VO2 critical temperature sensor. VO2 was deposited on an Al2O3 (1010) substrate and between two nickel (Ni) electrodes. A voltage can be applied across these electrodes to cause the τC of VO2 to decrease, causing the VO2 (M) film to go into an intermediate phase before fully transitioning into VO2 (R). During this intermediate phase, the measured current through the device was found to be linear with the change in temperature. At a voltage of 20 V, the τC is found to be ∼20 °C, enabling full range sensing capabilities from 20 to 68 °C. Another approach, which is based on the sensing ability resulting from an abrupt change of the dielectrically constant of VO2 during MIT instead of conductivity, was done by Yang et al. As mentioned in the previous section, optically stimulated applications of VO2 are promising due to the ultrafast transition mechanics, stability, as well as the broadband optical response of VO2-based devices. Hou et al. demonstrated the device stability and speed of response using a VO2 (M) nanowire on Au electrode setup (Figure 25a). It was reported that the device needs less than 1.6 s to detect IR (980 nm) and <1.0 s to recover (Figure 25b). The device was also reported to maintain responsivity for more than 500 cycles. Another design from Takeya et al. combined the photoresponsivity of VO2 film and the localized surface plasmon resonance of silver nanorods. The results indicated a correlation between the incident light transmission and resistivity within a wavelength of 400–900 nm. While the VO2 acted as a photosensitive component, the nanorod array introduced a wavelength and polarization sensitivity to the photodetector. Because the MIT of VO2 results in the change in the lattice structure and constant, it is also possible to induce MIT by causing changes to the lattice through mechanical force, which serves as the basis for VO2 flexible strain sensor application. Hu et al. showed that a VO2 strain sensor device could be fabricated by bonding one single VO2 nanobeam to a polystyrene (PS) substrate with silver paste and measuring the resistivity of the nanobeam as tensile and compressive stress is applied along the length of the nanobeam. In this study, VO2 (R) was not formed, and only the VO2 (M2) phase was formed due to the constraint of force applied (only 0.25% as compared to the required 2% at room temperature). However, the device showed remarkable potential when exhibiting a stepwise response to as small as 0.05% tensile or compressive strain.

A typical gas sensor design is demonstrated in Figure 25c. For gas sensor application, the semimetallic VO2 (B) phase is more commonly used than the insulating VO2 (M) phase to maintain sensing capability at room temperature. VO2, regardless of the phase, responds to humidity, ammonia (NH3), and nitrogen dioxide (NO2)., Compositing VO2 with carbon species such as single- or multiwalled carbon nanotubes (SWCNTs or MWCNTs) was reported by Evans et al. The setup was effective in creating a stable, responsive VO2–CNT gas sensor. Figure 25d shows excellent response and good recovery of VO2–SWCNT to different humidity levels. The resistive response was increased dramatically from 0.5 for pure VO2 (B) to 2.7 for VO2–SWCNT and 7.1 for VO2-MWCNT at 50% humidity. This p-type gas sensing response was also reported for NH3 in the same study despite the longer and lower recovery level recorded. Depending on the applications, property change across the MIT is not the only viable option to use VO2 in a functional device.

Electrical Switches, FETs, Oscillators, and Memristors

Different from sensing applications, the MIT of VO2 can be deliberately triggered with programmable duration and patterns to great advantage in electrical switching, FET, oscillator, and memory devices. Similar to optical switches in the previous section, by toggling VO2 across the MIT, it is possible to create an ON/OFF switching mechanism based on the difference in electrical resistance of VO2 (M) and VO2 (R). It has been demonstrated by Zhou et al. that a two-terminal VO2-based switching device can have ultrafast, reliable 2 orders of magnitude ON/OFF toggling ability within 2 ns. While the MIT in this report was induced by an applied current, an electrical switch activated by Joule heating was also reported in a separate study by Li et al. Mott FET is a gated FET device in which the conventional semiconductor channel is swapped with a Mott insulator, a material with the ability to switch from insulator to metal through external voltage to the gate. VO2, as a Mott insulator, is the prime material for novel Mott FET studies. An example of a typical Mott FET setup was reported by Yajima et al. in which a large current modulation can be observed at 315 K, indicating a positive-bias gate-controlled MIT near τC of VO2. Another novel Mott FET design was also fabricated by Shukla et al. with a VO2 as a source terminal. This design functioned well at room temperature (300 K) with reversible MIT triggered by the critical applied current. Taking advantage of the VO2-based Mott FET designs and combining it with a ferroelectric material, Zhang et al. fabricated a nonvolatile ferroelectric FET (FeFET) device with VO2 nanowires as the channel and Pb(Zr0.52Ti0.48)O3 (PZT) thin film as the dielectric gate (Figure 26a). The novel FeFET device was reported to achieve up to 85% resistance change under a gate voltage of 18 V (Figure 26b). Interestingly, the presence of the ferroelectric materials created a polarization effect after the applied voltage was removed, in which the channel resistance could attain up to 50%. Through this mechanism, it is possible to achieve multiple resistive states by the sweeping suitable gate voltage. To overcome the disadvantage of solid-gate oxide dielectric FET, such as current leakage, which might interfere with the MIT of VO2, ionic liquid (IL) and solid-state electrolyte gating have been the research interest for VO2 FET devices in recent years. However, the mechanism in which IL drives the MIT of VO2 is still a debate between different studies. Nanako et al. suggested that the underlying mechanism is the bulk carrier delocalization caused by the electrostatic effect. On a different train of thought, Jeong et al. attributed the transition to the field-induced creation of oxygen vacancies, rather than the purely electrostatic effect. Ji et al. and Shibuya et al., however, suggested that electrochemical protonation was the origin of the modulation of electrical property in VO2, similar to what was observed in the electrochromic setup in the previous section. An electronic oscillator is a common component in modern electronic circuitry which can produce periodic signals such as a square wave or a sine wave. Due to the periodicity of the output, it is often used to convert a direct current (DC) input into an alternate current output. The two main types of electronic oscillators are the linear (harmonic) and nonlinear (relaxation) oscillator. Because of the ability to undergo a nonlinear MIT, VO2 can be used as the basis for a nonlinear oscillator with a relaxation behavior stimulated by external electricity input. A VO2-based oscillator design by Leroy et al. and its I–V characteristic curve is shown in Figure 26c. The inset shows the schematic of the oscillator circuit in which a resistor (Rs) is connected to the VO2 device to produce a current-controlled negative differential resistance (NDR). The NDR portion happens when VO2 enters the transitive state between VO2 (M) and VO2 (R). It was reported that by controlling Rs, the VO2-based oscillator circuit can become self-sustaining, and the frequency can range from kHz to 1 MHz. Aside from the standard setup as shown in Figure 26c, studies have also been done in which two oscillators are coupled with a resistor, a capacitor, or FET in between. A memristor is a nonvolatile electronic memory device which has a programmable resistance. The resistance of a memristor is retained even after removal of the power and is dependent on the original applied voltage. It is crucial that the resistance can be reversed or reprogrammed. Thus, a two-terminal VO2 electrical device with the nonvolatile switching of resistance across the MIT can also be adapted into memristors. Bae et al. demonstrated a two-terminal memristor based on a single VO2 nanobeam. The nanobeam undergoes MIT when a bias of 3 and 5 V was applied for 0.25 s; the resistance of the device goes from an initial 1011 Ω to 109 and 108 Ω respectively. The resistance change can be reset with a zero-voltage bias. VO2 has also been utilized in other memory devices including a multistate free-standing VO2/TiO2 cantilever, resistive random-access memory (ReRAM) devices. and 3D memory array. Not included in the above discussion is the minor application of VO2 in field emitter and spintronic devices which are based on the abrupt drop in resistance across thermal- and magnetical-activated MIT, respectively. Studies on VO2/ZnO core–shell nanotetrapod thermal-activated field emitters were reported by Yin et al. in 2014. On the other hand, VO2-based spintronic devices and the behavior of the magnetoresistance of VO2 were reported in detail by Li et al., Choi et al., and Singh et al.

Mechanical Applications

The actuator is typically a component in a machine or a system which converts provided energy into mechanical motion. The concept of the actuator has been widely adapted into novel scientific research, especially in the field of microrobotics or micro-/nanoelectromechanics. VO2, which has a high theoretical work density (≈ 7 J cm–3) and fast response rate to external stimuli, is suitable for actuator applications. It offers the ability to offset disadvantageous low work density and the slow response rate of common actuator materials such as piezoelectric ceramics or polymers and CNT, respectively. In device fabrication, single crystal VO2 or composite bimorph of VO2 can be designed to respond to specific external stimuli such as light, heat, or electrical current. An example of a photodriven VO2 bimorph design was reported by Ma et al. in 2018. The VO2/CNC device was conceived by combining the carbon nanocoil (nanosprings twisted by hollow carbon nanofibers) core with a VO2 shell. When exposed to 980 nm radiation, the temperature of the spring increases unevenly, forming a temperature gradient from tip to end. This results in a transition gradient in which the tip becomes VO2(R) first and shrinks, creating the curvature. The VO2/CNC actuator delivers a large displacement-to-length ratio (∼0.4), fast response rate (9400 Hz), and long durability (>107 cycles). More recently, Shi et al. fabricated thermal-activated single-crystalline VO2 actuators (SCVAs) which were designed so that the τC of a single VO2 nanobeam is a gradient along the radial direction. When exposed to heat, one side of the fabricated W-doped VO2 nanobeam with lower τC would undergo MIT first and shrink, creating a bending as seen in Figure 27a. It was reported that this SCVAs design performed competitively with other reported VO2 bimorph actuator designs with an extremely high displacement-to-length ratio (∼1), high energy efficiency (∼0.83%), fast response rate in the order of kHz, and long durability (>107 cycles) (Figure 27b). VO2 electrothermal devices with joule heating activation for oscillator and microelectromechanical systems (MEMS) have also been fabricated with variable degrees of success. A resonator is a device that exhibits resonance at its eigenfrequency. With VO2, the eigenfrequency of a resonator can be dynamically controlled by thermally triggering the MIT. This specific frequency is positively related to Young’s modulus of VO2, which is widely different between the monoclinic phase (151 ± 2 GPa) and the rutile phase (218 ± 3 GPa). Studies have been made to determine the effects of elemental doping on the frequency modulating ability of VO2-based resonators. Rúa et al. compared the Cr-doped VO2 resonator with the undoped one and concluded that the doped sample had a higher frequency change due to a lower Young’s modulus. Other strategies to improve performance such as changing the shape from a simple cantilever have also been done by Manca et al.

Supercapacitors

VO2 (B) with its layered structure and multioxidation states is ideal as an electrode material with charge storage through insertion and fast surface Faradaic reaction. However, being in common with all metastable VO2 phases, VO2 (B) structural instability is not suitable for cyclic stability of SC application. Multiple studies have been done to combine VO2 (B) with various carbon composites to create stable electrode materials. An example of a VO2 and reduced graphene oxide (VO2 (B)/rGO) device by Liu et al. is shown in Figure 28. The schematic diagram of the all-solid-state sandwich-structured supercapacitor design is shown in Figure 28a, where symmetrical VO2 (B)/rGO and PVA/LiCl gel are used as electrodes and electrolytes, respectively. The performance of this design was reported to have a superior specific capacitance of 353 F g–1 at 1 A g–1 and a maximum power density of 7152 W kg–1 at an energy density of 3.13 Wh kg–1. By compositing VO2 (B) with rGO, 78% capacitance was retained after 10000 cycles, improving the stability of the device immensely (Figure 28b).

Magnetic Refrigeration

Magnetic refrigeration is a cooling technique that is based on the magnetocaloric effect (MCE). The MCE describes the phenomenon in which a suitable material can be heated up or cooled down when exposed to a changing magnetic field. Due to the changing magnetization when crossing the MIT, VO2 was first shown to be suitable for magnetic refrigeration application by Wu et al. in 2011 with a single crystalline nanorod fabrication technique. Although the potential was shown for VO2 in magnetic refrigeration applications, studies to further improve this are still limited.

Batteries

VO2 formed by edge-sharing VO6 octahedra with a unique bilayer structure exhibits a large lattice spacing that can accommodate Li+ (0.76 Å), Na+ (1.02 Å), K+ (1.38 Å), and Zn2+ (0.74 Å) insertion/extraction. Fan et al. designed and synthesized a binder-free VO2 cathode via biface VO2 arrays directly growing on a graphene foam (GF) network (Figure 29a). They constructed a geometric model of bilayered VO2 nanobelts through the growth direction and lattice spacings (Figure 29b). The relatively high stacking rate of the “steplike” VO6 octahedra along the [010] direction determines the preferred growth direction. As a result, the (001) facet of the VO2 nanobelt is the thinnest, and the interlayers between the (200) crystal planes provide a facile channel for Li+ and Na+ diffusion. Meanwhile, the graphene quantum dots (GQDs) coating on the VO2 surfaces can act as highly efficient surface protection to further enhance the Li+/Na+ storage. When as-prepared GF-supported GQD-anchored VO2 arrays (GVGs) are directly used as a LIBs/NIBs cathode, it exhibits two advantages: high ion diffusion sensitization and charge transport kinetics are beneficial to obtain high-rate capacities, and the homogeneous GQDs suppress VO2 dissolution which is in favor of retaining long-term cycles. The GVG electrode delivers a high specific capacity of 421 mAh g–1 at 1/3 C for Li+, which is much higher than that of the uncoated GF@VO2 electrode (391 mAh g–1). It can maintain 151 mAh g–1 even at 120 C,and 94% of the initial capacity can be retained after 1500 cycles (Figure 29c). When it is used as an NIB cathode, it exhibits a specific capacity of 306 mAh g–1 at 1/3 C and good capacity retention (Figure 29d). These results demonstrate that the GVG is an excellent electrode material for Li+/Na+ storage.

Zhang et al. first designed and synthesized a surface amorphized VO2 (B) nanorod (SA-VO2) with a crystalline core and a surface-amorphized shell heterostructure by an interfacial engineering strategy. The crystalline/amorphous heterointerface in SA-VO2 substantially narrows the bandgap, lowers the surface energy, and reduces the K+ diffusion barrier of VO2 (B) via DFT calculations. Therefore, the as-obtained SA-VO2 electrode exhibits a higher reversible capacity of 288.3 mAh g–1 (at 50 mA g–1), superior rate capacity (141.4 mAh g–1), and long-term cyclability (86% after 500 cycles at 500 mA g–1) (Figure 29e), while the VO2 only delivers a specific capacity of 147.2 mAh g–1 at 50 mA g–1 and maintains 16.5% capacity after 200 cycles at 500 mA g–1 (Figure 29f). Compared with oxygen-rich defect amorphous shell VO2, the crystalline/amorphous heterointerface SA-VO2 enhances the K+ storage capacity and enables rapid K+/electron transfer, which results in large capacity and outstanding rate capability (Figure 29g).

Mai et al. reported VO2 hollow microspheres with a high surface area and excellent structural stability via a facile and controllable ion-modulating approach. VO2 hollow microspheres deliver the best Li+ storage performance compared to six-armed microspindles and random nanowires. The highest surface area of VO2 hollow microspheres can provide efficient self-expansion, self-shrinkage buffering, and self-aggregation during lithiation/delithiation, which delivers 3 times higher capacity than that of random nanowires. In addition, they also synthesized highly homogeneous VO2 nanorods by a rapid and simple hydrothermal method for aqueous ZIBs cathode material (Figure 29h). The in situ XRD and ex-situ XPS/TEM results demonstrate that the VO2 undergoes a single-phase reaction during the discharge process, accompanying a phase transition process of VO2–Zn0.07VO2–Zn0.29VO2–Zn0.54VO2 with a unit cell volume expansion of 6.69%. On the contrary, the evolution of Zn0.54VO2–Zn0.25VO2–Zn0.09VO2–Zn0.04VO2 occurs during the Zn2+ deintercalation from the Zn0.54VO2. Meanwhile, detailed qualitative analysis verified that the VO2 unit cell expands in the a, b, and c directions sequentially during the discharge/charge processes. Satisfactorily, the VO2 nanorods deliver a high specific capacity of 325.6 mAh g–1 and excellent long cycle performance (86% after 3000 cycles), which is outstanding performance among the reported cathode materials of the aqueous ZIBs (Figure 29i,j).

HER, OER, and Water Splitting

VO2 is a well-known semiconductor material with a band gap of 0.7 eV, which is seldom considered as a candidate material as a catalyst or photocatalyst for the production of hydrogen. VO2 can be used as a photocatalyst for hydrogen evolution through phase engineering. Ajayan et al. synthesized the body-centered-cubic nanostructured VO2, which shows excellent photocatalytic activity with a hydrogen production rate up to 800 mmol m–2 h–1 from a mixture of water and ethanol under UV light at a power density of ∼27 mW cm–2. Furthermore, vanadium oxide composites have the great potential to accelerate water dissociation kinetics and reduce charge-transfer resistance., Tao et al. synthesized MoS2/VO2 hybrids by using a two-step hydrothermal method. The phase transition of VO2 exhibits a significant effect on hydrogen evolution properties of the heterostructures (an onset potential of 99 mV and a Tafel slope of 85 mV dec–1). The enhanced performance is mainly due to the faster electron transport as well as the strain effect on MoS2. Tu et al. fabricated Co3O4/VO2 heterogeneous nanosheet structures on carbon cloth (Co3O4/VO2/CC) by the combination of hydrothermal and electrodeposition methods. The Co3O4/VO2/CC composites gave good HER performance with a low overpotential of 108 mV at 10 mA cm–2 and a Tafel slope of 98 mV dec–1, which results from the abundant active sites, effective electron transport, and improved hydrogen binding energy. Najafi et al. prepared room temperature-stable metallic rutile VO2 nanosheets by the topochemical transformation of two-dimensional VSe2. By an O2 plasma pretreatment of the VSe2 nanosheets, the obtained VO2 nanosheets show a porous structure, which shows good HER and OER performances in either acidic or alkaline media. The symmetric two-electrode water splitting cell based on the porous VO2 nanosheets as both the anode and cathode delivers a current density of 10 mA cm–2 at cell voltages of 1.710 and 1.660 V in 0.5 M H2SO4 and 1 M KOH, respectively.

Structures and Synthesis

V2O5 has the highest oxygen state in vanadium–oxygen systems and is the most stable member of the series of vanadium oxides. V2O5 has multiple distinctive polymorphs, including α-V2O5 (orthorhombic), β-V2O5 (monoclinic or tetragonal), and γ-V2O5 (orthorhombic). Among them, the most common α-V2O5 is the thermodynamically stable phase (the unit cell structure belonging to the Pmnm space group with lattice parameters of a = 11.150 Å, b = 3.563 Å, and c = 4.370 Å), and the other two phases (β-V2O5 and γ-V2O5) can be transformed from the α-V2O5 phase under high pressure and high-temperature conditions. The orthorhombic structure of α-V2O5 is shown in Figure 30, in which each single layer of V2O5 consists of edge- and corner-sharing square pyramids, and the adjacent layers are bonded together along the c-axis by weak van der Waals bonds between the vanadium and oxygen of neighboring pyramids. Additionally, three different oxygens atoms, O1, O2, O3, have different coordinations depending on the position in each layer. The terminal/apical and bridging (corner-sharing) coordinated vanadyl oxygen atoms O1 and O2 have V–O bond lengths about of 1.54 and 1.77 Å, respectively; the triply coordinated O3 links three vanadium atoms via edge-sharing VO5 square pyramids, and the three corresponding V–O bond lengths are 1.88, 1.88, and 2.02 Å.

The outstanding characteristics of V2O5, such as a layered structure, a direct band gap in the visible-light region, high chemical and thermal stability, electrochemical safety, low cost, and easy preparation, make V2O5 a suitable material for electrochemical energy conversion and storage, catalysis,− solar cells, gas sensors, electrochromic devices, and optoelectronic devices. Compared with bulk V2O5, the nanostructured ones have higher surface to volume ratios, which is beneficial to improve various performances. Over the past few years, a variety of methods, such as sol–gel, hydrothermal, chemical vapor deposition, magnetron sputtering, and atomic layer deposition, have been developed to prepare V2O5 nanostructures.

The sol–gel method has been used to fabricate V2O5 thin films and nanopowders through V2O5 sols, which were prepared by ion exchange, alkoxide hydrolysis, peroxide-assisted hydrolysis, and melt-quenching. The disadvantage of using an ion exchange is the difficulty to control the vanadium concentration as it varies throughout the whole process. In addition, some foreign ions such as Na+ may remain in the gel after ion exchange. The alkoxide hydrolysis always involves some expensive raw materials, and the molten V2O5 quenching process will produce toxic gas.

Hence the synthesis of V2O5 sol by peroxide-assisted hydrolysis stands out due to its advantages of being environmental friendly, inexpensive, and requiring simple fabrication. Vanadium metal or commercial V2O5 powders are commonly used as a vanadium source that can be dissolved vigorously in a solution of hydrogen peroxide. According to Alonso et al. the dissolution of V2O5 into H2O2 produces unstable diperoxo [VO(O2)2]− and then is dissociated to the aqueous solution of [VO2]+ and [H2V10O28]4–. Etman et al. also found [H2V10O28]4– is the main species via real-time nuclear magnetic resonance. It is noted that, not related to the preparation method, the V2O5 sols are comprised of a fibrous structure, dissimilar from many inorganic sols that are typified by a random aggregate of particle structure. These fibrous structures can self-assemble into V2O5 nanofibers upon long-term aging through a coagulation mechanism. The obtained V2O5 sol was applied onto substrates via coating techniques, including spin coating, dip-coating, and spray process, and the subsequent drying and heat treatment are necessary to obtain V2O5 films.−Figure 31a highlight the evaporation, hydrolysis, and subsequent solidification procedure during the formation of V2O5 thin films by dip-coating.Figure 31b,c shows TEM diffraction patterns and corresponding FFT of V2O5 thin films formed from low concentration dilution (LCP) and high concentration dilution (HCP) using PEG-400 as an additive (HCP-PEG), respectively. V2O5 thin films formed from an LCP precursor show the formation of grains of orthorhombic V2O5 with defined grain boundaries (Figure 31b), while the HCP-PEG samples have a polycrystalline structure without uniformed grain (Figure 31c). Meanwhile, as shown in AFM images in Figure 31d–g, thin films formed from higher concentration precursors have a larger surface roughness (RS). Thermal treatment can further increase the RS due to the formation of crystallites on the thin film surface. The thin film formed from LCP-PEG precursor has the lowest RS of 0.2 nm. Liu et al. studied the substrate effect on the structure and electrical properties of nanocrystalline V2O5 thin films prepared by the sol–gel method. They found the annealed V2O5 film on the Si substrate exhibited more uniform rod-like morphology, and electrical measurements indicated the typical n-type semiconducting behavior. Senapati et al. prepared nanoscale V2O5 films having thicknesses ranging from 92 to 137 nm by spin coating V2O5 sol at different stages of aging. They reported the decrease of strain in the films with aging, and the electrical conductivity increased with aging due to the improved crystallinity of the films.

Besides thin films, the sol–gel method also provides good control over the size, morphology, doping, and chemical composition of V2O5 powders.− Li et al. reported the flower-like V2O5 powders prepared by coagulating V2O5 sol and subsequent annealing crystallization. V2O5/graphene hybrid aerogel was prepared by Wu et al. at ambient pressure through a simple one-pot sol–gel method from commercial V2O5 powder.Figure 32a illustrates the fabrication process and images of VOx nanofibers and graphene oxide sheets. First, graphene oxide (GO) aqueous solution was added to the V2O5 sol under vigorous stirring to induce hydrolysis and in situ recombination of the GO sheets and VOx oligomers. Then, a dark red VOx/GO hybrid gel was obtained after about 5 min because of the rapid formation of intermediate vanadium phases and the growth of nanofibers. After aging for 2 days, the gel gradually changed to deep green, and the V2O5 nanofibers are anchored and in situ grown on the graphene surfaces. The V2O5/graphene hybrid aerogel is so light that it can be lifted by a feather, but it is strong enough to support the weight of 100 g (Figure 32b).

The hydrothermal method has been widely used for the synthesis of a vast range of V2O5 nanostructures with a desired size and morphology, such as nanoparticles, nanowires, nanotube, nanosheets, and micro-/nanostructures., The importance of solubility of precursors, the pH value, the surfactant, as well as the hydrothermal temperature, reaction time, and solution filling factor are highlighted in many references. Li’s research group conducted extensive research on the hydrothermal treatment of V2O5 sol. They prepared V2O5 nanoparticles and ultralong nanobelts with the usage of an inorganic V2O5 sol precursor (Figure 33a,b)., The obtained single-crystalline V2O5 nanobelts have a large specific surface area, with width and thickness of 30–200 nm and length in millimeters or even longer (Figure 33b). Strong evidence suggested that the oriented attachment growth mechanism was responsible for the formation of V2O5 nanobelts. Pan et al. reported a one-step solvothermal method to form V2O5 hollow spheres without adding surfactants., As shown in Figure 33c, the SEM image of the V2O5 hollow spheres has a uniform size of around 1 μm in diameter. They investigated the time-dependent interior structural evolution by TEM and gave the possible growth mechanism of the V2O5 microspheres (Figure 33d): vanadium oxide nanoparticles are first generated by the hydrolysis of VOC2O4 and then aggregation to form solid microspheres in stage I. The solid spheres undergo the initial inside-out Ostwald ripening process and transform to the yolk–shelled structure (stage II). With extended solvothermal reaction, secondary Ostwald-ripening takes place on the preformed solid cores, resulting in the formation of a multishelled structure (stage III). Finally, completely hollow microspheres are obtained as a result of the thorough dissolution and recrystallization of the less stable interior architectures (stage IV). Hence, the interior structure of the VO2 microspheres could be effectively tailored by simply controlling the reaction duration and concentration of the precursor.

The fabrication of two-dimensional (2D) V2O5 nanosheets has been studied by Cao et al. As cathode materials for lithium-ion batteries, the resulting 2D V2O5 nanosheets (Figure 33e) exhibit remarkable electrochemical performances, including high reversible capacity, good cyclic stability, and great rate capability. 3D hierarchical vanadium oxide microstructures, including urchin-like microflowers (Figure 33f), have been successfully synthesized by Lou et al. through a solvothermal method. The morphologies of the microstructures can be easily tailored by varying the concentration of the vanadium oxalate solution, and the obtained V2O5 microflowers are highly porous with a surface area of 33.64 m2 g–1 giving high lithium storage capacity, and enhanced cycling stability and rate capability. V2O5 nanorings and microloops are rarely reported, but they are very interesting morphologies (Figure 33g). The cation-induced asymmetric strain is the main driving force in making a layered V2O5 coil into a ring structure.

CVD is widely used for depositing high-quality and high-performance solid materials. As shown in Figure 34a, CVD involves the transfer of precursor molecules, which are either liquid or gaseous to a reaction chamber by a carrier gas (step 1), and then it is followed by the reaction and/or decomposing on the surface of the substrate to produce the desired films (steps 2a, 3, and 4) or powders (step 2b), and the byproducts and unreacted precursor are transported out from the chamber at the end of the process (step 5a, 5b). There are several types of CVD systems such as atmospheric pressure CVD, aerosol-assisted chemical vapor deposition (AACVD), atomic layer CVD, plasma enhanced CVD, metal–organic CVD, and so forth. The morphology, size, crystal phase, and specific surface area of V2O5 can be affected by various parameters, namely the reaction time, substrate temperature, pressure, precursor properties, and reaction position during the CVD method. SEM images in Figure 34b display the effect of growth temperature on the morphological characteristics of V2O5 coatings. V2O5 grown at 350 and 375 °C has rod-like structures of nonuniform thickness and width, while at 400 °C pellet-like features of V2O5 are observed, and the morphology evolution could be due to the coexistence of both α-V2O5 and β-V2O5. Chun et al. prepared V2O5 nanosheets via the reaction of VCl3 vapor with oxygen in the CVD system without a vacuum system.Figure 34c shows representative optical images and their corresponding SEM images of the obtained V2O5 nanosheets and three distinguished shapes: hexagons, triangles, and truncated triangles. Wang et al. used the CVD method to control the morphologies of V2O5 by changing the reaction distance from the source position using vanadyl acetylacetonate (VO(acac)2) as the vanadium precursor. They found that VOx vapor and VO(acac)2 vapor existed simultaneously during the growth process, and the different supersaturation distributions of these two vapors led to three main growth areas. 1. V2O5 thin-films were formed at a high concentration and supersaturation of VOx in the region near the source material (Figure 34d); 2. nanowires with a length of about 10 μm and width about 200 nm were formed at a distance of 18 cm from the source due to the low vapor concentration of VOx (Figure 34e); 3. nanospheres with a diameter of about 200–500 nm were obtained when the source material was far away (about 30 cm) due to the high concentration and supersaturation of VO(acac)2 that was oxidized to V2O5 nanospheres (Figure 34f).

ALD is considered a specific type of CVD, which was first introduced in the 1960s and is currently receiving ever-growing attention as a method of choice for the growth of conformal coatings on nanostructures with high aspect ratios.Figure 35a depicts the typical ALD process, in which the precursor gases sequentially react with a surface to form an ultrathin film through a self-limiting process, and all byproducts and unreacted precursor molecules are purged out of the reactor. The primary advantages of ALD lie in subnanometer film thickness and conformality control that profit from the cyclic, self-saturating nature of ALD processes. Moreover, the ALD is unique as it is able to coat complex 3D structures with a high degree of uniformity and smoothness., Chen et al. successfully fabricated a multiwall carbon nanotube (MWCNT)/V2O5 core/shell sponge by ALD.Figure 35b shows the experimental flow schematically: the MWCNT sponge structure exhibits a very low density (∼7 mg/cm3) and high porosity (>99%), allowing for a high amount of active material loading; the V2O5 layer of about 16 nm is subsequently deposited on the MWCNT sponge by 1000 cycles of H2O-based ALD; finally, the MWCNT/V2O5 sponge is compressed and assembled in a coin cell battery, which enables de/lithiation in active material within a very short time. Figure 35c,d shows SEM images of the MWCNT sponge before and after ALD V2O5 coating, giving uniform and smooth V2O5 coating. Two different oxidants, O3 and H2O, have been studied during the ALD process, and it was found that as the ALD cycle numbers increased from 100 to 2500, the roughness for the O3-based films kept increasing from 0.7 to 10.4 nm while that for H2O-based films only increased from 0.4 to 1.9 nm (Figure 35e). Østreng et al. prepared V2O5 films by ALD using the β-diketonate VO(thd)2 and ozone as precursors. They found that the crystallographic orientation, optical properties, band gap, and surface roughness of the V2O5 films were correlated and could be varied by controlling the deposition temperature and film thickness.

PVD techniques (Figure 36a) involve evaporation and many different modes of physical sputter deposition, in which the primary source of the depositing species is a solid or liquid, as opposed to generally gaseous precursors in CVD. However, chemical reactions can and do occur in PVD systems, such as in the reactive sputtering deposition. The presence of the reactive gas (oxygen or nitrogen) in the chamber can significantly alter the PVD source. PVD possesses some unique advantages for the creation of uniform and dense solid thin films that strongly adhere to the substrates. Meanwhile, the thickness, composition, crystallinity, and crystal orientation of the thin film can be well controlled by changing the growth conditions with minimal risk of contamination due to the absence of organic reactants. Another advantage is to sequentially deposit several materials to form well-defined multilayer systems as well as special alloy compositions and structures. Magnetron sputtering is one of the most used PVD methods to fabricate a large range of materials, including metal oxides. The application of a negative voltage to the cathode will generate positively charged argon ions that can bombard the target ions to be ejected toward the substrate to form a film. Magnets are used in order to increase ion bombardment. This technique has been developed on an industrial scale to make large surface deposits with a wide variety of materials. V2O5 film consisted of fine long strip particles deposited by radio frequency magnetron sputtering, and the thickness of the V2O5 film was determined to be approximately 150 nm according to the cross-sectional SEM image (Figure 36b). V2O5 films underwent four different thermal transition behaviors to other vanadium oxides that were closely related to the oxygen proportion of the annealing ambient. Amorphous V2O5 film can be used as a hole injection layer in quantum dot light-emitting diodes, which exhibited a maximum luminance of 198.5 cd/m2, a turn-on voltage of 1.7 V, and a max external quantum efficiency of about 8.3%.

Pulsed laser deposition (PLD) is another PVD method, which has been preferred to grow different structures ranging from high-quality epitaxial thin films to various nanostructured layers. During PLD, a high-power pulsed laser beam is focused on a target of the desired composition. Material vaporized from the target is deposited as a thin film on a substrate that faces the target in an ultrahigh vacuum (UHV) environment. Polycrystalline V2O5 thin film in the desired orientation can be prepared by PLD, which has aligned nanorod morphology on a flexible stainless steel substrate (Figure 36c). Huotari et al. found the film surface morphology varied largely according to oxygen partial pressure: lower O2 partial pressures resulted in a denser and thinner film, while higher O2 partial pressures gave a film surfaces formed with randomly agglomerated nanoparticles or agglomerates with pillar-like morphology.

Electron-beam deposition (EBD) is another form of PVD where a target anode is bombarded with a high-energy electron beam that is given off from a charged tungsten filament under a high vacuum. The electron beam leads to joule heating and converts the target into the gaseous phase, which subsequently precipitates into the solid form on the desired substrate. Han et al. reported the growth of nanocolumnar V2O5 molecules that were aggregated with each other and collapsed after annealing treatment (Figure 36d). Most of the nanosized V2O5 columns’ structure could retain its original shape during the annealing process by changing the source from V2O5 to VO2. Highly oriented V2O5 thin films with nanosized grains were grown by EBD, and the film thickness was found to be in the range of 800–1200 nm that varied by adjusting the substrate position. Meanwhile, the mobility and carrier concentration of the oriented V2O5 thin films increased with the increase of V2O5 film thickness. Thermal evaporation is the vaporization of a material by heating to a temperature that the vapor pressure becomes appreciable, and the materials are sublimated from the target surface in a vacuum. By this method, heterojunctions, nanorods, nanoparticles, or highly crystalline V2O5 films have been studied in several reports. Wang et al. synthesized Ga-doped V2O5 nanorods by thermal evaporation at 850 °C and found interstitial Ga and Ga–O phases influence the photoluminescence properties of V2O5 nanorods. Berouaken et al. used thermal evaporation to prepare V2O5 nanoplatelets on the quartz crystal microbalance, followed by rapid thermal annealing. The obtained V2O5 nanoplatelets exhibited good sensing performance toward NH3 vapor at room temperature: a fast response time, a short recovery time, good stability, reproducibility, reversibility, and linearity. Velmurugan et al. prepared highly crystalline V2O5 films with a controlled thickness of about 530 nm and an average particle size of around 560 nm using a thermal evaporation process (Figure 36e). The films were fabricated in electrochemical microcapacitors and subjected to various electrochemical characterizations, which display improved reliability and excellent capacitance retention.

Batteries

V2O5 is a feasible cathode material for metal-ion storage due to its high output voltage and unique crystal structure with large interlayer spacing (4.4 Å). Mai et al. designed and synthesized a V2O5 hollow microclew (V2O5–HM) through a facile solvothermal assisted calcination method. The amorphous V2O5–HM precursor can convert into crystalline V2O5 through calcination (Figure 37a). Compared with crystalline V2O5 (V2O5-Ms) and V2O5 nanowires (V2O5–NWs), the V2O5–HMs exhibit the best Li+ storage performance (145.3 and 94.8 mAh g–1 at 0.67 and 65 C, respectively), which is due to the 3D hierarchical microstructure with intertangled nanowires (Figure 37b). This 3D hierarchical microstructure not only inherits fast electrolyte penetration, and short ionic and electronic transport pathways, but also significantly alleviates the strain during the Li+ intercalation/deintercalation. Meanwhile, compared to disordered V2O5 nanowires, a unique V2O5–HM structure effectively helps improve the tap density, which is more suitable for commercial applications (Figure 37c). V2O5 can be used as other metal ions (Na+/K+/Zn2+) electrode material except for LIB storage material. Chung group synthesized a nanosized V2O5/C composite cathode by ball milling the nanosized V2O5 with acetylene black and investigating the reaction mechanism in the NIB system. Generally, compared with other vanadium oxides (VO2, V2O3, et al.), V2O5 consists of the square-based pyramid with the highly distorted environment and exhibits the highest pre-edge intensity from the X-ray absorption near edge structure (XANES) result. Thus, the average vanadium oxidation state of the calcinated V2O5/C sample is +4.83 (Figure 37d), which may be due to the calcined carbon reduction (a slight difference with standard V2O5 in the pre-edge peak position). They used ex-situ XRD to demonstrate the major NaV2O5 with a minor Na2V2O5 phase formed at the first discharge state, accompanying a c lattice parameter increase by 9.09% and unit cell volume increase by 9.2%. At the subsequent charge, the NaV2O5 + Na2V2O5 will transform into NaV2O5 + V2O5 along with a V4+ → V5+ change (Figure 37e). The V2O5/C delivers an initial discharge capacity of 195 mAh g–1, and increases to 255 mAh g–1 at the 10th cycle corresponding to 1.7 Na+ inserts into per unit formula (Figure 37f). Besides LIB/NIB cathodes, V2O5 can be investigated as an aqueous zinc-ion battery (ZIB) cathode material, and the intercalation of water into the vanadium oxide can increase the interlayer distance, which is in favor of expanding the gallery for Zn2+ intercalation. Yang et al. reported V2O5·H2O/graphene (VOG) synthesized via a freeze-drying method. They investigated the critical role (“lubricating” effect) of structural H2O on the Zn2+ intercalation into bilayer V2O5·nH2O, and H2O-solvated Zn2+ possesses a largely reduced effective charge and improves electrochemical performance. VOG delivers a high capacity of 381 mAh g–1 at 60 mA g–1 and maintains 248 mAh g–1 at a high current density of 30 A g–1, which are much higher than those of most aqueous ZIB cathode materials (Figure 37g,h). The interlayer distances of VOG are 12.6, 13.5, and 10.4 Å at initial, discharge, and charge states by ex-situ XRD and MAS NMR (Figure 37i). These results demonstrate that water in vanadium oxide layers plays an important role in the performance of an aqueous ion battery.

To investigate the electrical conductivity and structural mechanism during lithium insertion/deinsertion of V2O5, Yoon et al. developed a 3D V2O5/rGO/CNT with short Li+ diffusion, and high continuous 3D conductive network, and investigated its structural mechanism during Li+ intercalation/deintercalation by in situ XRD/XANES analysis (Figure 38a). The 3D V2O5/rGO/CNT delivers a high discharge capacity of 100 mAh g–1 at 20 C, which is much higher than 2D V2O5/rGO (68 mAh g–1). There are numerous metastable phases of LixV2O5 during Li+ intercalation into V2O5. The α-Li0.26V2O5, ε-Li0.93V2O5, δ-Li1.27V2O5, γ-Li1.93V2O5, and ω-Li2.65V2O5 phase form in turn during the first discharge to 3.4, 3.3, 3.19, 2.28, and 2.01 V, respectively. In addition, the most reflection will return to the same position at a pristine state during the subsequent charge process, which can confirm the high structural reversibility of V2O5 in the ternary composite upon Li+ intercalation/deintercalation (Figure 38b). The 3D V2O5/rGO/CNT delivers a high discharge capacity of 100 mAh g–1 at 20 C, which is higher than 2D V2O5/rGO (68 mAh g–1) (Figure 38c). The preintercalation interlayer metal ions can act as pillars to increase the electronic conductivity, ion diffusion rate, and stability of layered vanadium oxides. Mai et al. designed and assembled Na0.33V2O5 (NVO) and V2O5 single nanowire devices to investigate the effect on the intrinsic electrical conductivity of Na+ intercalation. The conductivity of NVO is 5.9 × 104 S m–1, while the conductivity of V2O5 is 7.3 S m–1, which indicates that the electronic conductivity of V2O5 is greatly improved by the Na+ intercalation (Figure 38d). Stucky et al. fabricated a series of nanostructured Mn-doped V2O5 cathode materials and found that the larger Mn doping in the modified V2O5 structure can increase the cell volume, which facilitates high Li+ diffusion and improves the electronic conductivity (Figure 38e). The Mn0.01V1.99O5 delivers a high discharge capacity (251 mAh g–1 at 1 C) and excellent cycling stability (80% after 50 cycles), which is much higher than V2O5 (215 mAh g–1 vs. 70%) (Figure 38f,g). Fan et al. reported lightweight, freestanding V2O5 nanoarray-based positive electrodes (UGF-V2O5/PEDOT), which were prepared by growing a V2O5 nanobelt array directly on 3D ultrathin graphite foam (UGF), followed by coating the V2O5 with a mesoporous thin layer of the conducting polymer poly(3,4-ethylenedioxythiophene) (PEDOT) (Figure 38h). In addition, the PEDOT coating constructs an integrated conductive network for the V2O5, providing decreased electrode polarization, improved charge transfer kinetics, and a prolonged discharge plateau of V2O5 (Figure 38i,j), and therefore it can lead to an increased proportion of high-voltage capacity and energy density than that without PEDOT.

Supercapacitors

Among all types of vanadium oxides, V2O5 has attracted attention for the application of SCs due to its broad oxidation states, high specific capacitance, and low acquisition cost. Palani et al. fabricated RuO2 nanoparticle-decorated V2O5 nanoflakes by a solvothermal method.Figure 39a illustrates the corresponding schematic of the fabricated asymmetric cell that exhibited a high specific capacitance of 421 F g–1 at a current density of 1 A g–1 with excellent cyclic retention of 94.6% over 10000 cycles (Figure 39b). The symmetric device of V2O5||PVA-KOH||V2O5 was fabricated using thin flexible substrate by Velmurugan et al. (Figure 39c), where both annealed (A-500) and as-prepared (RT) V2O5 films were used as electrode material separately. A schematic representation of the structural image of the V2O5 is provided in Figure 39d. The CV curves in Figure 39e denote that the symmetric A-500 gives a larger area under the curve than the symmetric-RT, suggesting the improved performance of the annealed sample. The A-500 devices display a maximum energy density of 0.68 μWh cm–2, which is obviously much higher than that of the RT electrodes (0.05 μWh cm–2, Figure 39f). Moreover, the symmetric A-500 shows excellent cycle life up to 30000 cycles with a Coulombic efficiency of 99%. As shown in Figure 39g, the practical feasibility of the as-fabricated devices was demonstrated by lighting blue light-emitting diodes. Several groups reported the hybrid structure of V2O5 with carbon materials fabricated by different strategies for enhancing the SC property. For example, Sahu et al. synthesized graphene nanoribbon @V2O5 nanostrip composites to improve the conductive property of V2O5, which displays a high energy density of 42.09 Wh kg–1 and power density of 475 W kg–1. W. Sun et al. prepared a 3D monolithic aerogel composed of uniform carbon nanofibers/V2O5 core/shell nanostructures. The composite aerogel exhibits high specific capacitance (595.1 F g–1), excellent energy density (82.65 Wh kg–1), and good cycling behavior (>12000). Zhu et al. proposed a simple “liquid phase impregnation template” strategy to successfully synthesize hierarchically porous V2O5/C nanocomposites that exhibits a specific capacitance of 492.1 F g–1, as well as an energy density of 87.6 Wh kg–1. Yao et al. successfully synthesized SCs based on 3D networks hybrids of reduced graphene oxide and V2O5 nanobelts through a simple hydrothermal method. The rGO/V2O5 hybrid aerogel electrodes showed a high energy density of 249.7 W kg–1 and excellent long-term cycle stability (remaining 90.2% after 5000 cycles).

OER

V2O5 exhibited good OER performance due to the multivalent states of the V element, which can enrich active intermediates (*OH, *O, and *OOH) by regulating the valence electron structure of the V element., The OER performance could be enhanced by fabricating the composites with other materials. Lan et al. synthesized CoV2O6-V2O5/nitrogen-doped reduced graphene oxide composites (CoV2O6-V2O5/NRGO) by a one-pot hydrothermal method integrating polyoxovanadate, ethylenediamine (EN), and graphene oxide (GO) for the precursor and postcalcined process (Figure 40a). Without V2O5, the CoV2O6/NRGO delivered a relatively acceptable OER performance with an overpotential of 379 mV at a current density of 10 mA cm–2, which is comparable to that of IrO2 (337 mV). By adding V2O5, the OER performance could be enhanced with an overpotential of 239 mV at a current density of 10 mA cm–2 (Figure 40b). Furthermore, the CoV2O6–V2O5/NRGO exhibits a higher current density (47.08 mA cm–2) at an overpotential of 300 mV compared with CoV2O6/NRGO (0.45 mA cm–2) and IrO2 (3.95 mA cm–2) (Figure 40c). Meanwhile, the CoV2O6–V2O5/NRGO shows the fastest reaction kinetics with a Tafel slope of 49.7 mV dec–1, which could be due to the enhanced charge transport (Figure 40d). The CoV2O6–V2O5/NRGO gives good stability from the polarization curves, which are almost overlapping before and after 1000 cycles (Figure 40e). The theoretical calculation found that the existence of the hydrogen bond between V2O5 and intermediate HOO* of OER decreases the adsorption energy, which may be responsible for the low overpotential.

HER

In general, V2O5 is generally considered as an HER-inactive material due to the weak H* adsorption on V sites, which limits the formation of H* at the active site., Three main methods can be adopted to promote the HER performance of V2O5: defects creation, interfacial engineering, and forming composites. First of all, creating defects, especially vacancies, is regarded as an efficient way to enhance the HER performance.− Oxygen vacancy (VO) in transition metal oxides can accelerate the adsorption of the H* intermediates by activating the delocalized electrons of the metal center, which could lead to a better HER performance., Li et al. synthesized the V2O5 nanosheet arrays with VO on Co foam through a hydrothermal reaction. The highest VO concentration of 34.2% of the V2O5 nanosheet arrays on the Co foam (Co–V2O5–H) is easily obtained by controlling the pH = 1 of the NH4VO3 precursor solution. The Co–V2O5–H exhibits a low overpotential of 51 mV at a current density of 10 mA cm–2, which shows better performance compared to the V2O5 powder and low VO concentration samples (Figure 41a). Meanwhile, the catalyst shows a negligible potential drop at different current densities, indicating long-time stability (Figure 41b). The Co–V2O5–H is employed as both a cathode and anode to establish a two-electrode alkaline electrolyzer for overall water splitting, which can maintain a steady output current at a cell voltage of 1.6 V for 24 h (Figure 41c). Second, the interfacial engineering between V2O5 and transition metal also provides a route to improve the HER performance. Kim et al. directly grew the V2O5 particles on Ni foam via a one-step hydrothermal method. The as-prepared V2O5/Ni(OH)2@NF catalyst shows a low overpotential of 39 mV at a current density of 10 mA cm–2, which is comparable to Pt (35 mV @ 10 mA cm–2). The Ni(OH)2@NF samples show an overpotential of 188 mV at 10 mA cm–2, which indicates that the V2O5 plays an important role in the HER performance (Figure 41d). DFT calculation was performed to investigate the active sites of the V2O5/Ni(OH)2@NF catalyst. The exposed facet of V2O5 is the (010), (001), and (310) planes, which is confirmed by TEM (Figure 41e). The H adsorption energy of the V2O5 (001) and (310) surfaces is larger than that of V2O5 (010), which indicates that V2O5 (010) is the active surface. Furthermore, the ΔGH* values of the Ni-sites at the edges of interfaces in Ni(OH)2@Ni and V2O5@Ni and the O-site of V2O5 (corresponding to 2, 4, and 5, respectively, as shown in Figure 41f) are close to zero, which is nearly equal to that of the Pt(111)-surface. The calculation indicated the edges of the interfaces in V2O5/Ni(OH)2@NF play a significant role in the HER process. Moreover, the hierarchical V2O5@Ni3S2 hybrid nanoarray also exhibited a good overpotential of 95 mV at 10 mA cm–2, which is also due to the interfaces between V2O5 and Ni3S2. Third, V2O5 is also used as an additive with other materials to form a composite and further enhance the HER performance. By doping phosphorus and adding V2O5 into Pt/graphene, the prepared catalysts exhibit a good HER performance of the initial potential of 32 mV and a Tafel slope of 23 mV dec–1.

Photocatalysis

It is well documented that V2O5 has a typical narrow band gap (∼2.3 eV) and wide optical absorption range and is a high electron mobility semiconductor, which exhibits good photoresponsive properties by capturing visible light and is widely used in the electro-photocatalytic field, such as for hydrogen production, environmental pollutant degradation, etc.− Garcia et al. found that the morphology of V2O5 nano-/microparticles dramatically affected the hydrogen production by photocatalysis. The V2O5 with microbelts, nanoplates, and nanopillars morphology can be obtained by using sunflowers’ petals and the center of the sunflower as biodegradable templates during the synthesis. The nanopillar V2O5 delivers a maximum hydrogen generation rate in the presence of Na2SO3 as a sacrificial agent with 65.5 μmol g–1 h–1, which is higher than microbelts (26.3 μmol g–1 h–1) and nanoplates V2O5 (19.4 μmol g–1 h–1) (Figure 42a). The high hydrogen generation rate is attributed to the large surface area, high absorbance in the UV–vis range, high photocurrent, and high content of defects of the V2O5 nanopillars. By fabricating the nanocomposite with V2O5 containing 1 or 2 heterojunctions, the hydrogen generation rate will be further improved, which is due to the delayed electron–hole recombination. For example, graphitic carbon nitride nanosheets/V2O5 composites (578 μmol g–1 h–1), Na2Ti3O7/V2O5/g-C3N4 composites (11000 μmol g–1 h–1), Na2TiO3/V2O5/g-C3N4 composites (567 μmol g–1 h–1), and Nb-doped SnO2/V2O5 (1346 μmol g–1 h–1). Transition metal oxides (TiO2, ZnO, etc.) have been applied in the removal of environmental pollutants in water, which could completely decompose organic pollutants into CO2 and H2O by photocatalytic oxidation., Hollow V2O5 microspheres consisting of randomly packed platelets showed an enhanced UV light absorption compared to the commercial V2O5 powder, which led to the highest activity for degrading rhodamine B under UV light. Composites, consisting of 1D V2O5 nanorod and 2D carbon-based materials, are promising photocatalysts for environmental pollutant degradation. The V2O5 nanorods/graphene oxide and V2O5 nanorods/graphene nanocomposites showed good degradation performance of Victoria blue dye and methylene blue dye (>95% degradation within 90 min), respectively., Especially, the nanocomposites exhibited the best degradation performance under direct sunlight irradiation compared to UV and visible light. The g-C3N4 is also used to construct the heterojunctions with V2O5 for a high-performance degradation catalyst. The photoexcited electron in the conduction band of g-C3N4 shows a strong reducing ability, while the photoexcited hole on the valence band of V2O5 exhibits a strong oxidizing ability. Thus, the g-C3N4 /V2O5 heterojunctions exhibited efficient degradation performance of rhodamine B, methyl orange (MO), and methylene blue (MB) dyes under visible light (Figure 42b). Besides the graphene or g-C3N4, the V2O5 composites with other inorganic photocatalytic oxides also showed enhanced photocatalytic properties for degradation of organic pollutants, such as V2O5/BiVO4,, V2O5/CeO2, V2O5/TiO2,, and so on. The V2O5-based composite can degrade not only the dye molecules in solution, but also some small solvent molecules in the gas phase. The ternary V2O5/BiVO4/TiO2 nanocomposites exhibited a well-aligned band structure and increasing photoinduced charge carriers through the charges separation across their multiple interfaces, which resulted in good light absorption from the UV to the visible region and better photocatalytic activity for the decomposition of gaseous toluene compared to pure TiO2 and V2O5/BiVO4 under visible light irradiation (Figure 42c).

Electrochromism

Since Colton et al. did the pioneering work on the electrochromism of V2O5 in 1976–1977,, V2O5 has attracted increasing attention among transition metal oxides because it can exhibit both anodic and cathodic coloration. From 1975 to 1999, studies focused on the absorption-transmission spectra modulation in visible and infrared light, while, after 2000, V2O5 with various micro- and nanostructures were emerging, and researchers paid attention to improve some important key figures of merit to evaluate the electrochromic performance, such as the coloration efficiency, cycling life, switching time, and so on.− Recently, several reviews were published on the electrochromic application of V2O5 film, which provide more specific and detailed information on this topic.,,, For V2O5, the discoloration mechanism is explained as the result of injection/extraction of electrons and electrolyte cations and variation of valence change of vanadium ions, which is widely accepted.,

In order to improve the performance of V2O5 as an electrochromic device, many strategies have been designed. Especially, V2O5 nanostructures with small sizes and large specific surface areas are expected to facilitate the ion intercalation/deintercalation process, thereby enhancing the electrochromic properties. Panagopoulou et al. successfully prepared Mg-doped V2O5 thin films using RF sputtering, and found the 15 atom % Mg-doped films displayed optimal electrochromic properties with the fastest switching time of tc = 10/4 s (intercalation/deintercalation), the best coloration efficiency of 71.3 cm2 C1– at 560 nm, higher visible transmittance of 85%, and the highest contrast value between the coloration states of ΔT (34.4% @ 560 nm). Qi et al. fabricated flexible V2O5 nanosheets/graphene oxide films, which exhibit ultrafast coloring response time (1.6 s) and bleaching time (2 s) attributed to the reduced charge transport distances of the ultrathin nanosheet structure (4–40 nm). They also display an excellent transmittance contrast of 57.5% at 425 nm and reversible yellow/green/blue-gray multicolor changes. Tong et al. fabricated a 3D crystalline V2O5 nanorod architecture on ITO substrates by a colloidal crystal-assisted electrodeposition method. Such architecture exhibits a highly reversible Li-ion insertion/extraction process (Columbic efficiency up to 96.9%), five distinct color change, good transmittance modulation ΔT (38.48% @ 460 nm), and acceptable response times (8.8 s for coloration and 9.3 s for bleaching), making it a promising film electrode for electrochromic devices. Kim et al. prepared highly crystalline 2D V2O5 nanosheets by using single-layer V2CTx film as a sacrifice template (Figure 43a). The mean thickness and lateral size of V2CTx are 2.38 nm and 0.78 μm, respectively (Figure 43b). Figure 43c shows the SEM image of V2O5 nanosheets film after annealing of V2CTx at 350 °C. The 2D V2O5 nanosheets based electrochromic device has sharp multicolor transformations with a robust optical contrast from yellow to green to blue (Figure 43d). The corresponding optimal electrochromic performance shows a high optical contrast (53.98% @ 700 nm) and a fast response time (6.5 s for coloration and 5.0 s for bleaching).

V2O2

V2O2 is generally used as a tool to investigate the nature of metal–oxygen bonds, which is crucial to provide a proper rationalization of the relationship between structure and properties at an atomic scale. Stable V2O2 could not be synthesized by a traditional solid-state reaction or solution methods. However, it can be observed by IR spectroscopy on V, Ne, and O2 codeposited matrices. The calculation results show two possible vanadium bonding situations: 1) no bonds between the two vanadium atoms; 2) a short distance between the two vanadium atoms, which indicated multiple bonding (Figure 44a).− However, Himmel et al. found that there is a multiple vanadium–vanadium bond in V2O2 molecules. There are three V–V bonding and antibonding orbitals, which are occupied by 1.70, 1.58, 1.50 electrons and 0.42, 0.40, 0.49 electrons, respectively. Thus, 4.78 electrons are in V–V bonding orbitals, and 1.31 electrons are in V–V antibonding orbitals. Furthermore, five bond critical points (four points between the oxygen and the vanadium atoms and one point in the center between the vanadium atoms) and two ring critical points (between the center of the cluster and the oxygen atoms) can be identified in V2O2 molecules (Figure 44b), which confirms the presence of a V–V bond. The presence of the strong V–V bond may lead to unique optical and magnetic properties.

V4O9

V4O9 has an orthorhombic structure with a space group of Cmcm, which processes three types of VO polyhedra (Figure 45a). The VO5 pyramids and VO6 octahedra make pairs, which are connected by corner oxygen atoms from the VO4 tetrahedra. The valence of the vanadium ion in the octahedron and pyramid is +4, while that in the tetrahedron is +5. V4O9 is difficult to synthesize by a solid-state reaction from the mixture of binary V2O5 and V2O3 (or VO2), but it can be synthesized by the reduction of V2O5 using reducing agents of carbon, SO2, and sulfur., The amount of reducing agents dramatically affects the final products, which may consist of other vanadium oxides, such as V6O13 and VO2. Consequently, reducing V2O5 by the solvothermal method is a facile way to obtain V4O9. Different solvents (tetraethylene glycol, 2-propanol, and tetrahydrofuran) are used to synthesize V4O9 with different morphologies, such as nanoflakes, nanosheets, and so on.− Liang et al. investigated the aqueous zinc ion batteries of V4O9. It is found that the V4O9 exhibits fast zinc ion and electron diffusion, which is due to the unique tunnel structure and the V5+/V4+ mixed-valences induced metallic behavior. The V4O9 cathode shows a high reversible discharge capacity (420 mA h g–1 at 0.5 C). Even at a high current density of 50 C, it also exhibits an impressive discharge capacity of 234.4 mA h g–1, suggesting a fast Zn2+ storage ability of V4O9 (Figure 45b). Furthermore, V4O9 delivers an energy density of 175.8 W h kg–1 at a high power of 17625 W kg–1, which gives a high power density compared with other vanadium-based cathode materials in aqueous zinc ion batteries, such as V2O5, V6O13, LiV3O8, Na3V2(PO4)3, VO2, and VS2.

The 2D single layered V4O9 nanosheet assembled 3D microflowers exhibit good supercapacitor performance with a specific capacitance of 392 F g–1 at a current density of 0.5 A g–1 and 75% retained capacitance after 2000 cycles (Figure 45c). The flower-like structure assembled from ultrathin and well-separated nanosheets was unchanged during the charge–discharge cycles, which is responsible for the high capacitance and stability. Meanwhile, the V4O9 flower demonstrates a good ability to sense H2O2 and methanol with a detection limit of ∼0.1 μM and ∼60 μM, respectively (Figure 45d).

V6O13

The mixed-valence V6O13 attracts extensive attention because it can be used in a variety of ion batteries, such as Li+, Na+, Mg2+, Zn2+, and so on. As shown in Figure 46a, V6O13 is composed of alternating single and double vanadium oxide layers. There are two types of VO6 octahedra: V4+ occupied the yellow octahedra, and V5+ occupied the blue octahedra. All the octahedra are connected with corner O atoms, which form a tunnel-like structure., The solvothermal reaction is widely used to obtain V6O13 nanostructures, particles, and related composites, such as V6O13 nanogrooves, V6O13@hollow carbon microspheres, nest-like V6O13, V6O13 nanosheets, V6O13 nanowires, V6O13 nanorods, and so on (Figure 46b). Cao et al. recently developed a new strategy to synthesize V6O13 nanosheets by microwaves, which could reduce the thickness of V6O13 nanosheets greatly compared to that prepared by a hydrothermal method (Figure 46c,d).

Due to the tunnel-like structure and mixed-valence, V6O13 exhibits a metallic character at room temperature, which is beneficial for high-rate charge and discharge. When it is used as an LIB electrode material, 8 mol Li+ intercalated per formula unit endowing a high theoretical specific capacity of 417 mAh g–1 and energy density of 900 Wh kg–1. The Yu group synthesized a 3D V6O13 nanotextile with interconnected 1D nanogrooves via a facile solution-redox-based self-assembly route at room temperature (Figure 47a). They confirmed that the precursor concentration affected the mesh size in the textile structure. The 3D V6O13 delivers a high capacity of 326 mAh g–1 at 20 mA g–1 and maintains 80% capacity after 100 cycles at 500 mA g–1. The energy density can reach 780 Wh kg–1, which is much higher than those of commercialization cathodes (LiFePO4 and LiCoO2) (Figure 47b). The excellent electrochemical performance is due to the unique structure of 3D textiles, which can be maintained upon cycling and are beneficial for ion transport and cycle stability (Figure 47c). In addition, Mai et al. synthesized a novel ultrathin prelithiated V6O13 nanosheet by a secondary hydrothermal prelithiation process (Figure 47d). A single-nanosheet device was employed to in situ probe the intrinsic advantages of prelithiated nanosheets. Compared with nonlithiated V6O13 nanosheets, the ultrathin prelithiated V6O13 nanosheets exhibit a higher electrical conductivity and maintain the same conductance level after the Li+ intercalation (Figure 47e). Meanwhile, the specific capacity of the ultrathin prelithiated V6O13 nanosheets can be maintained at 98% after 150 cycles at a current density of 1000 mA g–1, which is much higher than 46% capacity of nonlithiated V6O13 nanosheets (Figure 47f). These results demonstrate that prelithiation is a strategy to obtain high-energy and long-cycling energy storage cathode materials.

Doping of various metal ions can improve the electrochemical performance and achieve a good capacity of the battery. The V6O13 with Al/Ga, Al/Fe, and Al/Na doping delivered an initial discharge specific capacity of 411.5 mA h g–1, 426.9 mA h g–1, and 514 mA h g–1 at 0.1 C, respectively. However, the electrochemical performances were poor at a high current density. V6O13 was also employed in multivalent ion (Zn2+, Mg2+) batteries.,,,, Choi et al. applied V6O13 as the ZIB cathode material and investigated its electrochemical behavior for Zn2+. In particular, they analyzed the effect of water in the electrolyte on the Zn2+ storage to investigate the physicochemical characteristics of V6O13. DFT calculation results demonstrate that the coordination environments of Zn show a big difference with/without water (Figure 47g,h). It will form octahedral coordination with water, but undercoordination without water. The Zn2+ storage of V6O13 increases with increasing water content in the electrolyte, and it can deliver a high capacity of 360 mAh g–1 and be maintained at 92% after 2000 cycles in an aqueous Zn(CF3SO3)2 (∼1 M) electrolyte (Figure 47i). Even at a high current density of 24 A g–1, it maintains a relatively high capacity of 145 mAh g–1. This work highlights that cointercalating water molecules play a vital role in enhancing the electrochemical performance of the aqueous ion storage system. Zhao et al. found that the K+ intercalated V6O13 exhibited a specific capacity of 367 mAh g–1 at 0.5 A g–1 and 198.8 mAh g–1 at 10 A g–1 (Figure 47j). Meanwhile, the capacity could be maintained at 90% after 2000 cycles at 10 A g–1. The DFT calculation suggested that K+ intercalation could significantly contribute to the reduction of the Zn2+ diffusion energy barrier (from 0.94 to 0.33 eV), which enables Zn ions to migrate away from the intercalation sites more easily (Figure 47k). Thus, K+ intercalated V6O13 showed good battery performance. Furthermore, by growing the V6O13 on carbon cloth and using the ZnSO4 as the electrolyte, the V6O13 cathode showed a capacity of 520 mAh g–1 (at a current density of 0.5 A g–1) and good cycle life (a stable capacity of 335 mAh g–1 over 1000 cycles) (Figure 47l).

Other Vanadium Oxides for Energy-Related Application

V5O12·6H2O is another layered monoclinic vanadium oxide separated by water pillars with a large interlayer spacing of 1.179 nm. Wang et al. added a small amount of platinum (Pt, 1.5 wt %) into the interlayer of V5O12·6H2O. The obtained V5O12·6H2O–Pt electrode delivers a specific capacity of 440 mAh g–1 at 500 mA g–1, and increases to 489 mAh g–1 for the third cycle, much higher than V5O12·6H2O electrodes (270 mAh g–1 at 500 mA g–1) (Figure 48a,b). They also demonstrated that the Pt additive makes no contribution, and is even counterproductive to conductivity, but facilitates a significant enhancement of pseudocapacitance. Therefore, it is clear that there is a strong relationship between Pt and the new phase Zn4SO4(OH)6·5H2O. The in situ XRD patterns show that obvious characteristic peaks shifted to lower angles during the discharge process and returned to the pristine state during charging, indicating the interlayer spacing gradual enlargement and recovery upon Zn2+ intercalation/deintercalation. Meanwhile, the formation/disappearance of the zinc hydroxyl complex is accompanied by Zn2+ insertion/extraction (Figure 48c).

Mai et al. artificially constructed a VOx cluster/reduced graphene oxide (rGO) cathode material with interfacially inserted Zn2+ repelling the pristine-bonded C atom into the plane of the rGO and constructing interfacial V–O–C bonds (Figure 49a). Meanwhile, the VOx consists of subnanoclusters (less than 1 nm per dimension) and partial nanoclusters (slightly larger than 1 nm). In combination with the electrons transferred to the rGO during the discharge process (Figure 49b–d), the reduced degree of defect is additional proof of the interfacial Zn2+ storage. As a result, they have discovered a new mechanism in which Zn2+ ions are stored mainly at the interface between VOx and rGO, which leads to anomalous valence changes compared to conventional mechanisms and exploits the storage capacity of the nonenergy storing active but highly conductive rGO. The obtained VOx-G heterostructure delivers a superior rate performance with a capacity of 174.4 mAh g–1 at an ultrahigh current density of 100 A g–1, with capacity retention of 39.4% for a 1000-fold increase in current density (Figure 49e). To the best of our knowledge, the rate performance is one of the best among ZIBs.

Besides the battery application, the nanostructured VOx materials/composites have been considered as a good catalyst for the HER due to the creation of the more active site and modification of the adsorption and desorption of H atoms. Guan et al. prepared 2D Zn-VOx-Co ultrathin nanosheets on carbon fiber paper by an electrodeposition method. The Zn-VOx-Co electrocatalysts contain the amorphous Co metal phase and crystalline Zn–Co alloy phase, which gives the materials a good HER performance (an overpotential of 46 mV at 10 mA cm–2 and a Tafel slope of 75 mV dec–1). Furthermore, Zhao et al. adopted the same method to fabricate a Ni(Cu)VOx catalyst by changing Zn and Co to Ni and Cu (Figure 50a). The Ni(Cu)VOx electrode displays a small overpotential of 21 mV at a current density of 10 mA cm–2 and a Tafel slope of 28 mV dec–1 (Figure 50b), which is comparable to the commercial 20% Pt/C catalyst (15 mV @ j = 10 mA cm–2 and 25 mV dec–1). Meanwhile, the Ni(Cu)VOx electrode also remains relatively stable for more than 100 h HER at 100 mA cm–2 (Figure 50c). The good HER performance is due to the existence of Ni–O–VOx sites, which promote the formation of highly disordered metallic Ni structures and further induce electron transfer from Ni to VOx.

The VOx based composites also exhibit good OER performance. Dong et al. synthesized 3D nanoflower-like VOx nanosheets (VOx/NiS/NF) by a hydrothermal method, which showed the OER performance with a low overpotential (330 mV at 50 mA cm–2) and a small Tafel slope (121 mV dec–1). Yang et al. synthesized a ternary Co-VOx-P catalyst with a nanoflower structure directly onto the Ni foam. The as-prepared materials demonstrated excellent catalytic performance for both the HER (an overpotential of 98 mV at 10 mA cm–2 and a Tafel slope of 59 mV dec–1) and OER (an overpotential of 230 mV at 100 mA cm–2 and a Tafel slope of 64 mV dec–1) under an alkaline environment.

Conclusions and Future Outlook

In the past two decades, there has been accelerated development of vanadium oxides due to the fact that they are the most promising candidates in versatile applications, such as batteries, energy saving smart windows, sensing, catalysts, optoelectronic devices, etc. In this review, we have discussed the V–O binary phase diagram, the structure and synthesis methods of five thermodynamically stable vanadium oxides (V2O3, V3O5, VO2, V3O7, V2O5) and some metastable vanadium oxides (V2O2, V4O9, V6O13) with selected applications on hydrogen evolution catalysis, supercapacitors, batteries, smart windows, and some other aspects. The battery, supercapacitor, and HER/OER performances of these vanadium oxides are summarized for comparison to provide an overview of the research in the field (Tables 1, 2, and 3).

While vanadium oxides are important materials for many applications, more detailed, mechanistic and systematical studies are needed to fully explore their potential as the bottleneck is increasingly related to the materials’ quality and device fabrications. We propose a few future works in this field which could be further developed in the following aspects (Figure 51):Searching the applications of vanadium oxides in underexplored areas. One example is biological thermal imaging by leveraging its thermal phase transition characteristics as the insulator-to-metal phase transition of some vanadium oxides gives sharply enhanced optical absorption above their critical temperature. It has potential applications in biochemistry, especially in the fields of bioimaging. Another example is that some nanostructured vanadium oxides also possess optimal physicochemical properties (e.g., optical, thermal, magnetic properties), which give the opportunity to be applied in intelligent medicine based on micro-/nanorobots.Development of high-level theoretical calculations to guide the rational design of vanadium oxides and their composites and to better understand the fundamentals of high-performance devices. The use of advanced computational tools and theories enables researchers to understand the origin of complex and interacting phenomena at multiple scales, which could accelerate our understanding of the fundamentals of high-performance devices and improve the operation and design of new materials systems. Additionally, machine learning is a useful toolkit for designing and exploring new vanadium oxides with desired properties. It can also aid the understanding of the complex correlations between structures and properties in vanadium oxides.Fabrication of ultrathin or two-dimensional vanadium oxides, which could be applied in some new electronic devices. Owing to the atomic-scale thickness of single layers, the 2D materials exhibit tunable electrical properties and bandgaps. Therefore, the 2D vanadium oxides may hold promise for a wide range of applications in low-power electronics, flexible electronics, optoelectronics, catalysis, batteries, and so on. Furthermore, the 2D vanadium oxides may also form novel 2D heterostructures with other ultrathin 2D nanomaterials, which would be of certain interest.Exploration of new approaches to stabilize the vanadium oxides related devices, especially thermal, light, moisture, and oxygen environmental stability. Long-term device stability is one of the most important challenges for all devices. Most vanadium oxides are sensitive to oxygen and moisture, which may lead to the degradation of the device performances. Furthermore, most stability measurements of the devices are performed under ambient conditions, which limits their application in high temperature, high humidity, and high light intensity conditions. Meanwhile, it is also important to understand the degradation mechanisms in the different types of devices based on vanadium oxides, which could be the key to improving the device stabilities.Exploring new green chemistry synthesis methods to eliminate the toxicity of vanadium oxides. Vanadium oxides cause a variety of toxic effects such as biochemical changes, neurobehavioral injury, and functional lesions in the liver and bones. Especially, vanadium oxides in breathing air can cause pulmonary problems and DNA damage in leukocytes. The toxicity is more related to the phase structure, stoichiometric ratio, concentration, particle size, and crystalline degree, which could be considered in all the processes of application of vanadium oxides. Therefore, exploring new green synthesis methods and avoiding risks to humans and the environment during vanadium oxides’ production, use, and disposal processes deserve more systematic and comprehensive studies. First, it is important to develop new synthesis protocols with minimum steps to prepare vanadium oxides, such as one-pot synthesis, completely enclosed-system synthesis, and so on. Second, the encapsulation of the devices to confine the toxicity needs to be considered when designing a new device. Third, the release of vanadium oxides to the atmosphere should be controlled during normal operation. For example, the marked decrease in toxicity is confirmed via silica coating on vanadium oxides due to the perception that the toxicity of vanadium oxide is closely related to the solubility and the robust silica barrier can isolate air and water to reduce the solubility.Realization of the bifunctional, trifunctional, or even multifunctional vanadium oxides to achieve integrated functionality. Different device integrations based on the same materials is expected to greatly reduce the cost and the incompatibility of different materials. Meanwhile, multifunctional materials would reduce the complexity of designing devices and promote the application in designated situations.

Searching the applications of vanadium oxides in underexplored areas. One example is biological thermal imaging by leveraging its thermal phase transition characteristics as the insulator-to-metal phase transition of some vanadium oxides gives sharply enhanced optical absorption above their critical temperature. It has potential applications in biochemistry, especially in the fields of bioimaging. Another example is that some nanostructured vanadium oxides also possess optimal physicochemical properties (e.g., optical, thermal, magnetic properties), which give the opportunity to be applied in intelligent medicine based on micro-/nanorobots.

Development of high-level theoretical calculations to guide the rational design of vanadium oxides and their composites and to better understand the fundamentals of high-performance devices. The use of advanced computational tools and theories enables researchers to understand the origin of complex and interacting phenomena at multiple scales, which could accelerate our understanding of the fundamentals of high-performance devices and improve the operation and design of new materials systems. Additionally, machine learning is a useful toolkit for designing and exploring new vanadium oxides with desired properties. It can also aid the understanding of the complex correlations between structures and properties in vanadium oxides.

Fabrication of ultrathin or two-dimensional vanadium oxides, which could be applied in some new electronic devices. Owing to the atomic-scale thickness of single layers, the 2D materials exhibit tunable electrical properties and bandgaps. Therefore, the 2D vanadium oxides may hold promise for a wide range of applications in low-power electronics, flexible electronics, optoelectronics, catalysis, batteries, and so on. Furthermore, the 2D vanadium oxides may also form novel 2D heterostructures with other ultrathin 2D nanomaterials, which would be of certain interest.

Exploration of new approaches to stabilize the vanadium oxides related devices, especially thermal, light, moisture, and oxygen environmental stability. Long-term device stability is one of the most important challenges for all devices. Most vanadium oxides are sensitive to oxygen and moisture, which may lead to the degradation of the device performances. Furthermore, most stability measurements of the devices are performed under ambient conditions, which limits their application in high temperature, high humidity, and high light intensity conditions. Meanwhile, it is also important to understand the degradation mechanisms in the different types of devices based on vanadium oxides, which could be the key to improving the device stabilities.

Exploring new green chemistry synthesis methods to eliminate the toxicity of vanadium oxides. Vanadium oxides cause a variety of toxic effects such as biochemical changes, neurobehavioral injury, and functional lesions in the liver and bones. Especially, vanadium oxides in breathing air can cause pulmonary problems and DNA damage in leukocytes. The toxicity is more related to the phase structure, stoichiometric ratio, concentration, particle size, and crystalline degree, which could be considered in all the processes of application of vanadium oxides. Therefore, exploring new green synthesis methods and avoiding risks to humans and the environment during vanadium oxides’ production, use, and disposal processes deserve more systematic and comprehensive studies. First, it is important to develop new synthesis protocols with minimum steps to prepare vanadium oxides, such as one-pot synthesis, completely enclosed-system synthesis, and so on. Second, the encapsulation of the devices to confine the toxicity needs to be considered when designing a new device. Third, the release of vanadium oxides to the atmosphere should be controlled during normal operation. For example, the marked decrease in toxicity is confirmed via silica coating on vanadium oxides due to the perception that the toxicity of vanadium oxide is closely related to the solubility and the robust silica barrier can isolate air and water to reduce the solubility.

Realization of the bifunctional, trifunctional, or even multifunctional vanadium oxides to achieve integrated functionality. Different device integrations based on the same materials is expected to greatly reduce the cost and the incompatibility of different materials. Meanwhile, multifunctional materials would reduce the complexity of designing devices and promote the application in designated situations.

Overall, we believe that vanadium oxides are great candidates for future applications in related fields and help solve key challenges in the global warming crisis. Moreover, the integration of vanadium oxides with multidisciplinary fields such as material science, device physics, civil engineering, mechanical design, and bioscience would continue to attract the interest of many scientists from different disciplines for new fascinating fields.

References 477

  1. 1.

    Moskalyk R. R.; Alfantazi A. M. Processing of Vanadium: a Review. Miner. Eng. 2003, 16, 793–805..

    Not in our library10.1016/s0892-6875(03)00213-9

  2. 2.

    Sutradhar M.; Da Silva J. A. L.; Pombeiro A. J. L. Chapter 1 Introduction: Vanadium, Its Compounds and Applications. In Vanadium Catalysis; The Royal Society of Chemistry, 2021; pp 1–11.

    Not in our library

  3. 3.

    Langeslay R. R.; Kaphan D. M.; Marshall C. L.; Stair P. C.; Sattelberger A. P.; Delferro M. Catalytic Applications of Vanadium: A Mechanistic Perspective. Chem. Rev. 2019, 119, 2128–2191..

    Not in our library10.1021/acs.chemrev.8b00245

  4. 4.

    Liu M.; Su B.; Tang Y.; Jiang X.; Yu A. Recent Advances in Nanostructured Vanadium Oxides and Composites for Energy Conversion. Adv. Energy Mater. 2017, 7, 1700885.

    Not in our library10.1002/aenm.201700885

  5. 5.

    Wu C.; Xie Y. Promising Vanadium Oxide and Hydroxide Nanostructures: from Energy Storage to Energy Saving. Energy Environ. Sci. 2010, 3, 1191–1206..

    Not in our library10.1039/c0ee00026d

  6. 6.

    Kianfar E. Recent Advances in Synthesis, Properties, and Applications of Vanadium Oxide Nanotube. Microchem. J. 2019, 145, 966–978..

    Not in our library10.1016/j.microc.2018.12.008

  7. 7.

    Wang Y.; Cao G. Synthesis and Enhanced Intercalation Properties of Nanostructured Vanadium Oxides. Chem. Mater. 2006, 18, 2787–2804..

    Not in our library10.1021/cm052765h

  8. 8.

    Wang Y.; Cao G. Developments in Nanostructured Cathode Materials for High-Performance Lithium-Ion Batteries. Adv. Mater. 2008, 20, 2251–2269..

    Not in our library10.1002/adma.200702242

  9. 9.

    Rehder D. The Future of/for Vanadium. Dalton Trans. 2013, 42, 11749–11761..

    Not in our library10.1039/c3dt50457c

  10. 10.

    Lamsal C.; Ravindra N. M. Vanadium Oxides: Synthesis, Properties, and Applications. In Semiconductors: Synthesis, Properties and Applications; Pech-Canul M. I., Ravindra N. M., Eds.; Springer International Publishing, 2019; pp 127–218.

    Not in our library

  11. 11.

    Cotton F. A.; Wilkinson G.; Murillo C. A.; Bochmann M. Advanced Inorganic Chemistry; John Wiley & Sons: New York, 1999.

    Not in our library

  12. 12.

    Prasadam V. P.; Bahlawane N.; Mattelaer F.; Rampelberg G.; Detavernier C.; Fang L.; Jiang Y.; Martens K.; Parkin I. P.; Papakonstantinou I. Atomic Layer Deposition of Vanadium Oxides: Process and Application Review. Mater. Today Chem. 2019, 12, 396–423..

    Not in our library10.1016/j.mtchem.2019.03.004

  13. 13.

    Wei J.; Ji H.; Guo W.; Nevidomskyy A. H.; Natelson D. Hydrogen Stabilization of Metallic Vanadium Dioxide in Single-Crystal Nanobeams. Nat. Nanotechnol. 2012, 7, 357–362..

    Not in our library10.1038/nnano.2012.70

  14. 14.

    Jang H. W.; Felker D. A.; Bark C. W.; Wang Y.; Niranjan M. K.; Nelson C. T.; Zhang Y.; Su D.; Folkman C. M.; Baek S. H.; et al. Metallic and Insulating Oxide Interfaces Controlled by Electronic Correlations. Science 2011, 331, 886–889..

    Not in our library10.1126/science.1198781

  15. 15.

    Lee J. H.; Kim J.-M.; Kim J.-H.; Jang Y.-R.; Kim J. A.; Yeon S.-H.; Lee S.-Y. Energy Storage: Toward Ultrahigh-Capacity V2O5 Lithium-Ion Battery Cathodes via One-Pot Synthetic Route from Precursors to Electrode Sheets. Adv. Mater. Interfaces 2016, 3, 1600173.

    Not in our library10.1002/admi.201600173

  16. 16.

    Cao A.-M.; Hu J.-S.; Liang H.-P.; Wan L.-J. Self-Assembled Vanadium Pentoxide (V2O5) Hollow Microspheres from Nanorods and Their Application in Lithium-Ion Batteries. Angew. Chem., Int. Ed. 2005, 44, 4391–4395..

    Not in our library10.1002/anie.200500946

  17. 17.

    Liu G.; Yu J. C.; Lu G. Q.; Cheng H.-M. Crystal Facet Engineering of Semiconductor Photocatalysts: Motivations, Advances and Unique Properties. Chem. Commun. 2011, 47, 6763–6783..

    Not in our library10.1039/c1cc10665a

  18. 18.

    Foo C. Y.; Sumboja A.; Tan D. J. H.; Wang J.; Lee P. S. Flexible and Highly Scalable V2O5-rGO Electrodes in an Organic Electrolyte for Supercapacitor Devices. Adv. Energy Mater. 2014, 4, 1400236.

    Not in our library10.1002/aenm.201400236

  19. 19.

    Wang S.; Owusu K. A.; Mai L.; Ke Y.; Zhou Y.; Hu P.; Magdassi S.; Long Y. Vanadium Dioxide for Energy Conservation and Energy Storage Applications: Synthesis and Performance Improvement. Appl. Energy 2018, 211, 200–217..

    Not in our library10.1016/j.apenergy.2017.11.039

  20. 20.

    Wu C.; Feng F.; Xie Y. Design of Vanadium Oxide Structures with Controllable Electrical Properties for Energy Applications. Chem. Soc. Rev. 2013, 42, 5157–5183..

    Not in our library10.1039/c3cs35508j

  21. 21.

    Xu X.; Xiong F.; Meng J.; Wang X.; Niu C.; An Q.; Mai L. Vanadium-Based Nanomaterials: A Promising Family for Emerging Metal-Ion Batteries. Adv. Funct. Mater. 2020, 30, 1904398.

    Not in our library10.1002/adfm.201904398

  22. 22.

    McNulty D.; Buckley D. N.; O’Dwyer C. Synthesis and Electrochemical Properties of Vanadium Oxide Materials and Structures as Li-Ion Battery Positive Electrodes. J. Power Sources 2014, 267, 831–873..

    Not in our library10.1016/j.jpowsour.2014.05.115

  23. 23.

    Armer C. F.; Yeoh J. S.; Li X.; Lowe A. Electrospun Vanadium-Based Oxides as Electrode Materials. J. Power Sources 2018, 395, 414–429..

    Not in our library10.1016/j.jpowsour.2018.05.076

  24. 24.

    Cheng Y.; Xia Y.; Chen Y.; Liu Q.; Ge T.; Xu L.; Mai L. Vanadium-Based Nanowires for Sodium-Ion Batteries. Nanotechnology 2019, 30, 192001.

    Not in our library10.1088/1361-6528/aaff82

  25. 25.

    Cheng F.; Chen J. Transition Metal Vanadium Oxides and Vanadate Materials for Lithium Batteries. J. Mater. Chem. 2011, 21, 9841–9848..

    Not in our library10.1039/c0jm04239k

  26. 26.

    Gonçalves J. M.; Ireno da Silva M.; Angnes L.; Araki K. Vanadium-Containing Electro and Photocatalysts for the Oxygen Evolution Reaction: a Review. J. Mater. Chem. A 2020, 8, 2171–2206..

    Not in our library10.1039/c9ta10857b

  27. 27.

    Ke Y.; Wang S.; Liu G.; Li M.; White T. J.; Long Y. Vanadium Dioxide: The Multistimuli Responsive Material and Its Applications. Small 2018, 14, 1802025.

    Not in our library10.1002/smll.201802025

  28. 28.

    Ke Y.; Yin Y.; Zhang Q.; Tan Y.; Hu P.; Wang S.; Tang Y.; Zhou Y.; Wen X.; Wu S.; et al. Adaptive Thermochromic Windows from Active Plasmonic Elastomers. Joule 2019, 3, 858–871..

    Not in our library10.1016/j.joule.2018.12.024

  29. 29.

    Chen D.; Li J.; Wu Q. Review of V2O5-Based Nanomaterials as Electrode for Supercapacitor. J. Nanopart. Res. 2019, 21, 201.

    Not in our library10.1007/s11051-019-4645-8

  30. 30.

    Qin H.; Liang S.; Chen L.; Li Y.; Luo Z.; Chen S. Recent Advances in Vanadium-based Nanomaterials and their Composites for Supercapacitors. Sustain. Energy Fuels 2020, 4, 4902–4933..

    Not in our library10.1039/d0se00897d

  31. 31.

    Mounasamy V.; Mani G. K.; Madanagurusamy S. Vanadium Oxide Nanostructures for Chemiresistive Gas and Vapour Sensing: a Review on State of the Art. Microchim. Acta 2020, 187, 253.

    Not in our library10.1007/s00604-020-4182-2

  32. 32.

    Morin F. J. Oxides Which Show a Metal-to-Insulator Transition at the Neel Temperature. Phys. Rev. Lett. 1959, 3, 34–36..

    Not in our library10.1103/physrevlett.3.34

  33. 33.

    Zylbersztejn A.; Mott N. F. Metal-Insulator Transition in Vanadium Dioxide. Phys. Rev. B 1975, 11, 4383–4395..

    Not in our library10.1103/physrevb.11.4383

  34. 34.

    Nadkarni G. S.; Shirodkar V. S. Experiment and Theory for Switching in Al/V2O5/Al Devices. Thin Solid Films 1983, 105, 115–129..

    Not in our library10.1016/0040-6090(83)90200-6

  35. 35.

    Kosuge K.; Takada T.; Kachi S. Phase Diagram and Magnetism of V2O3-V2O5 System. J. Phys. Soc. Jpn. 1963, 18, 318–319..

    Not in our library10.1143/jpsj.18.318

  36. 36.

    Heidemann A.; Kosuge K.; Ueda Y.; Kachi S. Hyperfine Interaction in V3O7. Phys. Status Solidi A 1977, 39, K37–K40..

    Not in our library10.1002/pssa.2210390152

  37. 37.

    Schwingenschlögl U.; Eyert V. The Vanadium Magnéli Phases VnO2n-1. Ann. Phys. 2004, 516, 475–510..

    Not in our library10.1002/andp.20045160901

  38. 38.

    Selbin J. The Chemistry of Oxovanadium(IV). Chem. Rev. 1965, 65, 153–175..

    Not in our library10.1021/cr60234a001

  39. 39.

    Chirayil T.; Zavalij P. Y.; Whittingham M. S. Hydrothermal Synthesis of Vanadium Oxides. Chem. Mater. 1998, 10, 2629–2640..

    Not in our library10.1021/cm980242m

  40. 40.

    Livage J. Hydrothermal Synthesis of Nanostructured Vanadium Oxides. Materials 2010, 3, 4175–4195..

    Not in our library10.3390/ma3084175

All 477 references

Reads

Read by people on this page

None yet.

Checked or read by AI through naulon

None yet.

A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.

Are you an author?

Is this your paper?

Claim it to add your note and be paid when agents cite it. Free.