1 Introduction

To achieve the goals of carbon peak and carbon neutrality, countries around the world need to deal with the challenge of the large and still growing base of CO2 emission [1, 2]. The world is actively releasing implementation plans to achieve peak CO2 emissions in key areas and sectors as well as a series of supporting measures to meet the goal of carbon peak and neutralization [3,4,5]. Among them, developing and utilizing more renewable energy from natural resources is considered the most effective solution. However, large-scale deployment of natural resources still has many inherent drawbacks, such as the intermittency of solar and wind [6,7,8]. Fortunately, there are many valuable technologies to overcome the shortage of clean energy from natural resources, such as rechargeable batteries, capacitors, solar cells, and fuel cells. As is known to all, the excellent performance of all the above devices is intimately and largely dependent on the remarkable properties of their materials. Thus, developing highly efficient materials to construct highly efficient devices is essential. Although these technologies harness clean energy from natural resources, large-scale clean energy utilization is still a long way to go. Therefore, more intensive investigations to explore new high-performance materials are urgently desired to construct highly efficient devices for using large-scale natural resources.

Non-stoichiometric tungsten oxides (WO3−x) have attracted increasing interest due to their unique properties [9,10,11,12,13]. It is widely known that W18O49 is an important material in the family of WO3−x, which has been extensively investigated owing to its earth abundance, highly tuneable composition, and high chemical stability. Especially, W18O49 nanowires (W18O49 NWs) are unique and versatile because they inherit all typical features of W18O49 bulk and also display a high aspect ratio, which is a benefit for photogenerated carriers to transfer along the axial direction. Meanwhile, the relatively large specific surface area and chemical stability of W18O49 NWs also make it an ideal building block for assembling various heterostructures, which allow W18O49 NWs materials applicable in smart windows [14, 15], electrochromic devices [16,17,18], photothermal therapy [19, 20], gases sensors [21,22,23], photodetector [24], photocatalysis [25,26,27], and so on. Recently, W18O49 NWs have been extensively studied in the energy conversion and storage area, including photovoltaic, fuel cells, rechargeable batteries, and capacitors [11]. Therefore, one-dimensional (1D) W18O49 NWs have been considered a substantially promising target material due to their unique anisotropic morphologies and abundant structure tuning capabilities. Since the pioneering work of preparing W18O49 NWs with diameters in tens of nanometers was reported, the research works on W18O49 NWs have achieved fast developments and brought substantial opportunities.

To date, there has been just one review that focuses on W18O49 for photocatalytic applications [28]. No review paper has focused on the applications of 1D W18O49 NWs for energy conversion and storage. In this review, we first provided a detailed discussion of the crystal structure of W18O49 and a brief introduction to the synthesis and growth mechanism of W18O49 NWs. Then the applications of W18O49 NWs in energy conversion applications, including photovoltaic, fuel cells, and fuel production, are summarized. Afterward, we reviewed the applications of W18O49 NWs in the electrochemical energy storage field, such as rechargeable batteries, capacitors, and electrochromic energy storage systems. Finally, the main future research directions on W18O49 NWs and energy-related applications were proposed. We hope this review will inspire more research and stimulate extensive future studies in this rapidly growing research field.

2 Crystal structure and synthesis of W18O49 NWs

2.1 Crystal structure of W18O49

W18O49 is a typical non-stoichiometric tungsten oxide, which can be refined in a pure form as a monoclinic structure type (P2/m) [29,30,31]. The crystal structure of W18O49 has been described in detail in previous literature [29, 30, 32], as shown in Fig. 1a, b. The crystal structure of W18O49 contains distorted WO6 octahedra, which links together by sharing corner. It is well known that the crystal structure of W18O49 has multiple valency tungsten ions and includes a considerable amount of oxygen deficiency [33,34,35]. To compensate for the oxygen deficiency in the crystal structure of W18O49, a part of edge sharing of the WO6 octahedra occurs in the existing corner sharing [33]. The possibilities of sharing modes between the WO6 octahedra building blocks are illustrated in Fig. 1c. Owing to the well-ordered corner/edge-sharing WO6 framework, the lattice network forms an open structure consisting of trigonal, quadrangular, and hexagonal tunnels [33]. As shown in Fig. 1d, it is clear that W18O49 has a large unit cell containing pore-like tunnels. These innate channels in the W18O49 structure can provide accommodation and diffusion pathways for cations [36, 37]. Specifically, W18O49 has strong anisotropic growth behavior along the [010] direction, which makes W18O49 easy to form 1D structure, such as nanofibers, nanowires [38, 39], and nanorods [40, 41]. Anisotropic growth-induced W18O49 NWs possess a larger surface-to-volume ratio and high aspect ratio than the bulk W18O49. Therefore, these unique structural characteristics of the W18O49 NWs provide advantages in several applications.

Fig. 1
figure 1

a Unit cells of W18O49. b Atomic structural model of monoclinic W18O49 showing WO6 octahedra. Reproduced with permission from Ref. [30]. Copyright 1981. Elsevier. c Possibilities of sharing modes between building blocks. d The continuous structure of W18O49 with hexagonal tunnels. Reproduced with permission from Ref. [33]. Copyright 2021, Springer

2.2 Synthesis and growth mechanism of W18O49 NWs

Substantial previous research results have confirmed that the physical and chemical properties of synthetic nanowire materials can be significantly improved or radically tailored when their sizes are reduced to the nanometer regime due to effects such as a high surface-to-volume ratio and quantum confinement. Therefore, the design and fabrication of high-quality nanowires are critical to exploring W18O49 NWs-based high-performance energy conversion and storage devices. Since the first synthesis of W18O49 NWs by infrared (IR) irradiation thermal evaporation approach, various methods have been successfully developed to synthesize W18O49 NWs, including solvothermal [42,43,44,45], thermal evaporation [46,47,48], hydrothermal [49,50,51], chemical vapor deposition [52,53,54], template-assisted [55] and microwave-assisted method [56], electron beam irradiation [57, 58], and so on. With the unremitting efforts of scientists, significant progress has been achieved in controlling the morphology, size, composition, and doping of W18O49 NWs. Nonetheless, the large-scale synthesis of W18O49 NWs with high quality using more environmentally friendly and low-cost methods remains a great challenge.

Generally speaking, different morphologies would result in different performances [59]. Therefore, it is vital to clarify the growth mechanism of W18O49 NWs for extending the applications. Although several methods have been reported for successfully preparing W18O49 NWs, relatively few articles have focused on the growth mechanism of the W18O49 NWs. This section will present the progress of the growth mechanisms of W18O49 NWs prepared by the thermal annealing and solvothermal methods. Thermal annealing methods are quick and modest for preparing W18O49 NWs. Currently, three kinds of growth mechanisms of W18O49 NWs using thermal annealing methods were reported, including WO2 intermediate layer growth mechanism, vapor–solid mechanism (V–S), and solid-state (S-S) growth mechanism. The WO2 intermediate layer growth mechanism may be divided into three steps, which are illustrated in Fig. 2a [60, 61]. First, the W particle surfaces were first oxidized into the WO2 layer. Secondly, the W–O–W chains are broken when the water vapor reacts with the surface of the WO2 layers, and then the W18O49 crystal nuclei form and grow along the [010] direction. Thirdly, as the oxidation reaction progresses, the W particles are exhausted while the long W18O49 NWs are obtained. The V–S may be divided into four steps, as summarized in Fig. 2b [62,63,64]: (1) a thin WOx layer was formed on the surface of the W film; (2) WOx·nH2O was generated on the outer WOx surface; (3) WOx·nH2O starts to decompose when the vapor pressure of volatile WOx·nH2O increases owing to the continuous oxidation of the tungsten film, and then the W18O49 crystal nuclei form and grow steadily while the WOx·nH2O decomposes; (4) as the reaction progresses, the long W18O49 nanowires are obtained. The S-S growth mechanism is illustrated in Fig. 2c [65]. First, H2O reacts with W to form WO3 and H2. Secondly, the WO3 is reduced by H2 to form W18O49. Thirdly, as the reactions (1) and (2) are repeated, long W18O49 NWs are obtained. Meanwhile, solvothermal is another widely used method for preparing W18O49 NWs. Although solvothermal methods have been commonly reported in the literature, relatively few studies have explored the mechanism. A possible growth mechanism of W18O49 NWs using the solvothermal method is shown in Fig. 2d [44]. It can be simulated as follows: (1) H2O between the H2WO4 layers is exchanged by alcohol or the intercalation of alcohol; (2) H2WO4 reacts with oleylamine to obtain tungstate oleylamine salt; (3) nucleation of W18O49 occurs and grows steadily while the tungstate oleylamine salt starts decomposing; (4) as the reaction progresses, W18O49 NWs with different diameters and lengths are obtained. Although numerous experiments have been conducted to understand the growth mechanism of W18O49 NWs during the last decade, the details of the growth mechanisms that occur during the synthesis of W18O49 NWs are still elusive. There is no doubt that more studies are needed to further clarify the underlying mechanism. Moreover, a precise and rapid growth mechanism is urgently desired to guide the large-scale synthesis of W18O49 NWs with controllable diameters and lengths.

Fig. 2
figure 2

Several typical growth mechanisms of W18O49 NWs. a WO2 intermediate layer growth mechanism, reproduced with permission from Ref. [61]. Copyright 2008. Elsevier. b V–S mechanism, reproduced with permission from Ref. [64]. Copyright 2009, American Chemical Society. c S-S growth mechanism, reproduced with permission from Ref. [65]. Copyright 2021. The Royal Society of Chemistry. d W18O49 NWs growth mechanism by solvothermal method. Reproduced with permission from Ref. [44]. Copyright 2014, Wiley-VCH

3 W18O49 NWs for energy conversion

3.1 W18O49 NWs for photovoltaic applications

Large-scale utilization of solar energy and technologies is the final solution to address the excess emissions of CO2. Photovoltaics (PV) or solar cells have been considered the most efficient way to utilize solar energy on a large scale [66,67,68]. Exploring and investigating new materials and technology is the intrinsic driving force for the gradual progress of PV [69, 70]. Over the past decades, scientists have struggled to search for efficient and low-cost materials to construct high-performance PV [71,72,73,74]. Pt is a well-known counter electrode catalyst for redox mediators in dye-sensitized solar cells (DSSCs) [75]. However, its scarcity and high cost limit its practical application [76,77,78]. Unfortunately, Pt is not always the best catalyst for all redox mediators. For instance, the catalytic effect of Pt catalysts is not very good when redox mediators are organic disulfide and alkali metal polysulfide [79,80,81]. Therefore, developing a low-cost, high-performance electrocatalyst to take the place of Pt is highly desirable in DSSCs.

1D W18O49 NWs have massive surface oxygen vacancies and high charge carrier density, which make W18O49 NWs very suitable as catalytic electrodes in DSSCs. Currently, W18O49 NWs with high surface oxygen vacancy (SOV) content were successfully synthesized and used as the catalyst for electrocatalytic reduction of redox mediators (I/I3) in DSSCs [82]. The scanning electron microscope (SEM) images of the W18O49 NWs before and after the SOV filling treatment are shown in Fig. 3a, b. It was found that there was a significant reduction in photoelectric conversion efficiencies (PCE) as the filling of SOVs time of the W18O49 NWs increased. It is shown in Fig. 3c that W18O49 NWs-based DSSCs achieved PCE of 7.8% due to the high SOVs content, which is close to that of Pt-based DSSCs. It was indicated that iodide reduction reaction activity was notably dependent on the SOVs of W18O49 NWs catalyst, which would serve as critical catalytic sites. The mechanism by which SOVs of W18O49 NWs regulate iodide reduction reactions and maintain the crystal phase and morphology of NWs was identified, as illustrated in Fig. 3d. It was confirmed that these findings would clarify the fundamental features of SOVs on metal oxides and contribute to the rational design of efficient catalysts and supports. Besides, the W18O49 NWs–reduced graphene oxide (rGO) composite was also prepared and applied in DSSCs. The first principles calculations of the adsorption energy between I3 molecule and W18O49 NWs demonstrated that W18O49 NWs had good catalytic activity. Cyclic voltammetry (CV), electrochemical impedance spectroscopy, and Tafel polarization tests also further confirmed that the W18O49–rGO presented high electrocatalytic activity for reducing I3 and low interface charge transfer impedance. Owing to the high electrocatalytic activity and low interface charge transfer impedance, the DSSCs based on W18O49–rGO achieved a PCE of 7.23%, which was comparable to the PCE of Pt-based DSSC (7.39%)[83]. Meanwhile, the mechanism on SOVs of W18O49 was also successfully verified in the cobalt complex and ferrocenium redox mediators [84]. It was revealed that the catalytic performance of W18O49 NWs for cobalt complex and ferrocenium redox mediators was also comparable to that of Pt. It was also found that the reduction reaction activity of the cobalt complex decreased slightly, whereas it increased slightly for ferrocenium after SOVs filling. In general, a clear relationship between the surface of electrocatalysts and the catalytic properties of different redox mediators will be helpful to the rational design of efficient catalysts in DSSCs. These works enriched the understanding of heterogeneous catalytic reactions on the surface of transition metal complexes for different redox mediators.

Fig. 3
figure 3

a, b SEM of the W18O49 with or without SOVs filling, c photocurrent density–voltage curves of DSSCs utilizing I/I3 as the redox mediator, d schematic mechanism of SOVs as an important catalytic site for electrocatalysis. Reproduced with permission from Ref. [82]. Copyright 2014, American Chemical Society

Moreover, W18O49 NWs exhibit excellent interfacial contact and excellent electrical conductivity caused by oxygen vacancies, which make them suitable as functional materials in organic solar cells (OSCs) and perovskite solar cells (PSCs). Specifically, W18O49 NWs film exhibit obviously improved ambient stability compared to poly(3,4-ethylenedioxythiophene)-poly (styrenesulfonate)(PEDOT:PSS) layer due to the inherently neutral and hydrophobic character of the W18O49 NWs. For instance, W18O49 NWs were successfully synthesized and used as an anode buffer layer in OSCs, which was confirmed as an alternative to conventional acidic PEDOT:PSS [85]. It was shown in Fig. 4a, b that the length of the as-prepared W18O49 was up to several micrometers and the diameter was below 30 nm. The energy level diagram of each material and the molecule structures of donors applied in fabricating OSCs are shown in Fig. 4c. When W18O49 NWs were used as the anode buffer layer in Fig. 4d–f, OSCs based on three typical polymer active layers PTB7:PC71BM, PTB7-Th:PC71BM, and PDBT-T1:PC71BM achieved higher PCEs of 8.23%, and 9.30%, and 9.09%, respectively. As a comparison, OSCs only obtained PCEs of 7.27%, 8.44% when PEDOT:PSS were used as the anode buffer layer. The results suggested that W18O49 NWs film was a promising candidate for organic solar cells’ anode buffer layer materials. In addition, W18O49 NWs were also successfully applied in fabricating the hole transport layer for all-inorganic CsPbBr3 PSCs [86]. It was revealed that W18O49 NWs with high work function promoted the hole extraction and reduced charge carrier recombination, resulting in a champion efficiency of 9.17% with a high Voc of 1.592 V for CsPbBr3 PSCs. The profound advantages, good stability, and applicability for potential ease manufacturing processes make W18O49 NWs a good candidate for the hole transport layer in OSCs and PSCs. Meanwhile, there is still a lack of relevant research in the preparation of large-area solar cells using W18O49 NWs, which would be the direction for scientists to work on in the future.

Fig. 4
figure 4

SEM images of the W18O49 with a high magnification and b low magnification. c Device architecture and energy levels of materials. J–V characteristics of OSCs based on the active layers of d PTB7:PC71BM, e PTB7-Th:PC71BM, f PDBT-T1:PC71BM. Reproduced with permission from Ref. [85]. Copyright 2017, American Chemical Society

3.2 W18O49 NWs for fuel cells

Three-dimensional structure (3D) can provide more reaction sites due to the higher spatial complexity. Thus, the 3D structural design strategy is a practical and versatile approach to improving the performance of materials. Therefore, the proper design of W18O49 NWs-based 3D electrodes could bring a promising path for fabricating novel oxygen reduction catalysts with high performance, low cost, and high CO tolerance for fuel cells. 3D structure Pt/W18O49 NWs/carbon paper electrodes were successfully designed and investigated as an oxygen reduction catalyst [87, 88]. Experimental results of the linear scan voltammogram showed that W18O49 NWs-based 3D electrodes presented higher electrocatalytic activity toward the oxygen reduction reaction than conventional Pt/C electrodes. Importantly, W18O49 NWs-based 3D electrodes also exhibited better CO tolerance. Based on the above catalyst design strategy, Pt/W18O49 NWs/carbon paper composite was prepared and successfully applied in proton exchange membrane fuel cells (PEMFCs) [87]. SEM images revealed that W18O49 NWs covered the carbon microfibers and formed a three-dimensional hierarchical electrode structure, as shown in Fig. 5a, b. Transmission electron microscope (TEM) images in Fig. 5c, d show that the W18O49 are very straight, with an average length of about 15 µm and diameters of 20–60 nm. As shown in Fig. 5e, the polarization curves revealed that PEMFCs based on Pt/W18O49 NWs/carbon paper achieved a high current density of 1.04 A·cm−2. In contrast, PEMFCs based on Pt/C electrode only obtained a low current density of 0.56 A·cm−2 under the same conditions.

Fig. 5
figure 5

SEM and TEM images of W18O49 NWs grown on carbon microfibers of carbon paper. a Low magnification SEM. b High magnification SEM image. c Low magnification TEM image of W18O49 NWs. d TEM image of W18O49 NWs. e Polarization curves and power density curves in single-cell PEMFC. Reproduced with permission from Ref. [87]. Copyright 2009, Elsevier

The combination of experimental and theoretical study is a very important approach to exploring new materials. The feasibility study of W18O49 NWs as catalyst supports for fuel cells was investigated in detail by theoretical calculation [89]. The theoretical calculation results revealed that W18O49 NWs with a 1.2 nm diameter showed potential as an oxygen reduction catalyst owing to their negative adsorption energy. Furthermore, the surface model of cobalt atom adsorbed onto the tungsten atoms on the W18O49 NWs (NWs–Co) surface revealed that it had high adsorption energy toward all molecules. Moreover, NWs–Co also could break the bond of oxygen in oxygen and hydrogen peroxide molecules to produce water as the main product. It was confirmed that the NWs–Co could oxidize carbon monoxide and methanol due to the high adsorption of NWs–Co toward these molecules. However, it can also bring mixed potential and thus reduce the performance of a single cell of direct methanol fuel cell (DMFC) when using it as a cathode catalyst. Fortunately, it was revealed that the adsorption of oxygen and hydrogen peroxide would be improved, and the adsorption of methanol and carbon monoxide would be reduced when the diameter of the W18O49 NWs was increased, or more cobalt atoms were added to the W18O49 NWs. Therefore, it is indicated that NW–Co had the potential to be used as catalyst support in the cathode for DMFC. Subsequently, the results of theoretical research were also proved successful in experiments [90, 91]. For instance, cobalt phthalocyanine-C/W18O49 NWs (CoPc-C/NWs) were prepared and successfully introduced as a non-platinum cathode catalyst for DMFCs [90]. It was revealed that the oxygen reduction reaction mechanism produced water as the main product when CoPc-C/NWs were used as catalysts, which had almost a 3.8-electron transfer number. It was demonstrated that the support of W18O49 NWs on CoPc enhanced the ORR activity. Single-cell DMFCs test showed that the power density was about 9.0 mW·cm−2 for CoPc-C/NWs. Importantly, CoPc-C/NWs also had excellent oxygen reduction activity in acidic media. Moreover, CoPc-C/NWs also did not react with methanol, which made it suitable as a feasible cathode catalyst for DMFC.

3.3 W18O49 NWs for fuel production

Direct conversion of CO2 to fuels and value-added chemicals is a promising strategy to achieve the goals of carbon peak and carbon neutrality [92,93,94]. One of the main challenges associated with CO2 reduction is exploring efficient catalytic materials for the activation of CO2 [95, 96]. 1D W18O49 NWs have massive surface oxygen vacancies and high charge carrier density, which make W18O49 NWs very suitable as catalytic materials for CO2 reduction.

W18O49 NWs with a 0.9 nm diameter were successfully synthesized and developed as an efficient photochemical catalyst for CO2 reduction [31]. Photochemical reduction CO2 experimental results showed that the average formation rate of methane was about 0.029 mmol·L−1 g−1 h−1 under irradiation with visible light. It was indicated from ultraviolet (UV)/visible (Vis) absorption spectroscopy and photoluminescence spectrum that the ultrathin W18O49 NWs consisted of a large number of oxygen vacancies, which was thought to be responsible for the high reactivity. The contrast test also confirmed that CO2 reduction activity was notably dependent on W18O49 SOVs.

Meanwhile, surface modification engineering was used to improve the photoreduction ability of W18O49 NWs. W18O49 NWs decorated with isolated Co atoms were successfully synthesized and applied as an efficient catalyst for CO2 photoreduction [97]. All experimental results and Density functional theory calculations confirmed that the energy band configuration of W18O49 NWs was greatly modified after the decoration of Co atoms, which should be fundamentally responsible for improving the redox capability of photoexcited electrons for CO2 reduction. The photocatalytic CO2 reduction performance revealed that CO and H2 generation rates were 21.18 and 6.49 mmol·g−1 h−1 for the first hour when the optimized W18O49@Co hybrid was used as the catalyst. As a comparison, the CO and H2 generation rates of W18O49 without Co were poor and nearly undetectable. Besides, element doping was employed to design the high-performance catalyst to overcome the shortcomings of fast electron–hole recombination, which would result in low efficiency in photocatalytic CO2 reduction. Currently, Cu+-doped W18O49 NWs were successfully synthesized and applied as an efficient catalyst for CO2 photoreduction [98]. The morphology, structure and the energy dispersive X-ray spectroscopy (EDX) mapping of Cu+ doped W18O49 NWs was presented in Fig. 6a–c. The SEM and TEM images showed that the W18O49 presented a wire-like morphology with 25–40 nm diameters. As shown in Fig. 6d–h, Cu, O, and W elements were uniformly distributed in each nanowire in terms of the selected area. It was revealed that the conduction edge of W18O49 changed to a more negative position, and the electron–hole recombination was also effectively inhibited after Cu+ doping. Moreover, Cu+-doped W18O49 NWs showed relatively poor hydrophilicity, which could reduce the active sites occupied by H2O molecules. All of these were beneficial for the occurrence of CO2 photoreduction. As shown in Fig. 6i, the photocatalytic CO2 reduction performance revealed that the CH4 generation rates were approximately 0.67 µmol·g−1 h−1 using W18O49 nanowires with a Cu+ doping amount of 2.5%, while generation rates of 0.16 µmol·g−1 h−1 were obtained when using pure W18O49 NWs. The best activity Cu+ doped W18O49 sample was roughly 4.0 times higher than that of pure W18O49 NWs. Therefore, transition metal doping and surface modification engineering strategies would provide an alternative for designing high-performance catalysts to reduce CO2 into fuels.

Fig. 6
figure 6

a SEM, b TEM, c high resolution transmission electron microscopy (HRTEM), d scanning transmission electron microscopy images, and eh EDX mapping of the Cu–W18O49-0.005 sample. i Photoreduction of CO2 into CH4 upon W18O49 and Cu–W18O49-x (x = 0.002, 0.005, 0.01, and 0.02) products. j Schematic illustration of the electronic band structures of W18O49 nanowires and the Cu–W18O49-0.005 Sample and illustration of photocatalytic CO2 reduction over Cu+-doped W18O49 using H2O as reducing agent. Reproduced with permission from Ref. [98]. Copyright 2020, Elsevier

Moreover, W18O49 NWs were used to assemble unique nanostructures for CO2 reduction. For instance, Au/TiO2/W18O49 NWs plasmonic heterostructure photocatalyst was successfully designed and constructed to overcome the disadvantages of poor efficiency and low selectivity for producing kinetically unfavorable hydrocarbons [99]. Owing to the interesting plasmon coupling between the Au and W18O49, the Au/TiO2/W18O49 can strongly harvest incident light to generate high-energy hot electrons. Meanwhile, the Au/TiO2/W18O49 can also adsorb intermediate products of CO and protons through the dual-heteroactive sites (Au–O–Ti and W–O–Ti) at their hetero-interface regions during the photocatalytic CO2 reduction process. The theoretical and experimental results all confirmed that the well-designed Au/TiO2/W18O49 plasmonic heterostructure could simultaneously confine high-energy hot electrons, protons, and CO during the photocatalytic CO2 reduction process, which resulted in high photocatalytic activity of 35.55 μmol·g−1 h−1 and high selectivity of 93.3% for CH4 production. Specifically, the CO generation rate was only 2.57 μmol·g−1 h−1. It was confirmed that the plasmonic active “hot spot”-confined photocatalysis strategy opened a new door for designing the new generation of plasmonic photocatalysts for high-efficiently converting CO2 into valuable fuels.

As is known to all, hydrogen is considered a promising substitute for fossil fuels owing to its high energy density and cleanliness [100,101,102,103]. W18O49 NWs and their composites were also applied as an efficient photocatalyst for hydrogen generation. For instance, plasmonic W18O49 NWs were successfully used as a photocatalyst to accelerate hydrogen generation from ammonia borane [104]. Under visible light irradiation, H2 evolution was 79.5 μmol in one hour with W18O49 NWs, compared to 14.8 μmol with commercial WO3. However, the H2 evolution efficiency of the single traditional semiconductor photocatalysts is still unsatisfactory. Therefore, the design of an ideal electrocatalyst for H2 evolution is highly required to provide high energy efficiency. Recently, a significant amount of effort has also been devoted to improving the performance of W18O49 NWs via structure engineering or electron regulating [105,106,107]. Obviously, structure control has been proven to be an effective method to improve the active area and optimize the performance of W18O49 NWs, such as novel architecture and morphologies [105, 108,109,110]. Meanwhile, heteroatom doping has emerged as an easy and mild approach to generate active sites and regulate electronic structure to enhance the conductivity or catalytic activity of W18O49. For instance, Mo element doping has been reported to efficiently optimize the electronic structure of W18O49 NWs and thus significantly boost the hydrogen evolution reaction rate [111]. The above approaches that can fundamentally optimize the properties of materials have been proven to be effective strategies for designing high-performance W18O49 NWs-based catalysts.

Meanwhile, the rational construction of heterostructure plasmonic materials has also been proven to be an effective strategy for designing high-performance W18O49 NWs catalysts with a broad absorption range, long-term stability, high charge-separation efficiency and strong redox ability for high-efficiency H2 generation. Currently, W18O49 NWs/TiO2 branched heterostructure was successfully designed and constructed by the solvothermal method and applied as a model system to investigate the kinetics process and catalytic activity for H2 generation under IR-light irradiation [112]. Figure 7a shows that the XRD patterns indicated the coexistence of characteristic diffraction peaks of monoclinic W18O49 and anatase TiO2 in the W18O49 NWs/TiO2. The SEM, TEM and EDX mapping images revealed that Ti was mainly distributed in the middle of the branched heterostructure and that W was dominant in the side region, which confirmed the structure of a TiO2 nanofibers (TiO2 NFs) backbone coated with W18O49 NWs branches, as illustrated in Fig. 7b–f. UV–Vis–IR absorption spectra in Fig. 7g suggested that the plasmonic W18O49 branches can strongly concentrate the incident light field with IR frequencies to generate energetic hot electrons surrounding the W18O49 NWs/TiO2 interface. It was also revealed that the transfer of a plasmon-induced hot electron from the W18O49 branches to the TiO2 backbones was completed within only about 200 fs, which is much faster than their relaxation process from the high-energy surface plasmon to the ground state (7–9 ps) in the W18O49 NWs. Thus, the generation and separation of a plasmonic hot electron in the W18O49/TiO2 were greatly promoted due to the ultrafast kinetics feature, which led to a remarkably enhanced catalytic activity for H2 generation compared to the activity of pure W18O49 NWs. As shown in Fig. 7h, the H2 generation rate was about 0.014 μmol·min−1 when TiO2 NFs were used as a catalyst. It is indicated that TiO2 NFs are almost inert for H2 generation. The H2 generation rate was about 0.50 μmol·min−1 when W18O49 NWs were used as a catalyst. The good catalytic activity of the W18O49 NWs is ascribed to their abundant surface-active sites and strong localized surface plasmon resonance (LSPR) absorption in the IR region. More impressively, the H2 generation rate was about 0.62 μmol·min−1 when W18O49 NWs were assembled onto TiO2 NFs. The H2 generation rate based on W18O49/TiO2 was almost about 1.24-fold compared with that of W18O49 NWs alone. It was indicated that the TiO2 NFs could act as a sensitizer to further improve the plasmon-driven catalytic activity of the W18O49 NWs. The mechanism of IR-driven transfer of plasmon-induced hot electron in a nonmetallic heterostructure for enhanced H2 generation was successfully verified, as illustrated in Fig. 7i. It is confirmed that the rational construction design is an ideal strategy for designing high-performance W18O49 NWs catalysts. Soon afterward, Yb3+/Er3+ doped NaYF4/W18O49 NWs heterostructures were also successfully synthesized and developed as an efficient photochemical catalyst for H2 generation [113]. Under 980 nm irradiation, H2 evolution was 2.11 μmol in 1 h with NaYF4:Yb–Er@W18O49 heterostructure. As a comparison, H2 evolution was negligible with NaYF4:Yb–Er NPs and H2 evolution was 0.60 μmol in 1 h with W18O49 NWs, which demonstrates the poor catalytic activity of unexcited plasmonic W18O49 NWs. As calculated, the NIR-driven catalyst of Yb3+/Er3+-doped NaYF4/W18O49 NWs heterostructure achieves a 3.5-fold increase in the catalytic H2 evolution from ammonia borane based on the process of plasmonic energy transfer, which is higher than that of pure W18O49 NWs. It was confirmed that the excited LSPR of W18O49 NWs could greatly enhance the catalytic H2 evolution, attributed to the plasmonic transfer process of “hot electrons”. This work concluded that the upconversion emission of Ln3+-doped NaYF4 nanoparticles at a suitable wavelength region can improve the LSPR excitation of the W18O49 NWs and boost the catalytic activity for H2 evolution on the surface of W18O49 NWs. Such work was considered to be the first demonstration of a high-energy hot electron transfer process in a plasmonic semiconductor heterostructure excited by low-energy IR photons, which stands out as one of the notable landmarks in expanding the photoresponse of current semiconductor photocatalysts from the UV (or Vis) to the IR light range for achieving full-spectrum-driven solar-to-fuel conversion.

Fig. 7
figure 7

a XRD patterns of a TiO2 NFs, b W18O49/TiO2, c W18O49 NWs. b, c SEM of TiO2 NFs and W18O49/TiO2. d TEM and the corresponding elemental mapping images of an individual W18O49/TiO2 branched heterostructure. e HRTEM image of a W18O49 NWs. f Selected area electron diffraction pattern of the W18O49/TiO2. g UV–Vis–IR absorption spectra of the as-fabricated samples. h Time-dependent H2 generation from NH3BH3 aqueous solution over different samples upon IR light irradiation (λ > 750 nm). i Schematic of the catalytic mechanism for H2 generation from NH3BH3 molecules over the plasmonic W18O49 NWs. Reproduced with permission from Ref. [112]. Copyright 2018, Wiley-VCH

Currently, the defect-rich structure and remarkable physicochemical properties make W18O49 NWs a potential candidate for energy conversion, as illustrated in Table 1. However, the efficiency of the W18O49 NWs-based catalysts still need to be improved. Obviously, W18O49 NWs need to couple with other semiconductor materials to construct hybrid materials to enhance the catalytic activity owing to the unsuitable band edge potential levels. Although surface modification engineering, element doping, and heterostructure assembly have been successfully applied to improve the efficiency of the W18O49 NWs-based catalysts, some heterojunctions of W18O49 NWs-based catalyst still exist and need to explore well, such as p–n junction, S-scheme, and Schottky barrier.

Table 1 Summary of W18O49 NWs-based catalyst for fuel production

4 W18O49 NWs for energy storage

4.1 W18O49 NWs for rechargeable batteries

Tungsten oxide has been regarded as a promising anode material for lithium-ion batteries due to its larger theoretical capacity and high density (693 mA·h·g−1 and 7.16 g·cm−3), compared to graphite (372 mA·h·g−1 and 2.26 g·cm−3) [114]. Generally speaking, anode materials based on conversion reaction mechanism always suffer from significant structural and volume variation during long-term cycle processes, leading to the electrodes’ poor cycling ability [115, 116]. The reduction of particle size and surface modification with the conductive material of electrode materials are effective approaches to overcome these problems. For instance, the preparation of nanowire structures has been proved to be an effective method. This section will present the recent progress of W18O49 NWs for rechargeable batteries, including anodes, cathodes, and separators.

The carbon layer on the surface of anode materials can act as a protective layer to prevent volume expansion and improve electronic conductivity. Carbon-coated ultrathin W18O49 NWs web was synthesized and applied as anode material for high-performance lithium-ion batteries [117]. The morphologies of W18O49 and W18O49@carbon nanowire are shown in Fig. 8a, b. It was clearly revealed from TEM images that the W18O49 NWs were composed of many individual thinner nanowires. The diameter of the W18O49 NWs was about 0.9 nm. W18O49/C-1 exhibit similar nanowire web morphology with W18O49 NWs, and a thin carbon layer with a thickness of about 0.6 nm was coated on the W18O49 NWs. As shown in Fig. 8c, the electrochemical test indicated that carbon-coated ultrathin W18O49 NWs web presented much better electrochemical performance than pure W18O49. It is revealed that the W18O49@carbon nanowire web electrode delivered a high lithium storage capacity of 889 mA·h·g−1 at 200 mA·g−1 after 250 cycles. It was confirmed that excellent electrochemical performance benefited from the incorporation of carbon and the unique ultrathin W18O49 nanowire web architecture. Meanwhile, ultrathin W18O49 NWs were used to improve the electrochemical performance of Si anode [118]. Schematic of the synthesis and the lithiation process of Si/W18O49 electrode is illustrated in Fig. 8d. TEM images showed that W18O49 NWs with a length of 200–400 nm and a diameter of 5–10 nm were homogeneously dispersed on the surface of Si nanoparticles, as shown in Fig. 8e. Based on the advantage of the uniform intertwists of ultrathin W18O49 NWs structure, W18O49 NWs/Si composite exhibited improved reversible capacity and rate capability compared to pure W18O49 NWs and Si. It is shown in Fig. 8f that the Si/W18O49 NWs-2 with high Si content of 62.5% in the composite presented a higher first capacity of 1379/2228 mA·h·g−1 at 400 mA·g−1, and retained a reversible capacity of 844 mA·h·g−1 over 100 cycles. It is confirmed that the uniform intertwists of ultrathin nanowires played an effective role in sustaining large lithiation/delithiation strain. This simple strategy will be helpful to design other Si-based and alloy-type anode materials with improved electrochemical performance.

Fig. 8
figure 8

a SEM and TEM images of W18O49 NWs. b SEM and TEM images of W18O49/C-1. c Cycling stability of W18O49, W18O49/C-1, W18O49/C-2 and W18O49/C-3 at a rate of 100 mA·g−1 and long-term cycling performance of W18O49/C-1 at a rate of 200 mA·g−1. Reproduced with permission from Ref. [117]. Copyright 2015. The Royal Society of Chemistry. d Schematic illustration of the synthesis and the lithiation process of the Si/W18O49 electrode. e TEM images of Si/W18O49. f Cycling performances and rate performances of W18O49 and Si/W18O49 electrodes. Reproduced with permission from Ref. [118]. Copyright 2017, Elsevier

Owing to the favorable defect structure and the novel properties of mixed valency, W18O49 was successfully used as functional material for Li–S batteries [119, 120]. Considerable efforts have been devoted to using W18O49 NWs for the improvement and optimization of electrode materials, whereas scarce studies have reported the use of W18O49 NWs to improve the electrode structure and cell configuration. Recently, the W18O49 NWs were reported as a barrier layer material to alleviate the undesirable shuttle effect for boosting the specific capacity and cycling stability of Li–S batteries [121]. The commercial sulfur cathode with 70% sulfur loading achieved an initial discharge capacity of 1142 mA·h·g−1 when W18O49 NWs/Super P was used as a barrier layer and retained the specific capacity at about 809 mA·h·g−1 after 50 cycles, while an initial specific discharge capacity of 935 mA·h·g−1 was obtained when pure Super P was used as the barrier layer. In addition, W18O49 NWs were also used as electrocatalysts to improve the electrochemical performance of graphite felt (GF) electrodes in vanadium redox flow batteries (VRFBs) [122]. It was revealed that hydrogen-treated W18O49 NWs (H–W18O49 NWs)-based VRFBs exhibit outstanding performance with 9.1 and 12.5% higher energy efficiency than the cells assembled with W18O49 NWs and treated GF, respectively, at a high current density of 80 mA·cm−2. It was confirmed that the superior performance of the H–W18O49 NWs electrocatalyst electrode could be attributed to the numerous oxygen vacancies, which can act as active sites for the VO2+/VO2+ redox reaction. Notably, the VRFBs assembled with the as-prepared hydrogen-treated W18O49 NWs-based electrode had high-capacity retention and excellent stability.

Overall, W18O49 NWs have been investigated as electrode materials for rechargeable batteries, which show good electrochemical performance. However, it is still a challenge to apply them as anodes for Li-ion batteries because of their large irreversible capacity, low rate capacity, low Coulombic efficiency, and bulky volume expansions during repeated charge intercalation/deintercalation processes. In addition, tungsten is a rare metal. It is not an element abundant on Earth, as required for energy storage applications. Therefore, it is suitable to use W18O49 NWs as a functional material for batteries, such as additives and catalysts.

4.2 W18O49 NWs for supercapacitors

Supercapacitors have attracted much attention due to their significant advantages, such as high power density and ultralong cycle life [123, 124]. Thus, supercapacitors have been considered one of the most effective solutions for the uninterrupted energy supply chain. Recently, many efforts have been devoted to achieving excellent comprehensive performance for supercapacitors, such as exploring new materials and introducing new reaction mechanisms [125,126,127]. Currently, monoclinic W18O49 was discovered as the emerging electrode material for capacitors [128, 129]. W18O49 possesses many tunnels in the crystal structure, including hexagonal tunnels, trigonal tunnels, and quadrilateral tunnels, which are helpful for ion intercalation. Moreover, the high oxygen vacancies content in W18O49 structure also can enhance the electrical conductivity by accelerating electron diffusion [34]. With a large surface-to-volume ratio in the 1D structure, massive exposed SOVs on the surface could provide abundant active sites, thus showing the great potential in applications of supercapacitors.

The W18O49 NWs/carbon felt (CF) was prepared by a facile solvothermal method and used for supercapacitor applications [130]. Owing to the three-dimensional porous nanostructure and sufficient oxygen deficiencies, the W18O49 NWs/CF exhibits a reduction in the resistance and fast reaction kinetics than WO3 NWs /CF. Therefore, W18O49 NWs/CF achieved a high capacity of 588.33 F·g−1 at 1 A·g−1. Particularly, W18O49 NWs/CF presented an excellent cycle performance that can maintain about 88% of its capacitance after 5000 cycles. W18O49 NWs/carbon cloth (CC) was synthesized by a simple solvothermal reaction and used as a positive electrode to construct flexible asymmetric supercapacitors [131]. The NWs/CC electrode showed good electrochemical performance with specific capacitance reaching up to 398 F·g−1 at 2 A·g−1. Moreover, the NWs/CC electrode presented a specific capacitance of 325 F·g−1 even at a much higher current density of 6 A·g−1. After 3000 cycles, the NWs/CC electrode can keep 92% of the initial specific capacitance. The as-fabricated flexible asymmetric supercapacitors (FASCs) based on NWs/CC obtained a high energy density of 28 Wh·kg−1 and 13 Wh·kg−1 at a power density of 745 W·kg−1 and 22.5 kW·kg−1, respectively. Specifically, the FASCs presented excellent cycle stability and achieved 81% of the specific capacitance after 10,000 cycles.

In addition to the above monovalent cation-based supercapacitors, W18O49 NWs were also successfully applied to design trivalent Al3+ ion intercalation supercapacitors [132]. As illustrated in Fig. 9a, b, a freestanding composite electrode consisting of uniformly distributed W18O49 NWs and single-walled carbon nanotubes (SCNTs) was successfully prepared and applied in Al3+ ion intercalation supercapacitors. It is shown in Fig. 9c that the freestanding SCNTs/W18O49 NWs-composite film electrode exhibits extremely high areal capacitances of 216 F·g−1 at 2 mA·cm−2 due to the highly efficient Al3+ ion intercalation into W18O49 NWs. To further evaluate the practical potential of composite film for supercapacitors, an Al-ion-based, flexible, asymmetric electrochemical capacitor was constructed. The capacitor presented a high volumetric energy density of 19.0 mWh·cm−3 at a high power density of 295 mW·cm−3. Impressively, the asymmetric supercapacitor exhibits excellent cycling stability, which retains 95.9% of its initial capacitance after 6000 cycles. Moreover, the as-prepared supercapacitor was successfully applied to power a flexible poly(3-hexylthiophene)-based electrochromic device (P-ECD), as shown in Fig. 9d, e. The results revealed that the P-ECDs could be switched using a fully charged supercapacitor at a high charge rate of 22 mA·cm−2, which present a promising capability as a power source for flexible electronic devices. Besides, W18O49 NWs–rGO was also used to fabricate an Al3+ ion asymmetric supercapacitor [133]. The as-fabricated supercapacitor delivered a high specific capacitance of 365.5 F·g−1 at 1 A·g−1 and exhibited excellent cycling stability with 96.7% capacitance retention at 12,000 cycles. More impressively, it demonstrates a high energy density of 28.5 Wh·kg−1 and a high power density of 751 W·kg−1.

Fig. 9
figure 9

Field emission SEM image and EDX elemental maps of the SCNTs)/W18O49 NWs-composite films, a surface and b cross section. c Areal capacitances of the composite electrodes under different current densities. d Schematic illustration of a wearable P-ECD integrated with a flexible Al3+-based supercapacitor as the power source. e Corresponding photographs of the wearable P-ECD at the colored and bleached states, respectively. Reproduced with permission from Ref. [132]. Copyright 2017, Wiley-VCH

4.3 W18O49 NWs for electrochromic energy storage systems

Electrochromic energy storage system is a new type of energy conversion and storage system [134, 135]. The system can demonstrate color alteration and the energy can also be stored directly inside the system. In general, an electrochromic energy storage system consists of two electrodes and an electrolyte. In recent years, two types of electrochromic energy storage systems have been reported, such as electrochromic batteries and electrochromic capacitors [136,137,138]. Currently, various materials have been developed to fabricate high-performance electrochromic energy storage systems, such as inorganic and conducting polymers [139,140,141,142]. To demonstrate color alteration and store energy directly inside the system, the selection of materials is highly important for fabricating a high-performance electrochromic energy storage system [143, 144].

W18O49 NWs were successfully applied to design smart supercapacitor, which possesses the capacity to sense changes in the level of stored energy by changing color with variations [145]. As illustrated in Fig. 10a, W18O49 NWs and polyaniline (PANI) were successfully employed as components to fabricate electrochromic supercapacitors. The W18O49 NWs exhibited high areal capacitances of 440, 360, 333, 315, and 302 F·g−1 at current densities of 2, 4, 6, 8, and 10 A·g−1, respectively. It was also indicated from the near-linear charge and discharge slopes that the hybrid electrochromic supercapacitor electrode presented excellent reversibility during the charge–discharge processes. It was revealed that the hybrid supercapacitor obtained an areal capacitance of 10 mF·cm−2. More impressively, it is illustrated in Fig. 10b that the smart supercapacitor can work in a widened window of 1.3 V while displaying variations in color schemes depending on the level of energy storage, which was distinguished from the conventional capacitor.

Fig. 10
figure 10

a Schematic of the process of fabricating a supercapacitor electrode composed of W18O49 NWs and a PANI layer. b Images of the supercapacitor electrode at several typical states demonstrating the stored energy conveyed through pattern color scheme recognition. Reproduced with permission from Ref. [145]. Copyright 2014, American Chemical Society. c Digital photo of W18O49 NWs/rGO and PANI-based electrochromic supercapacitor at different voltages. d Two electrochromic supercapacitors were integrated to power an LED. Reproduced with permission from Ref. [147]. Copyright 2021, Springer

In addition, W18O49 NWs-based hybrid film electrode was successfully prepared to fabricate electrochromic supercapacitor, which can recognize the level of stored energy with reversible color change [146]. The W18O49 NWs/rGO hybrid film electrode presented high conductivity, tunable resistances (23–39 Ω·sq−1), and transmittance (72–84%). Specifically, the W18O49 NWs/rGO electrode can maintain its functionalities after 4000 bending cycles. CV test indicated W18O49 NWs-based electrode clearly presented pseudocapacitive behavior, which would demonstrate considerable high areal capacitance. It was revealed that the as-fabricated film electrode achieved a high areal capacitance of 92 mF·cm−2 at current densities of 2 mA·cm−2. Meanwhile, the W18O49 NWs/rGO film electrode also presented excellent electrochromic performance, with fast switching speeds (8 s for coloration and 8.45 s for bleaching), and high coloration efficiency (46 cm2·C−1), and remarkable stability (96.41% of the original optical modulation). A symmetric electrochromic supercapacitor based on W18O49 NWs/rGO film electrode delivered a maximum areal capacitance of 48 mF·cm−2 and an energy density of 5.2 mWh·cm−2 with 0.391 mW·cm−2 power density. In particular, the device exhibited excellent mechanical flexibility and stability over 4000 cycles during the charge/discharge process. More impressively, the dual-functional supercapacitor exhibited a rapid and reversible response and high sustainability in optical modulation (96%) even under high current charge/discharge conditions, which is promising for real applications. Moreover, W18O49 NWs/rGO and PANI electrodes were also used to assemble an electrochromic supercapacitor [147]. W18O49 NWs/rGO film electrode exhibited a specific capacitance of 15.06 mF·cm−2 at current densities of 0.2 mA·cm−2. The PANI electrode presented a high specific capacitance of 62.26 mF·cm−2 at a discharge current of 0.2 mA·cm−2. As illustrated in Fig. 10c, d, the assembled device exhibited various color changes at the voltage of − 0.2 to 1.4 V, such as light green, dark green, light blue, and dark blue. The coloring efficiency reached 76.37 cm−2 C−1. It was believed that these creative supercapacitors equipped with smart functionalities could effectively meet the needs of humans in providing connectivity and delivering relevant information.

Obviously, W18O49 NWs have been successfully applied as electrode for supercapacitors and electrochromic supercapacitors, as illustrated in Table 2. However, the specific capacitance and cycle life of the W18O49 NWs-based supercapacitors still need to be improved. Although the method  of  using W18O49 NWs to couple with conductive materials to construct hybrid materials has been successfully applied to improve the capacitance of the W18O49 NWs-based electrode, other existing methods, such as surface modification engineering, element doping, and heterostructure assembly, still need to be well explored.

Table 2 Summary of W18O49 NWs-based materials for supercapacitors

5 Conclusion and perspectives

In conclusion, W18O49 NWs have demonstrated their enormous potential in the development of highly efficient energy conversion and storage systems due to their unique physicochemical properties and excellent electrochemical performances. Although considerable achievements are made, more efforts are still needed to further promote the fundamental understanding and practical applications of W18O49 NWs. Moreover, the synthesis of large-scale W18O49 NWs with high quality using more environmentally friendly and low-cost methods is still challenging, and there is undoubtedly plenty of room for improvement. Meanwhile, there are practical challenges that must be overcome to meet the energy needs of the next-generation power systems, including long-term cycling stability, superior energy and power capabilities, high rate performance, low price, and environmental friendliness. Many important concerns have to be considered, and further improvements must be made in the expansion of future W18O49 NWs-based materials for energy storage applications. Some prospects for further research of W18O49 NWs in energy conversion and storage are presented.

Firstly, future research should focus on ultralong W18O49 NWs. Ultralong W18O49 NWs with lengths larger than 100 μm and diameters smaller than 10 nm exhibit many superior properties to those of short-length W18O49 NWs. So developing new, low-cost, and environmentally friendly synthetic methods which are suitable for mass production of ultralong W18O49 NWs is necessary. The growth mechanisms of W18O49 NWs need to be further investigated; especially, the details of the growth mechanisms that occur during the synthesis of ultralong W18O49 NWs are still elusive.

Secondly, the low energy density of W18O49 NWs-based electrodes cannot meet the requirements of portable electronics. Therefore, further investigations should be aimed at improving their energy density. Besides, although numerous experiments have been conducted to understand the catalytic mechanism of W18O49 NWs during the last decade, a universally accepted catalytic mechanism is desired to guide the design and applications of W18O49 NWs.

Thirdly, W18O49 NWs can be explored for promising multifunctional devices. The successful experiences of W18O49 NWs applied in rechargeable batteries, capacitors, and electrochromic energy storage systems could be used for reference for fabricating new energy-related devices, such as photo-assisted rechargeable metal batteries and solar rechargeable batteries [148, 149].

Finally, large-scale production of W18O49 NWs needs to be studied for practical applications. The development of W18O49 NWs is just in the middle stage. The yield of W18O49 NWs for scalable fabrication is still insufficient, and producing W18O49 NWs with various structures is still costly. It is believed that the application of W18O49 NWs would be further extended from a simple laboratory process to device-oriented design and control. Therefore, environmentally friendly and low-cost preparative techniques for large-scale W18O49 NWs with high quality are urgently needed.