Large-scale production of tungsten trioxide nanoparticles for electrochromic application

Abstract

Large-scale, high yield production of tungsten trioxide (WO₃) particles was successfully realized by using a facile, high-temperature, catalyst-free, solid evaporation route with ammonium paratungstate tetrahydrate as raw material. The crystalline structure, size, and morphology of the as-synthesized WO₃ particles were systematically investigated by combined techniques of X-ray diffraction and electron microscopy, as a function of reaction temperature, deposition temperature, carrier gas flow rate, and size of the quartz tube. With low carrier gas flow rate (1–2 L min⁻¹), the WO₃ products collected from different deposition regions were found to exhibit a diversity of forms, such as thick rods, irregular, polyhedral, and octahedral particles, depending on the reaction temperature. At a reaction temperature of 1350 °C, increasing the carrier gas flow rate led to the formation of semi-spherical or quasi-spherical WO₃ particles with decreased size and improved size distribution. A reaction temperature of 1350 °C and an Ar flow rate of 6 L min⁻¹ yielded optimized quasi-spherical WO₃ nanoparticles with high size uniformity, which have been found to exhibit stable electrochromic performance with high colour contrast and H⁺ insertion ability. The WO₃ nanoparticles that can be effectively produced in high quantity are promising electrochromic candidates for potential applications in large-area smart windows.

---

1. Introduction

Electrochromic (EC) devices can reversibly change their optical properties by applying an external electrical voltage, making them potential applications in smart windows, high contrast displays, optical shutters, sunglasses and antiglare mirrors.^1–4 Particularly, smart windows for use in buildings are receiving considerable commercial interest since they can effectively save energy by regulating the throughput of solar light.^5,6 A typical EC device is a five-layer structure composed of a glass or plastic substrate (GS), a transparent conductor (TC), an electrochromic layer (EC), an ion conductor/electrolyte (IC), and an ion storage layer (IS), among which the EC layer that is constructed by EC materials is the principal component.^7–9 As an important oxide semiconductor, WO₃ is the most widely studied EC material and has received extensive attention in the past few decades due to its high coloration efficiency, high chemical and photoelectrochemical stability.^10–12

It is well-known that the intercalation and de-intercalation of electrons and ions (H⁺, Li⁺, Na⁺, or K⁺) lead to the coloration and bleaching processes of WO₃. In comparison with its bulk counterparts, therefore, WO₃ in nanometer scale with high surface activity and low dimensionality usually exhibits significantly enhanced EC performance (faster color-switching speed, higher coloration efficiency, and better cyclic stability) by reducing the ion diffusion path length.¹³ After years of efforts, a wide range of methods, including hydrothermal process,^14,15 laser albation,¹⁶ acid precipitation,¹⁷ flame spray process,¹⁸ thermal evaporation,^19–22 vapor deposition,^23–25 etc., have been utilized to prepare WO₃ nanomaterials in different forms. Till now, smart windows based on a series of WO₃ nanomaterials have been developed, but often suffer from high cost and low yield. Therefore, a facile, efficient process that can realize scale-up production of WO₃ EC nanomaterials is desirable. Among the various strategies for preparing WO₃ nanomaterials, chemical routes, such as sol–gel and hydrothermal methods, have advantages of simplicity and low cost yet complicated and time-consuming; laser ablation or flame spray process need special equipment and precise control. By contrast, evaporation or deposition routes are promising for large-scale production of WO₃ nanomaterials because they need only very simple equipment and could directly produce WO₃ with high crystallinity and purity. Up to now, however, evaporation routes have been widely adopted for the controllable growth of WO₃ products such as nanorods, nanowires as well as their arrays, but little work has been given to improve the yield of the final products.^25–27

The growth of WO₃ products via the evaporation method has been believed to follow the typical vapor-to-solid (VS) mechanism.¹⁹ During the VS growth, the tungsten-containing vapors are transferred from the precursor source by carrier gas to the low-temperature areas where nucleation, deposition, and growth of WO₃ occur. The reaction temperature determines the amount of vapors produced from the precursor, which in turn control the yield of the WO₃ products; the carrier gas flow rate could determine the residence time of WO₃ crystal vapors at high temperature zone and their supercooling degree at the deposition region and then, control the morphology and size of the WO₃ products. Therefore, it is desirable that increasing both the reaction temperature and the carrier gas flow rate could reduce the size of the WO₃ products and at the same time increase their yield.

In this work, we presented a facile, high-temperature, catalyst-free, solid evaporation route to prepare WO₃ nanoparticles in large scale and at high yield with commercial ammonium paratungstate tetrahydrate (APT·4H₂O) as precursor. The morphology, size, and crystalline structure of the final WO₃ particles have been systematically studied, as a function of reaction temperature, deposition temperature, carrier gas flow rate, and size of quartz tube. The EC performances of the as-obtained WO₃ nanoparticles have also been investigated.

2. Experimental details

WO₃ particles were prepared by the solid evaporation route with APT·4H₂O as precursor. The apparatus used for the production of WO₃ particles is schematically illustrated in Fig. 1. In a typical synthesis, 2 g of APT·4H₂O was firstly dispersed on an alumina plate, which was then located at the front end of the quartz tube inserted in the horizontal furnace. Upon reaching the reaction temperature (1300–1400 °C), the alumina plate containing the APT·4H₂O was manually transferred to the central area of the furnace. Subsequently, Ar gas was introduced and kept flushing at a flow rate in the range of 1–6 L min⁻¹ for 30 min. Two kinds of quartz tubes were utilized, i.e. small tube with diameter of 30 mm and large tube with diameter of 55 mm. Due to the temperature gradient from the center to the gas outlet end of the quartz tube, the as-prepared yellowish particles were collected from three different zones of the quartz tube by using homemade collectors. The zones that are 35–40 cm, 40–45 cm and 45–50 cm away from the tube center are labeled as zones A, B and C, respectively. The temperatures in zones A, B, and C, measured by a thermocouple, are in the ranges of 200–400 °C, 100–200 °C, and 40–100 °C, respectively. The controlling parameters of the evaporation process and the main characteristics of the resulting products are summarized in Table 1. After each reaction cycle, the as-deposited powders on the collectors were collected and their yields (refers to the mass ratio of the final products and the precursor) were calculated.

Fig. 1 Schematic illustration of the apparatus used for the production of WO₃ particles.

Table 1 Synthesis processing parameters and a summary of the morphology, size, and yield of the resulting products

Collecting area	Reaction temperature (°)	Ar gas flow rate (L min^−1)	Size of quartz tube (mm)	Reaction product	Morphology	Size range (nm)	Yield (mg)
Zone B	1300	2	30	WO3 (orthorhombic)	Polyhedral particles	20–800	350
	1350	1	30	WO3	Polyhedral particles	20–800	270
	1350	2	30	WO3 (orthorhombic)	Polyhedral particles	100–800	520
	1350	4	30	WO3	Quasi-spherical nanoparticles	Diameter: 20–600	500
	1350	6	30	WO3	Quasi-spherical nanoparticles	Diameter: 20–100	380
	1350	6	50	WO3	Quasi-spherical nanoparticles	Diameter: 20–100	430
	1400	2	30	WO3 (monoclinic)	Octahedral particles	Edge: 1000	500
Zone C	1300	2	30	WO3	Octahedral particles	Edge: 500–2000	490
					Ultrafine spherical particles	Diameter: <20
	1350	1	30	WO3	Octahedral particles	Edge: 500–1000	440
					Ultrafine spherical particles	Diameter: <20
	1350	2	30	WO3	Octahedral particles	Edge: 500–1000	710
	1350	4	30	WO3	Quasi-spherical nanoparticles	Diameter: 20–200	700
	1350	6	30	WO3	Quasi-spherical nanoparticles	Diameter: 20–100	860
	1350	6	50	WO3	Quasi-spherical nanoparticles	Diameter: 20–100	890
	1400	2	30	WO3	Octahedral particles	Edge: 500–1500	700
					Ultrafine spherical particles	<20

---

The crystal structure of the as-synthesized WO₃ particles was characterized by X-ray diffraction on a D/MAX-2500 X-ray diffractometer, with a Cu Kα radiation and a step size of 0.02° in the 2θ range of 20–60°. The morphology and structure of the WO₃ particles were analyzed by using a scanning electron microscope (FESEM, JSM-6700F, 15 kV), and a transmission electron microscope (TEM, JEOL 2000FX, operated at 120 kV). Size distribution of the WO₃ particles was analyzed by using a laser scattering particle analyzer (Microtrac S3500). The chemical composition of the WO₃ particles was appraised by X-ray photoelectron spectrometry (XPS, Kratos Axis Ultra DLD) using a monochromatic Al kR X-ray source (1486.6 eV). Pass energies of 160 and 40 eV were normally used for survey spectra and narrow scan spectra, respectively.

EC properties of the WO₃ nanoparticles were measured in a 0.5 M H₂SO₄ electrolyte solution by using Autolab Potentiostat (PGSTAT 302N). EC films of the WO₃ samples (hereafter referred to WO₃ EC films) were prepared by ways of spin coating. The WO₃ samples were dispersed in ethanol to form a suspension with a concentration of 10 g L⁻¹. The suspension was then spin coated onto transparent ITO coated glass with a dimension of 30 mm × 10 mm. The WO₃ EC films were dried under irradiation of high brightness light provided by a 1000 W filament lamp to avoid agglomeration of the WO₃ nanoparticles. Cyclic voltammetric (CV) curves of the WO₃ EC films were recorded at scanning rates ranging from 20 to 200 mV s⁻¹ between −0.8 and 1.0 V. Ag/AgCl and Pt foil were used as the reference electrode and the counter electrode, respectively.

3. Results and discussions

3.1 Influence of reaction temperature

Fig. 2 shows the typical SEM images of the samples collected from zones A, B, and C at reaction temperatures ranging from 1300 to 1400 °C with an Ar gas flow rate of 2 L min⁻¹. Obviously, the morphology of the as-prepared samples depends strongly on the reaction temperature and the collection locations. In zone A, few rod-shaped particles were collected, which become thicker with increased reaction temperature. When the reaction temperature reached 1400 °C, the final particles consist of tree-like structure supporting numerous thick rods. When the temperature was 1300 and 1350 °C, both the samples collected from zone B are mixture of polyhedral particles and semi-spherical particles. However, they both have monodisperse size distribution (Fig. S1 and S2, see ESI†). By comparison, the particles synthesized at 1350 °C show more uniform distribution than those synthesized at 1300 °C. At 1400 °C, the majority of the particles exhibit interesting octahedron-like morphology with sharp edges and corners and smooth surfaces. The size distribution of the octahedral particles is also monodisperse, with large average size of about 1.5 μm (Fig. S3†). In general, the sample synthesized at 1350 °C possesses not only the finest size but also the best size distribution. In zone C, all the samples are mixture of octahedron-like particles and ultrafine spherical particles, irrespective of the reaction temperature.

Fig. 2 SEM images of WO₃ particles synthesized at different reaction temperatures with an Ar gas flow rate of 2 L min⁻¹.

XRD results (Fig. 3) indicate that all the samples (collected in zone B) synthesized at different reaction temperature with an Ar gas flow rate of 2 L min⁻¹ are pure WO₃, but with different crystalline structure. The sharp, high-intensity feature of the diffraction peaks also suggests that the as-prepared WO₃ are highly crystallized. As shown in Fig. 3a and 3b, all the diffraction peaks of both the samples synthesized at 1300 °C and 1350 °C could be readily indexed as orthorhombic WO₃ (JCPDS 20-1324). When the reaction temperature is 1400 °C, the final sample is pure monoclinic WO₃ (JCPDS 43-1035), as seen in Fig. 3c. XPS was further used to investigate the chemical composition of the as-synthesized samples. Fig. 4a shows the XPS survey spectra of the sample synthesized at 1350 °C. All the peaks can be indexed to be W and O elements (the carbon peak is due to the carbon substrate), indicating high purity of the as-synthesized sample. A fitting curve of the W 4f level is shown in Fig. 4b. The peaks at binding energies of 35.6 and 37.7 eV match greatly well with the reported values for the W⁶⁺ oxidation state, indicating the presence of WO₃.²⁸

Fig. 3 XRD patterns of WO₃ particles synthesized at (a) 1300 °C, (b) 1350 °C, and (c) 1400 °C with an Ar gas flow rate of 2 L min⁻¹.

Fig. 4 (a) XPS survey spectra of WO₃ particles synthesized at 1300 °C with an Ar gas flow rate of 2 L min⁻¹ and (b) the corresponding W 4f core-level spectra.

3.2 Influence of Ar gas flow rate

The SEM images in Fig. 5 illustrate the influence of Ar gas flow rate on the morphology of the final WO₃ products in zones B and C synthesized at 1350 °C for 30 min. With a low Ar gas flow rate of 1 L min⁻¹, the sample collected in zone B is composed of polyhedral particles with inhomogeneous size. Here, the sample collected in zone C exhibits similar morphology to that synthesized at 1300 °C with Ar gas flow rate of 2 L min⁻¹, which consists of octahedral particles and ultrafine particles. Increasing Ar gas flow rate to 4 L min⁻¹, the morphology of the samples changed obviously and quasi-spherical particles with decreased size occur in both zones, as compared to the samples synthesized with Ar gas flow rates of 1 and 2 L min⁻¹. In zone B, some large quasi-spherical particles with diameters of 20–600 nm as well as few polyhedral particles can be clearly observed. In zone C, the sample consists mainly of quasi-spherical particles with diameter of 20–200 nm; in addition, few octahedron-like particles were found. The corresponding TEM image of the WO₃ particles in zone C synthesized with Ar gas flow rate of 4 L min⁻¹ further demonstrate clearly their quasi-spherical shape, as shown in Fig. 6a. The particle size continued to decrease with increased Ar gas flow rate. When the Ar gas flow rate was 6 L min⁻¹, WO₃ nanoparticles with small and uniform size were synthesized. Here, the WO₃ samples collected in zone B and zone C have very similar morphology and size, except that some ultrathin nanoparticles occurred in zone C. Size distribution analysis based on the SEM images indicates that both the samples from zone B and zone C are monodisperse, and their average diameters were calculated to be 68 and 60 nm, respectively (Fig. S4 and S5, see ESI†). Further TEM image analysis (Fig. 6b and c) indicates that, in both zone B and zone C, the as-synthesized WO₃ samples present diverse forms, such as quasi-spherical, semi-spherical, tetrahedral and octahedral shape, but a majority of them have quasi-spherical morphology with diameters in range of 20–100 nm, in accordance with the size distribution results.

Fig. 5 SEM images of WO₃ particles synthesized with different Ar gas flow rate at a reaction temperature of 1350 °C.

Fig. 6 (a) TEM image of WO₃ particles in zone C synthesized with Ar gas flow rate of 4 L min⁻¹; TEM images of WO₃ particles in (b) zone B and (c) zone C synthesized with an Ar gas flow rate of 6 L min⁻¹.

3.3 Influence of size of quartz tube

In order to increase the powder-collecting area, large quartz tube with diameter of 50 mm was introduced. As shown in Fig. 7, both the samples collected in zones B and C are mainly composed of quasi-spherical nanoparticles with uniform size distribution, which is extremely similar to those synthesized using small quartz tube with diameter of 30 mm (Fig. 5). By careful observation, it can be found that some ultrafine nanoparticles exist in both samples. In comparison with the synthesis using small quartz tube, much higher amount of powders can be collected when same amount of APT was used, indicating that production of WO₃ nanoparticles can be easily increased by enlarging the collecting area.

Fig. 7 SEM images of WO₃ nanoparticles synthesized using large quartz tube with an Ar gas flow rate of 6 L min⁻¹ at 1350 °C: (a) sample collected in zone B and (b) sample collected in zone C.

3.4 Growth mechanism

Because no catalyst was used in the present solid evaporation process, the formation of the WO₃ particles is thought to follow the traditional vapour-to-solid (V–S) growth mechanism, during which supersaturation, vapour nucleation, crystallization, and crystal growth take place.²⁹ Once transferred to the central area of the furnace, the APT·4H₂O would immediately be decomposed into WO₃ powders. Then, plenty of WO₃ vapours were generated due to the high reaction temperature. With the flushing of Ar gas, the WO₃ vapours were carried downstream to the quartz end, where nucleation and growth of WO₃ happened. In the evaporation process, the reaction temperature, deposition temperature, and Ar gas flow rate could determine the morphology and size of the final WO₃ powders because they influence the concentration, cooling rate, and crystallization period of the WO₃ vapours or nuclei. At high deposition temperature, for example, in zone A, thick WO₃ rods were fabricated. This can be ascribed to that there is large temperature gradient along the direction perpendicular to the substrate in zone A, leading to the preferential growth of WO₃ nuclei and finally the formation of WO₃ rods. Since only a very small fraction of powders are deposited on the collector in zone A, we paid more attention to the growth of the samples collected in zones B and C.

Fig. 8 provides a generalized illustration of the morphological evolution of WO₃ powders collected in zones B and C as a function of reaction temperature, collecting area (i.e. deposition temperature), and Ar gas flow rate. In zone B, at an Ar gas flow rate of 2 L min⁻¹, the final WO₃ samples underwent apparent morphological evolution with increased reaction temperature, from irregular particles to semi-spherical particles, and finally to octahedral particles, as shown in Fig. 8a. To explain this phenomenon, the evaporation, nucleation, and growth rate of the WO₃ nuclei or crystals at different temperatures as well as their growth habit should be considered.³⁰ At reaction temperatures of 1300 and 1350 °C, the competition between the nucleation rate and growth rate of the WO₃ nuclei or crystals prevails, leading to the formation of orthorhombic WO₃ particles with different morphologies. When growth of WO₃ crystals proceeded under condition that the nucleation rate is higher than the growth rate, quasi-spherical particles were formed; otherwise, irregular particles happened. The evaporation rate of the WO₃ source increased with increased reaction temperature. When the reaction temperature reached 1400 °C, the evaporation rate may become higher than the nucleation/growth rate of the WO₃ nuclei. Herein, the preferential growth habit of WO₃ crystals would play the leading role. With the WO₃ vapours continuously supplied, ideal, stable, monoclinic WO₃ octahedrons with low surface energies were finally formed by simply alternating stacking of the octahedral WO₆ units.^19,31

Fig. 8 Schematic illustration of the morphological evolution of WO₃ powders collected in zones B and C as a function of reaction temperature and Ar gas flow rate.

With the reaction temperature unchanged, the particles size of the final samples decreased with increased Ar gas flow rate, as shown in Fig. 8b. It is considered that, with increasing carrier gas flow rate, the nucleation rate of WO₃ crystals is increased due to the increased cooling rate of the WO₃ nuclei or vapors and the shortened residence time of the WO₃ vapours at high temperature zone. In contrast, the growth rate of the WO₃ crystals is decreased. As a result, the sizes of the WO₃ particles decreased with increased Ar gas flow rate. The few large particles synthesized with gas flow rate higher than 4 L min⁻¹ may origin from the self-agglomeration of WO₃ particles during their transferring from the high temperature zone to the deposition area. When the Ar gas flow rate was increased to 6 L min⁻¹, quasi-spherical WO₃ nanoparticles with diameters of 20–100 nm were formed. Here, the high Ar gas flow rate resulted in high supercooling degree, and hence the nucleation rate of the WO₃ crystals was higher than their growth rate, leading to the formation of quasi-spherical particles.

Differently from the apparent morphological evolution of the samples collected in zone B, all the final WO₃ samples collected in zone C at different reaction temperature with an Ar gas flow rate of 2 L min⁻¹ are mainly composed of octahedral particles, as shown in Fig. 8c. This should be attributed to the low deposition temperature in zone C. The low deposition temperature provide ideal nucleation place for the WO₃ vapours due to the higher supercooling degree. After nucleation, the growth of the WO₃ nuclei is directly related to their surface energies of crystallographic planes. Since the {111} planes have the lowest surface energies, stable WO₃ with exposed surfaces of {111} planes, namely WO₃ octahedron, are formed.²⁰ As seen from Fig. 8d, the size of the samples collected in zone C decreased with increasing Ar gas flow rate, and quasi-spherical particles were formed when the Ar gas flow rate was higher than 4 L min⁻¹. Here, due to the high Ar gas flow rate, the growth of the WO₃ crystals was dominated by the nucleation rate other than the surface energy. Therefore, quasi-spherical particles that are similar to those collected in zone B occurred. However, it should be noted that the sample from zone C have slightly better quality than that from zone B and some ultrafine nanoparticles also occurred in zone C. This should be attributed to the very high supercooling degree in zone C resulting from both the low deposition temperature and the high gas flow rate.

3.5 EC properties

As has been discussed above, a reaction temperature of 1350 °C and an Ar gas flow rate of 6 L min⁻¹ yielded optimized quasi-spherical WO₃ nanoparticles with uniform size distribution. The EC performance of the quasi-spherical WO₃ was investigated by means of cyclic voltammetry. The octahedral WO₃ particles (collected in both zones B and zone C) that were synthesized at a reaction temperature of 1400 and an Ar gas flow rate of 2 L min⁻¹ were chosen as a reference sample. For convenience, the octahedral WO₃ particles and the quasi-spherical WO₃ nanoparticles are labeled as WO₃–O and WO₃–S, respectively. Fig. 9 shows the CV curves of the WO₃–O and WO₃–S EC films at different cycles, recorded at a scanning rate of 50 mV s⁻¹. For both samples, the current response decreases slightly after 1000 cycles without a change in the shape of the CV curves, indicating their high cyclic stability in acidic environment. As shown in Fig. 9a, there are three oxidation peaks in cyclic voltammogram, indicating that three types of hydrogen intercalated sites exist in the WO₃–O EC film, i.e., reversible active sites, shallow trap sites (reversible), and deep trap sites (irreversible), which are typical CV features of polycrystalline WO₃.^32–34 In contrast, the WO₃–S EC film has only one oxidation peak caused by the shallow trap sites, see Fig. 9b. The digital photographs provide direct information with regard to the colour change of the EC films with varying voltages, as shown in the insets of Fig. 9. During each cycle, both the WO₃–O and WO₃–S EC films had a blue colour when they were cathodically polarized. As the cathodic potential reached −0.8 V, the WO₃–S EC film presented a deep blue coloration, while the WO₃–O EC film had a light blue coloration. When anodically polarized, the EC films were bleached and returned to original light yellowish. Therefore, it can be concluded that the WO₃–S EC film shows a larger light transmittance modulation than the WO₃–O EC film, making it suitable candidate for the application in smart window.

Fig. 9 CV curves of (a) WO₃–O and (b) WO₃–S EC films. Insets shows the digital photographs of the WO₃–O and WO₃–S EC films at bleached and colored states.

The coloration process of WO₃ is in accordance with the co-intercalation of electrons and H⁺ ions into WO₃ to form hydrogen tungsten bronzes (H_xW₁₈O₄₉) with blue coloration, and contrarily the bleaching process corresponds to the de-intercalation of electrons and H⁺ ions out from H_xW₁₈O₄₉ to electrolyte.^4,8,19 As is well-known, the integrated cathodic current density equates to the amount of H⁺ intercalated to form the H_xW₁₈O₄₉. The total cathodic charge for the WO₃–S was about 0.14 mC cm⁻², which is 4.5 times higher than that of the WO₃–O (0.0312 mC cm⁻²). This result signifies that, compared to the WO₃–O EC film, the WO₃–S EC film shows much higher charge-insertion density (Fig. 9), indicating faster H⁺ diffusion kinetics.

Fig. 10a shows the CV curves of the WO₃–S EC film recorded at different scan rates. At a scan rate of 50 mV s⁻¹, there is one main oxidation peak centered at about −0.01 V and one reduction peak at −0.3 V. With increasing scan rates, the oxidation peaks became broader and higher and shifted to a higher potential. The reduction peaks present the same phenomenon. When the current (i_p) at the oxidation peaks are plotted against the square root of scan rate (v^1/2), a linear relationship can be obtained, proving that both the coloration and bleaching (i.e., oxidation and reduction) processes proceeded with the H⁺ diffusion, as presented in Fig. 10b. The time constant of H⁺ diffusion process is a crucial parameter for EC devices since it determines the colour contrast, response time, and coloration efficiency. The H⁺ diffusion coefficients can be calculated using the Randles–Servcik equation of i_p = 2.72 × 10⁵n^3/2AD^1/2Cv^1/2, where n is the number of electrons and here it is assumed to be 1, C is the concentration of H⁺ (M), A is the electrode area (cm²), v is the scan rate (mV s⁻¹), i_p is the peak current (A), and D is the diffusion coefficient of H⁺.^35,36 For contrast, a scan rate of 50 mV s⁻¹ was accepted for both samples. According to the above equation, the H⁺ diffusion coefficients of the WO₃–S EC film are calculated to be 1.07 × 10⁻⁹ and 8.05 × 10⁻¹⁰ cm² s⁻¹ for the intercalation and deintercalation process, respectively, which are comparable to the reported values.^37,38 For the WO₃–S EC film, the values are calculated to be 6.8 × 10⁻¹¹ and 5.27 × 10⁻¹⁰ cm² s⁻¹ for the intercalation and deintercalation process, respectively. Obviously, the H⁺ diffusion coefficients of the WO₃–S EC film are much higher than those of the WO₃–O EC film.

Fig. 10 (a) CV curves of WO₃–S EC films with varying scan rates and (b) plot of cathodic and anodic peak current as a function of the square root of the scan rates.

It is generally considered that the electrochromism of tungsten oxide results from the intercalation and de-intercalation of the electrons from electrode and the cations (H⁺, Li⁺, and Na⁺ ions) from electrolyte, respectively.⁸ The EC performance of tungsten oxide is directly related to the diffusion of ions in the tungsten oxide-based EC film. As has been mentioned above, the WO₃–S EC film has not only higher H⁺ diffusion coefficients but also higher color contrast in comparison with the WO₃–O EC film. Two main factors should be considered. Firstly, the quasi-spherical WO₃ nanoparticles possess smaller and more uniform size as compared to the octahedral WO₃ particles, which can provide more active sites for the intercalation and de-intercalation of H⁺ ions. Secondly, the quasi-spherical WO₃ nanoparticles have superior assembly quality than the octahedral WO₃ particles in the EC films. SEM images (Fig. S6 and S7, see ESI†) of the EC films indicates that the quasi-spherical WO₃ nanoparticles in the WO₃–S film exhibit highly compact uniform assembly, while the octahendral nanoparticles (mixed with a small amount of ultrafine nanoparticles) are randomly distributed on the ITO substrate. The high-quality assembly of the quasi-spherical WO₃ nanoparticles could effectively reduce the ion diffusion path length. As a result, the WO₃–S EC film has higher H⁺ diffusion coefficiency than the WO₃–O EC film.

4. Conclusions

We presented the production of WO₃ particles in large quantity via a facile, high-temperature, catalyst-free, solid evaporation route with APT·4H₂O as raw material. The morphological evolution of the as-synthesized WO₃ particles has been systematically investigated, as a function of reaction temperature, deposition temperature, carrier gas flow rate, and size of the quartz tube. The ideal quasi-spherical WO₃ nanoparticles with uniform size distribution were synthesized with an Ar gas flow rate of 6 L min⁻¹ at 1350 °C. Based on the CV analysis, the WO₃ nanoparticles were found to exhibit stable electrochromic performance with high colour contrast, making them potential candidates as cathode materials for EC applications. The efficient, large-scale, and high-yield production of the WO₃ nanoparticles could help to develop large-scale EC devices.

Acknowledgements

This work is financially supported by the National Natural Science Foundation of China (no. 51202142 and 51202144), the Major State Basic Research Development Program of China (973 Program, no. 2014CB643306), the National High Technology Research and Development Program of China (863 Program, no. 2013AA041106), the Project of Science and Technology Commission of Shanghai Municipality (no. 12231204102), the Doctoral Fund of Ministry of Education of China (no. 20103121110003), the Research and Innovation Project of Shanghai Education Commission (no. 12YZ113), and the Natural Science Foundation of Shandong (no. ZR2011EMQ011), the Chenguang Scholar Project of Shanghai Education Commission (no. 11CG53). Dr Shibin Sun wishes to express sincere appreciation to Prof. Y. Q. Zhu in the University of Exeter for all the help and support during his stay in UK.

Notes and references

1. A. Llordes, G. Garcia, J. Gazquez and D. J. Milliron, Nature, 2013, 500, 323.
2. D. L. Chen, X. X. Hou, H. J. Wen, Y. Wang, H. L. Wang, X. J. Li, R. Zhang, H. X. Lu, H. L. Xu, S. K. Guan, J. Sun and L. Gao, Nanotechnology, 2010, 21, 035501.
3. S. G. Lee, D. Y. Lee, H. S. Lim, D. H. Lee, S. Lee and K. Cho, Adv. Mater., 2010, 22, 5013.
4. S. H. Lee, R. Deshpande, P. A. Parilla, K. M. Jones, B. To, A. H. Mahan and A. C. Dillon, Adv. Mater., 2006, 18, 763.
5. R. Sydam, M. Deepa and A. K. Srivastava, RSC Adv., 2012, 2, 9011.
6. X. J. Lv, J. W. Sun, B. Hu, M. Ouyang, Z. Y. Fu, P. J. Wang, G. F. Bian and C. Zhang, Nanotechnology, 2013, 24, 265705.
7. J. Zhang, X. L. Wang, X. H. Xia, C. D. Gu and J. P. Tu, Sol. Energy Mater. Sol. Cells, 2011, 95, 2107.
8. Z. H. Jiao, J. M. Wang, L. Ke, X. W. Liu, H. V. Demir, M. F. Yang and X. W. Sun, Electrochim. Acta, 2012, 63, 153.
9. X. T. Chang, S. B. Sun, Z. J. Li, X. Xu and Y. Y. Qiu, Appl. Surf. Sci., 2011, 257, 5726.
10. W. Morales, M. Cason, O. Aina, N. R. de Tacconi and K. Rajeshwar, J. Am. Chem. Soc., 2008, 130, 6318.
11. G. Hodes, D. Cahen and J. Manassen, Nature, 1976, 260, 312.
12. M. A. Butler, R. D. Nasby and R. K. Quinn, Solid State Commun., 1976, 19, 1011.
13. Z. H. Jiao, X. Wang, J. M. Wang, L. Ke, H. V. Demir, T. W. Koh and X. W. Sun, Chem. Commun., 2012, 48, 365.
14. G. C. Xi, J. H. Ye, Q. Ma, N. Su, H. Bai and C. Wang, J. Am. Chem. Soc., 2012, 134, 6508.
15. J. Zhang, J. P. Tu, G. H. Du, Z. M. Dong, Y. S. Wu, L. Chang, D. Xie, G. F. Cai and X. L. Wang, Sol. Energy Mater. Sol. Cells, 2013, 114, 31.
16. H. W. Zhang, G. T. Duan, Y. Li, X. X. Xu, Z. F. Dai and W. P. Cai, Cryst. Growth Des., 2012, 12, 2646.
17. S. Supothina, P. Seeharaj, S. Yoriya and M. Sriyudthsak, Ceram. Int., 2007, 33, 931.
18. D. Hidayat, A. Purwanto, W. N. Wang and K. Okuyama, Mater. Res. Bull., 2010, 45, 165.
19. Y. Baek and K. Yong, J. Phys. Chem. C, 2007, 111, 1213.
20. L. Fang, S. J. Baik, K. S. Lim, S. H. Yoo, M. S. Seo, S. J. Kang and J. W. Seo, Appl. Phys. Lett., 2010, 96, 193501.
21. K. J. Patel, C. J. Panchal, M. S. Desai and P. K. Mehta, Mater. Chem. Phys., 2010, 124, 884.
22. A. A. Mohammad and M. Gillet, Thin Solid Films, 2002, 408, 302.
23. S. Ashraf, R. Binions, C. S. Blackman and I. P. Parkin, Polyhedron, 2007, 26, 1493.
24. H. Wang, Y. Su, S. Chen and X. Quan, Mater. Res. Bull., 2013, 48, 1304.
25. Y. Zhang, Y. G. Chen, H. Liu, Y. Q. Zhou, R. Y. Li, M. Cai and X. L. Sun, J. Phys. Chem. C, 2009, 113, 1746.
26. F. Xu, A. Fahmi, Y. M. Zhao, Y. D. Xia and Y. Q. Zhu, Nanoscale, 2012, 4, 7031.
27. J. Zhou, L. Gong, S. Z. Deng, J. Chen, J. C. She, N. S. Xu, R. Yang and Z. L. Wang, Appl. Phys. Lett., 2005, 87, 223108.
28. S. Jeon and K. Yong, Nanotechnology, 2007, 18, 245602.
29. M. Gillet, R. Delamare and E. Gillet, J. Cryst. Growth, 2005, 279, 93.
30. Z. S. Houweling, J. W. Geus, M. de Jong, P.-P. R. M. L. Harks, K. H. M. van der Werf and R. E. I. Schropp, Mater. Chem. Phys., 2011, 131, 375.
31. H. B. Zhang, M. S. Yao, L. Y. Bai, W. C. Xiang, H. C. Jin, J. L. Li and F. L. Yuan, CrystEngComm, 2013, 15, 1432.
32. D. S. Dalavi, R. S. Devan, R. A. Patil, R. S. Patil, Y. R. Ma, S. B. Sadale, I. Y. Kim, J. H. Kim and P. S. Patil, J. Mater. Chem. C, 2013, 1, 3722.
33. J. Y. Li, J. F. Huang, J. P. Wu, L. Y. Cao, Q. J. Li and K. Yanagisawa, CrystEngComm, 2013, 15, 7904.
34. S. H. Baeck, K. S. Choi, T. F. Jaramillo, G. D. Stucky and E. W. McFarland, Adv. Mater., 2006, 15, 1269.
35. R. S. Crandall and B. W. Faughnan, Appl. Phys. Lett., 1975, 26, 120.
36. D. H. Kim, Sol. Energy Mater. Sol. Cells, 2012, 107, 81.
37. L. Liang, J. J. Zhang, Y. Y. Zhou, J. F. Xie, X. D. Zhang, M. L. Guan, B. C. Pan and Y. Xie, Sci. Rep, 2013, 3, 1936.
38. S. R. Bathe and P. S. Patil, Smart Mater. Struct., 2009, 18, 025004.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c3ra47733a

---