Enhanced Transmittance Modulation of SiO₂-Doped Crystalline WO₃ Films Prepared from a Polyethylene Oxide (PEO) Template

Abstract

Polyethylene oxide (PEO)-modified silicon dioxide (SiO₂)-doped crystalline tungsten trioxide (WO₃) films for use as electrochromic layers were prepared on indium tin oxide (ITO) glass by the sol–gel spin coating technique. The effects of the PEO template and SiO₂ on the electrochromic transmittance modulation ability of crystalline WO₃ films were investigated. Fourier transform infrared spectroscopy (FT-IR) spectra analysis indicated that PEO was decomposed after annealing at 500 °C for 3 h. X-ray diffraction (XRD) pattern analysis showed that both SiO₂ and PEO helped reduce the crystalline grain size of the WO₃ films. Atomic force microscope (AFM) images showed that the combined action of SiO₂ and PEO was helpful for achieving high surface roughness and a macroporous structure. An electrochromic test indicated that PEO-modified SiO₂-doped crystalline WO₃ films intercalated more charges (0.0165 C/cm²) than pure WO₃ crystalline films (0.0095 C/cm²). The above effects resulted in a good transmittance modulation ability (63.2% at 628 nm) of PEO-modified SiO₂-doped crystalline WO₃ films, which was higher than that of pure WO₃ crystalline films (9.4% at 628 nm).

1. Introduction

Tungsten trioxide (WO₃) is an important functional material, widely studied for applications in gas sensors [1], photochromism [2], photocatalysis [3], and electrochromism [4]. In recent years, the energy consumption of buildings for indoor temperature regulation has been increasing. Smart windows can reduce energy consumption by modulating light and heat inputs [5]. WO₃-based electrochromic layers have attracted increasing attention for smart windows applications. There are many methods for preparing WO₃ films, such as physical vapor deposition (PVD) [6], chemical vapor deposition (CVD) [7], sol–gel method [8], etc. However, both PVD and CVD have some characteristic drawbacks, such as high cost of equipment, high energy consumption, and vacuum dependency [9]. The sol–gel method can form large-area films in the atmosphere with less energy consumption and low equipment cost [10]. It is easy to realize doping uniformly at the molecular level by the sol–gel method. In addition, the pores formed by solvent evaporation during annealing can enhance the electrochromic properties of the films [11].

The electrochromic performance of WO₃ has been researched for several decades. As we know, the electrochromic performance is closely related to the morphology, structure, and crystallinity of the WO₃ films, and one effective method for enhancing the films’ electrochromic performance is to incorporate dopants such as TiO₂ [12], Fe₂O₃ [13], carbon [14], etc. It was found that crystalline WO₃ films have good redox stability [15], good durability [16], and good infrared-reflecting properties (good insulation) [17]. However, compared with amorphous WO₃ films, a drawback of crystalline WO₃ films is their low transmittance modulation ability [18,19], which limits their application.

To enhance the electrochromic performance, especially the transmittance modulation ability of crystalline WO₃ films, a feasible scheme is to develop large-surface area and macroporous crystalline WO₃ films. In previous studies, polyethylene oxide (PEO) was often used as a template for pore materials [20]. SiO₂ has been confirmed to help form pores [21] and reduce the crystalline grain size [22]. In our work, we tried to improve the transmittance modulation ability of crystalline WO₃ films by combining the advantages of PEO and SiO₂. As a result, we successfully prepared crystalline pure WO₃ films, SiO₂-doped crystalline WO₃ films, PEO-modified crystalline WO₃ films, and PEO-modified SiO₂-doped crystalline films by the sol–gel method. Besides, the FT-IR spectra, crystallization, surface morphology, and electrochromic properties of these films were investigated.

2. Materials and Methods

Tungsten powder (purity: 99.5%) and hydrogen peroxide (mass fraction: 30%) were mixed in a beaker (weight: 4 g, H₂O₂: 20 mL) to prepare a tungstate solution. When the tungsten powder completely reacted, the tungstate solution was condensed into 10 mL by heating, in order to remove superfluous H₂O₂. Next, four kinds of solution were prepared, as shown in

Table 1. The parameters of four kinds of solution. EW, control solution, pure tungstate solution; EWS, EWP, EWPS, experimental solutions (with SiO₂, PEO, SiO₂ and PEO, respectively).

Label	Absolute Ethanol/mL	Tungstate Solution/mL	Tetraethyl Orthosilicate (TEOS)/g	PEO polyethylene Oxide (Mw: 600,000)/g	Stirring Time/h	Heating Temperature/°C
EW	2.5	2.5	0	0	3	70
EWS	2.5	2.5	0.21	0
EWP	2.5	2.5	0	0.1
EWPS	2.5	2.5	0.21	0.1

. Solution EW (pure tungstate solution) was a control solution, and solutions EWS (with SiO₂), EWP (with PEO), EWPS (with SiO₂ and PEO) were experimental solutions. The tungstate solution concentration was around 1 mol/L. According to a previous study [22], a concentration of tetraethyl orthosilicate (TEOS) of 0.2 mol/L had a good inhibitory effect. PEO amounts had a significant effect on the structure and quality of the WO₃ films. In this experiment, the amounts of the above reactants were optimized and controlled for the control experiment. After stirring and heating, the four kinds of sol were obtained, as shown in Figure 1. All of them were transparent, colloid (Tyndall effect), and orange-yellow, which indicated that PEO and TEOS were uniformly dissolved in the sol.

Finally, all the films were prepared on ITO glass by spin coating (4000 revolutions per minute for 30 s). The area of the electrochromic layer was around 2 cm². The pure WO₃ films, SiO₂-doped WO₃ films, PEO-modified WO₃ films, and PEO-modified SiO₂-doped films annealed at 500 °C were named EW500, EWS500, EWP500, and EWPS500, respectively. Similarly, the PEO-modified SiO₂-doped films annealed at 100 and 300 °C were named EWPS100 and EWPS300, respectively.

The thickness of all films was measured by a probe surface profiler (Veeco Dektak 150, Veeco, NY, USA) and was about 200 ± 10 nm. FT-IR spectra of EWPS100, EWPS300, and EWPS500 were recorded by a FT-IR spectrophotometer (SHIMADZU IR Prestige-21, SHIMADZU, Tokyo, Japan) with an ITO glass acting as a blank. The crystalline structure of these films was determined by X-ray Diffraction using Cu Kα radiation (XRD, PANalytical Empyrean DY1577, PANalytical, Almelo, Netherlands). The surface morphology was measured by atomic force microscopy (AFM, Being Nano-Instruments BY3000, Being Nano-Instruments, Beijing, China). An amount of 0.8 mol/L of lithium perchlorate/propylene carbonate (LiClO₄/PC) electrolyte was used for the electrochromic test, and the electrode gap was 0.5 cm. The voltage supply and current measurement were supported by an electrochemical workstation (CH Instruments CHI600E, CH Instruments, Shanghai, China). The transmittance of these films at the initial state, colored state, and bleached state was measured by a spectrometer (Morpho PG2000, Morpho, Shanghai, China) with an ITO glass acting as blank.

3. Results and Discussion

Figure 2 shows the changes of the appearance of the EW500, EWS500, EWP500, and EWPS500 films as the annealing time increased. Because of the carbonization of PEO at high temperature, it indicates that the color of samples EWP500 and EWPS500 (PEO-doped) changed from transparent to grey after annealing for 15 min, when compared with EW500 and EWS500 (not PEO-doped), but such changes did not occur in samples EW500 and EWS500, because of the rapid evaporation of the organic solvents. Finally, the color of samples EWP500 and EWPS500 turned transparent again as the annealing time increased. In the process of PEO pyrolysis, CO₂ and water vapor, which acted on the film structure, escaped from the WO₃ film.

Figure 3 shows the FT-IR spectra of EWPS100, EWPS300, and EWPS500, which were annealed at different temperatures. The spectra (a), (b), and (c) of Figure 3 show broad bands located at 3500–3400 cm⁻¹, which were attributed to asymmetric and symmetric –OH stretching vibrations. In addition, another sharp band around 1625 cm⁻¹ was attributed to –OH bending vibration [23]. The –OH was mainly from physisorbed water and surface OH groups.

As spectra (a) and (b) of Figure 3 show, the bands located around 2900–3000 cm⁻¹ correspond to the symmetric and antisymmetric –C–H stretching vibration of aliphatic CH₃ and CH₂ from ethanol and PEO [24]. The bands located around 3000–3100 cm⁻¹ correspond to symmetric and antisymmetric =C–H stretching vibration, which was attributed to incomplete pyrolysis of PEO [25]. The bands around 1423 cm⁻¹ could be assigned to asymmetric –C–H bending vibration [26]. In spectrum (a), the sharp band around 1074 cm⁻¹ could be assigned to the asymmetric C–O–C stretching vibration from PEO [26]. In spectrum (a), the bands at 894 and 572 cm⁻¹ were attributed to symmetric and antisymmetric W(O)₂ stretching vibration and O–O stretching vibration, respectively [27]. In spectra (b) and (c), the bands around 965 cm⁻¹ were attributed to symmetric W=O stretching vibration, and the bands at 615 and 730 cm⁻¹ were attributed to asymmetric W–O–W stretching vibration [27,28]. The bands around 1094 cm⁻¹ were attributed to asymmetric Si–O–Si stretching vibration in spectrum (c) [29].

Therefore, the above FT-IR spectra analysis indicated that the organic components, including PEO and ethanol, were strongly reduced after film annealing at 500 °C for 3 h. PEO was decomposed into CO₂ and H₂O which play a significant role in the formation of large surface area and macroporous structures. Besides, the annealing process led to the formation of W–O–W and Si–O–Si structures.

Figure 4 shows the XRD patterns of EW500, EWS500, EWP500, EWPS500, and ITO glass, which were analyzed using the software Jade 6.0 and PDF #43-1035. As Figure 4 shows, the diffraction peaks of WO₃ could be observed in the patterns (a), (b), (c), and (d), which indicated that all these WO₃ films were crystallized. The main peak of WO₃ at 2θ ≈ 24.1° corresponds to the (200) planes, and its crystalline grain size was calculated using the Scherrer formula [30]:

$$D=\frac{0.89\mathsf{\gamma }}{\mathsf{\beta }\cos \mathsf{\theta }}$$

(1)

where D is the average size of the grain on the (200) plane, γ is the X-ray wavelength, which is 0.154056 nm, β is the full width at half maximum (FWHM) of the diffraction peak, θ is the diffraction angle. The grain sizes of these films are shown in Figure 5; the smallest grain size was found in the sample EWPS500. The results indicated that both the PEO and SiO₂ dopants are helpful for reducing the grain size; a smaller grain size indicates that there are more grain boundaries and defects in the films. Further, the XRD patterns in Figure 4 also indicated that the diffraction peak of SiO₂ resulted from the ITO glass rather than the SiO₂ dopant. SiO₂ is amorphous and is present at the grain boundaries [31,32], inhibiting the continuous growth of the grains; therefore, the grain sizes of samples EWPS500 and EWS500 were smaller than that of sample EW500.

Figure 6 and Figure 7 show the AFM 3D images (10,000 nm × 10,000 nm) of four kinds of samples and the surface roughness and increased surface area (∆S, compared with their projected area: 10,000 nm × 10,000 nm) of these samples, respectively. In Figure 6, it indicates that in samples EWP500 and EWPS500 (PEO-doped), the macroporous structure is large compared with samples EW500 and EWS500 (not PEO-doped), which indicated that the PEO template plays an important role in the formation of a rough surface and a macroporous structure, because PEO was decomposed into CO₂ and H₂O when annealing was carried out at 500 °C for 3 h. Figure 7 indicates that the SiO₂ dopant also helped to increase the roughness compared with the control film EW500. The surface area of all the experimental films was larger than that of the control film, and sample EWPS500 showed the best result. In this material system, the SiO₂ dopant cooperated with PEO to inhibit grain growth and increase surface roughness and surface area.

Figure 8 shows the current curves of different WO₃ films in a cycle of electrochromic test under ±2 V voltages. The injecting and extracting charges per unit area (∆Q) were calculated on the basis of the current of a cycle of electrochromic test. As we know, the electrochromic mechanism refers to a process of Li⁺ and electron migration [33]:

$${\mathrm{WO}}_3\left(\mathrm{colorless}\right)+x{\mathrm{e}}^{-}+x{\mathrm{Li}}^{+}\to {\mathrm{Li}}_x{\mathrm{WO}}_3(\mathrm{Blue})$$

(2)

Li⁺ is the main charge carrier in the LiClO₄/PC electrolyte. The above process finishes when the currents tend to 0 mA. Among the four samples, the ∆Q of sample EW500 were the least, which indicated that there was a low amount of Li_xWO₃ generated in sample EW500. The time corresponding to 90% of the total current change is defined as the electrochromic response time. In Figure 8, all films show a short electrochromic response time (around 7 s), which is acceptable and closed to a normal level [34].

Both the images and the transmittance of the initial state, colored state, and bleached state of these samples are shown in Figure 9. The transmittance modulation ability (∆T) is defined by the following formula at wavelength 628 nm:

$$\Delta T=|T_{\mathrm{c}}-T_{\mathrm{b}}|$$

(3)

In this formula, T_c and T_b are the transmittance of the colored state and bleached state at wavelength 628 nm, respectively. ∆T directly shows the optical contrasts in a certain percentage. In this experiment, the ITO glass (substrate) was used as a blank, whose transmittance was designated as 100% in the wavelength range of 400–900 nm. According to Figure 9, the ∆T of these samples were calculated, and the results were 9.4% (EW500), 27.8% (EWS500), 35.5% (EWP500), and 63.2% (EWPS500). It indicates that the ∆T of sample EWPS500, whose color was deep blue at the color state, was the highest among these samples. This indicates that sample EWPS500 could effectively reduce the passing of light.

In addition, coloration efficiency (CE) is another important parameter for evaluating the electrochromic performance, defined by the following formula [35,36]:

$$\mathrm{CE}=\frac{\Delta \mathrm{OD}}{\Delta Q}=\frac{\ln \frac{T_{\mathrm{b}}}{T_{\mathrm{c}}}}{\Delta Q}$$

(4)

In Formula (4), ∆OD is the optical density; ∆Q, T_c, and T_b were defined above. Figure 10 shows the different CE of the four samples. There was an increase in the CE of the WO₃ crystalline films when PEO and SiO₂ were used as the dopants. Both PEO and SiO₂ were efficient in enhancing the electrochromic performance, but the combined use of SiO₂ and PEO could further increase the CE from 23.23 cm²/C (EWS500) or 35.86 cm²/C (EWP500) to 89.69 cm²/C (EWPS500). Figure 11 shows the different ∆T of the first and the 500th cycle. It indicates that the transmittance modulation changed little (within 5%). Therefore, all four samples showed stable cycle characteristics, which is an advantage of the crystalline WO₃ films [15,16].

It is worth comparing the electrochromic transmittance modulation ability in our work with the past reported data.

Table 2. Comparisons with the past reported data: Electrochromic film, synthetic method, crystallinity, and ∆T at 628 nm.

Electrochromic Film	Synthetic Method	Crystallinity	∆T at 628 nm	Reference
EWPS500	Sol–gel method	Crystalline	63.2%	This work
PEG400 doped WO3	Sol–gel method	Crystalline	~14%	[19]
WO3	Thermal evaporation	Crystalline	~32%	[36]
WO3	Thermal evaporation	Amorphous	~66%	[36]
2%V2O5 doped WO3	RF magnetron sputtering	Amorphous	~49%	[37]

shows these comparisons. The results showed that the modulation ability of the crystalline WO₃ film was enhanced successfully by PEO modification and SiO₂ doping compared with other crystalline WO₃ films. Further, the transmittance modulation ability of EWPS500 appeared as good as that of the amorphous WO₃ film.

As mentioned above, the electrochromic performance is closely related to the morphology, structure, and crystallinity of the WO₃ films. The PEO template was decomposed into CO₂ and H₂O after annealing at 500 °C, which resulted in a high surface roughness and a macroporous structure. The high surface roughness indicated that there was a large surface area for reaction. Meanwhile, the pores on the structure provide channels for LiClO₄/PC electrolyte, so Li⁺ can easily enter the internal structure. Therefore, both EWP500 and EWPS500 showed better electrochromic transmittance modulation ability than EW500 and EWS500. In addition, the SiO₂ dopant could further increase the surface roughness, so the electrochromic transmittance modulation ability of EWPS500 was enhanced to the highest levels, compared with that of other samples. However, the influence of crystallinity cannot be ignored. As we know, the component of the grain boundaries is amorphous WO₃, which contains multiple defects. Li⁺ is easily intercalated into these defects, generating blue Li_xWO₃. The XRD analysis indicated that SiO₂ could inhibit the growth of the grains effectively, and the most direct proof of this phenomenon was the decrease of the grain size, which indicated that the grain boundaries were larger in samples EWS50, EWP500, and EWPS500. Sample EWPS500 had the above characteristics, so its transmittance modulation ability was the most outstanding among these samples.

4. Conclusions

PEO-modified and SiO₂-doped crystalline WO₃ films were prepared by the sol–gel spin coating technique. The combined action of PEO and SiO₂ increased the surface roughness and formed a macroporous structure on the films. In addition, both PEO and SiO₂ could reduce the crystalline grain size effectively. The above characteristics of morphology, structure, and crystallinity resulted in a good electrochromic transmittance modulation of the crystalline WO₃ films. The PEO-modified and SiO₂-doped crystalline WO₃ films could intercalate more charges (0.0165 C/cm²) than the pure WO₃ films (0.0095 C/cm²). Moreover, the transmittance modulation ability (∆T = 63.2% at 628 nm) of the former films was better than that of the latter films (∆T = 9.4% at 628 nm).