Hydrothermal synthesis and electrochromism of WO₃ nanocuboids

Abstract

Tungsten trioxide (WO₃) nanocuboids are preferentially developed through control over three important processing parameters: fluoroboric acid concentration, hydrothermal reaction time and temperature. Fluoroboric acid concentration has significant influence on the phase transformation from triclinic to monoclinic followed by intermediate metastable hexagonal tungsten trioxide and simultaneous formation of thin plates to cuboid shapes. Optimum temperature and time are essential to achieve cuboid morphology without crystalline water molecules. Orthorhombic tungstite (H₂WO₄) grows from virgin precursors below 180 °C temperature prior to the formation of triclinic WO₃. The monoclinic nanocuboid has average dimensions of 140 nm length, 120 nm width and 85 nm thickness but its surface area is reduced and its crystallinity is enhanced for times greater than the optimum reaction time. A probable hydrothermal assisted reaction has been proposed to justify the formation of morphology and phase selection. The band gap energy of WO₃ insignificantly varies with respect to processing conditions, with the lowest at 2.75 eV. The current density of 3.15 mA cm⁻² is attributed to the high-symmetry electrochemical reaction of the dip-coated nanocuboid WO₃/ITO electrode. Coloration and bleaching kinetics depict the proton assisted-bleaching as faster than the coloration phenomenon with the appearance of 72.2% electrochromic reversibility. The fabricated nanocuboid WO₃-coated ITO glass has fairly good optical transparency, electrochromic stability and reversibility.

---

1. Introduction

Tungsten trioxide (WO₃), an n-type semiconductor, has received renewed attention in recent times due to its intriguing physicochemical properties.¹ The nanostructured oxide semiconductor is of considerable interest because of its potential applications in devices such as gas sensors, solar cells, and photocatalysts.² However, its properties are not limited to electronic and optoelectronic devices; they are also applicable from condensed matter physics to solid state chemistry. Oxide nanoparticles with different morphologies, in particular, have held much interest due to their distinctive physical and chemical properties, which depend on their morphology and crystal structure.^3,4 Morphology management is a challenging task for the investigation of their application potentiality. Moreover, the lower metal oxide dimension enhances the electrical and optical performance to tune modern device properties. The complex three-dimensional (3D) nanostructures have also attracted much attention since this nanostructure possesses features of nanometer-scale building blocks and novel properties.³ In general, their cuboid morphology has high surface area and therefore higher surface energy over the spherical shape.⁵ Thus, control over the morphology of semiconductor materials is a critical issue and a fascinating research area. Over the past few decades, tungsten oxide semiconductor material has been extensively studied due to its high coloration efficiency and good cyclic stability among various electrochromic materials. Extensive literature review suggests that the different morphologies of WO₃ in nanometer scale have already been prepared for the functionality of nanomaterials and electrochromic devices.^6,7 Most of the morphology has been achieved by different solution-based methods such as sol–gel, co-precipitation and hydrothermal in addition to numerous vapor-phase deposition methods.^8–10 The hydrothermal method is one of the economic wet chemical routes over physical deposition techniques to prepare different morphologies of nanostructured WO₃ in the presence of structure-directing agents such as acids, surfactants and metal sulphates, chlorides and sulphides.^11–13 Recently, fluoroboric acid (HBF₄) has also been used to synthesize nanoplates and is designated as a structure directing agent.¹

A comparative study demonstrates that orthorhombic WO₃ rectangular slabs, nanowires and nanorods have been synthesized by using structure directing agents for the photodecomposition of Procion Red MX-5B dye.¹⁴ In another process, Su et al. synthesized both uniform orthorhombic and monoclinic WO₃ square nanoplates with the assistance of tartaric and citric acids, respectively, by a hydrothermal process.¹⁵ An aqueous sol transforms to WO₃·2H₂O gel through an ion exchange method that develops plate-shaped WO₃ after screen printing for NO₂ sensing.¹⁶ Another research group studied the synthesis of tungsten oxide nanoplates from tungsten(VI) ethoxide under low temperature with supercritical water in a continuous flow system. Rietveld refinement confirmed the presence of 93% hexagonal crystal structures and 7% triclinic structures of the WO₃ nanoplates. The flow rates and PEG as a surface modulator play important roles in controlling the ultimate morphology.¹⁷ A crystal seed-assisted hydrothermal method has also been employed to assemble plate- and brick-like nanostructured 3WO₃·H₂O films directly onto an FTO glass substrate through using Na₂SO₄ as the capping agent. However, the nanoplate exhibits a relatively higher current density of 0.2 mA cm⁻² than partially crystalline water containing nanobrick films for both intercalation and deintercalation processes over the same period.¹⁸ Another comparative study demonstrated that the capping agents Na₂SO₄, (NH₄)₂SO₄ and CH₃COONH₄ for oxidation of methanol and water splitting have significant effects on the formation of plate, wedge and sheet-like orthorhombic WO₃ nanostructures. The highest photoconversion efficiency of 0.3% was observed for sheet-like WO₃ prepared using CH₃COONH₄ under simulated solar illumination.¹⁹ Highly porous thick and opaque WO₃ films fabricated by the sol gel method exhibited a similar reversible electrochromic property for the 1^st and 1000^th cycles of cyclic voltammograms in 1 M LiClO₄ with 0.2 mA cm⁻²as current density.²⁰ A low-cost hydrothermal approach has been taken to prepare thick plate-like monoclinic tungsten oxide nanostructured film onto FTO glass substrates for the estimation of electrochemical properties for electrochromic display window application.²¹ The process temperature variation or annealing under different conditions led to the formation of different crystal structures, in which monoclinic spherical morphology demonstrated a high degree of photocatalytic activity for oxygen evolution through water splitting in comparison with the orthorhombic phase.²² There have been many reports on the synthesis of various two-dimensional plate-shaped particles, but limited literature has been found on three-dimensional nanostructures other than spherical morphology, which encourages concentration on this particular morphology. Techniques such as Langmuir–Blodgett, dip coating and sputtering have been used for the assembly of nanomaterials. These require complicated operating procedures and specific equipment with the addition of certain surfactants.^23–25 Although literature recommends hydrothermal coating as one of the methods for thin film fabrication, a low-cost method for large-scale fabrication of tungsten oxide nanostructures is still lacking. In this context, we have synthesized three-dimensional monoclinic nanocuboids of WO₃ through the hydrothermal method and have characterized them in terms of pure-phase crystal structure, surface area and morphology. A plausible reaction process had also been suggested that optimizes the critical processing parameters to control the monoclinic cuboid structure of WO₃. Furthermore, electrical and optical properties of optimized WO₃ nanocuboids have been examined to justify the electrochromic effect of dip-coated WO₃/ITO electrodes.

2. Experimental procedure

2.1. Preparation of WO₃ nanocuboids

Sodium tungstate (Na₂WO₄·2H₂O) was used as base precursor along with fluoroboric acid as a structure-directing agent to synthesize WO₃ nanocuboids. Prior to hydrothermal treatment, fluoroboric acid (HBF₄, 50% w/w) solution was added to a sodium tungstate aqueous solution and was constantly stirred on a magnetic stirrer at 300 rpm for 30 min to transform into a yellowish green precipitate. The solution together with the precipitate was then transferred to a 50 ml Teflon beaker, placed inside an autoclave (high-pressure metal bomb) that was sealed tightly and kept at certain temperature for a predetermined time in a hot air oven. After autoclaving, the precipitate together with the solvent was centrifuged at 13 000 rpm to remove the excess HBF₄. Hot water followed by isopropanol was used for washing the precipitate. The residue after centrifuging was freeze-dried at a temperature of −52 °C and a vacuum of 10 Torr. Solute concentration and effective volume of the reaction chamber were also important considerations for the aforementioned process. The experiments were carried out by varying the concentration of HBF₄, time duration and reaction temperature to optimize the crystal structure and morphology. A series of experiments were conducted at various solute concentrations, a constant temperature of 180 °C and a constant time of 4 hours. Similarly, additional experiments were carried out to determine the effect of hydrothermal duration and temperature. The powders were characterized in terms of crystallinity, crystal structure, and morphology by different physicochemical techniques.

2.2. Characterization of WO₃ nanocuboids

X-ray diffraction (XRD) patterns for the powders were obtained using a Philips X-Ray diffractometer with Ni filtered CuK_α radiation (λ = 1.5418 Å). Rietveld refinement of the optimized data was analyzed using the FullProf Suite program and compared with the PCR database. Differential scanning calorimetry and thermogravimetric analysis (DSC–TG) of WO₃ nanocuboids after synthesis for 4 and 6 hours was carried out using a Netzsch (Germany) STA449C/4/MFC/G apparatus. FESEM images of hydrothermally treated samples were acquired to determine the morphological changes with respect to HBF₄ concentration, duration and temperature variation using the NOVA NANOSEM FEI450 system. For this, the samples were mounted on a double-sided carbon tape attached to an SEM stub and sputter-coated with gold for 120 s. The specific surface area of the powders was measured using nitrogen as the adsorbate in a BET apparatus (Quantachrome Autosorb, USA). The morphology of the WO₃ nanocuboids was further studied by transmission electron microscope (JEOL JEM-2100). Diffuse reflectance measurement was done through a Shimadzu spectrophotometer (UV-2450) to evaluate the band gap energy for WO₃ nanocuboids. Room temperature diffuse reflection percentage was measured in the wavelength region of 200–700 nm.

2.3. Fabrication and electrochemical measurement of WO₃-coated ITO

WO₃ nanocuboid particles were homogeneously dispersed in ethanol and dip-coated onto a transparent conducting oxide (TCO) substrate to evaluate the electrochromic effect of a WO₃/ITO electrode with a 2 cm² area (dimensions: 2 cm × 1 cm). A stable WO₃ suspension was prepared by ultrasonication of a solution containing 0.5 g of WO₃ nanopowder in 4 ml absolute ethanol. Commercial grade 84% optically transparent conducting Sn-doped indium oxide (ITO) glass substrate was used as the TCO substrate to fabricate the working electrode. The conductive ITO glass substrate was cleaned by ultrasonication through a successive immersion in distilled water, acetone and ethanol prior to WO₃ coating. The dip-coated samples were dried at 60 °C for 30 min.²⁶ The electrochemical properties of the WO₃/ITO electrode were determined by Cyclic Voltammetry (CV), chronoamperometry (CA) and chronocoulometry (CC) techniques. The direct optical transmittance was also measured using a UV-vis spectrophotometer with a bare ITO glass substrate as a reference electrode. The electrochemical intercalation and deintercalation of electrons and H⁺ ions were carried out using Biologic Science Instruments SP-50 controlled by a personal computer installed with EC-lab software in the three-electrode cell configuration using 1 M H₂SO₄ electrolyte solution with platinum (Pt) as a counter electrode, saturated Ag/AgCl as a reference electrode and as-prepared WO₃ films as the working electrode.

3. Results and discussion

3.1. Influence of processing conditions

Powder diffraction patterns confirmed the phase content, crystal structure and crystallinity of the WO₃ nanopowders. The room temperature XRD patterns for all of the powders are shown in Fig. 1. Initially, experiments are designed chiefly to determine the primary crystal structure and phase of the products formed upon changing the HBF₄ molar concentration (M) from stoichiometric 2 M to the higher concentration of 5 M. The other parameters such as temperature (180 °C) and time (4 hours) are kept constant during this experiment. The XRD pattern of synthesized powders under these conditions is shown in Fig. 1a. The stoichiometric HBF₄ concentration (2 M) begins to form the triclinic tungsten oxide phase mixed with the partial monoclinic phase (JCPDS card no. 830949 and 431035). The pure triclinic tungsten oxide (JCPDS card no. 830949) phase is recorded with increasing concentration of 2.5 M HBF₄. There are no peaks of a monoclinic phase found in this concentration. A pure hexagonal tungsten oxide phase (JCPDS card no. 752187) starts to form at 3 M HBF₄ concentration. In the later stage, coexistence of hexagonal and monoclinic phases is found for the 3.5 M concentration. This depicts the plausible conversion of hexagonal to monoclinic phase. The pure monoclinic crystalline tungsten oxide (JCPDS card no. 720677) phase is observed upon further increasing the concentration (4 M to 5 M), as shown in the figure. The above investigation shows that WO₃ transition follows triclinic to hexagonal followed by monoclinic, which is similar to the observation of Zadeh et al.^1,27 Both stable triclinic and monoclinic WO₃ phases are formed by varying the HBF₄ concentration in the present studied system. However, the hexagonal phase occurs as only a metastable phase and has a tendency to form a stable monoclinic phase with further HBF₄ concentration increments, particularly beyond 4 M.²⁸

Fig. 1 XRD patterns for the effect of HBF₄ molar concentration at 180 °C for 4 hours (a), effect of time at 180 °C with 4 M HBF₄ concentration (b), and effect of temperature for 6 hours at 4 M HBF₄ concentration (c), where * = triclinic, # = monoclinic, ^ = hexagonal, @ = orthorhombic tungstite crystal structure.

The growth phenomenon depends on time under isothermal conditions irrespective of any process condition.²⁹ Hence, it is necessary to understand the effect of time on the formation of the pure phase along with the desired morphology. A time variation is done starting from 0–8 hours while keeping the other parameters such as HBF₄ concentration (4 M) and temperature (180 °C) constant, as shown in Fig. 1b. The molar concentration of HBF₄ is chosen to be 4 M as formation of only the monoclinic phase is favored for HBF₄ concentrations greater than 4 M. Zero hour is defined as before insertion of the closed vessel in the hydrothermal process chamber. Immediate HBF₄ addition results in orthorhombic tungstic acid (Tungstite, JCPDS card no. 840886) along with some impurity peaks. The pure orthorhombic tungstite (Tungstite, JCPDS card no. 840886) phase is formed at only 2 hours duration. Interestingly, the crystalline monoclinic phase has been detected for the duration of 4 hours and the phase remains constant beyond this time limit. This result depicts the periodic formation of orthorhombic tungstite (H₂WO₄) and monoclinic WO₃ crystal structures by varying the hydrothermal duration. Process control variables suggest that the optimized conditions for obtaining pure monoclinic tungsten oxide is 4 M HBF₄ concentration, 4 hours duration and 180 °C temperature. However, these conditions provide low crystalline cuboid morphology and subsequent band-gap energy due to the content of crystalline water.³⁰ The presence of water content is further confirmed by thermal analysis in the next section. In this result, 6 hours processing time is again considered as the optimum hydrothermal time. However, no significant change in the crystallinity is observed after the hydrothermal duration of 6 hours. Every synthesis requires a minimum thermal energy to overcome the potential energy barrier for the completion of the targeted reaction. To optimize this, additional studies on varying the reaction temperature (160 °C to 200 °C) have also been carried out at constant conditions such as 4 M HBF₄ concentration and 6 hours reaction duration. Fig. 1c clearly represents orthorhombic tungstic acid at 160 °C. However, a mixture of orthorhombic tungstic acid and monoclinic tungsten oxide results at 170 °C, which indicates the initialization of monoclinic crystal phase at such temperature. Additionally, incrementing the temperature up to 200 °C only enhances the crystallinity of a monoclinic phase or in another sense decreases the surface area. Thus, optimized conditions for the synthesis of WO₃ nanocuboids from Na₂WO₄·2H₂O and HBF₄ include temperature of 180 °C, time of 6 hours and HBF₄ concentration of 4 M.

3.2. Rietveld refinement for optimum WO₃ nanocuboids

In a recent article, Ma et al.¹ started from the aforesaid precursors and reported an identical XRD pattern up to a maximum 2θ value of 50° and concluded the structure to be triclinic as per JCPDS – 321395. In comparison, the present XRD pattern for extended diffraction angle 15–80° (2θ) consists of broader peaks at higher angles of 53–57° (2θ). This peak in particular matches well with the monoclinic instead of triclinic tungsten oxide crystal structure, as confirmed from the Rietveld profile fitting method (Fig. 2). Tungsten oxide is calculated as a pure monoclinic structure (PCR file code: wo3_p21n_80056, a = 7.39773, b = 7.49056, c = 7.64700, β = 88.7102° with space group P21/n). The crystal structure model of WO₃ (PCR file code: 80056) is used as a starting point for the refinement. The characteristic reflections of WO₃ are pointed out as (200), (020) and (002). It is observed that the reflections are merged into a broader peak centered at the (020) position, which may be due to the overlap of the strongly broadened peaks. The Rietveld refinement has the lowest indices values, which are recorded as R_p = 23.2, R_wp = 29.9 and χ² = 1.34, respectively.³¹

Fig. 2 Rietveld refinement of WO₃ nanocuboids prepared at conditions of 180 °C, 6 hours and 4 M HBF₄.

3.3. Thermal analysis of WO₃ nanocuboids

Thermal analysis of WO₃ nanocuboids for both 4 hours and 6 hours is presented in Fig. 3a and b, respectively. The thermogram of 4 hours illustrates the instability of the compound until 280 °C. A total of 3.5% weight loss is observed from room temperature to 280 °C which corresponds to the presence of water molecules in the WO₃ nanocuboid crystal system. In the DSC plot, the endothermic peak near 80 °C corresponds to the loss of physisorbed water. A broad exothermic peak at around 550 °C is associated with the further crystallization of WO₃. However, the thermogram of 6 hours reveals high thermal stability of the compound until 600 °C as there is no weight loss, although plausible crystallization at high temperature has been depicted by a small exothermic hump. This study predicts that the synthesized particles have a tendency to a high degree of crystallinity at high temperature as the powder synthesized at low temperature has a tendency to form a semi-crystalline structure.³² Phase changes of specific cuboid WO₃ morphology during hydrothermal reaction follows orthorhombic tungstite → monoclinic WO₃ transformation, which has already been confirmed through crystal phase and thermal analysis at optimum HBF₄ molar concentration. The aforementioned process conversion can be represented through eqn (1)–(3).

Na₂WO₄·2H₂O + 2HBF₄ → H₂WO₄ (orthorhombic) + 2NaBF₄ + 2H₂O (1)

H₂WO₄ (orthorhombic) → WO₃·xH₂O (monoclinic) + H₂O (2)

WO₃·xH₂O → WO₃ + xH₂O (3)

Fig. 3 TG-DSC plot of WO₃ nanocuboids at (a) 4 hours and (b) 6 hours.

The value of x is found as 0.5 and is confirmed from the thermogravimetric data presented in Fig. 3.

3.4. Morphological analysis of WO₃ nanocuboids

The reaction sequence, concentration of HBF₄, temperature and time are varied to obtain different morphologies, which are represented in FESEM images in Fig. 4–6. Discrete small dots appear due to gold coating. Detail analysis suggests that the 4 M HBF₄ concentration develops an adequate crystal structure, crystallinity and desired morphology. Hence, lower and upper limits such as 3.5 M and 4.5 M HBF₄ concentration have been considered for fine tuning and detailed analysis. Fig. 4a represents three particles prepared at 3.5 M, 4 M and 4.5 M HBF₄ concentrations, respectively. At low concentration, formation of partially agglomerated spherical particles along with plate-shaped particles is preferred instead of formation of nanocuboids. However, at the optimum 4 M concentration, nanocuboid morphology is obtained, as shown in Fig. 4b. From the image, typical dimensions of the cuboid appear to be 130 nm length, 110 nm width and 85 nm thickness. However, further concentration increment has no distinct influence on the morphology, as shown in Fig. 4c. Fig. 5 shows the effect of hydrothermal duration on the morphology of WO₃ and the plot of specific surface area versus time to determine the growth phenomenon affected by the surface activity. A hydrothermal duration of 2 hours initially develops uniform plate-like morphology with an average particle size of 400 nm and an orthorhombic tungstite crystal structure. Herein, the selected hydrothermal method follows the expected development of WO₃ crystal from aqueous metal salt solutions under critical pressure at a pertinent temperature and solution pH through the structure directing agents. The WO₃ nanoparticle formation mechanism with particular growth direction depends on the solubility of the metal oxides and the reaction kinetics during the synthesis. It significantly alters the critical point of temperature assistance to form high pressure and changes the resultant dielectric properties of aqueous media.³³ At the initial stage, tungstite (H₂WO₄) formation occurs with a structure of layers of WO₆ octahedra sharing four equatorial oxygen atoms which are linked through hydrogen bonds derived from interaction between water molecules and oxygen present in the axial position of the octahedra. Plate-shaped particles are formed due to inhibition of the (010) crystal plane, which is the normal growth direction of the H₂WO₄ structure.¹ Due to insufficient BF₄⁻ anion concentration, structural orientation of plates cannot be induced by adding the BF₄⁻ anion to the hydrogen bond system to occupy the apexes with a boron atom at the center. Thus, the growth axis is not restricted but rather prefers growth of the structure forming cuboids. Moreover, with further time increments up to 4 hours; the cuboid-like WO₃ particles are mixed with a fraction of the plate-like particles. This reveals that the growth mechanism follows a plate- to cuboid-like morphology. Nearly homogeneous cuboid-like morphology has been observed for 6 hours without formation of plates. A gradual decrease in surface area is observed with the time increment because of the agglomeration of particles, which reveals that due to the highly active surface, the particles tend to stabilize themselves following the agglomeration process. Fig. 6 represents the effects of lower and upper temperature range from the critical temperature of 180 °C. Plate-like particles and more agglomerated cuboid-like particles are confirmed at 170 °C and 190 °C temperatures, respectively. Fig. 7 represents the TEM micrograph including the HRTEM image and SAED pattern for WO₃ nanocuboids. Fig. 7a shows soft agglomerated distinct nanocuboid morphology of WO₃ nanoparticles with an average particle length of 140 nm and width of 120 nm. The topographical features reveal that the particle edges are imperfect, which is attributed to the effect of certain tilting angles of individual cuboid particles, whereas the TEM of plate morphology exhibits perfect edges, as reported by Su et al.¹⁵ Fig. 7b represents the HR-TEM image of a single nanocuboid. The image clearly shows the resolved lattice fringes in the visible range, indicating its single crystalline nature. The lattice distance is found to be 0.36 nm which can be readily indexed to the (002) plane of monoclinic WO₃ nanocuboids. The value calculated also resembles the value calculated from the X-ray diffraction pattern, which indicates the tentative growth of the particle along this crystal plane. Fig. 7c shows the selected area electron diffraction (SAED) pattern of a single crystalline WO₃ nanocuboid. In general, h = 0, k = 0 and l = x (where x = 1, 2 or 4) planes restrict the 3D growth pattern and develop 2D plate-shaped morphology.³⁴ However, the TEM illustration depicts that the highly crystalline WO₃ exhibits an excellent growth phenomenon along the [002] direction to form cuboid morphology.

Fig. 4 FESEM images under experimental conditions of time 6 hours and temperature 180 °C at (a) 3.5 M, (b) 4 M and (c) 4.5 M HBF₄ concentrations.

Fig. 5 FESEM images under experimental conditions of 4 M HBF₄ concentration and temperature 180 °C for (a) 2 hours, (b) 4 hours, (c) 6 hours and (d) 8 hours time duration and (e) BET surface area vs. time plot.

Fig. 6 FESEM images under experimental conditions of 4 M HBF₄ concentration for time 6 hours at (a) 170 °C and (b) 190 °C temperature.

Fig. 7 TEM analysis of WO₃ nanocuboids: (a) morphology, (b) d-spacing and (c) SAED pattern.

3.5. Band gap of WO₃ nanocuboids

Band-gap energy of the selected powder is measured from the UV-vis diffuse reflectance spectra. The measured diffuse reflectance spectra (Fig. 8) have been used for the estimation of band-gap energy from the Tauc plot, which is the square root of the Kubelka–Munk function multiplied by the photon energy and plotted against the photon energy (E_photon = hλ), as shown in the inset of Fig. 8. The Kubelka–Munk unit of absorption is calculated from the following equation:

Fig. 8 UV-DRS of 2 M HBF₄, 3 M HBF₄, 4 hours and 6 hours. Inset represents the Tauc plot.

KMU = (1 − R)²/2R, where R = reflectance.¹⁰ Band gap energy is calculated after synthesis at different conditions. For example, 2 M HBF₄ concentration at 180 °C for 4 hours with triclinic WO₃ structure, 3 M HBF₄ concentration at 180 °C for 4 hours with hexagonal WO₃ structure, 4 M HBF₄ concentration at 180 °C for both 4 and 6 hours with monoclinic WO₃ structure are considered for the band-gap measurements. The band gap energy of triclinic and hexagonal crystal structures is calculated to be 3.25 eV and 2.85 eV, respectively, as illustrated by the 3.09 eV value for a hexagonal structure in previous report.¹¹ A nearly equal range of band-gap energy of 2.75 eV is found for both 4 and 6 hours treated monoclinic WO₃ nanocuboids without a significant effect of crystalline water.

3.6. Electrochromic response of WO₃ nanocuboids

The optimized WO₃ nanocuboid coated onto the ITO conducting substrate (WO₃/ITO) is used as the working electrode for electrochemical analysis. The fabricated dip-coated WO₃/ITO electrodes are semi-transparent in nature. The direct optical transmittance of the fabricated electrode has a transparency of about 52%. In order to evaluate the electrochromic properties, typical cyclic voltammograms (CV) have been taken between −1.0 and +1.0 V at a scan rate of 100 mV s⁻¹ using 1 M H₂SO₄ as the working electrolyte. Fig. 9a shows the cyclic voltammograms recorded for WO₃/ITO electrodes at the 1^st, 100^th and 500^th cycles. It can be seen that the shapes of the CV curves remain unchanged after the 100^th and 500^th cycles with negligible increased current response, which indicates good cycling stability of the WO₃/ITO electrodes. This is probably due to the good distribution of the particles on the ITO surface after drying of the WO₃ nanocuboid dip-coated ITO electrodes. The topographical FESEM image shows the distribution and texture of particles on the ITO glass surface. It is evident from the image that the particles are relatively well distributed and interconnected. However, the small (∼200 nm) pin holes may provide additional ion transportation phenomena and subsequent electrochromic behavior. The electrochromic mechanism of WO₃/ITO electrodes in the H⁺ electrolyte can be well expressed as follows:

WO₃ + xe⁻+ xH⁺ ↔ H_xWO₃ (4)

Fig. 9 (a) Cyclic voltammograms of WO₃/ITO film at 100 mV s⁻¹ for 1^st, 100^th and 500^th cycle. (b) Topographical image of dip-coated WO₃ nanocuboids onto ITO glass substrate.

Unique coloration and bleaching processes have been observed for WO₃ nanocuboid coated ITO electrodes during the intercalation and deintercalation of electrons from the WO₃/ITO electrode and H⁺ ions from the electrolyte. As shown in the CV curves, the current is found to move negatively with decreasing voltage, corresponding to co-intercalation of electrons and H⁺ into WO₃ to form hydrogen tungsten bronze.²³ Cathodic and anodic current peaks are observed near −0.25 V and 0.1 V, respectively. The current density observed for the WO₃/ITO electrode is near 3.15 mA cm⁻². Most importantly, a high symmetry of the anodic and cathodic peak is found, which reveals the better electrochromic behavior of WO₃ nanocuboids. A reversible color change of the WO₃/ITO electrodes with varying voltages has been observed. The initial WO₃/ITO electrode is light green in color, but upon application of −0.5 V, the electrode color changes to light blue which then further turns to darker blue at −1.0 V. Bleaching of the electrode occurs at positive voltages of 0.5 V and 1.0 V. Thus, we can conclude that during the deintercalation of H⁺ ions, the color of the electrodes bleaches gradually with increasing voltage before the original color is finally recovered. This observation predicts that positive voltage bleaching of the electrodes occurs. The estimation of coloration and bleaching time for the WO₃/ITO electrode is known from the current time transients. A typical CA graph is shown in Fig. 10 for the coloration and bleaching of WO₃/ITO electrodes. During the measurement of CA, experimental increments of voltage are allowed to sweep from the rest potential of 0.0 V to −1 V for a period of 10 s to obtain the coloration phenomenon of the electrodes. Similarly, in reverse, a positive voltage of 1 V is applied for the next 10 seconds, which results in bleaching of electrodes. Interesting coloration and bleaching kinetics are recorded from this analysis. Diffusion of H⁺ to the electrolyte (bleaching process) is faster than that from the electrolyte (coloration process) and hence needs 3.18 seconds and 5.05 seconds, respectively. The faster bleaching kinetics than the coloration kinetics is governed possibly by space charge transfer through the electrode and potential barrier at the electrolyte–WO₃ interface, respectively, which controls the two processes diversely.

Fig. 10 Chronoamperometry (CA) measurement for WO₃/ITO film.

In order to study the amount of charge intercalation of e⁻ and H⁺ ions from an electrolyte with respect to time, chronocoulometry measurement of WO₃/ITO electrode is carried out at voltage sweep range between −1.0 V and +1.0 V. The CC measurement for intercalation and deintercalation process is carried out for the step of 10 seconds each for the forward and backward scan. Fig. 11 represents a typical CC plot of the WO₃/ITO electrode for the first cycle, and the inset figure represents the CC plot for first 15 cycles. During the forward scan of the WO₃/ITO electrode, the diffusion process leads to the charge intercalation by reducing W⁶⁺ to W⁵⁺ states. In the reverse, removal of intercalated charge occurs during the backward bleaching process due to oxidation of W⁵⁺ to W⁶⁺ states. The charge intercalation and deintercalation quantified data from the plot have been used to calculate the reversibility of the coloration/bleaching processes. The reversibility of the film can be calculated from the following relation:

where Q_di and Q_i refers to the amount of charge deintercalated and intercalated in the WO₃/ITO films. The percentage of electrochromic reversibility of the WO₃/ITO electrode for first cycle is 72.2%. There is a slight increment in the reversibility after first CC cycle which remains almost constant until the 15^th cycle. The reversibility difference between the 1^st and 2^nd cycle is approximately 5%. From CV, the electrode exhibits high stability until the 500^th cycle; therefore, the reversibility obviously remains the same. UV-visible transmittance spectroscopy is performed for the WO₃/ITO electrodes in the wavelength range of 200–900 nm after performing the electrochemical analysis. Fig. 12 shows the room temperature optical transmittance data for WO₃/ITO colored and WO₃/ITO bleached electrodes. The merit of optical performance is to quantify the optical density change (ΔOD) and the coloration efficiency of the WO₃/ITO films. The optical transmittance has been found to decrease with the coloration of the films and then to increase upon the deintercalation process. The transmittance value of WO₃/ITO electrodes at a visible wavelength of 550 nm is utilized for determining the optical density changes. The ΔOD calculation is done as per the following relation:³⁵where T_b and T_c are the transmittance value in the bleached and colored state, respectively. The achieved transmittance efficiency for both bleached and colored monoclinic WO₃ thin film is well comparable with reported data.³⁶ The coloration and bleaching of WO₃/ITO electrode is carried out by applying a potential step of −1.0 V to +1.0 V for a fixed time. One of the prime parameter to characterize an electrochromic material is coloration efficiency (CE) of the electrodes. In order to quantify the efficiency during the induction of charge in an electrochromic device, CE is represented as the change in optical density (ΔOD) per unit charge insertion per unit area. It is given by the following relation mentioned below:where Q_i is the amount of charge intercalated in the WO₃/ITO electrodes to cause changes in optical density (ΔOD) and A is the area of the electrode. Thus, the coloration efficiency as calculated for the WO₃/ITO electrode has been found to be 60.17 cm² C⁻¹. The result obtained is in accordance with reports by other groups.^37,38

Fig. 11 Chronocoulometry (CC) measurement for WO₃/ITO film (inset represents the CC for the first 15 cycles).

Fig. 12 Optical transmittance spectra versus wavelength of colored and bleached films.

4. Conclusion

Initial variation of HBF₄ concentration at a temperature of 180 °C and time of 4 h follows triclinic WO₃ → hexagonal WO₃ → monoclinic WO₃ phase transformation to obtain WO₃ nanocuboids from sodium tungstate. However, monoclinic and cuboid tungsten trioxide develops through orthorhombic tungstite → monoclinic WO₃ at an optimum fluoroboric acid concentration. Hydrothermal time, temperature and structure directing reagent concentration have significant influence in achieving desired phase and morphology. The nanocuboids exhibit optimum band gap and adherence to develop dip-coated WO₃/ITO electrodes. Tungsten trioxide-coated ITO electrodes have remarkable electrochemical response, electrochromic efficiency, reversibility, and optical switching characteristics. The developed process and achievement of the electrochromic response is considerably appreciable, which might have economic viability for fabricating tailor-made smart glass.

References

1. J. Ma, J. Zhang, S. Wang, T. Wang, J. Lian, X. Duan and W. Zheng, J. Phys. Chem. C, 2011, 115, 18157.
2. L. Liang, J. Zhang, Y. Zhou, J. Xie, X. Zhang, M. Guan, B. Pan and Y. Xie, Sci. Rep., 2013, 3, 1936.
3. X. He, C. Hu, Q. Yi, X. Wang, H. Hua and X. Li, Catal. Lett., 2012, 142, 637.
4. S. Vallejos, T. Stoycheva, P. Umek, C. Navio, R. Synders, C. Bittencourt, E. Llobet, C. Blackman, S. Moniz and X. Correig, Chem. Commun., 2011, 47, 565.
5. B. X. Li, G. X. Rong, Y. Xie, L. F. Huang and C. Q. Feng, Inorg. Chem., 2006, 45, 6404.
6. Q. Sun, J. Luo, Z. Xie, J. Wang and X. Su, Mater. Lett., 2008, 62, 2992.
7. A. Wolcott, T. R. Kuykendall, W. Chen, S. Chen and J. Z. Zhang, J. Phys. Chem. B, 2006, 110, 25288.
8. Z. Zhao and M. Miyauchi, J. Phys. Chem. C, 2009, 113, 6539.
9. Y. H. Jun, C. Y. Qi, Y. Fang, P. Y. Hua, H. X. Wu, Z. Ding and T. D. Sheng, Chin. Phys. B, 2011, 20, 036103.
10. B. Yang, P. R. F. Barnes, Y. Zhang and V. Luca, Catal. Lett., 2007, 118, 280.
11. H. Hassani, E. Marzbanrad, C. Zamani and B. Raissi, J. Mater. Sci.: Mater. Electron., 2011, 22, 1264.
12. A. Novinrooz, M. Sharbatdaran and H. Noorkojouri, Cent. Eur. J. Phys., 2005, 3(3), 456.
13. X. Song, Y. Zhao and Y. Zheng, Mater. Lett., 2006, 60, 3405.
14. S. Rajagopal, D. Nataraj, D. Mangalaraj, Y. Djaoued, J. Robichand and O. Y. Khyzhun, Nanoscale Res. Lett., 2009, 4, 1335.
15. X. Su, F. Xiao, Y. Li, J. Jian, Q. Sun and J. Wang, Mater. Lett., 2010, 64, 1232.
16. Y. Choi, G. Sakai, K. Shimanoe, N. Miura and N. Yamazoe, Sens. Actuators, B, 2003, 95, 258.
17. M. G. Fabra, P. Dunne, D. Grant, P. Gooden and E. Lester, Chem. Eng. J., 2013, 226, 22.
18. Z. Jiao, X. Wang, J. Wang, L. Ke, H. V. Demir, T. W. Koh and X. W. Sun, Chem. Commun., 2012, 48, 365.
19. Z. Jiao, J. Wang, L. Ke, X. W. Sun and H. V. Demir, ACS Appl. Mater. Interfaces, 2011, 3, 229.
20. G. Leftheriotis, S. Papaefthimiou, P. Yianoulis, A. Siokou and D. Kefalas, Appl. Surf. Sci., 2003, 218, 275.
21. Z. Jiao, X. W. Sun, J. Wang, L. Ke and H. V. Demir, J. Phys. D: Appl. Phys., 2010, 43, 285501.
22. G. Xin, W. Guo and T. Ma, Appl. Surf. Sci., 2009, 256, 165.
23. X. Chang, S. Sun, Z. Li, X. Xu and Y. Qui, Appl. Surf. Sci., 2011, 257, 5726.
24. D. S. Hinczewski, M. Hinczewski, I. Sorar, F. Z. Tepehan and G. G. Tepehan, Sol. Energy Mater., 2008, 92, 821.
25. N. M. G. Parreira, N. J. M. Carvalho and A. Cavaleiro, Thin Solid Films, 2006, 510, 191.
26. J. K. Kim, K. Shin, S. M. Cho, T. Lee and J. H. Park, Energy Environ. Sci., 2011, 4, 1465.
27. H. Zheng, J. Ou, M. S. Strano, R. B. Kaner, A. Mitchell and K. K. Zadeh, Adv. Funct. Mater., 2011, 21, 2175.
28. I. M. Szilagyi, L. Wang, P. Gouma, C. Balazsi, J. Madarasz and G. Pokol, Mater. Res. Bull., 2009, 44, 505.
29. D. Sarkar, M. Chu, S. Cho, Y. Kim and B. Basu, J. Am. Ceram. Soc., 2009, 92, 2877.
30. P. Reiss, M. Protiere and L. Li, Small, 2009, 5, 154.
31. S. Pokhrel, J. Birkenstock, M. Schowalter, A. Rosenauer and L. Madler, Cryst. Growth Des., 2010, 10, 632.
32. F. Huang, H. Zhang and J. F. Banfield, Nano Lett., 2003, 3, 373.
33. H. Hayashi and Y. Hakuta, Materials, 2010, 3, 3794.
34. D. Chen, H. Wang, R. Zhang, L. Gao, Y. Sugahara and A. Yasumori, J. Ceram. Process. Res., 2008, 9, 596.
35. 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., 2013, 1, 3722.
36. C. K. Wang, C. K. Lin, C. L. Wu, S. C. Wang and J. L. Huang, Electrochim. Acta, 2013, 112, 24.
37. F. Lin, J. Cheng, C. Engtrakul, A. C. Dillon, D. Nordlund, R. G. Moore, T. C. Weng, S. K. R. Williams and R. M. Richards, J. Mater. Chem., 2012, 22, 16817.
38. M. Deepa, R. Sharma, A. Basu and S. A. Agnihotry, Electrochim. Acta, 2005, 50, 3545.

---